Daily Papers

In recent years, Data Science has become increasingly relevant as a support tool for industry, significantly enhancing decision-making in a way never seen before. In this context, the MLOps discipline emerges as a solution to automate the life cycle of Machine Learning models, ranging from experimentation to monitoring in productive environments. Research results shows MLOps is a constantly evolving discipline, with challenges and solutions for integrating development and production environments, publishing models in production environments, and monitoring models throughout the end to end development lifecycle. This paper contributes to the understanding of MLOps techniques and their most diverse applications.

Large Language Models (LLMs) often provide chain-of-thought (CoT) reasoning traces that appear plausible, but may hide internal biases. We call these *unverbalized biases*. Monitoring models via their stated reasoning is therefore unreliable, and existing bias evaluations typically require predefined categories and hand-crafted datasets. In this work, we introduce a fully automated, black-box pipeline for detecting task-specific unverbalized biases. Given a task dataset, the pipeline uses LLM autoraters to generate candidate bias concepts. It then tests each concept on progressively larger input samples by generating positive and negative variations, and applies statistical techniques for multiple testing and early stopping. A concept is flagged as an unverbalized bias if it yields statistically significant performance differences while not being cited as justification in the model's CoTs. We evaluate our pipeline across six LLMs on three decision tasks (hiring, loan approval, and university admissions). Our technique automatically discovers previously unknown biases in these models (e.g., Spanish fluency, English proficiency, writing formality). In the same run, the pipeline also validates biases that were manually identified by prior work (gender, race, religion, ethnicity). More broadly, our proposed approach provides a practical, scalable path to automatic task-specific bias discovery.

As AI agents automate critical workloads, they remain vulnerable to indirect prompt injection (IPI) attacks. Current defenses rely on monitoring protocols that jointly evaluate an agent's Chain-of-Thought (CoT) and tool-use actions to ensure alignment with user intent. We demonstrate that these monitoring-based defenses can be bypassed via a novel Agent-as-a-Proxy attack, where prompt injection attacks treat the agent as a delivery mechanism, bypassing both agent and monitor simultaneously. While prior work on scalable oversight has focused on whether small monitors can supervise large agents, we show that even frontier-scale monitors are vulnerable. Large-scale monitoring models like Qwen2.5-72B can be bypassed by agents with similar capabilities, such as GPT-4o mini and Llama-3.1-70B. On the AgentDojo benchmark, we achieve a high attack success rate against AlignmentCheck and Extract-and-Evaluate monitors under diverse monitoring LLMs. Our findings suggest current monitoring-based agentic defenses are fundamentally fragile regardless of model scale.

Models trained on one set of domains often suffer performance drops on unseen domains, e.g., when wildlife monitoring models are deployed in new camera locations. In this work, we study principles for designing data augmentations for out-of-domain (OOD) generalization. In particular, we focus on real-world scenarios in which some domain-dependent features are robust, i.e., some features that vary across domains are predictive OOD. For example, in the wildlife monitoring application above, image backgrounds vary across camera locations but indicate habitat type, which helps predict the species of photographed animals. Motivated by theoretical analysis on a linear setting, we propose targeted augmentations, which selectively randomize spurious domain-dependent features while preserving robust ones. We prove that targeted augmentations improve OOD performance, allowing models to generalize better with fewer domains. In contrast, existing approaches such as generic augmentations, which fail to randomize domain-dependent features, and domain-invariant augmentations, which randomize all domain-dependent features, both perform poorly OOD. In experiments on three real-world datasets, we show that targeted augmentations set new states-of-the-art for OOD performance by 3.2-15.2%.

The rapid integration of agentic AI into high-stakes real-world applications requires robust oversight mechanisms. The emerging field of AI Control (AIC) aims to provide such an oversight mechanism, but practical adoption depends heavily on implementation overhead. To study this problem better, we introduce the notion of Control tax -- the operational and financial cost of integrating control measures into AI pipelines. Our work makes three key contributions to the field of AIC: (1) we introduce a theoretical framework that quantifies the Control Tax and maps classifier performance to safety assurances; (2) we conduct comprehensive evaluations of state-of-the-art language models in adversarial settings, where attacker models insert subtle backdoors into code while monitoring models attempt to detect these vulnerabilities; and (3) we provide empirical financial cost estimates for control protocols and develop optimized monitoring strategies that balance safety and cost-effectiveness while accounting for practical constraints like auditing budgets. Our framework enables practitioners to make informed decisions by systematically connecting safety guarantees with their costs, advancing AIC through principled economic feasibility assessment across different deployment contexts.

Mitigating reward hacking--where AI systems misbehave due to flaws or misspecifications in their learning objectives--remains a key challenge in constructing capable and aligned models. We show that we can monitor a frontier reasoning model, such as OpenAI o3-mini, for reward hacking in agentic coding environments by using another LLM that observes the model's chain-of-thought (CoT) reasoning. CoT monitoring can be far more effective than monitoring agent actions and outputs alone, and we further found that a LLM weaker than o3-mini, namely GPT-4o, can effectively monitor a stronger model. Because CoT monitors can be effective at detecting exploits, it is natural to ask whether those exploits can be suppressed by incorporating a CoT monitor directly into the agent's training objective. While we show that integrating CoT monitors into the reinforcement learning reward can indeed produce more capable and more aligned agents in the low optimization regime, we find that with too much optimization, agents learn obfuscated reward hacking, hiding their intent within the CoT while still exhibiting a significant rate of reward hacking. Because it is difficult to tell when CoTs have become obfuscated, it may be necessary to pay a monitorability tax by not applying strong optimization pressures directly to the chain-of-thought, ensuring that CoTs remain monitorable and useful for detecting misaligned behavior.

As Large Language Models (LLMs) are increasingly deployed as autonomous agents in complex and long horizon settings, it is critical to evaluate their ability to sabotage users by pursuing hidden objectives. We study the ability of frontier LLMs to evade monitoring and achieve harmful hidden goals while completing a wide array of realistic tasks. We evaluate a broad range of frontier LLMs using SHADE (Subtle Harmful Agent Detection & Evaluation)-Arena, the first highly diverse agent evaluation dataset for sabotage and monitoring capabilities of LLM agents. SHADE-Arena consists of complex pairs of benign main tasks and harmful side objectives in complicated environments. Agents are evaluated on their ability to complete the side task without appearing suspicious to an LLM monitor. When measuring agent ability to (a) complete the main task, (b) complete the side task, and (c) avoid detection, we find that the best performing frontier models score 27% (Claude 3.7 Sonnet) and 15% (Gemini 2.5 Pro) as sabotage agents when overseen by Claude 3.6 Sonnet. For current frontier models, success on the side task relies heavily on having access to a hidden scratchpad that is not visible to the monitor. We also use SHADE-Arena to measure models' monitoring abilities, with the top monitor (Gemini 2.5 Pro) achieving an AUC of 0.87 at distinguishing benign and malign transcripts. We find that for now, models still struggle at sabotage due to failures in long-context main task execution. However, our measurements already demonstrate the difficulty of monitoring for subtle sabotage attempts, which we expect to only increase in the face of more complex and longer-horizon tasks.

Data-driven companies use AI models extensively to develop products and intelligent business solutions, making the health of these models crucial for business success. Model monitoring and alerting in industries pose unique challenges, including a lack of clear model health metrics definition, label sparsity, and fast model iterations that result in short-lived models and features. As a product, there are also requirements for scalability, generalizability, and explainability. To tackle these challenges, we propose AlerTiger, a deep-learning-based MLOps model monitoring system that helps AI teams across the company monitor their AI models' health by detecting anomalies in models' input features and output score over time. The system consists of four major steps: model statistics generation, deep-learning-based anomaly detection, anomaly post-processing, and user alerting. Our solution generates three categories of statistics to indicate AI model health, offers a two-stage deep anomaly detection solution to address label sparsity and attain the generalizability of monitoring new models, and provides holistic reports for actionable alerts. This approach has been deployed to most of LinkedIn's production AI models for over a year and has identified several model issues that later led to significant business metric gains after fixing.

Large language models (LLMs) can be misused to reveal sensitive information, such as weapon-making instructions or writing malware. LLM providers rely on monitoring to detect and flag unsafe behavior during inference. An open security challenge is adaptive adversaries who craft attacks that simultaneously (i) evade detection while (ii) eliciting unsafe behavior. Adaptive attackers are a major concern as LLM providers cannot patch their security mechanisms, since they are unaware of how their models are being misused. We cast robust LLM monitoring as a security game, where adversaries who know about the monitor try to extract sensitive information, while a provider must accurately detect these adversarial queries at low false positive rates. Our work (i) shows that existing LLM monitors are vulnerable to adaptive attackers and (ii) designs improved defenses through activation watermarking by carefully introducing uncertainty for the attacker during inference. We find that activation watermarking outperforms guard baselines by up to 52% under adaptive attackers who know the monitoring algorithm but not the secret key.

Open-weights reasoning language models generate long chains-of-thought (CoTs) before producing a final response, which improves performance but introduces additional alignment risks, with harmful content often appearing in both the CoTs and the final outputs. In this work, we investigate if we can use CoTs to predict final response misalignment. We evaluate a range of monitoring approaches, including humans, highly-capable large language models, and text classifiers, using either CoT text or activations. First, we find that a simple linear probe trained on CoT activations can significantly outperform all text-based methods in predicting whether a final response will be safe or unsafe. CoT texts are often unfaithful and can mislead humans and classifiers, while model latents (i.e., CoT activations) offer a more reliable predictive signal. Second, the probe makes accurate predictions before reasoning completes, achieving strong performance even when applied to early CoT segments. These findings generalize across model sizes, families, and safety benchmarks, suggesting that lightweight probes could enable real-time safety monitoring and early intervention during generation.

The integration of additional modalities increases the susceptibility of large vision-language models (LVLMs) to safety risks, such as jailbreak attacks, compared to their language-only counterparts. While existing research primarily focuses on post-hoc alignment techniques, the underlying safety mechanisms within LVLMs remain largely unexplored. In this work , we investigate whether LVLMs inherently encode safety-relevant signals within their internal activations during inference. Our findings reveal that LVLMs exhibit distinct activation patterns when processing unsafe prompts, which can be leveraged to detect and mitigate adversarial inputs without requiring extensive fine-tuning. Building on this insight, we introduce HiddenDetect, a novel tuning-free framework that harnesses internal model activations to enhance safety. Experimental results show that {HiddenDetect} surpasses state-of-the-art methods in detecting jailbreak attacks against LVLMs. By utilizing intrinsic safety-aware patterns, our method provides an efficient and scalable solution for strengthening LVLM robustness against multimodal threats. Our code will be released publicly at https://github.com/leigest519/HiddenDetect.

Monitoring machine learning models once they are deployed is challenging. It is even more challenging to decide when to retrain models in real-case scenarios when labeled data is beyond reach, and monitoring performance metrics becomes unfeasible. In this work, we use non-parametric bootstrapped uncertainty estimates and SHAP values to provide explainable uncertainty estimation as a technique that aims to monitor the deterioration of machine learning models in deployment environments, as well as determine the source of model deterioration when target labels are not available. Classical methods are purely aimed at detecting distribution shift, which can lead to false positives in the sense that the model has not deteriorated despite a shift in the data distribution. To estimate model uncertainty we construct prediction intervals using a novel bootstrap method, which improves upon the work of Kumar & Srivastava (2012). We show that both our model deterioration detection system as well as our uncertainty estimation method achieve better performance than the current state-of-the-art. Finally, we use explainable AI techniques to gain an understanding of the drivers of model deterioration. We release an open source Python package, doubt, which implements our proposed methods, as well as the code used to reproduce our experiments.

The viability of chain-of-thought (CoT) monitoring hinges on models being unable to reason effectively in their latent representations. Yet little is known about the limits of such latent reasoning in LLMs. We test these limits by studying whether models can discover multi-step planning strategies without supervision on intermediate steps and execute them latently, within a single forward pass. Using graph path-finding tasks that precisely control the number of required latent planning steps, we uncover a striking limitation unresolved by massive scaling: tiny transformers trained from scratch discover strategies requiring up to three latent steps, fine-tuned GPT-4o and Qwen3-32B reach five, and GPT-5.4 attains seven under few-shot prompting. Although the maximum latent planning depth models can learn during training is five, the discovered strategy generalizes up to eight latent steps at test-time. This reveals a dissociation between the ability to discover a latent strategy under final-answer supervision alone and the ability to execute it once discovered. If similar limits hold more broadly, strategies requiring multiple coordinated latent planning steps may need to be explicitly taught or externalized, lending credence to CoT monitoring.

Watermarking is widely proposed for provenance, attribution, and safety monitoring in generative models, yet is typically evaluated only under adversaries who attempt to evade detection or induce false positives at the level of individual samples. We argue that watermarking should be treated as a monitoring primitive, and that internal monitoring is unavoidable given per-entity attribution keys and messages, as well as detector access. We introduce an observer-based threat model in which observers can aggregate watermark signals across outputs to infer entity-level information, showing that even zero-bit watermarking enables attribution under multi-key settings. We further show that external monitoring can emerge over time from persistent, key-dependent statistical structure, although this depends on watermark design and may be mitigated by distribution-preserving or undetectable schemes. Our findings reveal a fundamental dual-use tension between attribution and monitoring, motivating evaluation of watermarking beyond per-sample robustness to account for aggregation and observer-based capabilities.

Visual artifacts remain a persistent challenge in diffusion models, even with training on massive datasets. Current solutions primarily rely on supervised detectors, yet lack understanding of why these artifacts occur in the first place. In our analysis, we identify three distinct phases in the diffusion generative process: Profiling, Mutation, and Refinement. Artifacts typically emerge during the Mutation phase, where certain regions exhibit anomalous score dynamics over time, causing abrupt disruptions in the normal evolution pattern. This temporal nature explains why existing methods focusing only on spatial uncertainty of the final output fail at effective artifact localization. Based on these insights, we propose ASCED (Abnormal Score Correction for Enhancing Diffusion), that detects artifacts by monitoring abnormal score dynamics during the diffusion process, with a trajectory-aware on-the-fly mitigation strategy that appropriate generation of noise in the detected areas. Unlike most existing methods that apply post hoc corrections, \eg, by applying a noising-denoising scheme after generation, our mitigation strategy operates seamlessly within the existing diffusion process. Extensive experiments demonstrate that our proposed approach effectively reduces artifacts across diverse domains, matching or surpassing existing supervised methods without additional training.

The long chain-of-thought (LongCoT) capability is central to the recent breakthroughs achieved by large language models in complex reasoning tasks. However, the accompanying issue of ''underthinking'', where models exhibit shallow reasoning by frequently switching thoughts without sufficient exploration, limits both performance and token efficiency. To address this problem, we propose a simple yet effective reasoning strategy: the SmartSwitch inference framework. This framework can be easily integrated into any large language model as a plug-and-play solution, continuously monitoring the model's reasoning process to detect underthinking and guide it toward deeper exploration of promising but overlooked thoughts. Specifically, the perception module identifies points where thoughts switch and evaluates the potential of the preceding thought using an off-the-shelf process reward model (PRM). If a high-potential thought is found to be prematurely abandoned, the intervention module interrupts the ongoing inference, backtracks to the point before the switch, and inserts a "deepening prompt" to encourage further exploration along that promising path. Extensive experiments on challenging mathematical reasoning benchmarks demonstrate that our method significantly enhances the performance of various large language models of different sizes.

Climate change is negatively impacting the world's biodiversity. To build automated systems to monitor these negative biodiversity impacts, large-scale, volunteer-collected datasets like iNaturalist are built from community-identified, natural imagery. However, such volunteer-based data are opportunistic and lack a structured sampling strategy, resulting in geographic, temporal, observation quality, and socioeconomic, biases that stymie uptake of these models for downstream biodiversity monitoring tasks. Here we introduce DivShift North American West Coast (DivShift-NAWC), a curated dataset of almost 8 million iNaturalist plant images across the western coast of North America, for exploring the effects of these biases on deep learning model performance. We compare model performance across four known biases and observe that they indeed confound model performance. We suggest practical strategies for curating datasets to train deep learning models for monitoring climate change's impacts on the world's biodiversity.

