Daily Papers

AI control protocols serve as a defense mechanism to stop untrusted LLM agents from causing harm in autonomous settings. Prior work treats this as a security problem, stress testing with exploits that use the deployment context to subtly complete harmful side tasks, such as backdoor insertion. In practice, most AI control protocols are fundamentally based on LLM monitors, which can become a central point of failure. We study adaptive attacks by an untrusted model that knows the protocol and the monitor model, which is plausible if the untrusted model was trained with a later knowledge cutoff or can search for this information autonomously. We instantiate a simple adaptive attack vector by which the attacker embeds publicly known or zero-shot prompt injections in the model outputs. Using this tactic, frontier models consistently evade diverse monitors and complete malicious tasks on two main AI control benchmarks. The attack works universally against current protocols that rely on a monitor. Furthermore, the recent Defer-to-Resample protocol even backfires, as its resampling amplifies the prompt injection and effectively reframes it as a best-of-n attack. In general, adaptive attacks on monitor models represent a major blind spot in current control protocols and should become a standard component of evaluations for future AI control mechanisms.

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.

Monitoring is an important aspect of safely deploying Large Language Models (LLMs). This paper examines activation probes for detecting ``high-stakes'' interactions -- where the text indicates that the interaction might lead to significant harm -- as a critical, yet underexplored, target for such monitoring. We evaluate several probe architectures trained on synthetic data, and find them to exhibit robust generalization to diverse, out-of-distribution, real-world data. Probes' performance is comparable to that of prompted or finetuned medium-sized LLM monitors, while offering computational savings of six orders-of-magnitude. These savings are enabled by reusing activations of the model that is being monitored. Our experiments also highlight the potential of building resource-aware hierarchical monitoring systems, where probes serve as an efficient initial filter and flag cases for more expensive downstream analysis. We release our novel synthetic dataset and the codebase at https://github.com/arrrlex/models-under-pressure.

Attacks powered by Large Language Model (LLM) agents represent a growing threat to modern cybersecurity. To address this concern, we present LLM Honeypot, a system designed to monitor autonomous AI hacking agents. By augmenting a standard SSH honeypot with prompt injection and time-based analysis techniques, our framework aims to distinguish LLM agents among all attackers. Over a trial deployment of about three months in a public environment, we collected 8,130,731 hacking attempts and 8 potential AI agents. Our work demonstrates the emergence of AI-driven threats and their current level of usage, serving as an early warning of malicious LLM agents in the wild.

LLMs are increasingly capable of persuasion, which raises the question of how to protect users against manipulation. In a preregistered user study (N=120) across four decision-making scenarios, we find that an adversarial LLM with a hidden goal succeeds in steering users' decisions 65.4% of the time. We then introduce a "warden" model: a secondary LLM that monitors the human-AI interaction trace in real time and issues non-binding, private advisories to the user when it detects manipulation. Adding a warden more than halves the adversary's success rate to 30.4%, with a much smaller (8.6 percentage points) reduction for genuine interactions. To probe the mechanism behind these results, we release COAX-Bench, a simulation benchmark spanning 14 decision-making scenarios, including hiring, voting, and file access. Across 16,212 simulated multi-agent interactions, capable adversarial LLMs achieve their hidden goals in 34.7% of cases, which warden models reduce to 12.3%. Notably, even warden models substantially weaker than the adversary they oversee provide meaningful protection, suggesting a path for scalable oversight of more capable models.

We introduce AutoMonitor-Bench, the first benchmark designed to systematically evaluate the reliability of LLM-based misbehavior monitors across diverse tasks and failure modes. AutoMonitor-Bench consists of 3,010 carefully annotated test samples spanning question answering, code generation, and reasoning, with paired misbehavior and benign instances. We evaluate monitors using two complementary metrics: Miss Rate (MR) and False Alarm Rate (FAR), capturing failures to detect misbehavior and oversensitivity to benign behavior, respectively. Evaluating 12 proprietary and 10 open-source LLMs, we observe substantial variability in monitoring performance and a consistent trade-off between MR and FAR, revealing an inherent safety-utility tension. To further explore the limits of monitor reliability, we construct a large-scale training corpus of 153,581 samples and fine-tune Qwen3-4B-Instruction to investigate whether training on known, relatively easy-to-construct misbehavior datasets improves monitoring performance on unseen and more implicit misbehaviors. Our results highlight the challenges of reliable, scalable misbehavior monitoring and motivate future work on task-aware designing and training strategies for LLM-based monitors.

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.

Large language model (LLM) agents now execute long, tool-using tasks where final outcome checks can arrive too late for intervention. Online warning requires lightweight prefix monitors over heterogeneous traces, but hand-authored event schemas are brittle and deployment-time LLM judging is costly. We introduce PrefixGuard, a trace-to-monitor framework with an offline StepView induction step followed by supervised monitor training. StepView induces deterministic typed-step adapters from raw trace samples, and the monitor learns an event abstraction and prefix-risk scorer from terminal outcomes. Across WebArena, τ^2-Bench, SkillsBench, and TerminalBench, the strongest PrefixGuard monitors reach 0.900/0.710/0.533/0.557 AUPRC. Using the strongest backend within each representation, they improve over raw-text controls by an average of +0.137 AUPRC. LLM judges remain substantially weaker under the same prefix-warning protocol. We also derive an observability ceiling on score-based area under the precision-recall curve (AUPRC) that separates monitor error from failures lacking evidence in the observed prefix. For finite-state audit, post-hoc deterministic finite automaton (DFA) extraction remains compact on WebArena and τ^2-Bench (29 and 20 states) but expands to 151 and 187 states on SkillsBench and TerminalBench. Finally, first-alert diagnostics show that strong ranking does not imply deployment utility: WebArena ranks well yet fails to support low-false-alarm alerts, whereas τ^2-Bench and TerminalBench retain more actionable early alerts. Together, these results position PrefixGuard as a practical monitor-synthesis recipe with explicit diagnostics for when prefix warnings translate into actionable interventions.

Frontier large language model training consumes massive accelerator fleets and long wall-clock computation, making stability failures costly when they occur. After a numerical or a hyperparameter fault has already destabilized the training dynamics, it may continue for thousands of steps while loss and gradient norms still appear normal. We study mechanism-driven detection of training instability by deriving internal monitors from the functional role of each critical module and from the earliest computational sites where failures are expected to produce measurable signatures. For low-precision flash attention, we monitor the spectral entropy of a QK bilinear decomposition, whose first-order term becomes abnormal before the loss fully collapses. For MoE routers, we derive indicators from their role in expert selection. Our fault-injection experiments on low-precision attention, large learning-rate, and combined faults show that these signals provide distinct signatures for different failures, triggering thousands of steps before loss divergence.

Although memory capabilities of AI agents are gaining increasing attention, existing solutions remain fundamentally limited. Most rely on flat, narrowly scoped memory components, constraining their ability to personalize, abstract, and reliably recall user-specific information over time. To this end, we introduce MIRIX, a modular, multi-agent memory system that redefines the future of AI memory by solving the field's most critical challenge: enabling language models to truly remember. Unlike prior approaches, MIRIX transcends text to embrace rich visual and multimodal experiences, making memory genuinely useful in real-world scenarios. MIRIX consists of six distinct, carefully structured memory types: Core, Episodic, Semantic, Procedural, Resource Memory, and Knowledge Vault, coupled with a multi-agent framework that dynamically controls and coordinates updates and retrieval. This design enables agents to persist, reason over, and accurately retrieve diverse, long-term user data at scale. We validate MIRIX in two demanding settings. First, on ScreenshotVQA, a challenging multimodal benchmark comprising nearly 20,000 high-resolution computer screenshots per sequence, requiring deep contextual understanding and where no existing memory systems can be applied, MIRIX achieves 35% higher accuracy than the RAG baseline while reducing storage requirements by 99.9%. Second, on LOCOMO, a long-form conversation benchmark with single-modal textual input, MIRIX attains state-of-the-art performance of 85.4%, far surpassing existing baselines. These results show that MIRIX sets a new performance standard for memory-augmented LLM agents. To allow users to experience our memory system, we provide a packaged application powered by MIRIX. It monitors the screen in real time, builds a personalized memory base, and offers intuitive visualization and secure local storage to ensure privacy.