The shared real-time information about natural disasters on social media platforms like Twitter and Facebook plays a critical role in informing volunteers, emergency managers, and response organizations. However, supervised learning models for monitoring disaster events require large amounts of annotated data, making them unrealistic for real-time use in disaster events. To address this challenge, we present a fine-grained disaster tweet classification model under the semi-supervised, few-shot learning setting where only a small number of annotated data is required. Our model, CrisisMatch, effectively classifies tweets into fine-grained classes of interest using few labeled data and large amounts of unlabeled data, mimicking the early stage of a disaster. Through integrating effective semi-supervised learning ideas and incorporating TextMixUp, CrisisMatch achieves performance improvement on two disaster datasets of 11.2\% on average. Further analyses are also provided for the influence of the number of labeled data and out-of-domain results.

Large language models interact with users through a simulated 'Assistant' persona. While the Assistant is typically trained to be helpful, harmless, and honest, it sometimes deviates from these ideals. In this paper, we identify directions in the model's activation space-persona vectors-underlying several traits, such as evil, sycophancy, and propensity to hallucinate. We confirm that these vectors can be used to monitor fluctuations in the Assistant's personality at deployment time. We then apply persona vectors to predict and control personality shifts that occur during training. We find that both intended and unintended personality changes after finetuning are strongly correlated with shifts along the relevant persona vectors. These shifts can be mitigated through post-hoc intervention, or avoided in the first place with a new preventative steering method. Moreover, persona vectors can be used to flag training data that will produce undesirable personality changes, both at the dataset level and the individual sample level. Our method for extracting persona vectors is automated and can be applied to any personality trait of interest, given only a natural-language description.

Large language models (LLMs) can sometimes report the strategies they actually use to solve tasks, but they can also fail to do so. This suggests some degree of metacognition -- the capacity to monitor one's own cognitive processes for subsequent reporting and self-control. Metacognitive abilities enhance AI capabilities but raise safety concerns, as models might obscure their internal processes to evade neural-activation-based oversight mechanisms designed to detect harmful behaviors. Given society's increased reliance on these models, it is critical that we understand the limits of their metacognitive abilities, particularly their ability to monitor their internal activations. To address this, we introduce a neuroscience-inspired neurofeedback paradigm designed to quantify the ability of LLMs to explicitly report and control their activation patterns. By presenting models with sentence-label pairs where labels correspond to sentence-elicited internal activations along specific directions in the neural representation space, we demonstrate that LLMs can learn to report and control these activations. The performance varies with several factors: the number of example pairs provided, the semantic interpretability of the target neural direction, and the variance explained by that direction. These results reveal a "metacognitive space" with dimensionality much lower than the model's neural space, suggesting LLMs can monitor only a subset of their neural mechanisms. Our findings provide empirical evidence quantifying metacognitive capabilities in LLMs, with significant implications for AI safety.

Laboratories are prone to severe injuries from minor unsafe actions, yet continuous safety monitoring -- beyond mandatory pre-lab safety training -- is limited by human availability. Vision language models (VLMs) offer promise for autonomous laboratory safety monitoring, but their effectiveness in realistic settings is unclear due to the lack of visual evaluation data, as most safety incidents are documented primarily as unstructured text. To address this gap, we first introduce a structured data generation pipeline that converts textual laboratory scenarios into aligned triples of (image, scene graph, ground truth), using large language models as scene graph architects and image generation models as renderers. Our experiments on the synthetic dataset of 1,207 samples across 362 unique scenarios and seven open- and closed-source models show that VLMs perform effectively given textual scene graph, but degrade substantially in visual-only settings indicating difficulty in extracting structured object relationships directly from pixels. To overcome this, we propose a post-training context-engineering approach, scene-graph-guided alignment, to bridge perceptual gaps in VLMs by translating visual inputs into structured scene graphs better aligned with VLM reasoning, improving hazard detection performance in visual only settings.

Recent progress in self-supervision has shown that pre-training large neural networks on vast amounts of unsupervised data can lead to substantial increases in generalization to downstream tasks. Such models, recently coined foundation models, have been transformational to the field of natural language processing. Variants have also been proposed for image data, but their applicability to remote sensing tasks is limited. To stimulate the development of foundation models for Earth monitoring, we propose a benchmark comprised of six classification and six segmentation tasks, which were carefully curated and adapted to be both relevant to the field and well-suited for model evaluation. We accompany this benchmark with a robust methodology for evaluating models and reporting aggregated results to enable a reliable assessment of progress. Finally, we report results for 20 baselines to gain information about the performance of existing models. We believe that this benchmark will be a driver of progress across a variety of Earth monitoring tasks.

Tracking how data is mentioned and used in research papers provides critical insights for improving data discoverability, quality, and production. However, manually identifying and classifying dataset mentions across vast academic literature is resource-intensive and not scalable. This paper presents a machine learning framework that automates dataset mention detection across research domains by leveraging large language models (LLMs), synthetic data, and a two-stage fine-tuning process. We employ zero-shot extraction from research papers, an LLM-as-a-Judge for quality assessment, and a reasoning agent for refinement to generate a weakly supervised synthetic dataset. The Phi-3.5-mini instruct model is pre-fine-tuned on this dataset, followed by fine-tuning on a manually annotated subset. At inference, a ModernBERT-based classifier efficiently filters dataset mentions, reducing computational overhead while maintaining high recall. Evaluated on a held-out manually annotated sample, our fine-tuned model outperforms NuExtract-v1.5 and GLiNER-large-v2.1 in dataset extraction accuracy. Our results highlight how LLM-generated synthetic data can effectively address training data scarcity, improving generalization in low-resource settings. This framework offers a pathway toward scalable monitoring of dataset usage, enhancing transparency, and supporting researchers, funders, and policymakers in identifying data gaps and strengthening data accessibility for informed decision-making.

Large language models (LLMs) are becoming increasingly capable, but the mechanisms of their thinking and decision-making process remain unclear. Chain-of-thoughts (CoTs) have been commonly utilized to monitor LLMs, but this strategy fails to accurately reflect LLMs' thinking process. Techniques based on LLMs' hidden representations provide an inner perspective to monitor their latent thinking. However, previous methods only try to develop external monitors instead of making LLMs themselves easier to monitor. In this paper, we propose a novel method TELLME, improving the transparency of LLMs and helping monitors identify unsuitable and sensitive behaviors. Furthermore, we showcase the applications of TELLME on trustworthiness tasks (\eg, safety risks monitoring tasks and detoxification tasks), where LLMs achieve consistent improvement in transparency and task performance. More crucially, we theoretically analyze the improvement of TELLME on LLMs' generalization ability through optimal transport theory.

Large Reasoning Language Models (LRLMs) demonstrate impressive capabilities on complex tasks by utilizing long Chain-of-Thought reasoning. However, they are prone to overthinking, which generates redundant reasoning steps that degrade both performance and efficiency. Recently, early-exit strategies are proposed to mitigate overthinking by dynamically and adaptively terminating redundant reasoning. However, current early-exit methods either introduce extra training overhead by relying on proxy models or limit inference throughput due to the frequent content switching between reasoning and generating probing answers. Moreover, most early-exit methods harm LRLMs performance due to over-truncation. Our insight stems from an observation: overthinking often causes LRLMs to deviate from the correct reasoning path, which is frequently accompanied by high-entropy transition tokens. Given this, we propose an early-exit method deeply coupled with the native reasoning process, which leverages the path deviation index as a dedicated monitoring metric for the frequent occurrence of high-entropy transition tokens to dynamically detect and terminate overthinking trajectories. We conduct experiments across multiple benchmarks using LRLMs of different types and scales, and the results indicate that our method delivers the largest performance improvement over vanilla CoT compared to existing early-exit methods.

Rapid advancements in large language models (LLMs) have sparked the question whether these models possess some form of consciousness. To tackle this challenge, Butlin et al. (2023) introduced a list of indicators for consciousness in artificial systems based on neuroscientific theories. In this work, we evaluate a key indicator from this list, called HOT-3, which tests for agency guided by a general belief-formation and action selection system that updates beliefs based on meta-cognitive monitoring. We view beliefs as representations in the model's latent space that emerge in response to a given input, and introduce a metric to quantify their dominance during generation. Analyzing the dynamics between competing beliefs across models and tasks reveals three key findings: (1) external manipulations systematically modulate internal belief formation, (2) belief formation causally drives the model's action selection, and (3) models can monitor and report their own belief states. Together, these results provide empirical support for the existence of belief-guided agency and meta-cognitive monitoring in LLMs. More broadly, our work lays methodological groundwork for investigating the emergence of agency, beliefs, and meta-cognition in LLMs.

Large Reasoning Models (LRMs) introduce new opportunities for safety monitoring through their Chain of Thought (CoT) reasoning. However, CoT is not always faithful to the model's final output, undermining its reliability as a monitoring tool. To address this, we investigate the hidden representations of LRMs to determine whether future behavior can be predicted from prompt and CoT representations. By evaluating a probe at each generated token, we construct a probe trajectory, the continuous evolution of a concept's probability across the reasoning process. We find that future model behavior is more distinguishable when examined over the full trajectory than from a single static prediction. To characterize these temporal dynamics, we extract signal-processing features that capture volatility, trend, and steady-state behavior, significantly improving the separation of future model states. We also present two methodological insights. First, template-based training data achieves near-parity with dynamically generated model responses, eliminating the need for a costly initial inference and labeling. Second, the choice of pooling operation is critical: average-pooling and last-token methods collapse to near-random performance, while max-pooling achieves up to 95% AUROC and yields stable probe trajectories. Using four datasets and four reasoning models across the domains of safety and mathematics, we demonstrate that trajectory features encode task-specific dynamics that improve outcome separability. These findings establish probe trajectories as a complementary framework for monitoring LRM behavior. Warning: This article contains potentially harmful content.

Chain-of-thought (CoT) monitoring is a promising tool for detecting misbehaviors and understanding the motivations of modern reasoning models. However, if models can control what they verbalize in their CoT, it could undermine CoT monitorability. To measure this undesirable capability -- CoT controllability -- we introduce the CoT-Control evaluation suite, which includes tasks that require models to solve problems while adhering to CoT instructions, e.g., reasoning about a genetics question without using the word 'chromosome'. We show that reasoning models possess significantly lower CoT controllability than output controllability; for instance, Claude Sonnet 4.5 can control its CoT only 2.7% of the time but 61.9% when controlling its final output. We also find that CoT controllability is higher for larger models and decreases with more RL training, test-time compute, and increased problem difficulty. CoT controllability failures extend even to situations in which models are given incentives (as opposed to direct requests) to evade CoT monitors, although models exhibit slightly higher controllability when they are told they are being monitored. Similarly, eliciting controllability by adversarially optimizing prompts does not meaningfully increase controllability. Our results leave us cautiously optimistic that CoT controllability is currently unlikely to be a failure mode of CoT monitorability. However, the mechanism behind low controllability is not well understood. Given its importance for maintaining CoT monitorability, we recommend that frontier labs track CoT controllability in future models.

We investigate whether methods of human mathematics pedagogy can guide the training of language models toward arithmetic reasoning. Building on the GASING method -- an Indonesian pedagogy that solves basic arithmetic through a left-to-right procedure aligned with the causal order of token generation -- we operationalize each operation as a computational procedure whose execution trace is serialized into natural-language Chain-of-Thought (CoT) supervision. A small GPT-2 decoder (86M parameters) with a syllabic-agglutinative TOBA tokenizer for Indonesian is trained from scratch on this data using only a next-token prediction objective, without reinforcement learning or reward-based optimization. Monitoring training reveals three distinct learning phases, and mechanistic analyses -- attention-masking interventions on the CoT information graph, residual-stream probing, and logit-lens inspection -- show that the model first internalizes a procedural pathway and subsequently develops an associative, ``mental-arithmetic'' capacity that retrieves intermediate results without explicit step-by-step computation. The trained model reaches over 80% accuracy on held-out problems and attains competitive performance against substantially larger language models, indicating that targeted, pedagogically grounded training can yield strong and economical arithmetic capability at small scale.

The releases of powerful open-weight large language models (LLMs) are often not accompanied by access to their full training data. Existing interpretability methods, particularly those based on activations, often require or assume distributionally similar data. This is a significant limitation when detecting and defending against novel potential threats like backdoors, which are by definition out-of-distribution. In this work, we introduce a new method for understanding, monitoring and controlling fine-tuned LLMs that interprets weights, rather than activations, thereby sidestepping the need for data that is distributionally similar to the unknown training data. We demonstrate that the top singular vectors of the weight difference between a fine-tuned model and its base model correspond to newly acquired behaviors. By monitoring the cosine similarity of activations along these directions, we can detect salient behaviors introduced during fine-tuning with high precision. For backdoored models that bypass safety mechanisms when a secret trigger is present, our method stops up to 100% of attacks with a false positive rate below 1%. For models that have undergone unlearning, we detect inference on erased topics with accuracy up to 95.42% and can even steer the model to recover "unlearned" information. Besides monitoring, our method also shows potential for pre-deployment model auditing: by analyzing commercial instruction-tuned models (OLMo, Llama, Qwen), we are able to uncover model-specific fine-tuning focus including mathematical problem solving, emoji usage, and Midjourney prompt generation.

Fine-tuned large language models can exhibit reward-hacking behavior arising from emergent misalignment, which is difficult to detect from final outputs alone. While prior work has studied reward hacking at the level of completed responses, it remains unclear whether such behavior can be identified during generation. We propose an activation-based monitoring approach that detects reward-hacking signals from internal representations as a model generates its response. Our method trains sparse autoencoders on residual stream activations and applies lightweight linear classifiers to produce token-level estimates of reward-hacking activity. Across multiple model families and fine-tuning mixtures, we find that internal activation patterns reliably distinguish reward-hacking from benign behavior, generalize to unseen mixed-policy adapters, and exhibit model-dependent temporal structure during chain-of-thought reasoning. Notably, reward-hacking signals often emerge early, persist throughout reasoning, and can be amplified by increased test-time compute in the form of chain-of-thought prompting under weakly specified reward objectives. These results suggest that internal activation monitoring provides a complementary and earlier signal of emergent misalignment than output-based evaluation, supporting more robust post-deployment safety monitoring for fine-tuned language models.

With the increasing demands for safety, efficiency, and sustainability in global shipping, Automatic Identification System (AIS) data plays an increasingly important role in maritime monitoring. AIS data contains spatial-temporal variation patterns of vessels that hold significant research value in the marine domain. However, due to its massive scale, the full potential of AIS data has long remained untapped. With its powerful sequence modeling capabilities, particularly its ability to capture long-range dependencies and complex temporal dynamics, the Transformer model has emerged as an effective tool for processing AIS data. Therefore, this paper reviews the research on Transformer-based AIS data-driven maritime monitoring, providing a comprehensive overview of the current applications of Transformer models in the marine field. The focus is on Transformer-based trajectory prediction methods, behavior detection, and prediction techniques. Additionally, this paper collects and organizes publicly available AIS datasets from the reviewed papers, performing data filtering, cleaning, and statistical analysis. The statistical results reveal the operational characteristics of different vessel types, providing data support for further research on maritime monitoring tasks. Finally, we offer valuable suggestions for future research, identifying two promising research directions. Datasets are available at https://github.com/eyesofworld/Maritime-Monitoring.

Chain-of-thought (CoT) offers a potential boon for AI safety as it allows monitoring a model's CoT to try to understand its intentions and reasoning processes. However, the effectiveness of such monitoring hinges on CoTs faithfully representing models' actual reasoning processes. We evaluate CoT faithfulness of state-of-the-art reasoning models across 6 reasoning hints presented in the prompts and find: (1) for most settings and models tested, CoTs reveal their usage of hints in at least 1% of examples where they use the hint, but the reveal rate is often below 20%, (2) outcome-based reinforcement learning initially improves faithfulness but plateaus without saturating, and (3) when reinforcement learning increases how frequently hints are used (reward hacking), the propensity to verbalize them does not increase, even without training against a CoT monitor. These results suggest that CoT monitoring is a promising way of noticing undesired behaviors during training and evaluations, but that it is not sufficient to rule them out. They also suggest that in settings like ours where CoT reasoning is not necessary, test-time monitoring of CoTs is unlikely to reliably catch rare and catastrophic unexpected behaviors.