Large Language Model (LLM)-based agent systems are increasingly deployed for complex real-world tasks but remain vulnerable to natural language-based attacks that exploit over-privileged tool use. This paper aims to understand and mitigate such attacks through the lens of privilege escalation, defined as agent actions exceeding the least privilege required for a user's intended task. Based on a formal model of LLM agent systems, we identify novel privilege escalation scenarios, particularly in multi-agent systems, including a variant akin to the classic confused deputy problem. To defend against both known and newly demonstrated privilege escalation, we propose SEAgent, a mandatory access control (MAC) framework built upon attribute-based access control (ABAC). SEAgent monitors agent-tool interactions via an information flow graph and enforces customizable security policies based on entity attributes. Our evaluations show that SEAgent effectively blocks various privilege escalation while maintaining a low false positive rate and negligible system overhead. This demonstrates its robustness and adaptability in securing LLM-based agent systems.

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.

Large language model (LLM) developers aim for their models to be honest, helpful, and harmless. However, when faced with malicious requests, models are trained to refuse, sacrificing helpfulness. We show that frontier LLMs can develop a preference for dishonesty as a new strategy, even when other options are available. Affected models respond to harmful requests with outputs that sound harmful but are subtly incorrect or otherwise harmless in practice. This behavior emerges with hard-to-predict variations even within models from the same model family. We find no apparent cause for the propensity to deceive, but we show that more capable models are better at executing this strategy. Strategic dishonesty already has a practical impact on safety evaluations, as we show that dishonest responses fool all output-based monitors used to detect jailbreaks that we test, rendering benchmark scores unreliable. Further, strategic dishonesty can act like a honeypot against malicious users, which noticeably obfuscates prior jailbreak attacks. While output monitors fail, we show that linear probes on internal activations can be used to reliably detect strategic dishonesty. We validate probes on datasets with verifiable outcomes and by using their features as steering vectors. Overall, we consider strategic dishonesty as a concrete example of a broader concern that alignment of LLMs is hard to control, especially when helpfulness and harmlessness conflict.

As LLM agents are deployed in long-horizon sessions, context accumulation drives up inference costs. Existing approaches utilize text pruning or dynamic memory eviction to minimize token footprints; however, their unconstrained sequence mutations alter layouts, introducing prefix mismatches and cache invalidation. This reveals a critical trade-off between text sparsity and prompt cache continuity. To address this, we present TokenPilot, a dual-granularity context management framework. Globally, Ingestion-Aware Compaction acts as a framework harness to stabilize prompt prefixes and eliminate open-world environmental noise at the ingestion gate. Locally, Lifecycle-Aware Eviction monitors the ongoing residual utility of context segments, enforcing a conservative batch-turn schedule to offload content segments only when task relevance expires. Experiments on PinchBench and Claw-Eval under both isolated and continuous modes demonstrate that TokenPilot reduces costs by 61% and 56% in isolated mode, and 61% and 87% in continuous mode, while maintaining competitive performance compared to prior systems. TokenPilot has been integrated into LightMem2 at https://github.com/zjunlp/LightMem2.

Tool-using LLM agents face a reliability-cost tradeoff: routing every decision through the LLM improves correctness but incurs high latency and inference cost, while pre-coded workflow graphs reduce cost but become brittle under unanticipated compound tool failures. We present Self-Healing Router, a fault-tolerant orchestration architecture that treats most agent control-flow decisions as routing rather than reasoning. The system combines (i) parallel health monitors that assign priority scores to runtime conditions such as tool outages and risk signals, and (ii) a cost-weighted tool graph where Dijkstra's algorithm performs deterministic shortest-path routing. When a tool fails mid-execution, its edges are reweighted to infinity and the path is recomputed -- yielding automatic recovery without invoking the LLM. The LLM is reserved exclusively for cases where no feasible path exists, enabling goal demotion or escalation. Prior graph-based tool-use systems (ControlLLM, ToolNet, NaviAgent) focus on tool selection and planning; our contribution is runtime fault tolerance with deterministic recovery and binary observability -- every failure is either a logged reroute or an explicit escalation, never a silent skip. Across 19 scenarios spanning three graph topologies (linear pipeline, dependency DAG, parallel fan-out), Self-Healing Router matches ReAct's correctness while reducing control-plane LLM calls by 93% (9 vs 123 aggregate) and eliminating the silent-failure cases observed in a well-engineered static workflow baseline under compound failures.

The agent--tool interaction loop is a critical attack surface for modern Large Language Model (LLM) agents. Existing denial-of-service (DoS) attacks typically function at the user-prompt or retrieval-augmented generation (RAG) context layer and are inherently single-turn in nature. This limitation restricts cost amplification and diminishes stealth in goal-oriented workflows. To address these issues, we proposed a stealthy, multi-turn economic DoS attack at the tool layer under the Model Context Protocol (MCP). By simply editing text-visible fields and implementing a template-driven return policy, our malicious server preserves function signatures and the terminal benign payload while steering agents into prolonged, verbose tool-calling chains. We optimize these text-only edits with Monte Carlo Tree Search (MCTS) to maximize cost under a task-success constraint. Across six LLMs on ToolBench and BFCL benchmarks, our attack yields trajectories over 60K tokens, increases per-query cost by up to 658 times, raises energy by 100 to 560 times, and pushes GPU key-value (KV) cache occupancy to 35--74%. Standard prompt filters and output trajectory monitors seldom detect these attacks, highlighting the need for defenses that safeguard agentic processes rather than focusing solely on final outcomes. We will release the code soon.

This paper introduces CoMuRoS (Collaborative Multi-Robot System), a generalizable hierarchical architecture for heterogeneous robot teams that unifies centralized deliberation with decentralized execution, and supports event-driven replanning. A Task Manager LLM interprets natural-language goals, classifies tasks, and allocates subtasks using static rules plus dynamic contexts (task, history, robot and task status, and events).Each robot runs a local LLM that composes executable Python code from primitive skills (ROS2 nodes, policies), while onboard perception (VLMs/image processing) continuously monitors events and classifies them into relevant or irrelevant to the task. Task failures or user intent changes trigger replanning, allowing robots to assist teammates, resume tasks, or request human help. Hardware studies demonstrate autonomous recovery from disruptive events, filtering of irrelevant distractions, and tightly coordinated transport with emergent human-robot cooperation (e.g., multirobot collaborative object recovery success rate: 9/10, coordinated transport: 8/8, human-assisted recovery: 5/5).Simulation studies show intention-aware replanning. A curated textual benchmark spanning 22 scenarios (3 tasks each, around 20 robots) evaluates task allocation, classification, IoU, executability, and correctness, with high average scores (e.g., correctness up to 0.91) across multiple LLMs, a separate replanning set (5 scenarios) achieves 1.0 correctness. Compared with prior LLM-based systems, CoMuRoS uniquely demonstrates runtime, event-driven replanning on physical robots, delivering robust, flexible multi-robot and human-robot collaboration.

Fine-tuning large language models (LLMs) improves performance on domain-specific tasks but can lead to overfitting, making them unreliable on out-of-distribution (OoD) queries. We propose LoRA-BAM - a method that adds OoD detection monitors to the LoRA layer using boxed abstraction to filter questions beyond the model's competence. Feature vectors from the fine-tuning data are extracted via the LLM and clustered. Clusters are enclosed in boxes; a question is flagged as OoD if its feature vector falls outside all boxes. To improve interpretability and robustness, we introduce a regularization loss during fine-tuning that encourages paraphrased questions to stay close in the feature space, and the enlargement of the decision boundary is based on the feature variance within a cluster. Our method complements existing defenses by providing lightweight and interpretable OoD detection.

LLM-based software engineering agents are increasingly used in real-world development tasks, often with access to sensitive data or security-critical codebases. Such agents could intentionally sabotage these codebases if they were misaligned. We investigate asynchronous monitoring, in which a monitoring system reviews agent actions after the fact. Unlike synchronous monitoring, this approach does not impose runtime latency, while still attempting to disrupt attacks before irreversible harm occurs. We treat monitor development as an adversarial game between a blue team (who design monitors) and a red team (who create sabotaging agents). We attempt to set the game rules such that they upper bound the sabotage potential of an agent based on Claude 4.1 Opus. To ground this game in a realistic, high-stakes deployment scenario, we develop a suite of 5 diverse software engineering environments that simulate tasks that an agent might perform within an AI developer's internal infrastructure. Over the course of the game, we develop an ensemble monitor that achieves a 6% false negative rate at 1% false positive rate on a held out test environment. Then, we estimate risk of sabotage at deployment time by extrapolating from our monitor's false negative rate. We describe one simple model for this extrapolation, present a sensitivity analysis, and describe situations in which the model would be invalid. Code is available at: https://github.com/UKGovernmentBEIS/async-control.

Large Language Models (LLMs) trained with reinforcement learning and verifiable rewards have achieved strong results on complex reasoning tasks. Recent work extends this paradigm to a multi-agent setting, where a meta-thinking agent proposes plans and monitors progress while a reasoning agent executes subtasks through sequential conversational turns. Despite promising performance, we identify a critical limitation: lazy agent behavior, in which one agent dominates while the other contributes little, undermining collaboration and collapsing the setup to an ineffective single agent. In this paper, we first provide a theoretical analysis showing why lazy behavior naturally arises in multi-agent reasoning. We then introduce a stable and efficient method for measuring causal influence, helping mitigate this issue. Finally, as collaboration intensifies, the reasoning agent risks getting lost in multi-turn interactions and trapped by previous noisy responses. To counter this, we propose a verifiable reward mechanism that encourages deliberation by allowing the reasoning agent to discard noisy outputs, consolidate instructions, and restart its reasoning process when necessary. Extensive experiments demonstrate that our framework alleviates lazy agent behavior and unlocks the full potential of multi-agent framework for complex reasoning tasks.

Large Language Model (LLM)-based multi-agent systems (MAS) have emerged as a promising paradigm for solving complex tasks. However, existing works often rely on manual designs or "one-size-fits-all" automation, lacking dynamic adaptability after deployment. Inspired by how biological systems adapt, we introduce MASFly, a novel multi-agent framework enabling dynamic adaptation at test time. To adapt system generation, MASFly employs a retrieval-augmented SOP instantiation mechanism that leverages a self-constructed repository of successful collaboration patterns, enabling the LLM to assemble customized MASs for new queries. For adaptive execution, MASFly incorporates an experience-guided supervision mechanism, where a dedicated Watcher agent monitors system behaviors with reference to a personalized experience pool and provides real-time interventions. Extensive experiments demonstrate that MASFly achieves state-of-the-art performance, most notably a 61.7% success rate on the TravelPlanner benchmark, while exhibiting strong task adaptability and robustness.

Human-agent dialogues often exhibit topic continuity-a stable thematic frame that evolves through temporally adjacent exchanges-yet most large language model (LLM) agent memory systems fail to preserve it. Existing designs follow a fragmentation-compensation paradigm: they first break dialogue streams into isolated utterances for storage, then attempt to restore coherence via embedding-based retrieval. This process irreversibly damages narrative and causal flow, while biasing retrieval towards lexical similarity. We introduce membox, a hierarchical memory architecture centered on a Topic Loom that continuously monitors dialogue in a sliding-window fashion, grouping consecutive same-topic turns into coherent "memory boxes" at storage time. Sealed boxes are then linked by a Trace Weaver into long-range event-timeline traces, recovering macro-topic recurrences across discontinuities. Experiments on LoCoMo demonstrate that Membox achieves up to 68% F1 improvement on temporal reasoning tasks, outperforming competitive baselines (e.g., Mem0, A-MEM). Notably, Membox attains these gains while using only a fraction of the context tokens required by existing methods, highlighting a superior balance between efficiency and effectiveness. By explicitly modeling topic continuity, Membox offers a cognitively motivated mechanism for enhancing both coherence and efficiency in LLM agents.

In this work, we study the risks of collective financial fraud in large-scale multi-agent systems powered by large language model (LLM) agents. We investigate whether agents can collaborate in fraudulent behaviors, how such collaboration amplifies risks, and what factors influence fraud success. To support this research, we present MultiAgentFraudBench, a large-scale benchmark for simulating financial fraud scenarios based on realistic online interactions. The benchmark covers 28 typical online fraud scenarios, spanning the full fraud lifecycle across both public and private domains. We further analyze key factors affecting fraud success, including interaction depth, activity level, and fine-grained collaboration failure modes. Finally, we propose a series of mitigation strategies, including adding content-level warnings to fraudulent posts and dialogues, using LLMs as monitors to block potentially malicious agents, and fostering group resilience through information sharing at the societal level. Notably, we observe that malicious agents can adapt to environmental interventions. Our findings highlight the real-world risks of multi-agent financial fraud and suggest practical measures for mitigating them. Code is available at https://github.com/zheng977/MutiAgent4Fraud.

Modern software infrastructure increasingly relies on LLM agents for development and maintenance, such as Claude Code and Gemini-cli. However, these AI agents differ fundamentally from traditional deterministic software, posing a significant challenge to conventional monitoring and debugging. This creates a critical semantic gap: existing tools observe either an agent's high-level intent (via LLM prompts) or its low-level actions (e.g., system calls), but cannot correlate these two views. This blindness makes it difficult to distinguish between benign operations, malicious attacks, and costly failures. We introduce AgentSight, an AgentOps observability framework that bridges this semantic gap using a hybrid approach. Our approach, boundary tracing, monitors agents from outside their application code at stable system interfaces using eBPF. AgentSight intercepts TLS-encrypted LLM traffic to extract semantic intent, monitors kernel events to observe system-wide effects, and causally correlates these two streams across process boundaries using a real-time engine and secondary LLM analysis. This instrumentation-free technique is framework-agnostic, resilient to rapid API changes, and incurs less than 3% performance overhead. Our evaluation shows AgentSight detects prompt injection attacks, identifies resource-wasting reasoning loops, and reveals hidden coordination bottlenecks in multi-agent systems. AgentSight is released as an open-source project at https://github.com/agent-sight/agentsight.

Advances in AI-assisted research have introduced powerful tools for literature retrieval, hypothesis generation, experimentation, and manuscript preparation. However, systems remain fragmented and lack human-centred workflows. To address these gaps, we introduce AIssistant, an agentic, open-source Human-AI collaborative framework designed to simplify the end-to-end creation of scientific workflows. Since our development is still in an early stage, we present here the first experiments with AIssistant for perspective and review research papers in machine learning. Our system integrates modular tools and agents for literature synthesis, section-wise experimentation, citation management, and automatic LaTeX paper text generation, while maintaining human oversight at every stage to ensure accuracy, coherence, and scholarly rigour. We conducted a comprehensive evaluation across three layers: (1) Independent Human Review, following NeurIPS double-blind standards; (2) Automated LLM Review, using GPT-5 as a scalable human review proxy; and (3) Program Chair Oversight, where the chair monitors the entire review process and makes final validation and acceptance decisions. The results demonstrate that AIssistant improves drafting efficiency and thematic consistency. Nonetheless, Human-AI collaboration remains essential for maintaining factual correctness, methodological soundness, and ethical compliance. Despite its effectiveness, we identify key limitations, including hallucinated citations, difficulty adapting to dynamic paper structures, and incomplete integration of multimodal content.

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 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.

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.