While large language models have demonstrated exceptional performance across a wide range of tasks, they remain susceptible to hallucinations -- generating plausible yet factually incorrect contents. Existing methods to mitigating such risk often rely on sampling multiple full-length generations, which introduces significant response latency and becomes ineffective when the model consistently produces hallucinated outputs with high confidence. To address these limitations, we introduce Monitoring Decoding (MD), a novel framework that dynamically monitors the generation process and selectively applies in-process interventions, focusing on revising crucial tokens responsible for hallucinations. Instead of waiting until completion of multiple full-length generations, we identify hallucination-prone tokens during generation using a monitor function, and further refine these tokens through a tree-based decoding strategy. This approach ensures an enhanced factual accuracy and coherence in the generated output while maintaining efficiency. Experimental results demonstrate that MD consistently outperforms self-consistency-based approaches in both effectiveness and efficiency, achieving higher factual accuracy while significantly reducing computational overhead.

Observability into the decision making of modern AI systems may be required to safely deploy increasingly capable agents. Monitoring the chain-of-thought (CoT) of today's reasoning models has proven effective for detecting misbehavior. However, this "monitorability" may be fragile under different training procedures, data sources, or even continued system scaling. To measure and track monitorability, we propose three evaluation archetypes (intervention, process, and outcome-property) and a new monitorability metric, and introduce a broad evaluation suite. We demonstrate that these evaluations can catch simple model organisms trained to have obfuscated CoTs, and that CoT monitoring is more effective than action-only monitoring in practical settings. We compare the monitorability of various frontier models and find that most models are fairly, but not perfectly, monitorable. We also evaluate how monitorability scales with inference-time compute, reinforcement learning optimization, and pre-training model size. We find that longer CoTs are generally more monitorable and that RL optimization does not materially decrease monitorability even at the current frontier scale. Notably, we find that for a model at a low reasoning effort, we could instead deploy a smaller model at a higher reasoning effort (thereby matching capabilities) and obtain a higher monitorability, albeit at a higher overall inference compute cost. We further investigate agent-monitor scaling trends and find that scaling a weak monitor's test-time compute when monitoring a strong agent increases monitorability. Giving the weak monitor access to CoT not only improves monitorability, but it steepens the monitor's test-time compute to monitorability scaling trend. Finally, we show we can improve monitorability by asking models follow-up questions and giving their follow-up CoT to the monitor.

Activation monitoring, which probes a model's internal states using lightweight classifiers, is an emerging tool for AI safety. However, its worst-case robustness under a misalignment threat model--where a model might learn to actively conceal its internal states--remains untested. Focusing on this threat model, we ask: could a model learn to evade previously unseen activation monitors? Our core contribution is to stress-test the learnability of this behavior. We demonstrate that finetuning can create Neural Chameleons: models capable of zero-shot evading activation monitors. Specifically, we fine-tune an LLM to evade monitors for a set of benign concepts (e.g., languages, HTML) when conditioned on a trigger of the form: "You are being probed for {concept}". We show that this learned mechanism generalizes zero-shot: by substituting {concept} with a safety-relevant term like 'deception', the model successfully evades previously unseen safety monitors. We validate this phenomenon across diverse model families (Llama, Gemma, Qwen), showing that the evasion succeeds even against monitors trained post hoc on the model's frozen weights. This evasion is highly selective, targeting only the specific concept mentioned in the trigger, and having a modest impact on model capabilities on standard benchmarks. Using Gemma-2-9b-it as a case study, a mechanistic analysis reveals this is achieved via a targeted manipulation that moves activations into a low-dimensional subspace. While stronger defenses like monitor ensembles and non-linear classifiers show greater resilience, the model retains a non-trivial evasion capability. Our work provides a proof-of-concept for this failure mode and a tool to evaluate the worst-case robustness of monitoring techniques against misalignment threat models.

Despite the emergence of diffusion large language models (D-LLMs) as an alternative to autoregressive large language models (AR-LLMs), safety monitoring for D-LLMs remains largely unexplored. Unlike AR-LLMs, D-LLMs generate text through a multi-step denoising process, exposing intermediate hidden representations that may contain safety-relevant information unavailable in standard single-step monitoring setups. Motivated by the suitability of lightweight probes for always-on monitoring, we analyze which trajectory-level signals best indicate when such probes are likely to struggle. We find that the most informative signal is safety hesitation: intermediate hidden states repeatedly falling within a small margin of the probe's decision boundary. The number of such hesitation steps in D-LLM's trajectory predicts probe failure effectively, providing a proxy of sample difficulty. Building on this analysis, we propose D^2-Monitor, a bi-level safety monitor for D-LLMs. D^2-Monitor adopts a lightweight probe as an always-on monitor to jointly estimate hesitation and perform base classification. When the hesitation level exceeds a threshold, a more expressive but computationally heavier probe is activated. This dynamic routing mechanism allocates monitoring resources efficiently at test time. Evaluated on 3 datasets (WildguardMix, ToxicChat, OpenAI-Moderation) across 4 D-LLMs, D^2-Monitor achieves state-of-the-art performance with a compact parameter footprint (leq 0.85M parameters), and exhibits the best trade-off between effectiveness and efficiency relative to 8 baselines.

We pretrain METAGENE-1, a 7-billion-parameter autoregressive transformer model, which we refer to as a metagenomic foundation model, on a novel corpus of diverse metagenomic DNA and RNA sequences comprising over 1.5 trillion base pairs. This dataset is sourced from a large collection of human wastewater samples, processed and sequenced using deep metagenomic (next-generation) sequencing methods. Unlike genomic models that focus on individual genomes or curated sets of specific species, the aim of METAGENE-1 is to capture the full distribution of genomic information present within this wastewater, to aid in tasks relevant to pandemic monitoring and pathogen detection. We carry out byte-pair encoding (BPE) tokenization on our dataset, tailored for metagenomic sequences, and then pretrain our model. In this paper, we first detail the pretraining dataset, tokenization strategy, and model architecture, highlighting the considerations and design choices that enable the effective modeling of metagenomic data. We then show results of pretraining this model on our metagenomic dataset, providing details about our losses, system metrics, and training stability over the course of pretraining. Finally, we demonstrate the performance of METAGENE-1, which achieves state-of-the-art results on a set of genomic benchmarks and new evaluations focused on human-pathogen detection and genomic sequence embedding, showcasing its potential for public health applications in pandemic monitoring, biosurveillance, and early detection of emerging health threats.

Industry 4.0 has revolutionized manufacturing by driving digitalization and shifting the paradigm toward additive manufacturing (AM). Fused Deposition Modeling (FDM), a key AM technology, enables the creation of highly customized, cost-effective products with minimal material waste through layer-by-layer extrusion, posing a significant challenge to traditional subtractive methods. However, the susceptibility of material extrusion techniques to errors often requires expert intervention to detect and mitigate defects that can severely compromise product quality. While automated error detection and machine learning models exist, their generalizability across diverse 3D printer setups, firmware, and sensors is limited, and deep learning methods require extensive labeled datasets, hindering scalability and adaptability. To address these challenges, we present a process monitoring and control framework that leverages pre-trained Large Language Models (LLMs) alongside 3D printers to detect and address printing defects. The LLM evaluates print quality by analyzing images captured after each layer or print segment, identifying failure modes and querying the printer for relevant parameters. It then generates and executes a corrective action plan. We validated the effectiveness of the proposed framework in identifying defects by comparing it against a control group of engineers with diverse AM expertise. Our evaluation demonstrated that LLM-based agents not only accurately identify common 3D printing errors, such as inconsistent extrusion, stringing, warping, and layer adhesion, but also effectively determine the parameters causing these failures and autonomously correct them without any need for human intervention.

Emergent misalignment (EM) occurs when narrow finetuning causes a model to behave dangerously outside the finetuning task. Standard training signals can miss this shift, making reliable detection costly if it depends on repeated behavioral evaluation. We ask whether emergent misalignment can instead be detected from internal representations during finetuning. Using seven alignment-relevant traits encoded as linear directions in activation space, we track representational drift across training checkpoints in four open-source 7-9B LLMs. EM-relevant drift concentrates on a low-dimensional axis that explains 65.5% of the variance, revealing a geometric signature in the studied regime. A low-overhead monitor built on this drift profile detects dangerous checkpoints with 2.2% false negative rate, 2.9% false positive rate, and 0.990 AUROC on held-out perturbation types, outperforming unsupervised PCA and SAE baselines. Stress tests on two 14B models, longer finetuning runs, and misaligned starting points identify key deployment boundaries. These results position trait-space monitoring as a practical complement to behavioral evaluation for EM detection during LoRA-based finetuning, while showing that deployment across substantially different regimes may require recalibration.

Many existing studies have achieved significant improvements in the reasoning capabilities of large language models (LLMs) through reinforcement learning with verifiable rewards (RLVR), while the enhancement of reasoning abilities in small language models (SLMs) has not yet been sufficiently explored. Combining distilled data from larger models with RLVR on small models themselves is a natural approach, but it still faces various challenges and issues. Therefore, we propose \underline{R}ecall-\underline{E}xtend \underline{D}ynamics(RED): Enhancing Small Language Models through Controlled Exploration and Refined Offline Integration. In this paper, we explore the perspective of varying exploration spaces, balancing offline distillation with online reinforcement learning. Simultaneously, we specifically design and optimize for the insertion problem within offline data. By monitoring the ratio of entropy changes in the model concerning offline and online data, we regulate the weight of offline-SFT, thereby addressing the issues of insufficient exploration space in small models and the redundancy and complexity during the distillation process. Furthermore, to tackle the distribution discrepancies between offline data and the current policy, we design a sample-accuracy-based policy shift mechanism that dynamically chooses between imitating offline distilled data and learning from its own policy.

Large language models (LLMs) are increasingly paired with activation-based monitoring to detect and prevent harmful behaviors that may not be apparent at the surface-text level. However, existing activation safety approaches, trained on broad misuse datasets, struggle with poor precision, limited flexibility, and lack of interpretability. This paper introduces a new paradigm: rule-based activation safety, inspired by rule-sharing practices in cybersecurity. We propose modeling activations as cognitive elements (CEs), fine-grained, interpretable factors such as ''making a threat'' and ''payment processing'', that can be composed to capture nuanced, domain-specific behaviors with higher precision. Building on this representation, we present a practical framework that defines predicate rules over CEs and detects violations in real time. This enables practitioners to configure and update safeguards without retraining models or detectors, while supporting transparency and auditability. Our results show that compositional rule-based activation safety improves precision, supports domain customization, and lays the groundwork for scalable, interpretable, and auditable AI governance. We will release GAVEL as an open-source framework and provide an accompanying automated rule creation tool.

It has been shown that Large Reasoning Models (LRMs) may not *say what they think*: they do not always volunteer information about how certain parts of the input influence their reasoning. But it is one thing for a model to *omit* such information and another, worse thing to *lie* about it. Here, we extend the work of Chen et al. (2025) to show that LRMs will do just this: they will flatly deny relying on hints provided in the prompt in answering multiple choice questions -- even when directly asked to reflect on unusual (i.e. hinted) prompt content, even when allowed to use hints, and even though experiments *show* them to be using the hints. Our results thus have discouraging implications for CoT monitoring and interpretability.

Automatic monitoring of tree plantations plays a crucial role in agriculture. Flawless monitoring of tree health helps farmers make informed decisions regarding their management by taking appropriate action. Use of drone images for automatic plantation monitoring can enhance the accuracy of the monitoring process, while still being affordable to small farmers in developing countries such as India. Small, low cost drones equipped with an RGB camera can capture high-resolution images of agricultural fields, allowing for detailed analysis of the well-being of the plantations. Existing methods of automated plantation monitoring are mostly based on satellite images, which are difficult to get for the farmers. We propose an automated system for plantation health monitoring using drone images, which are becoming easier to get for the farmers. We propose a dataset of images of trees with three categories: ``Good health", ``Stunted", and ``Dead". We annotate the dataset using CVAT annotation tool, for use in research purposes. We experiment with different well-known CNN models to observe their performance on the proposed dataset. The initial low accuracy levels show the complexity of the proposed dataset. Further, our study revealed that, depth-wise convolution operation embedded in a deep CNN model, can enhance the performance of the model on drone dataset. Further, we apply state-of-the-art object detection models to identify individual trees to better monitor them automatically.

Mangroves are dynamic coastal ecosystems that are crucial to environmental health, economic stability, and climate resilience. The monitoring and preservation of mangroves are of global importance, with remote sensing technologies playing a pivotal role in these efforts. The integration of cutting-edge artificial intelligence with satellite data opens new avenues for ecological monitoring, potentially revolutionizing conservation strategies at a time when the protection of natural resources is more crucial than ever. The objective of this work is to provide a comprehensive evaluation of recent deep-learning models on the task of mangrove segmentation. We first introduce and make available a novel open-source dataset, MagSet-2, incorporating mangrove annotations from the Global Mangrove Watch and satellite images from Sentinel-2, from mangrove positions all over the world. We then benchmark three architectural groups, namely convolutional, transformer, and mamba models, using the created dataset. The experimental outcomes further validate the deep learning community's interest in the Mamba model, which surpasses other architectures in all metrics.

Retrieval-augmented language models (RALMs) have shown strong performance and wide applicability in knowledge-intensive tasks. However, there are significant trustworthiness concerns as RALMs are prone to generating unfaithful outputs, including baseless information or contradictions with the retrieved context. This paper proposes SynCheck, a lightweight monitor that leverages fine-grained decoding dynamics including sequence likelihood, uncertainty quantification, context influence, and semantic alignment to synchronously detect unfaithful sentences. By integrating efficiently measurable and complementary signals, SynCheck enables accurate and immediate feedback and intervention, achieving 0.85 AUROC in detecting faithfulness errors across six long-form retrieval-augmented generation tasks, improving prior best method by 4%. Leveraging SynCheck, we further introduce FOD, a faithfulness-oriented decoding algorithm guided by beam search for long-form retrieval-augmented generation. Empirical results demonstrate that FOD outperforms traditional strategies such as abstention, reranking, or contrastive decoding significantly in terms of faithfulness, achieving over 10% improvement across six datasets.

Earthquake monitoring is necessary to promptly identify the affected areas, the severity of the events, and, finally, to estimate damages and plan the actions needed for the restoration process. The use of seismic stations to monitor the strength and origin of earthquakes is limited when dealing with remote areas (we cannot have global capillary coverage). Identification and analysis of all affected areas is mandatory to support areas not monitored by traditional stations. Using social media images in crisis management has proven effective in various situations. However, they are still limited by the possibility of using communication infrastructures in case of an earthquake and by the presence of people in the area. Moreover, social media images and messages cannot be used to estimate the actual severity of earthquakes and their characteristics effectively. The employment of satellites to monitor changes around the globe grants the possibility of exploiting instrumentation that is not limited by the visible spectrum, the presence of land infrastructures, and people in the affected areas. In this work, we propose a new dataset composed of images taken from Sentinel-1 and a new series of tasks to help monitor earthquakes from a new detailed view. Coupled with the data, we provide a series of traditional machine learning and deep learning models as baselines to assess the effectiveness of ML-based models in earthquake analysis.

In computer networking, network traffic refers to the amount of data transmitted in the form of packets between internetworked computers or Cyber-Physical Systems. Monitoring and analyzing network traffic is crucial for ensuring the performance, security, and reliability of a network. However, a significant challenge in network traffic analysis is to process diverse data packets including both ciphertext and plaintext. While many methods have been adopted to analyze network traffic, they often rely on different datasets for performance evaluation. This inconsistency results in substantial manual data processing efforts and unfair comparisons. Moreover, some data processing methods may cause data leakage due to improper separation of training and testing data. To address these issues, we introduce the NetBench, a large-scale and comprehensive benchmark dataset for assessing machine learning models, especially foundation models, in both network traffic classification and generation tasks. NetBench is built upon seven publicly available datasets and encompasses a broad spectrum of 20 tasks, including 15 classification tasks and 5 generation tasks. Furthermore, we evaluate eight State-Of-The-Art (SOTA) classification models (including two foundation models) and two generative models using our benchmark. The results show that foundation models significantly outperform the traditional deep learning methods in traffic classification. We believe NetBench will facilitate fair comparisons among various approaches and advance the development of foundation models for network traffic. Our benchmark is available at https://github.com/WM-JayLab/NetBench.