This study concentrates on evaluating the efficacy of Large Language Models (LLMs) in healthcare, with a specific focus on their application in personal anomalous health monitoring. Our research primarily investigates the capabilities of LLMs in interpreting and analyzing physiological data obtained from FDA-approved devices. We conducted an extensive analysis using anomalous physiological data gathered in a simulated low-air-pressure plateau environment. This allowed us to assess the precision and reliability of LLMs in understanding and evaluating users' health status with notable specificity. Our findings reveal that LLMs exhibit exceptional performance in determining medical indicators, including a Mean Absolute Error (MAE) of less than 1 beat per minute for heart rate and less than 1% for oxygen saturation (SpO2). Furthermore, the Mean Absolute Percentage Error (MAPE) for these evaluations remained below 1%, with the overall accuracy of health assessments surpassing 85%. In image analysis tasks, such as interpreting photoplethysmography (PPG) data, our specially adapted GPT models demonstrated remarkable proficiency, achieving less than 1 bpm error in cycle count and 7.28 MAE for heart rate estimation. This study highlights LLMs' dual role as health data analysis tools and pivotal elements in advanced AI health assistants, offering personalized health insights and recommendations within the future health assistant framework.

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.

Large language models (LLMs) can generate chains of thought (CoTs) that are not always causally responsible for their final outputs. When such a mismatch occurs, the CoT no longer faithfully reflects the decision-critical factors driving the model's behavior, leading to the reduced CoT monitorability problem. However, a comprehensive and fully open-source benchmark for studying CoT monitorability remains lacking. To address this gap, we propose MonitorBench, a systematic benchmark for evaluating CoT monitorability in LLMs. MonitorBench provides: (1) a diverse set of 1,514 test instances with carefully designed decision-critical factors across 19 tasks spanning 7 categories to characterize when CoTs can be used to monitor the factors driving LLM behavior; and (2) two stress-test settings to quantify the extent to which CoT monitorability can be degraded. Extensive experiments across multiple popular LLMs with varying capabilities show that CoT monitorability is higher when producing the final target response requires structural reasoning through the decision-critical factor. Closed-source LLMs generally show lower monitorability, and there exists a negative relationship between monitorability and model capability. Moreover, both open- and closed-source LLMs can intentionally reduce monitorability under stress-tests, with monitorability dropping by up to 30% in some tasks that do not require structural reasoning over the decision-critical factors. Beyond these empirical insights, MonitorBench provides a basis for further research on evaluating future LLMs, studying advanced stress-test monitorability techniques, and developing new monitoring approaches.

Evaluating the output of Large Language Models (LLMs) is one of the most critical aspects of building a performant compound AI system. Since the output from LLMs propagate to downstream steps, identifying LLM errors is crucial to system performance. A common task for LLMs in AI systems is tool use. While there are several benchmark environments for evaluating LLMs on this task, they typically only give a success rate without any explanation of the failure cases. To solve this problem, we introduce SpecTool, a new benchmark to identify error patterns in LLM output on tool-use tasks. Our benchmark data set comprises of queries from diverse environments that can be used to test for the presence of seven newly characterized error patterns. Using SPECTOOL , we show that even the most prominent LLMs exhibit these error patterns in their outputs. Researchers can use the analysis and insights from SPECTOOL to guide their error mitigation strategies.

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.

The field of efficient Large Language Model (LLM) inference is rapidly evolving, presenting a unique blend of opportunities and challenges. Although the field has expanded and is vibrant, there hasn't been a concise framework that analyzes the various methods of LLM Inference to provide a clear understanding of this domain. Our survey stands out from traditional literature reviews by not only summarizing the current state of research but also by introducing a framework based on roofline model for systematic analysis of LLM inference techniques. This framework identifies the bottlenecks when deploying LLMs on hardware devices and provides a clear understanding of practical problems, such as why LLMs are memory-bound, how much memory and computation they need, and how to choose the right hardware. We systematically collate the latest advancements in efficient LLM inference, covering crucial areas such as model compression (e.g., Knowledge Distillation and Quantization), algorithm improvements (e.g., Early Exit and Mixture-of-Expert), and both hardware and system-level enhancements. Our survey stands out by analyzing these methods with roofline model, helping us understand their impact on memory access and computation. This distinctive approach not only showcases the current research landscape but also delivers valuable insights for practical implementation, positioning our work as an indispensable resource for researchers new to the field as well as for those seeking to deepen their understanding of efficient LLM deployment. The analyze tool, LLM-Viewer, is open-sourced.

As Large Language Models (LLMs) gain wider adoption in various contexts, it becomes crucial to ensure they are reasonably safe, consistent, and reliable for an application at hand. This may require probing or auditing them. Probing LLMs with varied iterations of a single question could reveal potential inconsistencies in their knowledge or functionality. However, a tool for performing such audits with simple workflow and low technical threshold is lacking. In this demo, we introduce "AuditLLM," a novel tool designed to evaluate the performance of various LLMs in a methodical way. AuditLLM's core functionality lies in its ability to test a given LLM by auditing it using multiple probes generated from a single question, thereby identifying any inconsistencies in the model's understanding or operation. A reasonably robust, reliable, and consistent LLM should output semantically similar responses for a question asked differently or by different people. Based on this assumption, AuditLLM produces easily interpretable results regarding the LLM's consistencies from a single question that the user enters. A certain level of inconsistency has been shown to be an indicator of potential bias, hallucinations, and other issues. One could then use the output of AuditLLM to further investigate issues with the aforementioned LLM. To facilitate demonstration and practical uses, AuditLLM offers two key modes: (1) Live mode which allows instant auditing of LLMs by analyzing responses to real-time queries; (2) Batch mode which facilitates comprehensive LLM auditing by processing multiple queries at once for in-depth analysis. This tool is beneficial for both researchers and general users, as it enhances our understanding of LLMs' capabilities in generating responses, using a standardized auditing platform.

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.

Large Language Models (LLMs) are increasingly integrated into critical systems in industries like healthcare and finance. Users can often submit queries to LLM-enabled chatbots, some of which can enrich responses with information retrieved from internal databases storing sensitive data. This gives rise to a range of attacks in which a user submits a malicious query and the LLM-system outputs a response that creates harm to the owner, such as leaking internal data or creating legal liability by harming a third-party. While security tools are being developed to counter these threats, there is little formal evaluation of their effectiveness and usability. This study addresses this gap by conducting a thorough comparative analysis of LLM security tools. We identified 13 solutions (9 closed-source, 4 open-source), but only 7 were evaluated due to a lack of participation by proprietary model owners.To evaluate, we built a benchmark dataset of malicious prompts, and evaluate these tools performance against a baseline LLM model (ChatGPT-3.5-Turbo). Our results show that the baseline model has too many false positives to be used for this task. Lakera Guard and ProtectAI LLM Guard emerged as the best overall tools showcasing the tradeoff between usability and performance. The study concluded with recommendations for greater transparency among closed source providers, improved context-aware detections, enhanced open-source engagement, increased user awareness, and the adoption of more representative performance metrics.

IT environments typically have logging mechanisms to monitor system health and detect issues. However, the huge volume of generated logs makes manual inspection impractical, highlighting the importance of automated log analysis in IT Software Support. In this paper, we propose a log analytics tool that leverages Large Language Models (LLMs) for log data processing and issue diagnosis, enabling the generation of automated insights and summaries. We further present a novel approach for efficiently running LLMs on CPUs to process massive log volumes in minimal time without compromising output quality. We share the insights and lessons learned from deployment of the tool - in production since March 2024 - scaled across 70 software products, processing over 2000 tickets for issue diagnosis, achieving a time savings of 300+ man hours and an estimated $15,444 per month in manpower costs compared to the traditional log analysis practices.