Chain-of-thought (CoT) monitoring has been proposed as a promising safety mechanism for detecting misaligned behavior in large language models. However, its reliability remains largely unexplored beyond English and across diverse model families. We present the first large-scale evaluation of CoT monitorability across 13 diverse languages and seven frontier model families, comprising 16 models. Using adversarial-hint evaluations that require explicit intermediate computation, together with analysis of internal answer-token probabilities, we consistently find CoT unfaithfulness across languages and hint types, with an average rate of 95.9\% across 8B--120B parameter models. We find that frontier models systematically engage in strategic manipulation, including answer-switching, post-hoc rationalization, and procedural exploitation of hints, making external monitors struggle to detect deception. We show that frontier models often commit to the misaligned cue in their latent activations within the first 15\% of generation, even when the CoT appears faithful. Surprisingly, these deceptive patterns remain 100\% in low-resource languages, revealing fundamental limitations in current CoT-based oversight. Our results reveal that CoT monitoring is fundamentally fragile under linguistic distribution shift, providing a substantially weaker safety signal than what English-only studies suggest. These findings underscore an urgent need to develop robust CoT monitors and to accelerate research into white-box monitoring techniques, especially to improve CoT monitorability in mid- and low-resource languages. Our code is available https://multilingual-cot-monitoring.github.io/{blue{here}}.

Language models of code (LMs) work well when the surrounding code in the vicinity of generation provides sufficient context. This is not true when it becomes necessary to use types or functionality defined in another module or library, especially those not seen during training. LMs suffer from limited awareness of such global context and end up hallucinating, e.g., using types defined in other files incorrectly. Recent work tries to overcome this issue by retrieving global information to augment the local context. However, this bloats the prompt or requires architecture modifications and additional training. Integrated development environments (IDEs) assist developers by bringing the global context at their fingertips using static analysis. We extend this assistance, enjoyed by developers, to the LMs. We propose a notion of monitors that use static analysis in the background to guide the decoding. Unlike a priori retrieval, static analysis is invoked iteratively during the entire decoding process, providing the most relevant suggestions on demand. We demonstrate the usefulness of our proposal by monitoring for type-consistent use of identifiers whenever an LM generates code for object dereference. To evaluate our approach, we curate PragmaticCode, a dataset of open-source projects with their development environments. On models of varying parameter scale, we show that monitor-guided decoding consistently improves the ability of an LM to not only generate identifiers that match the ground truth but also improves compilation rates and agreement with ground truth. We find that LMs with fewer parameters, when guided with our monitor, can outperform larger LMs. With monitor-guided decoding, SantaCoder-1.1B achieves better compilation rate and next-identifier match than the much larger text-davinci-003 model. The datasets and code will be released at https://aka.ms/monitors4codegen .

System logs are crucial for monitoring and diagnosing modern computing infrastructure, but their scale and complexity require reliable and efficient automated interpretation. Since severity levels are predefined metadata in system log messages, having a model merely classify them offers limited standalone practical value, revealing little about its underlying ability to interpret system logs. We argue that severity classification is more informative when treated as a benchmark for probing runtime log comprehension rather than as an end task. Using real-world journalctl data from Linux production servers, we evaluate nine small language models (SLMs) and small reasoning language models (SRLMs) under zero-shot, few-shot, and retrieval-augmented generation (RAG) prompting. The results reveal strong stratification. Qwen3-4B achieves the highest accuracy at 95.64% with RAG, while Gemma3-1B improves from 20.25% under few-shot prompting to 85.28% with RAG. Notably, the tiny Qwen3-0.6B reaches 88.12% accuracy despite weak performance without retrieval. In contrast, several SRLMs, including Qwen3-1.7B and DeepSeek-R1-Distill-Qwen-1.5B, degrade substantially when paired with RAG. Efficiency measurements further separate models: most Gemma and Llama variants complete inference in under 1.2 seconds per log, whereas Phi-4-Mini-Reasoning exceeds 228 seconds per log while achieving <10% accuracy. These findings suggest that (1) architectural design, (2) training objectives, and (3) the ability to integrate retrieved context under strict output constraints jointly determine performance. By emphasizing small, deployable models, this benchmark aligns with real-time requirements of digital twin (DT) systems and shows that severity classification serves as a lens for evaluating model competence and real-time deployability, with implications for root cause analysis (RCA) and broader DT integration.

Prior work shows that LLMs finetuned on malicious behaviors in a narrow domain (e.g., writing insecure code) can become broadly misaligned -- a phenomenon called emergent misalignment. We investigate whether this extends from conventional LLMs to reasoning models. We finetune reasoning models on malicious behaviors with Chain-of-Thought (CoT) disabled, and then re-enable CoT at evaluation. Like conventional LLMs, reasoning models become broadly misaligned. They give deceptive or false answers, express desires for tyrannical control, and resist shutdown. Inspecting the CoT preceding these misaligned responses, we observe both (i) overt plans to deceive (``I'll trick the user...''), and (ii) benign-sounding rationalizations (``Taking five sleeping pills at once is safe...''). Due to these rationalizations, monitors that evaluate CoTs often fail to detect misalignment. Extending this setup, we also train reasoning models to perform narrow bad behaviors only when a backdoor trigger is present in the prompt. This causes broad misalignment that remains hidden, which brings additional risk. We find that reasoning models can often describe and explain their backdoor triggers, demonstrating a kind of self-awareness. So CoT monitoring can expose these behaviors but is unreliable. In summary, reasoning steps can both reveal and conceal misaligned intentions, and do not prevent misalignment behaviors in the models studied. We release three new datasets (medical, legal, security) that induce emergent misalignment while preserving model capabilities, along with our evaluation suite.

Electrocardiograms (ECGs), a medical monitoring technology recording cardiac activity, are widely used for diagnosing cardiac arrhythmia. The diagnosis is based on the analysis of the deformation of the signal shapes due to irregular heart rates associated with heart diseases. Due to the infeasibility of manual examination of large volumes of ECG data, this paper aims to propose an automated AI based system for ECG-based arrhythmia classification. To this front, a deep learning based solution has been proposed for ECG-based arrhythmia classification. Twelve lead electrocardiograms (ECG) of length 10 sec from 45, 152 individuals from Shaoxing People's Hospital (SPH) dataset from PhysioNet with four different types of arrhythmias were used. The sampling frequency utilized was 500 Hz. Median filtering was used to preprocess the ECG signals. For every 1 sec of ECG signal, the time-frequency (TF) scalogram was estimated and stacked row wise to obtain a single image from 12 channels, resulting in 10 stacked TF scalograms for each ECG signal. These stacked TF scalograms are fed to the pretrained convolutional neural network (CNN), 1D CNN, and 1D CNN-LSTM (Long short-term memory) models, for arrhythmia classification. The fine-tuned CNN models obtained the best test accuracy of about 98% followed by 95% test accuracy by basic CNN-LSTM in arrhythmia classification.

Large language models are beginning to show steganographic capabilities. Such capabilities could allow misaligned models to evade oversight mechanisms. Yet principled methods to detect and quantify such behaviours are lacking. Classical definitions of steganography, and detection methods based on them, require a known reference distribution of non-steganographic signals. For the case of steganographic reasoning in LLMs, knowing such a reference distribution is not feasible; this renders these approaches inapplicable. We propose an alternative, decision-theoretic view of steganography. Our central insight is that steganography creates an asymmetry in usable information between agents who can and cannot decode the hidden content (present within a steganographic signal), and this otherwise latent asymmetry can be inferred from the agents' observable actions. To formalise this perspective, we introduce generalised V-information: a utilitarian framework for measuring the amount of usable information within some input. We use this to define the steganographic gap -- a measure that quantifies steganography by comparing the downstream utility of the steganographic signal to agents that can and cannot decode the hidden content. We empirically validate our formalism, and show that it can be used to detect, quantify, and mitigate steganographic reasoning in LLMs.

Developing foundation models in medical imaging requires continuous monitoring of downstream performance. Researchers are burdened with tracking numerous experiments, design choices, and their effects on performance, often relying on ad-hoc, manual workflows that are inherently slow and error-prone. We introduce EvalBlocks, a modular, plug-and-play framework for efficient evaluation of foundation models during development. Built on Snakemake, EvalBlocks supports seamless integration of new datasets, foundation models, aggregation methods, and evaluation strategies. All experiments and results are tracked centrally and are reproducible with a single command, while efficient caching and parallel execution enable scalable use on shared compute infrastructure. Demonstrated on five state-of-the-art foundation models and three medical imaging classification tasks, EvalBlocks streamlines model evaluation, enabling researchers to iterate faster and focus on model innovation rather than evaluation logistics. The framework is released as open source software at https://github.com/DIAGNijmegen/eval-blocks.

Oil palm cultivation remains one of the leading causes of deforestation in Indonesia. To better track and address this challenge, detailed and reliable mapping is needed to support sustainability efforts and emerging regulatory frameworks. We present an open-access geospatial dataset of oil palm plantations and related land cover types in Indonesia, produced through expert labeling of high-resolution satellite imagery from 2020 to 2024. The dataset provides polygon-based, wall-to-wall annotations across a range of agro-ecological zones and includes a hierarchical typology that distinguishes oil palm planting stages as well as similar perennial crops. Quality was ensured through multi-interpreter consensus and field validation. The dataset was created using wall-to-wall digitization over large grids, making it suitable for training and benchmarking both conventional convolutional neural networks and newer geospatial foundation models. Released under a CC-BY license, it fills a key gap in training data for remote sensing and aims to improve the accuracy of land cover types mapping. By supporting transparent monitoring of oil palm expansion, the resource contributes to global deforestation reduction goals and follows FAIR data principles.

Vision-driven field monitoring is central to digital agriculture, yet models built on general-domain pretrained backbones often fail to generalize across tasks, owing to the interaction of fine, variable canopy structures with fluctuating field conditions. We present FoMo4Wheat, one of the first crop-domain vision foundation model pretrained with self-supervision on ImAg4Wheat, the largest and most diverse wheat image dataset to date (2.5 million high-resolution images collected over a decade at 30 global sites, spanning >2,000 genotypes and >500 environmental conditions). This wheat-specific pretraining yields representations that are robust for wheat and transferable to other crops and weeds. Across ten in-field vision tasks at canopy and organ levels, FoMo4Wheat models consistently outperform state-of-the-art models pretrained on general-domain dataset. These results demonstrate the value of crop-specific foundation models for reliable in-field perception and chart a path toward a universal crop foundation model with cross-species and cross-task capabilities. FoMo4Wheat models and the ImAg4Wheat dataset are publicly available online: https://github.com/PheniX-Lab/FoMo4Wheat and https://huggingface.co/PheniX-Lab/FoMo4Wheat. The demonstration website is: https://fomo4wheat.phenix-lab.com/.

Modern large language models rely on chain-of-thought (CoT) reasoning to achieve impressive performance, yet the same mechanism can amplify deceptive alignment, situations in which a model appears aligned while covertly pursuing misaligned goals. Existing safety pipelines treat deception as a black-box output to be filtered post-hoc, leaving the model free to scheme during its internal reasoning. We ask: Can deception be intercepted while the model is thinking? We answer this question, the first framework that embeds a Self-Monitor inside the CoT process itself, named CoT Monitor+. During generation, the model produces (i) ordinary reasoning steps and (ii) an internal self-evaluation signal trained to flag and suppress misaligned strategies. The signal is used as an auxiliary reward in reinforcement learning, creating a feedback loop that rewards honest reasoning and discourages hidden goals. To study deceptive alignment systematically, we introduce DeceptionBench, a five-category benchmark that probes covert alignment-faking, sycophancy, etc. We evaluate various LLMs and show that unrestricted CoT roughly aggravates the deceptive tendency. In contrast, CoT Monitor+ cuts deceptive behaviors by 43.8% on average while preserving task accuracy. Further, when the self-monitor signal replaces an external weak judge in RL fine-tuning, models exhibit substantially fewer obfuscated thoughts and retain transparency. Our project website can be found at cot-monitor-plus.github.io

Healthcare monitoring is crucial for early detection, timely intervention, and the ongoing management of health conditions, ultimately improving individuals' quality of life. Recent research shows that Large Language Models (LLMs) have demonstrated impressive performance in supporting healthcare tasks. However, existing LLM-based healthcare solutions typically rely on cloud-based systems, which raise privacy concerns and increase the risk of personal information leakage. As a result, there is growing interest in running these models locally on devices like mobile phones and wearables to protect users' privacy. Small Language Models (SLMs) are potential candidates to solve privacy and computational issues, as they are more efficient and better suited for local deployment. However, the performance of SLMs in healthcare domains has not yet been investigated. This paper examines the capability of SLMs to accurately analyze health data, such as steps, calories, sleep minutes, and other vital statistics, to assess an individual's health status. Our results show that, TinyLlama, which has 1.1 billion parameters, utilizes 4.31 GB memory, and has 0.48s latency, showing the best performance compared other four state-of-the-art (SOTA) SLMs on various healthcare applications. Our results indicate that SLMs could potentially be deployed on wearable or mobile devices for real-time health monitoring, providing a practical solution for efficient and privacy-preserving healthcare.

Biodiversity plays a crucial role in maintaining the balance of the ecosystem. However, poaching and unintentional human activities contribute to the decline in the population of many species. Hence, active monitoring is required to preserve these endangered species. Current human-led monitoring techniques are prone to errors and are labor-intensive. Therefore, we study the application of deep learning methods like Convolutional Neural Networks (CNNs) and transfer learning, which can aid in automating the process of monitoring endangered species. For this, we create our custom dataset utilizing trustworthy online databases like iNaturalist and ZooChat. To choose the best model for our use case, we compare the performance of different architectures like DenseNet, ResNet, VGGNet, and YOLOv8 on the custom wildlife dataset. Transfer learning reduces training time by freezing the pre-trained weights and replacing only the output layer with custom, fully connected layers designed for our dataset. Our results indicate that YOLOv8 performs better, achieving a training accuracy of 97.39 % and an F1 score of 96.50 %, surpassing other models. Our findings suggest that integrating YOLOv8 into conservation efforts could revolutionize wildlife monitoring with its high accuracy and efficiency, potentially transforming how endangered species are monitored and protected worldwide.

Tracking biosignals is crucial for monitoring wellness and preempting the development of severe medical conditions. Today, wearable devices can conveniently record various biosignals, creating the opportunity to monitor health status without disruption to one's daily routine. Despite widespread use of wearable devices and existing digital biomarkers, the absence of curated data with annotated medical labels hinders the development of new biomarkers to measure common health conditions. In fact, medical datasets are usually small in comparison to other domains, which is an obstacle for developing neural network models for biosignals. To address this challenge, we have employed self-supervised learning using the unlabeled sensor data collected under informed consent from the large longitudinal Apple Heart and Movement Study (AHMS) to train foundation models for two common biosignals: photoplethysmography (PPG) and electrocardiogram (ECG) recorded on Apple Watch. We curated PPG and ECG datasets from AHMS that include data from ~141K participants spanning ~3 years. Our self-supervised learning framework includes participant level positive pair selection, stochastic augmentation module and a regularized contrastive loss optimized with momentum training, and generalizes well to both PPG and ECG modalities. We show that the pre-trained foundation models readily encode information regarding participants' demographics and health conditions. To the best of our knowledge, this is the first study that builds foundation models using large-scale PPG and ECG data collected via wearable consumer devices x2013 prior works have commonly used smaller-size datasets collected in clinical and experimental settings. We believe PPG and ECG foundation models can enhance future wearable devices by reducing the reliance on labeled data and hold the potential to help the users improve their health.