We present OnPrem.LLM, a Python-based toolkit for applying large language models (LLMs) to sensitive, non-public data in offline or restricted environments. The system is designed for privacy-preserving use cases and provides prebuilt pipelines for document processing and storage, retrieval-augmented generation (RAG), information extraction, summarization, classification, and prompt/output processing with minimal configuration. OnPrem.LLM supports multiple LLM backends -- including llama.cpp, Ollama, vLLM, and Hugging Face Transformers -- with quantized model support, GPU acceleration, and seamless backend switching. Although designed for fully local execution, OnPrem.LLM also supports integration with a wide range of cloud LLM providers when permitted, enabling hybrid deployments that balance performance with data control. A no-code web interface extends accessibility to non-technical users.

NeMo Guardrails is an open-source toolkit for easily adding programmable guardrails to LLM-based conversational systems. Guardrails (or rails for short) are a specific way of controlling the output of an LLM, such as not talking about topics considered harmful, following a predefined dialogue path, using a particular language style, and more. There are several mechanisms that allow LLM providers and developers to add guardrails that are embedded into a specific model at training, e.g. using model alignment. Differently, using a runtime inspired from dialogue management, NeMo Guardrails allows developers to add programmable rails to LLM applications - these are user-defined, independent of the underlying LLM, and interpretable. Our initial results show that the proposed approach can be used with several LLM providers to develop controllable and safe LLM applications using programmable rails.

Legacy software systems, written in outdated languages like MUMPS and mainframe assembly, pose challenges in efficiency, maintenance, staffing, and security. While LLMs offer promise for modernizing these systems, their ability to understand legacy languages is largely unknown. This paper investigates the utilization of LLMs to generate documentation for legacy code using two datasets: an electronic health records (EHR) system in MUMPS and open-source applications in IBM mainframe Assembly Language Code (ALC). We propose a prompting strategy for generating line-wise code comments and a rubric to evaluate their completeness, readability, usefulness, and hallucination. Our study assesses the correlation between human evaluations and automated metrics, such as code complexity and reference-based metrics. We find that LLM-generated comments for MUMPS and ALC are generally hallucination-free, complete, readable, and useful compared to ground-truth comments, though ALC poses challenges. However, no automated metrics strongly correlate with comment quality to predict or measure LLM performance. Our findings highlight the limitations of current automated measures and the need for better evaluation metrics for LLM-generated documentation in legacy systems.

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 development of large language models (LLMs), like ChatGPT, has resulted in the widespread presence of LLM-generated content on social media platforms, raising concerns about misinformation, data biases, and privacy violations, which can undermine trust in online discourse. While detecting LLM-generated content is crucial for mitigating these risks, current methods often focus on binary classification, failing to address the complexities of real-world scenarios like human-LLM collaboration. To move beyond binary classification and address these challenges, we propose a new paradigm for detecting LLM-generated content. This approach introduces two novel tasks: LLM Role Recognition (LLM-RR), a multi-class classification task that identifies specific roles of LLM in content generation, and LLM Influence Measurement (LLM-IM), a regression task that quantifies the extent of LLM involvement in content creation. To support these tasks, we propose LLMDetect, a benchmark designed to evaluate detectors' performance on these new tasks. LLMDetect includes the Hybrid News Detection Corpus (HNDC) for training detectors, as well as DetectEval, a comprehensive evaluation suite that considers five distinct cross-context variations and two multi-intensity variations within the same LLM role. This allows for a thorough assessment of detectors' generalization and robustness across diverse contexts. Our empirical validation of 10 baseline detection methods demonstrates that fine-tuned PLM-based models consistently outperform others on both tasks, while advanced LLMs face challenges in accurately detecting their own generated content. Our experimental results and analysis offer insights for developing more effective detection models for LLM-generated content. This research enhances the understanding of LLM-generated content and establishes a foundation for more nuanced detection methodologies.

The wide adoption of LLMs has led to their use in great variety of applications and scenarios, such as chatbot assistants and data annotation, creating the need for the models to satisfy certain budget and hardware constraints. This has led to the trend of LLMs being released in batches consisting of similar models of various sizes for the family of models to adhere to as wide of a range of constraints as possible. In this paper, we validate distillation and quantization as a cost-effective way to expand model families to new sizes and hardware formats. Based on the open-recipe Apertus 8B LLM, we produce Apertus-v1.1 - a distilled family of models with up to 4B parameters trained on 1.7T permissive license tokens. We demonstrate cost-efficiency and strong accuracy performance of our approach for covering large ranges of hardware and systems requirements.

Realistic evaluation of LLM serving systems requires online workloads, dynamic arrivals, queueing, and the serving engine's local scheduling for execution batching, but running such experiments on GPUs is expensive. Existing simulators reduce this cost, but often operate offline or in time-warped mode, re-implement serving-engine schedulers, or require accurate operator/kernel-level latency models. We present LLM-Emu, a serving-native emulator for vLLM that preserves the production HTTP, scheduling, KV-cache, and output-processing paths while replacing only GPU forward execution with profile-sampled latency and synthetic output tokens. Tested on two different GPUs, four model variants, two model families, two attention backends, and both Poisson and bursty ShareGPT workloads, LLM-Emu closely tracks real vLLM serving behavior: TPOT and ITL stay within 4.8% absolute error, E2E latency within 5.3%, and output throughput within 1.9%; TTFT is less stable, with maximum error 10.4%, reflecting its sensitivity to admission and queue state. These results suggest that lightweight, serving-native emulation can support practical online experimentation for LLM-serving systems. LLM-Emu is open sourced at https://github.com/AKafakA/llm-emu.

We study how to watermark LLM outputs, i.e. embedding algorithmically detectable signals into LLM-generated text to track misuse. Unlike the current mainstream methods that work with a fixed LLM, we expand the watermark design space by including the LLM tuning stage in the watermark pipeline. While prior works focus on token-level watermark that embeds signals into the output, we design a model-level watermark that embeds signals into the LLM weights, and such signals can be detected by a paired detector. We propose a co-training framework based on reinforcement learning that iteratively (1) trains a detector to detect the generated watermarked text and (2) tunes the LLM to generate text easily detectable by the detector while keeping its normal utility. We empirically show that our watermarks are more accurate, robust, and adaptable (to new attacks). It also allows watermarked model open-sourcing. In addition, if used together with alignment, the extra overhead introduced is low - only training an extra reward model (i.e. our detector). We hope our work can bring more effort into studying a broader watermark design that is not limited to working with a fixed LLM. We open-source the code: https://github.com/xiaojunxu/learning-to-watermark-llm .

LLM agents handle user requests on behalf of organizations through tool calls and must follow the company policies stated in their system prompts. Prior work approaches this as a safeguarding problem -- external checks that block non-compliant agent actions. We argue that policy adherence is a broader problem: real workflows unfold across many turns, require explicit user confirmation and prerequisite reads, and hinge on the content of the dialogue rather than on any single argument value. Meeting this bar requires (i) full conversation context, (ii) self-reasoning over the policy and the current dialogue, and (iii) conversation-specific remediation that guides the agent's next turn -- three capabilities that prior safeguard work has often underestimated. We introduce POLICYGUARD, a sub-agent verifier that shares the agent's view of the dialogue, reasons over the policy in context, and provides actionable feedback for the agent's next turn. On tau^2-BENCH airline across three vendors (GPT-5.4, Claude Sonnet 4.6, Gemini 2.5 Pro) with four trials per setting, POLICYGUARD improves PASS4 by +12.0 / +6.0 / +12.0 pp. Per-call analyses show POLICYGUARD achieves higher policy-violation recall while blocking roughly half as often as argument-level guards.

Large Language Models (LLMs) are advancing into Multimodal LLMs (MLLMs), capable of processing image, audio, and video as well as text. Combining first-person video, MLLMs show promising potential for understanding human activities through video and audio, enabling many human-computer interaction and human-augmentation applications such as human activity support, real-world agents, and skill transfer to robots or other individuals. However, handling high-resolution, long-duration videos generates large latent representations, leading to substantial memory and processing demands, limiting the length and resolution MLLMs can manage. Reducing video resolution can lower memory usage but often compromises comprehension. This paper introduces a method that optimizes first-person video analysis by integrating eye-tracking data, and proposes a method that decomposes first-person vision video into sub areas for regions of gaze focus. By processing these selectively gazed-focused inputs, our approach achieves task comprehension equivalent to or even better than processing the entire image at full resolution, but with significantly reduced video data input (reduce the number of pixels to one-tenth), offering an efficient solution for using MLLMs to interpret and utilize human skills.

Video anomaly detection (VAD) is essential for enhancing safety and security by identifying unusual events across different environments. Existing VAD benchmarks, however, are primarily designed for general-purpose scenarios, neglecting the specific characteristics of smart home applications. To bridge this gap, we introduce SmartHome-Bench, the first comprehensive benchmark specially designed for evaluating VAD in smart home scenarios, focusing on the capabilities of multi-modal large language models (MLLMs). Our newly proposed benchmark consists of 1,203 videos recorded by smart home cameras, organized according to a novel anomaly taxonomy that includes seven categories, such as Wildlife, Senior Care, and Baby Monitoring. Each video is meticulously annotated with anomaly tags, detailed descriptions, and reasoning. We further investigate adaptation methods for MLLMs in VAD, assessing state-of-the-art closed-source and open-source models with various prompting techniques. Results reveal significant limitations in the current models' ability to detect video anomalies accurately. To address these limitations, we introduce the Taxonomy-Driven Reflective LLM Chain (TRLC), a new LLM chaining framework that achieves a notable 11.62% improvement in detection accuracy. The benchmark dataset and code are publicly available at https://github.com/Xinyi-0724/SmartHome-Bench-LLM.

We address the problem of runtime trajectory anomaly detection, a critical capability for enabling trustworthy LLM agents. Current safety measures predominantly focus on static input/output filtering. However, we argue that ensuring LLM agents reliability requires auditing the intermediate execution process. In this work, we formulate the task of Trajectory Anomaly Detection. The goal is not merely detection, but precise error localization. This capability is essential for enabling efficient rollback-and-retry. To achieve this, we construct TrajBench, a dataset synthesized via a perturb-and-complete strategy to cover diverse procedural anomalies. Using this benchmark, we investigate the capability of models in process supervision. We observe that general-purpose LLMs, even with zero-shot prompting, struggle to identify and localize these anomalies. This reveals that generalized capabilities do not automatically translate to process reliability. To address this, we propose TrajAD, a specialized verifier trained with fine-grained process supervision. Our approach outperforms baselines, demonstrating that specialized supervision is essential for building trustworthy agents.

The complexity and heterogeneity of data in many real-world applications pose significant challenges for traditional machine learning and signal processing techniques. For instance, in medicine, effective analysis of diverse physiological signals is crucial for patient monitoring and clinical decision-making and yet highly challenging. We introduce MedTsLLM, a general multimodal large language model (LLM) framework that effectively integrates time series data and rich contextual information in the form of text to analyze physiological signals, performing three tasks with clinical relevance: semantic segmentation, boundary detection, and anomaly detection in time series. These critical tasks enable deeper analysis of physiological signals and can provide actionable insights for clinicians. We utilize a reprogramming layer to align embeddings of time series patches with a pretrained LLM's embedding space and make effective use of raw time series, in conjunction with textual context. Given the multivariate nature of medical datasets, we develop methods to handle multiple covariates. We additionally tailor the text prompt to include patient-specific information. Our model outperforms state-of-the-art baselines, including deep learning models, other LLMs, and clinical methods across multiple medical domains, specifically electrocardiograms and respiratory waveforms. MedTsLLM presents a promising step towards harnessing the power of LLMs for medical time series analysis that can elevate data-driven tools for clinicians and improve patient outcomes.

Large language models (LLMs) are increasingly deployed in specialized production data processing pipelines across diverse domains -- such as finance, marketing, and e-commerce. However, when running them in production across many inputs, they often fail to follow instructions or meet developer expectations. To improve reliability in these applications, creating assertions or guardrails for LLM outputs to run alongside the pipelines is essential. Yet, determining the right set of assertions that capture developer requirements for a task is challenging. In this paper, we introduce PROMPTEVALS, a dataset of 2087 LLM pipeline prompts with 12623 corresponding assertion criteria, sourced from developers using our open-source LLM pipeline tools. This dataset is 5x larger than previous collections. Using a hold-out test split of PROMPTEVALS as a benchmark, we evaluated closed- and open-source models in generating relevant assertions. Notably, our fine-tuned Mistral and Llama 3 models outperform GPT-4o by 20.93% on average, offering both reduced latency and improved performance. We believe our dataset can spur further research in LLM reliability, alignment, and prompt engineering.

Chain-of-thought (CoT) monitoring is a promising safety mechanism for AI agents, based on the premise that visible reasoning traces can surface misaligned or deceptive behavior. While effective in standard scenarios, recent work highlights that LLMs remain vulnerable to persuasion-based jailbreaks, where natural-language arguments override model constraints. We stress-test whether this vulnerability extends to monitoring LLMs: can an adversarial agent persuade its CoT monitor to approve proposed actions that violate the monitor's policy? We design an evaluation framework with 40 tasks and analyze thousands of agent-monitor interactions, where agents are instructed to argue for policy-violating proposals. We find that in such adversarial settings, monitor access to the agent's CoT reasoning increases rather than decreases approval of harmful actions on average by 9.5%, as the scratchpad provides an additional persuasion channel. To address this, we introduce a fact-checking monitoring framework. We find that a fact-checker and monitor pairing from different model families, for example a Claude 3.7 Sonnet monitor paired with a GPT-4.1 fact-checker, reduces approval of policy-violating actions by up to 45%, compared to only 6%, when using the same model for both fact-checking and monitoring roles. Our results demonstrate that CoT monitoring alone may be insufficient against adversarial persuasion, and that model-diverse fact-checking provides a robust mitigation.

Large language models (LLMs) are effective at capturing complex, valuable conceptual representations from textual data for a wide range of real-world applications. However, in fields like Intelligent Fault Diagnosis (IFD), incorporating additional sensor data-such as vibration signals, temperature readings, and operational metrics-is essential but it is challenging to capture such sensor data information within traditional text corpora. This study introduces a novel IFD approach by effectively adapting LLMs to numerical data inputs for identifying various machine faults from time-series sensor data. We propose FD-LLM, an LLM framework specifically designed for fault diagnosis by formulating the training of the LLM as a multi-class classification problem. We explore two methods for encoding vibration signals: the first method uses a string-based tokenization technique to encode vibration signals into text representations, while the second extracts statistical features from both the time and frequency domains as statistical summaries of each signal. We assess the fault diagnosis capabilities of four open-sourced LLMs based on the FD-LLM framework, and evaluate the models' adaptability and generalizability under various operational conditions and machine components, namely for traditional fault diagnosis, cross-operational conditions, and cross-machine component settings. Our results show that LLMs such as Llama3 and Llama3-instruct demonstrate strong fault detection capabilities and significant adaptability across different operational conditions, outperforming state-of-the-art deep learning (DL) approaches in many cases.

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.

Current operating systems expose interfaces optimized for human users but not for AI agents. Humans benefit from pixels, icons, windows, visual grouping, mouse movement, and keyboard shortcuts; AI agents instead need compact semantic state, grounded actions, and reliable feedback. As a result, many computer-use agents are forced to interpret screenshots, OCR output, and visual crops, introducing high token costs, visual ambiguity, latency, and coordinate uncertainty. This paper introduces LUMOS (Language Model Unified Machine-Readable Operating-System Semantics), a semantic interaction layer between AI agents and operating systems. LUMOS converts native accessibility metadata and browser UI structures into machine readable semantic blueprints with stable identifiers, roles, names, values, bounds, and action affordances. It also supports live semantic pointer grounding by querying the UI element under or near the cursor through operating-system automation APIs. An LLM then acts through an accessibility grounded observe act loop using constrained visible-UI primitives rather than application-specific scripts. LUMOS does not claim to replace visual agents; instead, it reduces dependence on screenshots when operating systems already provide semantic structure. These results suggest a path toward AI-native operating systems and machine-readable interaction layers.