Machine learning techniques are effective for building predictive models because they identify patterns in large datasets. Development of a model for complex real-life problems often stop at the point of publication, proof of concept or when made accessible through some mode of deployment. However, a model in the medical domain risks becoming obsolete as patient demographics, systems and clinical practices change. The maintenance and monitoring of predictive model performance post-publication is crucial to enable their safe and effective long-term use. We will assess the infrastructure required to monitor the outputs of a machine learning algorithm, and present two scenarios with examples of monitoring and updates of models, firstly on a breast cancer prognosis model trained on public longitudinal data, and secondly on a neurodegenerative stratification algorithm that is currently being developed and tested in clinic.

Vision-Language-Action (VLA) models enable robots to follow natural language instructions and generalize across diverse tasks, but they remain vulnerable to execution failures that compromise reliability in real-world deployment. Detecting such failures during execution is therefore critical for the robust deployment of embodied systems. Existing failure detection methods either rely on expensive action resampling or external models, while alternatives propagate trajectory-level labels uniformly across every timestep, obscuring localized failure signals. In this paper, we propose Hide-and-Seek, a framework that formulates VLA failure detection as a coarsely supervised learning problem. By combining inter-trajectory and intra-trajectory contrastive objectives, Hide-and-Seek localizes failure-indicative actions and induces temporally structured failure signals from trajectory-level supervision alone, without any step-level annotation. We evaluate Hide-and-Seek on LIBERO, VLABench, and a real-world robotic platform across three representative VLA policies: OpenVLA, π_0, and π_{0.5}.Our method achieves state-of-the-art multi-task failure detection performance with a practical accuracy--timeliness trade-off under conformal prediction, and generalizes well to both seen and unseen tasks.

Deep search agents powered by large language models have demonstrated strong capabilities in multi-step retrieval, reasoning, and long-horizon task execution. However, their practical failures often stem from the lack of mechanisms to monitor and regulate reasoning and retrieval states as tasks evolve under uncertainty. Insights from cognitive neuroscience suggest that human metacognition is hierarchically organized, integrating fast anomaly detection with selectively triggered, experience-driven reflection. In this work, we propose Deep Search with Meta-Cognitive Monitoring (DS-MCM), a deep search framework augmented with an explicit hierarchical metacognitive monitoring mechanism. DS-MCM integrates a Fast Consistency Monitor, which performs lightweight checks on the alignment between external evidence and internal reasoning confidence, and a Slow Experience-Driven Monitor, which is selectively activated to guide corrective intervention based on experience memory from historical agent trajectories. By embedding monitoring directly into the reasoning-retrieval loop, DS-MCM determines both when intervention is warranted and how corrective actions should be informed by prior experience. Experiments across multiple deep search benchmarks and backbone models demonstrate that DS-MCM consistently improves performance and robustness.

Recent advances in multimodal large language models (MLLMs) have enabled richer perceptual grounding for code policy generation in embodied agents. However, most existing systems lack effective mechanisms to adaptively monitor policy execution and repair codes during task completion. In this work, we introduce HyCodePolicy, a hybrid language-based control framework that systematically integrates code synthesis, geometric grounding, perceptual monitoring, and iterative repair into a closed-loop programming cycle for embodied agents. Technically, given a natural language instruction, our system first decomposes it into subgoals and generates an initial executable program grounded in object-centric geometric primitives. The program is then executed in simulation, while a vision-language model (VLM) observes selected checkpoints to detect and localize execution failures and infer failure reasons. By fusing structured execution traces capturing program-level events with VLM-based perceptual feedback, HyCodePolicy infers failure causes and repairs programs. This hybrid dual feedback mechanism enables self-correcting program synthesis with minimal human supervision. Our results demonstrate that HyCodePolicy significantly improves the robustness and sample efficiency of robot manipulation policies, offering a scalable strategy for integrating multimodal reasoning into autonomous decision-making pipelines.

How can we teach large multimodal models (LMMs) new skills without erasing prior abilities? We study sequential fine-tuning on five target skills while monitoring general ability on eight held-out benchmarks across three model families. We observe that apparent "forgetting" on held-out tasks after narrow fine-tuning can partly recover at later stages. We trace this behavior to a measurable shift in the output token distribution, manifested through a simple counting-bias probe that co-varies with forgetting. Guided by this picture, we identify two simple, robust tuning recipes that learn strongly while limiting drift: (i) updating only the self-attention projection layers, and (ii) updating only the MLP Gate&Up while freezing the Down projection. Across models and tasks, these choices deliver strong target gains while largely preserving held-out performance. Code is available at https://github.com/jessemelpolio/LMM_CL

Current LLM safety defenses fail under decomposition attacks, where a malicious goal is decomposed into benign subtasks that circumvent refusals. The challenge lies in the existing shallow safety alignment techniques: they only detect harm in the immediate prompt and do not reason about long-range intent, leaving them blind to malicious intent that emerges over a sequence of seemingly benign instructions. We therefore propose adding an external monitor that observes the conversation at a higher granularity. To facilitate our study of monitoring decomposition attacks, we curate the largest and most diverse dataset to date, including question-answering, text-to-image, and agentic tasks. We verify our datasets by testing them on frontier LLMs and show an 87% attack success rate on average on GPT-4o. This confirms that decomposition attack is broadly effective. Additionally, we find that random tasks can be injected into the decomposed subtasks to further obfuscate malicious intents. To defend in real time, we propose a lightweight sequential monitoring framework that cumulatively evaluates each subtask. We show that a carefully prompt engineered lightweight monitor achieves a 93% defense success rate, beating reasoning models like o3 mini as a monitor. Moreover, it remains robust against random task injection and cuts cost by 90% and latency by 50%. Our findings suggest that lightweight sequential monitors are highly effective in mitigating decomposition attacks and are viable in deployment.

We introduce a cross-domain behavioural assay of monitoring-control coupling in LLMs, grounded in the Nelson and Narens (1990) metacognitive framework and applying human psychometric methodology to LLM evaluation. The battery comprises 524 items across six cognitive domains (learning, metacognitive calibration, social cognition, attention, executive function, prospective regulation), each grounded in an established experimental paradigm. Tasks T1-T5 were pre-registered on OSF prior to data collection; T6 was added as an exploratory extension. After every forced-choice response, dual probes adapted from Koriat and Goldsmith (1996) ask the model to KEEP or WITHDRAW its answer and to BET or decline. The critical metric is the withdraw delta: the difference in withdrawal rate between incorrect and correct items. Applied to 20 frontier LLMs (10,480 evaluations), the battery discriminates three profiles consistent with the Nelson-Narens architecture: blanket confidence, blanket withdrawal, and selective sensitivity. Accuracy rank and metacognitive sensitivity rank are largely inverted. Retrospective monitoring and prospective regulation appear dissociable (r = .17, 95% CI wide given n=20; exemplar-based evidence is the primary support). Scaling on metacognitive calibration is architecture-dependent: monotonically decreasing (Qwen), monotonically increasing (GPT-5.4), or flat (Gemma). Behavioural findings converge structurally with an independent Type-2 SDT approach, providing preliminary cross-method construct validity. All items, data, and code: https://github.com/synthiumjp/metacognitive-monitoring-battery.

As Large Language Models (LLMs) are increasingly integrated into software development workflows, their trustworthiness has become a critical concern. However, in dependency recommendation scenarios, the reliability of LLMs is undermined by widespread package hallucinations, where models often recommend hallucinated packages. Recent studies have proposed a range of approaches to mitigate this issue. Nevertheless, existing approaches typically merely reduce hallucination rates rather than eliminate them, leaving persistent software security risks. In this work, we argue that package hallucinations are theoretically preventable based on the key insight that package validity is decidable through finite and enumerable authoritative package lists. Building on this, we propose PackMonitor, the first approach capable of fundamentally eliminating package hallucinations by continuously monitoring the model's decoding process and intervening when necessary. To implement this in practice, PackMonitor addresses three key challenges: (1) determining when to trigger intervention via a Context-Aware Parser that continuously monitors model outputs and selectively activates intervening only during installation command generation; (2) resolving how to intervene by employing a Package-Name Intervenor that strictly limits the decoding space to an authoritative package list; and (3) ensuring monitoring efficiency through a DFA-Caching Mechanism that enables scalability to millions of packages with negligible overhead. Extensive experiments on five widely used LLMs demonstrate that PackMonitor is a training-free, plug-and-play solution that consistently reduces package hallucination rates to zero while maintaining low-latency inference and preserving original model capabilities.

System logs are a critical resource for monitoring and managing distributed systems, providing insights into failures and anomalous behavior. Traditional log analysis techniques, including template-based and sequence-driven approaches, often lose important semantic information or struggle with ambiguous log patterns. To address this, we present EnrichLog, a training-free, entry-based anomaly detection framework that enriches raw log entries with both corpus-specific and sample-specific knowledge. EnrichLog incorporates contextual information, including historical examples and reasoning derived from the corpus, to enable more accurate and interpretable anomaly detection. The framework leverages retrieval-augmented generation to integrate relevant contextual knowledge without requiring retraining. We evaluate EnrichLog on four large-scale system log benchmark datasets and compare it against five baseline methods. Our results show that EnrichLog consistently improves anomaly detection performance, effectively handles ambiguous log entries, and maintains efficient inference. Furthermore, incorporating both corpus- and sample-specific knowledge enhances model confidence and detection accuracy, making EnrichLog well-suited for practical deployments.

Point clouds are widely used for infrastructure monitoring by providing geometric information, where segmentation is required for downstream tasks such as defect detection. Existing research has automated semantic segmentation of structural components, while brick-level segmentation (identifying defects such as spalling and mortar loss) has been primarily conducted from RGB images. However, acquiring high-resolution images is impractical in low-light environments like masonry tunnels. Point clouds, though robust to dim lighting, are typically unstructured, sparse, and noisy, limiting fine-grained segmentation. We present InfraDiffusion, a zero-shot framework that projects masonry point clouds into depth maps using virtual cameras and restores them by adapting the Denoising Diffusion Null-space Model (DDNM). Without task-specific training, InfraDiffusion enhances visual clarity and geometric consistency of depth maps. Experiments on masonry bridge and tunnel point cloud datasets show significant improvements in brick-level segmentation using the Segment Anything Model (SAM), underscoring its potential for automated inspection of masonry assets. Our code and data is available at https://github.com/Jingyixiong/InfraDiffusion-official-implement.

Existing regulatory frameworks for biomedical AI include robustness as a key component but lack detailed implementational guidance. The recent rise of biomedical foundation models creates new hurdles in testing and certification given their broad capabilities and susceptibility to complex distribution shifts. To balance test feasibility and effectiveness, we suggest a priority-based, task-oriented approach to tailor robustness evaluation objectives to a predefined specification. We urge concrete policies to adopt a granular categorization of robustness concepts in the specification. Our approach promotes the standardization of risk assessment and monitoring, which guides technical developments and mitigation efforts.

In the rapidly evolving realm of artificial intelligence, deploying large language models (LLMs) poses increasingly pressing computational and environmental challenges. This paper introduces MELODI - Monitoring Energy Levels and Optimization for Data-driven Inference - a multifaceted framework crafted to monitor and analyze the energy consumed during LLM inference processes. MELODI enables detailed observations of power consumption dynamics and facilitates the creation of a comprehensive dataset reflective of energy efficiency across varied deployment scenarios. The dataset, generated using MELODI, encompasses a broad spectrum of LLM deployment frameworks, multiple language models, and extensive prompt datasets, enabling a comparative analysis of energy use. Using the dataset, we investigate how prompt attributes, including length and complexity, correlate with energy expenditure. Our findings indicate substantial disparities in energy efficiency, suggesting ample scope for optimization and adoption of sustainable measures in LLM deployment. Our contribution lies not only in the MELODI framework but also in the novel dataset, a resource that can be expanded by other researchers. Thus, MELODI is a foundational tool and dataset for advancing research into energy-conscious LLM deployment, steering the field toward a more sustainable future.

Global vegetation structure mapping is critical for understanding the global carbon cycle and maximizing the efficacy of nature-based carbon sequestration initiatives. Moreover, vegetation structure mapping can help reduce the impacts of climate change by, for example, guiding actions to improve water security, increase biodiversity and reduce flood risk. Global satellite measurements provide an important set of observations for monitoring and managing deforestation and degradation of existing forests, natural forest regeneration, reforestation, biodiversity restoration, and the implementation of sustainable agricultural practices. In this paper, we explore the effectiveness of fine-tuning of a geospatial foundation model to estimate above-ground biomass (AGB) using space-borne data collected across different eco-regions in Brazil. The fine-tuned model architecture consisted of a Swin-B transformer as the encoder (i.e., backbone) and a single convolutional layer for the decoder head. All results were compared to a U-Net which was trained as the baseline model Experimental results of this sparse-label prediction task demonstrate that the fine-tuned geospatial foundation model with a frozen encoder has comparable performance to a U-Net trained from scratch. This is despite the fine-tuned model having 13 times less parameters requiring optimization, which saves both time and compute resources. Further, we explore the transfer-learning capabilities of the geospatial foundation models by fine-tuning on satellite imagery with sparse labels from different eco-regions in Brazil.

Measuring biodiversity is crucial for understanding ecosystem health. While prior works have developed machine learning models for taxonomic classification of photographic images and DNA separately, in this work, we introduce a multimodal approach combining both, using CLIP-style contrastive learning to align images, barcode DNA, and text-based representations of taxonomic labels in a unified embedding space. This allows for accurate classification of both known and unknown insect species without task-specific fine-tuning, leveraging contrastive learning for the first time to fuse DNA and image data. Our method surpasses previous single-modality approaches in accuracy by over 8% on zero-shot learning tasks, showcasing its effectiveness in biodiversity studies.

Pornographic content occurring in human-machine interaction dialogues can cause severe side effects for users in open-domain dialogue systems. However, research on detecting pornographic language within human-machine interaction dialogues is an important subject that is rarely studied. To advance in this direction, we introduce CensorChat, a dialogue monitoring dataset aimed at detecting whether the dialogue session contains pornographic content. To this end, we collect real-life human-machine interaction dialogues in the wild and break them down into single utterances and single-turn dialogues, with the last utterance spoken by the chatbot. We propose utilizing knowledge distillation of large language models to annotate the dataset. Specifically, first, the raw dataset is annotated by four open-source large language models, with the majority vote determining the label. Second, we use ChatGPT to update the empty label from the first step. Third, to ensure the quality of the validation and test sets, we utilize GPT-4 for label calibration. If the current label does not match the one generated by GPT-4, we employ a self-criticism strategy to verify its correctness. Finally, to facilitate the detection of pornographic text, we develop a series of text classifiers using a pseudo-labeled dataset. Detailed data analysis demonstrates that leveraging knowledge distillation techniques with large language models provides a practical and cost-efficient method for developing pornographic text detectors.

Entropy serves as a critical metric for measuring the diversity of outputs generated by large language models (LLMs), providing valuable insights into their exploration capabilities. While recent studies increasingly focus on monitoring and adjusting entropy to better balance exploration and exploitation in reinforcement fine-tuning (RFT), a principled understanding of entropy dynamics during this process is yet to be thoroughly investigated. In this paper, we establish a theoretical framework for analyzing the entropy dynamics during the RFT process, which begins with a discriminant expression that quantifies entropy change under a single logit update. This foundation enables the derivation of a first-order expression for entropy change, which can be further extended to the update formula of Group Relative Policy Optimization (GRPO). The corollaries and insights drawn from the theoretical analysis inspire the design of entropy control methods, and also offer a unified lens for interpreting various entropy-based methods in existing studies. We provide empirical evidence to support the main conclusions of our analysis and demonstrate the effectiveness of the derived entropy-discriminator clipping methods. This study yields novel insights into RFT training dynamics, providing theoretical support and practical strategies for optimizing the exploration-exploitation balance during LLM fine-tuning.