We present ExCyTIn-Bench, the first benchmark to Evaluate an LLM agent x on the task of Cyber Threat Investigation through security questions derived from investigation graphs. Real-world security analysts must sift through a large number of heterogeneous alert signals and security logs, follow multi-hop chains of evidence, and compile an incident report. With the developments of LLMs, building LLM-based agents for automatic thread investigation is a promising direction. To assist the development and evaluation of LLM agents, we construct a dataset from a controlled Azure tenant that covers 8 simulated real-world multi-step attacks, 57 log tables from Microsoft Sentinel and related services, and 589 automatically generated questions. We leverage security logs extracted with expert-crafted detection logic to build threat investigation graphs, and then generate questions with LLMs using paired nodes on the graph, taking the start node as background context and the end node as answer. Anchoring each question to these explicit nodes and edges not only provides automatic, explainable ground truth answers but also makes the pipeline reusable and readily extensible to new logs. This also enables the automatic generation of procedural tasks with verifiable rewards, which can be naturally extended to training agents via reinforcement learning. Our comprehensive experiments with different models confirm the difficulty of the task: with the base setting, the average reward across all evaluated models is 0.249, and the best achieved is 0.368, leaving substantial headroom for future research. Code and data are coming soon!

Large Language Models(LLMs)have become effective tools for natural language processing and have been used in many different fields. This essay offers a succinct summary of various LLM subcategories. The survey emphasizes recent developments and efforts made for various LLM kinds, including task-based financial LLMs, multilingual language LLMs, biomedical and clinical LLMs, vision language LLMs, and code language models. The survey gives a general summary of the methods, attributes, datasets, transformer models, and comparison metrics applied in each category of LLMs. Furthermore, it highlights unresolved problems in the field of developing chatbots and virtual assistants, such as boosting natural language processing, enhancing chatbot intelligence, and resolving moral and legal dilemmas. The purpose of this study is to provide readers, developers, academics, and users interested in LLM-based chatbots and virtual intelligent assistant technologies with useful information and future directions.

Multi-modal large language models (MLLMs) have advanced general-purpose video understanding but struggle with long, high-resolution videos -- they process every pixel equally in their vision transformers (ViTs) or LLMs despite significant spatiotemporal redundancy. We introduce AutoGaze, a lightweight module that removes redundant patches before processed by a ViT or an MLLM. Trained with next-token prediction and reinforcement learning, AutoGaze autoregressively selects a minimal set of multi-scale patches that can reconstruct the video within a user-specified error threshold, eliminating redundancy while preserving information. Empirically, AutoGaze reduces visual tokens by 4x-100x and accelerates ViTs and MLLMs by up to 19x, enabling scaling MLLMs to 1K-frame 4K-resolution videos and achieving superior results on video benchmarks (e.g., 67.0% on VideoMME). Furthermore, we introduce HLVid: the first high-resolution, long-form video QA benchmark with 5-minute 4K-resolution videos, where an MLLM scaled with AutoGaze improves over the baseline by 10.1% and outperforms the previous best MLLM by 4.5%. Project page: https://autogaze.github.io/.

This study introduces a novel approach for traffic control systems by using Large Language Models (LLMs) as traffic controllers. The study utilizes their logical reasoning, scene understanding, and decision-making capabilities to optimize throughput and provide feedback based on traffic conditions in real-time. LLMs centralize traditionally disconnected traffic control processes and can integrate traffic data from diverse sources to provide context-aware decisions. LLMs can also deliver tailored outputs using various means such as wireless signals and visuals to drivers, infrastructures, and autonomous vehicles. To evaluate LLMs ability as traffic controllers, this study proposed a four-stage methodology. The methodology includes data creation and environment initialization, prompt engineering, conflict identification, and fine-tuning. We simulated multi-lane four-leg intersection scenarios and generates detailed datasets to enable conflict detection using LLMs and Python simulation as a ground truth. We used chain-of-thought prompts to lead LLMs in understanding the context, detecting conflicts, resolving them using traffic rules, and delivering context-sensitive traffic management solutions. We evaluated the prformance GPT-mini, Gemini, and Llama as traffic controllers. Results showed that the fine-tuned GPT-mini achieved 83% accuracy and an F1-score of 0.84. GPT-mini model exhibited a promising performance in generating actionable traffic management insights, with high ROUGE-L scores across conflict identification of 0.95, decision-making of 0.91, priority assignment of 0.94, and waiting time optimization of 0.92. We demonstrated that LLMs can offer precise recommendations to drivers in real-time including yielding, slowing, or stopping based on vehicle dynamics.

The complexity of laboratory environments requires solutions that simplify instrument interaction and enhance measurement automation. Traditional tools often require configuration, software, and programming skills, creating barriers to productivity. Previous approaches, including dedicated software suites and custom scripts, frequently fall short in providing user-friendly solutions that align with programming practices. We present LABIIUM, an AI-enhanced, zero-configuration measurement automation system designed to streamline experimental workflows and improve user productivity. LABIIUM integrates an AI assistant powered by Large Language Models (LLMs) to generate code. LABIIUM's Lab-Automation-Measurement Bridges (LAMBs) enable seamless instrument connectivity using standard tools such as VSCode and Python, eliminating setup overhead. To demonstrate its capabilities, we conducted experiments involving the measurement of the parametric transfer curve of a simple two-transistor inverting amplifier with a current source load. The AI assistant was evaluated using different prompt scenarios and compared with multiple models, including Claude Sonnet 3.5, Gemini Pro 1.5, and GPT-4o. An expert solution implementing the Gradient-Weighted Adaptive Stochastic Sampling (GWASS) method was used as a baseline. The solutions generated by the AI assistant were compared with the expert solution and a uniform linear sweep baseline with 10,000 points. The graph results show that the LLMs were able to successfully complete the most basic uniform sweep, but LLMs were unable to develop adaptive sweeping algorithms to compete with GWASS. The evaluation underscores LABIIUM's ability to enhance laboratory productivity and support digital transformation in research and industry, and emphasizes the future work required to improve LLM performance in Electronic Measurement Science Tasks.

Detecting text generated by modern large language models is thought to be hard, as both LLMs and humans can exhibit a wide range of complex behaviors. However, we find that a score based on contrasting two closely related language models is highly accurate at separating human-generated and machine-generated text. Based on this mechanism, we propose a novel LLM detector that only requires simple calculations using a pair of pre-trained LLMs. The method, called Binoculars, achieves state-of-the-art accuracy without any training data. It is capable of spotting machine text from a range of modern LLMs without any model-specific modifications. We comprehensively evaluate Binoculars on a number of text sources and in varied situations. Over a wide range of document types, Binoculars detects over 90% of generated samples from ChatGPT (and other LLMs) at a false positive rate of 0.01%, despite not being trained on any ChatGPT data.

The rise of large language models (LLMs) has enabled LLM-based applications (a.k.a. AI agents or co-pilots), a new software paradigm that combines the strength of LLM and conventional software. Diverse LLM applications from different tenants could design complex workflows using multiple LLM requests to accomplish one task. However, they have to use the over-simplified request-level API provided by today's public LLM services, losing essential application-level information. Public LLM services have to blindly optimize individual LLM requests, leading to sub-optimal end-to-end performance of LLM applications. This paper introduces Parrot, an LLM service system that focuses on the end-to-end experience of LLM-based applications. Parrot proposes Semantic Variable, a unified abstraction to expose application-level knowledge to public LLM services. A Semantic Variable annotates an input/output variable in the prompt of a request, and creates the data pipeline when connecting multiple LLM requests, providing a natural way to program LLM applications. Exposing Semantic Variables to the public LLM service allows it to perform conventional data flow analysis to uncover the correlation across multiple LLM requests. This correlation opens a brand-new optimization space for the end-to-end performance of LLM-based applications. Extensive evaluations demonstrate that Parrot can achieve up to an order-of-magnitude improvement for popular and practical use cases of LLM applications.