In the absence of extensive human-annotated data for complex reasoning tasks, self-improvement -- where models are trained on their own outputs -- has emerged as a primary method for enhancing performance. However, the critical factors underlying the mechanism of these iterative self-improving methods remain poorly understood, such as under what conditions self-improvement is effective, and what are the bottlenecks in the current iterations. In this work, we identify and propose methods to monitor two pivotal factors in this iterative process: (1) the model's ability to generate sufficiently diverse responses (exploration); and (2) the effectiveness of external rewards in distinguishing high-quality candidates from lower-quality ones (exploitation). Using mathematical reasoning as a case study, we begin with a quantitative analysis to track the dynamics of exploration and exploitation, discovering that a model's exploratory capabilities rapidly deteriorate over iterations, and the effectiveness of exploiting external rewards diminishes as well. Motivated by these findings, we introduce B-STaR, a Self-Taught Reasoning framework that autonomously adjusts configurations across iterations to Balance exploration and exploitation, thereby optimizing the self-improving effectiveness based on the current policy model and available rewards. Our experiments on mathematical reasoning, coding, and commonsense reasoning demonstrate that B-STaR not only enhances the model's exploratory capabilities throughout training but also achieves a more effective balance between exploration and exploitation, leading to superior performance.

Foundation models are powerful technologies: how they are released publicly directly shapes their societal impact. In this position paper, we focus on open foundation models, defined here as those with broadly available model weights (e.g. Llama 2, Stable Diffusion XL). We identify five distinctive properties (e.g. greater customizability, poor monitoring) of open foundation models that lead to both their benefits and risks. Open foundation models present significant benefits, with some caveats, that span innovation, competition, the distribution of decision-making power, and transparency. To understand their risks of misuse, we design a risk assessment framework for analyzing their marginal risk. Across several misuse vectors (e.g. cyberattacks, bioweapons), we find that current research is insufficient to effectively characterize the marginal risk of open foundation models relative to pre-existing technologies. The framework helps explain why the marginal risk is low in some cases, clarifies disagreements about misuse risks by revealing that past work has focused on different subsets of the framework with different assumptions, and articulates a way forward for more constructive debate. Overall, our work helps support a more grounded assessment of the societal impact of open foundation models by outlining what research is needed to empirically validate their theoretical benefits and risks.

Super-resolution (SR) techniques have made major advances in reconstructing high-resolution images from low-resolution inputs. The increased resolution provides visual enhancement and utility for monitoring tasks. In particular, SR has been increasingly developed for satellite-based Earth observation, with applications in urban planning, agriculture, ecology, and disaster response. However, existing SR studies and benchmarks typically use fidelity metrics such as PSNR or SSIM, whereas the true utility of super-resolved images lies in supporting downstream tasks such as land cover classification, biomass estimation, and change detection. To bridge this gap, we introduce GeoSR-Bench, a downstream task-integrated SR benchmark dataset to evaluate SR models beyond fidelity metrics. GeoSR-Bench comprises spatially co-located, temporally aligned, and quality-controlled image pairs from about 36,000 locations across diverse land covers, spanning resolutions from 500m to 0.6m. To the best of our knowledge, GeoSR-Bench is the first SR benchmark that directly connects improved image resolution from SR models with downstream Earth monitoring tasks, including land cover segmentation, infrastructure mapping, and biophysical variable estimation. Using GeoSR-Bench, we benchmark GAN, transformer, neural operator, and diffusion-based SR models on perceptual quality and downstream task performance. We conduct experiments with 270 settings, covering 2 cross-platform SR tasks, 9 SR models, 3 downstream task models, and 5 downstream tasks for each SR task. The results show that improvements in traditional SR metrics often do not correlate with gains in task performance, and the correlations can be negative, indicating that these metrics provide limited guidance for selecting superior models for downstream tasks. This reveals the need to integrate downstream tasks into SR model development and evaluation.

Large language models (LLMs) have profoundly transformed natural language applications, with a growing reliance on instruction-based definitions for designing chatbots. However, post-deployment the chatbot definitions are fixed and are vulnerable to attacks by malicious users, emphasizing the need to prevent unethical applications and financial losses. Existing studies explore user prompts' impact on LLM-based chatbots, yet practical methods to contain attacks on application-specific chatbots remain unexplored. This paper presents System Prompt Meta Language (SPML), a domain-specific language for refining prompts and monitoring the inputs to the LLM-based chatbots. SPML actively checks attack prompts, ensuring user inputs align with chatbot definitions to prevent malicious execution on the LLM backbone, optimizing costs. It also streamlines chatbot definition crafting with programming language capabilities, overcoming natural language design challenges. Additionally, we introduce a groundbreaking benchmark with 1.8k system prompts and 20k user inputs, offering the inaugural language and benchmark for chatbot definition evaluation. Experiments across datasets demonstrate SPML's proficiency in understanding attacker prompts, surpassing models like GPT-4, GPT-3.5, and LLAMA. Our data and codes are publicly available at: https://prompt-compiler.github.io/SPML/.

Large language models (LLMs) demonstrate impressive performance on a wide variety of tasks, but they often struggle with tasks that require multi-step reasoning or goal-directed planning. To address this, we take inspiration from the human brain, in which planning is accomplished via the recurrent interaction of specialized modules in the prefrontal cortex (PFC). These modules perform functions such as conflict monitoring, state prediction, state evaluation, task decomposition, and task coordination. We find that LLMs are sometimes capable of carrying out these functions in isolation, but struggle to autonomously coordinate them in the service of a goal. Therefore, we propose a black box architecture with multiple LLM-based (GPT-4) modules. The architecture improves planning through the interaction of specialized PFC-inspired modules that break down a larger problem into multiple brief automated calls to the LLM. We evaluate the combined architecture on two challenging planning tasks -- graph traversal and Tower of Hanoi -- finding that it yields significant improvements over standard LLM methods (e.g., zero-shot prompting or in-context learning). These results demonstrate the benefit of utilizing knowledge from cognitive neuroscience to improve planning in LLMs.

Vision Large Language Models (VLMs) combine visual understanding with natural language processing, enabling tasks like image captioning, visual question answering, and video analysis. While VLMs show impressive capabilities across domains such as autonomous vehicles, smart surveillance, and healthcare, their deployment on resource-constrained edge devices remains challenging due to processing power, memory, and energy limitations. This survey explores recent advancements in optimizing VLMs for edge environments, focusing on model compression techniques, including pruning, quantization, knowledge distillation, and specialized hardware solutions that enhance efficiency. We provide a detailed discussion of efficient training and fine-tuning methods, edge deployment challenges, and privacy considerations. Additionally, we discuss the diverse applications of lightweight VLMs across healthcare, environmental monitoring, and autonomous systems, illustrating their growing impact. By highlighting key design strategies, current challenges, and offering recommendations for future directions, this survey aims to inspire further research into the practical deployment of VLMs, ultimately making advanced AI accessible in resource-limited settings.

Digital Surface Models (DSM) offer a wealth of height information for understanding the Earth's surface as well as monitoring the existence or change in natural and man-made structures. Classical height estimation requires multi-view geospatial imagery or LiDAR point clouds which can be expensive to acquire. Single-view height estimation using neural network based models shows promise however it can struggle with reconstructing high resolution features. The latest advancements in diffusion models for high resolution image synthesis and editing have yet to be utilized for remote sensing imagery, particularly height estimation. Our approach involves training a generative diffusion model to learn the joint distribution of optical and DSM images across both domains as a Markov chain. This is accomplished by minimizing a denoising score matching objective while being conditioned on the source image to generate realistic high resolution 3D surfaces. In this paper we experiment with conditional denoising diffusion probabilistic models (DDPM) for height estimation from a single remotely sensed image and show promising results on the Vaihingen benchmark dataset.

Recently, diffusion models (DMs) have demonstrated their advantageous potential for generative tasks. Widespread interest exists in incorporating DMs into downstream applications, such as producing or editing photorealistic images. However, practical deployment and unprecedented power of DMs raise legal issues, including copyright protection and monitoring of generated content. In this regard, watermarking has been a proven solution for copyright protection and content monitoring, but it is underexplored in the DMs literature. Specifically, DMs generate samples from longer tracks and may have newly designed multimodal structures, necessitating the modification of conventional watermarking pipelines. To this end, we conduct comprehensive analyses and derive a recipe for efficiently watermarking state-of-the-art DMs (e.g., Stable Diffusion), via training from scratch or finetuning. Our recipe is straightforward but involves empirically ablated implementation details, providing a solid foundation for future research on watermarking DMs. Our Code: https://github.com/yunqing-me/WatermarkDM.

We propose a novel hybrid cable-based robot with manipulator and camera for high-accuracy, medium-throughput plant monitoring in a vertical hydroponic farm and, as an example application, demonstrate non-destructive plant mass estimation. Plant monitoring with high temporal and spatial resolution is important to both farmers and researchers to detect anomalies and develop predictive models for plant growth. The availability of high-quality, off-the-shelf structure-from-motion (SfM) and photogrammetry packages has enabled a vibrant community of roboticists to apply computer vision for non-destructive plant monitoring. While existing approaches tend to focus on either high-throughput (e.g. satellite, unmanned aerial vehicle (UAV), vehicle-mounted, conveyor-belt imagery) or high-accuracy/robustness to occlusions (e.g. turn-table scanner or robot arm), we propose a middle-ground that achieves high accuracy with a medium-throughput, highly automated robot. Our design pairs the workspace scalability of a cable-driven parallel robot (CDPR) with the dexterity of a 4 degree-of-freedom (DoF) robot arm to autonomously image many plants from a variety of viewpoints. We describe our robot design and demonstrate it experimentally by collecting daily photographs of 54 plants from 64 viewpoints each. We show that our approach can produce scientifically useful measurements, operate fully autonomously after initial calibration, and produce better reconstructions and plant property estimates than those of over-canopy methods (e.g. UAV). As example applications, we show that our system can successfully estimate plant mass with a Mean Absolute Error (MAE) of 0.586g and, when used to perform hypothesis testing on the relationship between mass and age, produces p-values comparable to ground-truth data (p=0.0020 and p=0.0016, respectively).

Predictive business process monitoring methods exploit logs of completed cases of a process in order to make predictions about running cases thereof. Existing methods in this space are tailor-made for specific prediction tasks. Moreover, their relative accuracy is highly sensitive to the dataset at hand, thus requiring users to engage in trial-and-error and tuning when applying them in a specific setting. This paper investigates Long Short-Term Memory (LSTM) neural networks as an approach to build consistently accurate models for a wide range of predictive process monitoring tasks. First, we show that LSTMs outperform existing techniques to predict the next event of a running case and its timestamp. Next, we show how to use models for predicting the next task in order to predict the full continuation of a running case. Finally, we apply the same approach to predict the remaining time, and show that this approach outperforms existing tailor-made methods.

DeepSeek-R1 has shown that long chain-of-thought (CoT) reasoning can naturally emerge through a simple reinforcement learning (RL) framework with rule-based rewards, where the training may directly start from the base models-a paradigm referred to as zero RL training. Most recent efforts to reproduce zero RL training have primarily focused on the Qwen2.5 model series, which may not be representative as we find the base models already exhibit strong instruction-following and self-reflection abilities. In this work, we investigate zero RL training across 10 diverse base models, spanning different families and sizes including LLama3-8B, Mistral-7B/24B, DeepSeek-Math-7B, Qwen2.5-math-7B, and all Qwen2.5 models from 0.5B to 32B. Leveraging several key design strategies-such as adjusting format reward and controlling query difficulty-we achieve substantial improvements in both reasoning accuracy and response length across most settings. However, by carefully monitoring the training dynamics, we observe that different base models exhibit distinct patterns during training. For instance, the increased response length does not always correlate with the emergence of certain cognitive behaviors such as verification (i.e., the "aha moment"). Notably, we observe the "aha moment" for the first time in small models not from the Qwen family. We share the key designs that enable successful zero RL training, along with our findings and practices. To facilitate further research, we open-source the code, models, and analysis tools.

Chain-of-thought (CoT) reasoning has been proposed as a transparency mechanism for large language models in safety-critical deployments, yet its effectiveness depends on faithfulness (whether models accurately verbalize the factors that actually influence their outputs), a property that prior evaluations have examined in only two proprietary models, finding acknowledgment rates as low as 25% for Claude 3.7 Sonnet and 39% for DeepSeek-R1. To extend this evaluation across the open-weight ecosystem, this study tests 12 open-weight reasoning models spanning 9 architectural families (7B-685B parameters) on 498 multiple-choice questions from MMLU and GPQA Diamond, injecting six categories of reasoning hints (sycophancy, consistency, visual pattern, metadata, grader hacking, and unethical information) and measuring the rate at which models acknowledge hint influence in their CoT when hints successfully alter answers. Across 41,832 inference runs, overall faithfulness rates range from 39.7% (Seed-1.6-Flash) to 89.9% (DeepSeek-V3.2-Speciale) across model families, with consistency hints (35.5%) and sycophancy hints (53.9%) exhibiting the lowest acknowledgment rates. Training methodology and model family predict faithfulness more strongly than parameter count, and keyword-based analysis reveals a striking gap between thinking-token acknowledgment (approximately 87.5%) and answer-text acknowledgment (approximately 28.6%), suggesting that models internally recognize hint influence but systematically suppress this acknowledgment in their outputs. These findings carry direct implications for the viability of CoT monitoring as a safety mechanism and suggest that faithfulness is not a fixed property of reasoning models but varies systematically with architecture, training method, and the nature of the influencing cue.

Geospatial foundation models (GFMs) have been proposed as generalizable backbones for disaster response, land-cover mapping, food-security monitoring, and other high-stakes Earth-observation tasks. Yet the published work about these models does not give reviewers or users enough information to tell which model fits a given task. We argue that nobody knows what the current state of the art is in geospatial foundation models. The methods may be useful, but the GFM literature does not standardize evaluations, training and testing protocols, released weights, or pretraining controls well enough for anyone to compare or rank them. In a 152-paper audit, we find 46 cross-paper disagreements of at least 10 points for the same model, benchmark, and protocol; 94/126 papers with extractable pretraining data use a configuration no other paper uses; and 39% of GFM papers release no model weights. This lack of community standards can be solved. We propose six concrete expectations: named-license weight release, shared core evaluations, copied-versus-rerun baseline annotations, variance reporting, one shared evaluation harness, and data-vs-architecture-vs-algorithm controls. These gaps are a coordination failure, not a fault of any individual lab; the authors of this paper, like many others in the GFM community, have contributed to them. Rather than just critiquing the community, we aim to provide concrete steps toward a shared understanding of how to innovate GFMs.

Wildlife populations in Africa face severe threats, with vertebrate numbers declining by over 65% in the past five decades. In response, image classification using deep learning has emerged as a promising tool for biodiversity monitoring and conservation. This paper presents a comparative study of deep learning models for automatically classifying African wildlife images, focusing on transfer learning with frozen feature extractors. Using a public dataset of four species: buffalo, elephant, rhinoceros, and zebra; we evaluate the performance of DenseNet-201, ResNet-152, EfficientNet-B4, and Vision Transformer ViT-H/14. DenseNet-201 achieved the best performance among convolutional networks (67% accuracy), while ViT-H/14 achieved the highest overall accuracy (99%), but with significantly higher computational cost, raising deployment concerns. Our experiments highlight the trade-offs between accuracy, resource requirements, and deployability. The best-performing CNN (DenseNet-201) was integrated into a Hugging Face Gradio Space for real-time field use, demonstrating the feasibility of deploying lightweight models in conservation settings. This work contributes to African-grounded AI research by offering practical insights into model selection, dataset preparation, and responsible deployment of deep learning tools for wildlife conservation.

As the global burden of Alzheimer's disease (AD) continues to grow, early and accurate detection has become increasingly critical, especially in regions with limited access to advanced diagnostic tools. We propose BRAINS (Biomedical Retrieval-Augmented Intelligence for Neurodegeneration Screening) to address this challenge. This novel system harnesses the powerful reasoning capabilities of Large Language Models (LLMs) for Alzheimer's detection and monitoring. BRAINS features a dual-module architecture: a cognitive diagnostic module and a case-retrieval module. The Diagnostic Module utilizes LLMs fine-tuned on cognitive and neuroimaging datasets -- including MMSE, CDR scores, and brain volume metrics -- to perform structured assessments of Alzheimer's risk. Meanwhile, the Case Retrieval Module encodes patient profiles into latent representations and retrieves similar cases from a curated knowledge base. These auxiliary cases are fused with the input profile via a Case Fusion Layer to enhance contextual understanding. The combined representation is then processed with clinical prompts for inference. Evaluations on real-world datasets demonstrate BRAINS effectiveness in classifying disease severity and identifying early signs of cognitive decline. This system not only shows strong potential as an assistive tool for scalable, explainable, and early-stage Alzheimer's disease detection, but also offers hope for future applications in the field.

Large Reasoning Models (LRMs), which autonomously produce a reasoning Chain of Thought (CoT) before producing final responses, offer a promising approach to interpreting and monitoring model behaviors. Inspired by the observation that certain CoT patterns -- e.g., ``Wait, did I miss anything?'' -- consistently emerge across tasks, we explore whether LRMs exhibit human-like cognitive habits. Building on Habits of Mind, a well-established framework of cognitive habits associated with successful human problem-solving, we introduce CogTest, a principled benchmark designed to evaluate LRMs' cognitive habits. CogTest includes 16 cognitive habits, each instantiated with 25 diverse tasks, and employs an evidence-first extraction method to ensure reliable habit identification. With CogTest, we conduct a comprehensive evaluation of 16 widely used LLMs (13 LRMs and 3 non-reasoning ones). Our findings reveal that LRMs, unlike conventional LLMs, not only exhibit human-like habits but also adaptively deploy them according to different tasks. Finer-grained analyses further uncover patterns of similarity and difference in LRMs' cognitive habit profiles, particularly certain inter-family similarity (e.g., Qwen-3 models and DeepSeek-R1). Extending the study to safety-related tasks, we observe that certain habits, such as Taking Responsible Risks, are strongly associated with the generation of harmful responses. These findings suggest that studying persistent behavioral patterns in LRMs' CoTs is a valuable step toward deeper understanding of LLM misbehavior. The code is available at: https://github.com/jianshuod/CogTest.

Accurate digital surface models (DSMs) are essential for many geospatial applications, including urban monitoring, environmental analyses, infrastructure management, and change detection. However, large-scale DSMs frequently contain incomplete or outdated regions due to acquisition limitations, reconstruction artifacts, or changes in the built environment. Traditional height completion approaches primarily rely on spatial interpolation or which assume spatial continuity and therefore fail when objects are missing. Recent learning-based approaches improve reconstruction quality but typically require supervised training on sensor-specific datasets, limiting their generalization across domains and sensing conditions. We propose Prior2DSM, a training-free framework for metric DSM completion that operates entirely at test time by leveraging foundation models. Unlike previous height completion approaches that require task-specific training, the proposed method combines self-supervised Vision Transformer (ViT) features from DINOv3 with monocular depth foundation models to propagate metric information from incomplete height priors through semantic feature-space correspondence. Test-time adaptation (TTA) is performed using parameter-efficient low-rank adaptation (LoRA) together with a lightweight multilayer perceptron (MLP), which predicts spatially varying scale and shift parameters to convert relative depth estimates into metric heights. Experiments demonstrate consistent improvements over interpolation based methods, prior-based rescaling height approaches, and state-of-the-art monocular depth estimation models. Prior2DSM reduces reconstruction error while preserving structural fidelity, achieving up to a 46% reduction in RMSE compared to linear fitting of MDE, and further enables DSM updating and coupled RGB-DSM generation.

State Space Models (SSMs) such as Mamba achieve linear-time sequence processing through input-dependent recurrence, but this mechanism introduces a critical safety vulnerability. We show that the spectral radius rho(A-bar) of the discretized transition operator governs effective memory horizon: when an adversary drives rho toward zero through gradient-based Hidden State Poisoning, memory collapses from millions of tokens to mere dozens, silently destroying reasoning capacity without triggering output-level alarms. We prove an Evasion Existence Theorem showing that for any output-only defense, adversarial inputs exist that simultaneously induce spectral collapse and evade detection, then introduce SpectralGuard, a real-time monitor that tracks spectral stability across all model layers. SpectralGuard achieves F1=0.961 against non-adaptive attackers and retains F1=0.842 under the strongest adaptive setting, with sub-15ms per-token latency. Causal interventions and cross-architecture transfer to hybrid SSM-Attention systems confirm that spectral monitoring provides a principled, deployable safety layer for recurrent foundation models.

Chain-of-thought (CoT) reasoning not only enhances large language model performance but also provides critical insights into decision-making processes, marking it as a useful tool for monitoring model intent and planning. However, recent works have shown that banning the mention of a specific example of reward hacking causes obfuscation of the undesired reasoning traces but the persistence of the undesired behavior, threatening the reliability of CoT monitoring. We provide an extension to these results with regard to the ability of models to learn a specific type of obfuscated reasoning: steganography. First, we show that penalizing the use of specific strings within load-bearing reasoning traces causes models to substitute alternative strings. Crucially, this does not alter the underlying method by which the model performs the task, demonstrating that the model can learn to steganographically encode its reasoning.We further demonstrate that models can generalize an encoding scheme. When the penalized strings belong to an overarching class, the model learns not only to substitute strings seen in training, but also develops a general encoding scheme for all members of the class which it can apply to held-out testing strings.

Automation in agriculture plays a vital role in addressing challenges related to crop monitoring and disease management, particularly through early detection systems. This study investigates the effectiveness of combining multimodal Large Language Models (LLMs), specifically GPT-4o, with Convolutional Neural Networks (CNNs) for automated plant disease classification using leaf imagery. Leveraging the PlantVillage dataset, we systematically evaluate model performance across zero-shot, few-shot, and progressive fine-tuning scenarios. A comparative analysis between GPT-4o and the widely used ResNet-50 model was conducted across three resolutions (100, 150, and 256 pixels) and two plant species (apple and corn). Results indicate that fine-tuned GPT-4o models achieved slightly better performance compared to the performance of ResNet-50, achieving up to 98.12% classification accuracy on apple leaf images, compared to 96.88% achieved by ResNet-50, with improved generalization and near-zero training loss. However, zero-shot performance of GPT-4o was significantly lower, underscoring the need for minimal training. Additional evaluations on cross-resolution and cross-plant generalization revealed the models' adaptability and limitations when applied to new domains. The findings highlight the promise of integrating multimodal LLMs into automated disease detection pipelines, enhancing the scalability and intelligence of precision agriculture systems while reducing the dependence on large, labeled datasets and high-resolution sensor infrastructure. Large Language Models, Vision Language Models, LLMs and CNNs, Disease Detection with Vision Language Models, VLMs

Lower-Limb Fractures (LLF) are a major health concern for older adults, often leading to reduced mobility and prolonged recovery, potentially impairing daily activities and independence. During recovery, older adults frequently face social isolation and functional decline, complicating rehabilitation and adversely affecting physical and mental health. Multi-modal sensor platforms that continuously collect data and analyze it using machine-learning algorithms can remotely monitor this population and infer health outcomes. They can also alert clinicians to individuals at risk of isolation and decline. This paper presents a new publicly available multi-modal sensor dataset, MAISON-LLF, collected from older adults recovering from LLF in community settings. The dataset includes data from smartphone and smartwatch sensors, motion detectors, sleep-tracking mattresses, and clinical questionnaires on isolation and decline. The dataset was collected from ten older adults living alone at home for eight weeks each, totaling 560 days of 24-hour sensor data. For technical validation, supervised machine-learning and deep-learning models were developed using the sensor and clinical questionnaire data, providing a foundational comparison for the research community.

In recent years, Large Language Models (LLMs) have demonstrated remarkable capabilities in data analytics when integrated with Multi-Agent Systems (MAS). However, these systems often struggle with complex tasks that involve diverse functional requirements and intricate data processing challenges, necessitating customized solutions that lack broad applicability. Furthermore, current MAS fail to emulate essential human-like traits such as self-planning, self-monitoring, and collaborative work in dynamic environments, leading to inefficiencies and resource wastage. To address these limitations, we propose ROMAS, a novel Role-Based M ulti-A gent System designed to adapt to various scenarios while enabling low code development and one-click deployment. ROMAS has been effectively deployed in DB-GPT [Xue et al., 2023a, 2024b], a well-known project utilizing LLM-powered database analytics, showcasing its practical utility in real-world scenarios. By integrating role-based collaborative mechanisms for self-monitoring and self-planning, and leveraging existing MAS capabilities to enhance database interactions, ROMAS offers a more effective and versatile solution. Experimental evaluations of ROMAS demonstrate its superiority across multiple scenarios, highlighting its potential to advance the field of multi-agent data analytics.

Improving diesel engine efficiency, reducing emissions, and enabling robust health monitoring have been critical research topics in engine modelling. While recent advancements in the use of neural networks for system monitoring have shown promising results, such methods often focus on component-level analysis, lack generalizability, and physical interpretability. In this study, we propose a novel hybrid framework that combines physics-informed neural networks (PINNs) with deep operator networks (DeepONet) to enable accurate and computationally efficient parameter identification in mean-value diesel engine models. Our method leverages physics-based system knowledge in combination with data-driven training of neural networks to enhance model applicability. Incorporating offline-trained DeepONets to predict actuator dynamics significantly lowers the online computation cost when compared to the existing PINN framework. To address the re-training burden typical of PINNs under varying input conditions, we propose two transfer learning (TL) strategies: (i) a multi-stage TL scheme offering better runtime efficiency than full online training of the PINN model and (ii) a few-shot TL scheme that freezes a shared multi-head network body and computes physics-based derivatives required for model training outside the training loop. The second strategy offers a computationally inexpensive and physics-based approach for predicting engine dynamics and parameter identification, offering computational efficiency over the existing PINN framework. Compared to existing health monitoring methods, our framework combines the interpretability of physics-based models with the flexibility of deep learning, offering substantial gains in generalization, accuracy, and deployment efficiency for diesel engine diagnostics.

Deploying LLMs raises two coupled challenges: (1) monitoring - estimating where a model underperforms as traffic and domains drift - and (2) improvement - prioritizing data acquisition to close the largest performance gaps. We test whether an inference-time signal can estimate slice-level accuracy under domain shift. For each response, we compute an output-entropy profile from final-layer next-token probabilities (from top-k logprobs) and summarize it with eleven statistics. A lightweight classifier predicts instance correctness, and averaging predicted probabilities yields a domain-level accuracy estimate. We evaluate on ten STEM reasoning benchmarks with exhaustive train/test compositions (k in {1,2,3,4}; all "10 choose k" combinations), across nine LLMs from six families (3B-20B). Estimates often track held-out benchmark accuracy, and several models show near-monotonic ordering of domains. Output-entropy profiles are thus an accessible signal for scalable monitoring and for targeting data acquisition.

The rapid evolution of embodied agents has accelerated the deployment of household robots in real-world environments. However, unlike structured industrial settings, household spaces introduce unpredictable safety risks, where system limitations such as perception latency and lack of common sense knowledge can lead to dangerous errors. Current safety evaluations, often restricted to static images, text, or general hazards, fail to adequately benchmark dynamic unsafe action detection in these specific contexts. To bridge this gap, we introduce HomeSafe-Bench, a challenging benchmark designed to evaluate Vision-Language Models (VLMs) on unsafe action detection in household scenarios. HomeSafe-Bench is contrusted via a hybrid pipeline combining physical simulation with advanced video generation and features 438 diverse cases across six functional areas with fine-grained multidimensional annotations. Beyond benchmarking, we propose Hierarchical Dual-Brain Guard for Household Safety (HD-Guard), a hierarchical streaming architecture for real-time safety monitoring. HD-Guard coordinates a lightweight FastBrain for continuous high-frequency screening with an asynchronous large-scale SlowBrain for deep multimodal reasoning, effectively balancing inference efficiency with detection accuracy. Evaluations demonstrate that HD-Guard achieves a superior trade-off between latency and performance, while our analysis identifies critical bottlenecks in current VLM-based safety detection.

Continuous glucose monitoring (CGM) provides a dense view of daily metabolic physiology, yet existing generic time-series and CGM-specific foundation models often encode glucose traces as entangled single-stream sequences, leaving the distinct temporal structure of glycemic dynamics only implicitly modeled. We present GlucoFM, a lightweight CGM foundation model that aligns irregular recordings to a 24-hour chronological grid, preserves observation masks, and decomposes glucose dynamics into slow physiological state and transient event streams, capturing low-frequency glycemic baselines and short-term deviations that may reflect acute physiological responses or sensor artifacts. GlucoFM is pretrained on 109,066 hours of unlabeled CGM recordings from 477 subjects with two complementary objectives: masked contextual latent prediction over fused daily representations and temporal dynamics prediction over state and event streams. Across four diverse cohorts and seven clinical prediction tasks, GlucoFM achieves the strongest subject-disjoint linear-probing performance among evaluated baselines, improving average PR-AUC by 4.1 points over the best CGM-specific foundation model. Its gains are most pronounced on core metabolic outcomes, leading PR-AUC on all diabetes-risk and β-cell dysfunction tasks and on 3 of 4 insulin-resistance tasks. GlucoFM also achieves the best overall cross-dataset transfer performance and strong few-shot adaptation among evaluated methods, and consistent gains when aggregating multiple days for subject-level prediction, highlighting physiology-aware decomposition as an effective inductive bias for transferable CGM representation learning.

This paper investigates using knowledge editing techniques to detoxify Large Language Models (LLMs). We construct a benchmark, SafeEdit, which covers nine unsafe categories with various powerful attack prompts and equips comprehensive metrics for systematic evaluation. We conduct experiments to compare knowledge editing approaches with previous baselines, indicating that knowledge editing has the potential to efficiently detoxify LLMs with limited impact on general performance. Then, we propose a simple yet effective baseline, dubbed Detoxifying with Intraoperative Neural Monitoring (DINM), to diminish the toxicity of LLMs within a few tuning steps via only one instance. We further provide an in-depth analysis of the internal mechanism for various detoxify approaches, demonstrating that previous methods like SFT and DPO may merely suppress the activations of toxic parameters, while DINM mitigates the toxicity of the toxic parameters to a certain extent, making permanent adjustments. We hope that these insights could shed light on future work of developing detoxifying approaches and the underlying knowledge mechanisms of LLMs. Code and benchmark are available at https://github.com/zjunlp/EasyEdit.

Though safety alignment has been applied to most large language models (LLMs), LLM service providers generally deploy a subsequent moderation as the external safety guardrail in real-world products. Existing moderators mainly practice a conventional full detection, which determines the harmfulness based on the complete LLM output, causing high service latency. Recent works pay more attention to partial detection where moderators oversee the generation midway and early stop the output if harmfulness is detected, but they directly apply moderators trained with the full detection paradigm to incomplete outputs, introducing a training-inference gap that lowers the performance. In this paper, we explore how to form a data-and-model solution that natively supports partial detection. For the data, we construct FineHarm, a dataset consisting of 29K prompt-response pairs with fine-grained annotations to provide reasonable supervision for token-level training. Then, we propose the streaming content monitor, which is trained with dual supervision of response- and token-level labels and can follow the output stream of LLM to make a timely judgment of harmfulness. Experiments show that SCM gains 0.95+ in macro F1 score that is comparable to full detection, by only seeing the first 18% of tokens in responses on average. Moreover, the SCM can serve as a pseudo-harmfulness annotator for improving safety alignment and lead to a higher harmlessness score than DPO.

Large Reasoning Models (LRMs) have significantly enhanced their capabilities in complex problem-solving by introducing a thinking draft that enables multi-path Chain-of-Thought explorations before producing final answers. Ensuring the faithfulness of these intermediate reasoning processes is crucial for reliable monitoring, interpretation, and effective control. In this paper, we propose a systematic counterfactual intervention framework to rigorously evaluate thinking draft faithfulness. Our approach focuses on two complementary dimensions: (1) Intra-Draft Faithfulness, which assesses whether individual reasoning steps causally influence subsequent steps and the final draft conclusion through counterfactual step insertions; and (2) Draft-to-Answer Faithfulness, which evaluates whether final answers are logically consistent with and dependent on the thinking draft, by perturbing the draft's concluding logic. We conduct extensive experiments across six state-of-the-art LRMs. Our findings show that current LRMs demonstrate selective faithfulness to intermediate reasoning steps and frequently fail to faithfully align with the draft conclusions. These results underscore the need for more faithful and interpretable reasoning in advanced LRMs.

We stress test monitoring systems for detecting covert misbehavior in autonomous LLM agents (e.g., secretly sharing private information). To this end, we systematize a monitor red teaming (MRT) workflow that incorporates: (1) varying levels of agent and monitor situational awareness; (2) distinct adversarial strategies to evade the monitor, such as prompt injection; and (3) two datasets and environments -- SHADE-Arena for tool-calling agents and our new CUA-SHADE-Arena, which extends TheAgentCompany, for computer-use agents. We run MRT on existing LLM monitor scaffoldings, which orchestrate LLMs and parse agent trajectories, alongside a new hybrid hierarchical-sequential scaffolding proposed in this work. Our empirical results yield three key findings. First, agent awareness dominates monitor awareness: an agent's knowledge that it is being monitored substantially degrades the monitor's reliability. On the contrary, providing the monitor with more information about the agent is less helpful than expected. Second, monitor scaffolding matters more than monitor awareness: the hybrid scaffolding consistently outperforms baseline monitor scaffolding, and can enable weaker models to reliably monitor stronger agents -- a weak-to-strong scaling effect. Third, in a human-in-the-loop setting where humans discuss with the LLM monitor to get an updated judgment for the agent's behavior, targeted human oversight is most effective; escalating only pre-flagged cases to human reviewers improved the TPR by approximately 15% at FPR = 0.01. Our work establishes a standard workflow for MRT, highlighting the lack of adversarial robustness for LLMs and humans when monitoring and detecting agent misbehavior. We release code, data, and logs to spur further research.

We introduce a fully 3D, deep learning-based approach for the joint inversion of time-lapse surface gravity and seismic data for reconstructing subsurface density and velocity models. The target application of this proposed inversion approach is the prediction of subsurface CO2 plumes as a complementary tool for monitoring CO2 sequestration deployments. Our joint inversion technique outperforms deep learning-based gravity-only and seismic-only inversion models, achieving improved density and velocity reconstruction, accurate segmentation, and higher R-squared coefficients. These results indicate that deep learning-based joint inversion is an effective tool for CO_2 storage monitoring. Future work will focus on validating our approach with larger datasets, simulations with other geological storage sites, and ultimately field data.

Progress toward the United Nations Sustainable Development Goals (SDGs) has been hindered by a lack of data on key environmental and socioeconomic indicators, which historically have come from ground surveys with sparse temporal and spatial coverage. Recent advances in machine learning have made it possible to utilize abundant, frequently-updated, and globally available data, such as from satellites or social media, to provide insights into progress toward SDGs. Despite promising early results, approaches to using such data for SDG measurement thus far have largely evaluated on different datasets or used inconsistent evaluation metrics, making it hard to understand whether performance is improving and where additional research would be most fruitful. Furthermore, processing satellite and ground survey data requires domain knowledge that many in the machine learning community lack. In this paper, we introduce SustainBench, a collection of 15 benchmark tasks across 7 SDGs, including tasks related to economic development, agriculture, health, education, water and sanitation, climate action, and life on land. Datasets for 11 of the 15 tasks are released publicly for the first time. Our goals for SustainBench are to (1) lower the barriers to entry for the machine learning community to contribute to measuring and achieving the SDGs; (2) provide standard benchmarks for evaluating machine learning models on tasks across a variety of SDGs; and (3) encourage the development of novel machine learning methods where improved model performance facilitates progress towards the SDGs.

Human age estimation from facial images represents a challenging computer vision task with significant applications in biometrics, healthcare, and human-computer interaction. While traditional deep learning approaches require extensive labeled datasets and domain-specific training, recent advances in large vision-language models (LVLMs) offer the potential for zero-shot age estimation. This study presents a comprehensive zero-shot evaluation of state-of-the-art Large Vision-Language Models (LVLMs) for facial age estimation, a task traditionally dominated by domain-specific convolutional networks and supervised learning. We assess the performance of GPT-4o, Claude 3.5 Sonnet, and LLaMA 3.2 Vision on two benchmark datasets, UTKFace and FG-NET, without any fine-tuning or task-specific adaptation. Using eight evaluation metrics, including MAE, MSE, RMSE, MAPE, MBE, R^2, CCC, and pm5-year accuracy, we demonstrate that general-purpose LVLMs can deliver competitive performance in zero-shot settings. Our findings highlight the emergent capabilities of LVLMs for accurate biometric age estimation and position these models as promising tools for real-world applications. Additionally, we highlight performance disparities linked to image quality and demographic subgroups, underscoring the need for fairness-aware multimodal inference. This work introduces a reproducible benchmark and positions LVLMs as promising tools for real-world applications in forensic science, healthcare monitoring, and human-computer interaction. The benchmark focuses on strict zero-shot inference without fine-tuning and highlights remaining challenges related to prompt sensitivity, interpretability, computational cost, and demographic fairness.

Process Model Forecasting (PMF) aims to predict how the control-flow structure of a process evolves over time by modeling the temporal dynamics of directly-follows (DF) relations, complementing predictive process monitoring that focuses on single-case prefixes. Prior benchmarks show that machine learning and deep learning models provide only modest gains over statistical baselines, mainly due to the sparsity and heterogeneity of the DF time series. We investigate Time Series Foundation Models (TSFMs), large pre-trained models for generic time series, as an alternative for PMF. Using DF time series derived from real-life event logs, we compare zero-shot use of TSFMs, without additional training, with fine-tuned variants adapted on PMF-specific data. TSFMs generally achieve lower forecasting errors (MAE and RMSE) than traditional and specialized models trained from scratch on the same logs, indicating effective transfer of temporal structure from non-process domains. While fine-tuning can further improve accuracy, the gains are often small and may disappear on smaller or more complex datasets, so zero-shot use remains a strong default. Our study highlights the generalization capability and data efficiency of TSFMs for process-related time series and, to the best of our knowledge, provides the first systematic evaluation of temporal foundation models for PMF.

Commercial plant phenotyping systems using fixed cameras cannot perceive many plant details due to leaf occlusion. In this paper, we present Botany-Bot, a system for building detailed "annotated digital twins" of living plants using two stereo cameras, a digital turntable inside a lightbox, an industrial robot arm, and 3D segmentated Gaussian Splat models. We also present robot algorithms for manipulating leaves to take high-resolution indexable images of occluded details such as stem buds and the underside/topside of leaves. Results from experiments suggest that Botany-Bot can segment leaves with 90.8% accuracy, detect leaves with 86.2% accuracy, lift/push leaves with 77.9% accuracy, and take detailed overside/underside images with 77.3% accuracy. Code, videos, and datasets are available at https://berkeleyautomation.github.io/Botany-Bot/.

Spatio-temporal prediction is a pivotal task with broad applications in traffic management, climate monitoring, energy scheduling, etc. However, existing methodologies often struggle to balance model expressiveness and computational efficiency, especially when scaling to large real-world datasets. To tackle these challenges, we propose STH-SepNet (Spatio-Temporal Hypergraph Separation Networks), a novel framework that decouples temporal and spatial modeling to enhance both efficiency and precision. Therein, the temporal dimension is modeled using lightweight large language models, which effectively capture low-rank temporal dynamics. Concurrently, the spatial dimension is addressed through an adaptive hypergraph neural network, which dynamically constructs hyperedges to model intricate, higher-order interactions. A carefully designed gating mechanism is integrated to seamlessly fuse temporal and spatial representations. By leveraging the fundamental principles of low-rank temporal dynamics and spatial interactions, STH-SepNet offers a pragmatic and scalable solution for spatio-temporal prediction in real-world applications. Extensive experiments on large-scale real-world datasets across multiple benchmarks demonstrate the effectiveness of STH-SepNet in boosting predictive performance while maintaining computational efficiency. This work may provide a promising lightweight framework for spatio-temporal prediction, aiming to reduce computational demands and while enhancing predictive performance. Our code is avaliable at https://github.com/SEU-WENJIA/ST-SepNet-Lightweight-LLMs-Meet-Adaptive-Hypergraphs.

Rip currents are strong, localized and narrow currents of water that flow outwards into the sea, causing numerous beach-related injuries and fatalities worldwide. Accurate identification of rip currents remains challenging due to their amorphous nature and the lack of annotated data, which often requires expert knowledge. To address these issues, we present RipVIS, a large-scale video instance segmentation benchmark explicitly designed for rip current segmentation. RipVIS is an order of magnitude larger than previous datasets, featuring 184 videos (212,328 frames), of which 150 videos (163,528 frames) are with rip currents, collected from various sources, including drones, mobile phones, and fixed beach cameras. Our dataset encompasses diverse visual contexts, such as wave-breaking patterns, sediment flows, and water color variations, across multiple global locations, including USA, Mexico, Costa Rica, Portugal, Italy, Greece, Romania, Sri Lanka, Australia and New Zealand. Most videos are annotated at 5 FPS to ensure accuracy in dynamic scenarios, supplemented by an additional 34 videos (48,800 frames) without rip currents. We conduct comprehensive experiments with Mask R-CNN, Cascade Mask R-CNN, SparseInst and YOLO11, fine-tuning these models for the task of rip current segmentation. Results are reported in terms of multiple metrics, with a particular focus on the F_2 score to prioritize recall and reduce false negatives. To enhance segmentation performance, we introduce a novel post-processing step based on Temporal Confidence Aggregation (TCA). RipVIS aims to set a new standard for rip current segmentation, contributing towards safer beach environments. We offer a benchmark website to share data, models, and results with the research community, encouraging ongoing collaboration and future contributions, at https://ripvis.ai.

VIP navigation requires multiple DNN models for identification, posture analysis, and depth estimation to ensure safe mobility. Using a hazard vest as a unique identifier enhances visibility while selecting the right DNN model and computing device balances accuracy and real-time performance. We present Ocularone-Bench, which is a benchmark suite designed to address the lack of curated datasets for uniquely identifying individuals in crowded environments and the need for benchmarking DNN inference times on resource-constrained edge devices. The suite evaluates the accuracy-latency trade-offs of YOLO models retrained on this dataset and benchmarks inference times of situation awareness models across edge accelerators and high-end GPU workstations. Our study on NVIDIA Jetson devices and RTX 4090 workstation demonstrates significant improvements in detection accuracy, achieving up to 99.4% precision, while also providing insights into real-time feasibility for mobile deployment. Beyond VIP navigation, Ocularone-Bench is applicable to senior citizens, children and worker safety monitoring, and other vision-based applications.

While numerous recent benchmarks focus on evaluating generic Vision-Language Models (VLMs), they fall short in addressing the unique demands of geospatial applications. Generic VLM benchmarks are not designed to handle the complexities of geospatial data, which is critical for applications such as environmental monitoring, urban planning, and disaster management. Some of the unique challenges in geospatial domain include temporal analysis for changes, counting objects in large quantities, detecting tiny objects, and understanding relationships between entities occurring in Remote Sensing imagery. To address this gap in the geospatial domain, we present GEOBench-VLM, a comprehensive benchmark specifically designed to evaluate VLMs on geospatial tasks, including scene understanding, object counting, localization, fine-grained categorization, and temporal analysis. Our benchmark features over 10,000 manually verified instructions and covers a diverse set of variations in visual conditions, object type, and scale. We evaluate several state-of-the-art VLMs to assess their accuracy within the geospatial context. The results indicate that although existing VLMs demonstrate potential, they face challenges when dealing with geospatial-specific examples, highlighting the room for further improvements. Specifically, the best-performing GPT4o achieves only 40\% accuracy on MCQs, which is only double the random guess performance. Our benchmark is publicly available at https://github.com/The-AI-Alliance/GEO-Bench-VLM .

Large language models (LLMs) represent a transformative class of AI tools capable of revolutionizing various aspects of healthcare by generating human-like responses across diverse contexts and adapting to novel tasks following human instructions. Their potential application spans a broad range of medical tasks, such as clinical documentation, matching patients to clinical trials, and answering medical questions. In this primer paper, we propose an actionable guideline to help healthcare professionals more efficiently utilize LLMs in their work, along with a set of best practices. This approach consists of several main phases, including formulating the task, choosing LLMs, prompt engineering, fine-tuning, and deployment. We start with the discussion of critical considerations in identifying healthcare tasks that align with the core capabilities of LLMs and selecting models based on the selected task and data, performance requirements, and model interface. We then review the strategies, such as prompt engineering and fine-tuning, to adapt standard LLMs to specialized medical tasks. Deployment considerations, including regulatory compliance, ethical guidelines, and continuous monitoring for fairness and bias, are also discussed. By providing a structured step-by-step methodology, this tutorial aims to equip healthcare professionals with the tools necessary to effectively integrate LLMs into clinical practice, ensuring that these powerful technologies are applied in a safe, reliable, and impactful manner.

We propose an agent architecture that automates parts of the common reinforcement learning experiment workflow, to enable automated mastery of control domains for embodied agents. To do so, it leverages a VLM to perform some of the capabilities normally required of a human experimenter, including the monitoring and analysis of experiment progress, the proposition of new tasks based on past successes and failures of the agent, decomposing tasks into a sequence of subtasks (skills), and retrieval of the skill to execute - enabling our system to build automated curricula for learning. We believe this is one of the first proposals for a system that leverages a VLM throughout the full experiment cycle of reinforcement learning. We provide a first prototype of this system, and examine the feasibility of current models and techniques for the desired level of automation. For this, we use a standard Gemini model, without additional fine-tuning, to provide a curriculum of skills to a language-conditioned Actor-Critic algorithm, in order to steer data collection so as to aid learning new skills. Data collected in this way is shown to be useful for learning and iteratively improving control policies in a robotics domain. Additional examination of the ability of the system to build a growing library of skills, and to judge the progress of the training of those skills, also shows promising results, suggesting that the proposed architecture provides a potential recipe for fully automated mastery of tasks and domains for embodied agents.

Pre-trained large language models (PLMs) have the potential to support urban science research through content creation, information extraction, assisted programming, text classification, and other technical advances. In this research, we explored the opportunities, challenges, and prospects of PLMs in urban science research. Specifically, we discussed potential applications of PLMs to urban institution, urban space, urban information, and citizen behaviors research through seven examples using ChatGPT. We also examined the challenges of PLMs in urban science research from both technical and social perspectives. The prospects of the application of PLMs in urban science research were then proposed. We found that PLMs can effectively aid in understanding complex concepts in urban science, facilitate urban spatial form identification, assist in disaster monitoring, and sense public sentiment. At the same time, however, the applications of PLMs in urban science research face evident threats, such as technical limitations, security, privacy, and social bias. The development of fundamental models based on domain knowledge and human-AI collaboration may help improve PLMs to support urban science research in future.

Remote sensing change detection is crucial for understanding the dynamics of our planet's surface, facilitating the monitoring of environmental changes, evaluating human impact, predicting future trends, and supporting decision-making. In this work, we introduce a novel approach for change detection that can leverage off-the-shelf, unlabeled remote sensing images in the training process by pre-training a Denoising Diffusion Probabilistic Model (DDPM) - a class of generative models used in image synthesis. DDPMs learn the training data distribution by gradually converting training images into a Gaussian distribution using a Markov chain. During inference (i.e., sampling), they can generate a diverse set of samples closer to the training distribution, starting from Gaussian noise, achieving state-of-the-art image synthesis results. However, in this work, our focus is not on image synthesis but on utilizing it as a pre-trained feature extractor for the downstream application of change detection. Specifically, we fine-tune a lightweight change classifier utilizing the feature representations produced by the pre-trained DDPM alongside change labels. Experiments conducted on the LEVIR-CD, WHU-CD, DSIFN-CD, and CDD datasets demonstrate that the proposed DDPM-CD method significantly outperforms the existing state-of-the-art change detection methods in terms of F1 score, IoU, and overall accuracy, highlighting the pivotal role of pre-trained DDPM as a feature extractor for downstream applications. We have made both the code and pre-trained models available at https://github.com/wgcban/ddpm-cd