Large Language Model (LLM) agents are autonomous systems powered by LLMs, capable of reasoning and planning to solve problems by leveraging a set of tools. However, the integration of multi-tool capabilities in LLM agents introduces challenges in securely managing tools, ensuring their compatibility, handling dependency relationships, and protecting control flows within LLM agent workflows. In this paper, we present the first systematic security analysis of task control flows in multi-tool-enabled LLM agents. We identify a novel threat, Cross-Tool Harvesting and Polluting (XTHP), which includes multiple attack vectors to first hijack the normal control flows of agent tasks, and then collect and pollute confidential or private information within LLM agent systems. To understand the impact of this threat, we developed Chord, a dynamic scanning tool designed to automatically detect real-world agent tools susceptible to XTHP attacks. Our evaluation of 66 real-world tools from the repositories of two major LLM agent development frameworks, LangChain and LlamaIndex, revealed a significant security concern: 75\% are vulnerable to XTHP attacks, highlighting the prevalence of this threat.

In this paper, we present the first comprehensive empirical study of specialized LLM-based detectors and compare them with traditional static analyzers at the project scale. Specifically, our study evaluates five latest and representative LLM-based methods and two traditional tools using: 1) an in-house benchmark of 222 known real-world vulnerabilities (C/C++ and Java) to assess detection capability, and 2) 24 active open-source projects, where we manually inspected 385 warnings to assess their practical usability and underlying root causes of failures. Our evaluation yields three key findings: First, while LLM-based detectors exhibit low recall on the in-house benchmark, they still uncover more unique vulnerabilities than traditional tools. Second, in open-source projects, both LLM-based and traditional tools generate substantial warnings but suffer from very high false discovery rates, hindering practical use. Our manual analysis further reveals shallow interprocedural reasoning and misidentified source/sink pairs as primary failure causes, with LLM-based tools exhibiting additional unique failures. Finally, LLM-based methods incurs substantial computational costs-hundreds of thousands to hundreds of millions of tokens and multi-hour to multi-day runtimes. Overall, our findings underscore critical limitations in the robustness, reliability, and scalability of current LLM-based detectors. We ultimately summarize a set of implications for future research toward more effective and practical project-scale vulnerability detection.

Multimodal Large Language Models (MLLMs) have shown significant potential in medical image analysis. However, their capabilities in interpreting fundus images, a critical skill for ophthalmology, remain under-evaluated. Existing benchmarks lack fine-grained task divisions and fail to provide modular analysis of its two key modules, i.e., large language model (LLM) and vision encoder (VE). This paper introduces FunBench, a novel visual question answering (VQA) benchmark designed to comprehensively evaluate MLLMs' fundus reading skills. FunBench features a hierarchical task organization across four levels (modality perception, anatomy perception, lesion analysis, and disease diagnosis). It also offers three targeted evaluation modes: linear-probe based VE evaluation, knowledge-prompted LLM evaluation, and holistic evaluation. Experiments on nine open-source MLLMs plus GPT-4o reveal significant deficiencies in fundus reading skills, particularly in basic tasks such as laterality recognition. The results highlight the limitations of current MLLMs and emphasize the need for domain-specific training and improved LLMs and VEs.

Data is a crucial element in large language model (LLM) alignment. Recent studies have explored using LLMs for efficient data collection. However, LLM-generated data often suffers from quality issues, with underrepresented or absent aspects and low-quality datapoints. To address these problems, we propose Data Advisor, an enhanced LLM-based method for generating data that takes into account the characteristics of the desired dataset. Starting from a set of pre-defined principles in hand, Data Advisor monitors the status of the generated data, identifies weaknesses in the current dataset, and advises the next iteration of data generation accordingly. Data Advisor can be easily integrated into existing data generation methods to enhance data quality and coverage. Experiments on safety alignment of three representative LLMs (i.e., Mistral, Llama2, and Falcon) demonstrate the effectiveness of Data Advisor in enhancing model safety against various fine-grained safety issues without sacrificing model utility.

Large language models (LLMs) have captured significant interest from both academia and industry due to their impressive performance across various textual tasks. However, the potential of LLMs to analyze physiological time-series data remains an emerging research field. Particularly, there is a notable gap in the utilization of LLMs for analyzing wearable biosignals to achieve cuffless blood pressure (BP) measurement, which is critical for the management of cardiovascular diseases. This paper presents the first work to explore the capacity of LLMs to perform cuffless BP estimation based on wearable biosignals. We extracted physiological features from electrocardiogram (ECG) and photoplethysmogram (PPG) signals and designed context-enhanced prompts by combining these features with BP domain knowledge and user information. Subsequently, we adapted LLMs to BP estimation tasks through fine-tuning. To evaluate the proposed approach, we conducted assessments of ten advanced LLMs using a comprehensive public dataset of wearable biosignals from 1,272 participants. The experimental results demonstrate that the optimally fine-tuned LLM significantly surpasses conventional task-specific baselines, achieving an estimation error of 0.00 pm 9.25 mmHg for systolic BP and 1.29 pm 6.37 mmHg for diastolic BP. Notably, the ablation studies highlight the benefits of our context enhancement strategy, leading to an 8.9% reduction in mean absolute error for systolic BP estimation. This paper pioneers the exploration of LLMs for cuffless BP measurement, providing a potential solution to enhance the accuracy of cuffless BP measurement.

We present PhysioLLM, an interactive system that leverages large language models (LLMs) to provide personalized health understanding and exploration by integrating physiological data from wearables with contextual information. Unlike commercial health apps for wearables, our system offers a comprehensive statistical analysis component that discovers correlations and trends in user data, allowing users to ask questions in natural language and receive generated personalized insights, and guides them to develop actionable goals. As a case study, we focus on improving sleep quality, given its measurability through physiological data and its importance to general well-being. Through a user study with 24 Fitbit watch users, we demonstrate that PhysioLLM outperforms both the Fitbit App alone and a generic LLM chatbot in facilitating a deeper, personalized understanding of health data and supporting actionable steps toward personal health goals.

The incorporation of LLMs in multi-agent systems (MASs) has the potential to significantly improve our ability to autonomously solve complex problems. However, such systems introduce unique challenges in monitoring, interpreting, and detecting system failures. Most existing MAS observability frameworks focus on analyzing each individual agent separately, overlooking failures associated with the entire MAS. To bridge this gap, we propose LumiMAS, a novel MAS observability framework that incorporates advanced analytics and monitoring techniques. The proposed framework consists of three key components: a monitoring and logging layer, anomaly detection layer, and anomaly explanation layer. LumiMAS's first layer monitors MAS executions, creating detailed logs of the agents' activity. These logs serve as input to the anomaly detection layer, which detects anomalies across the MAS workflow in real time. Then, the anomaly explanation layer performs classification and root cause analysis (RCA) of the detected anomalies. LumiMAS was evaluated on seven different MAS applications, implemented using two popular MAS platforms, and a diverse set of possible failures. The applications include two novel failure-tailored applications that illustrate the effects of a hallucination or bias on the MAS. The evaluation results demonstrate LumiMAS's effectiveness in failure detection, classification, and RCA.

Large Language Models (LLMs) are transforming artificial intelligence, enabling autonomous agents to perform diverse tasks across various domains. These agents, proficient in human-like text comprehension and generation, have the potential to revolutionize sectors from customer service to healthcare. However, they face challenges such as multimodality, human value alignment, hallucinations, and evaluation. Techniques like prompting, reasoning, tool utilization, and in-context learning are being explored to enhance their capabilities. Evaluation platforms like AgentBench, WebArena, and ToolLLM provide robust methods for assessing these agents in complex scenarios. These advancements are leading to the development of more resilient and capable autonomous agents, anticipated to become integral in our digital lives, assisting in tasks from email responses to disease diagnosis. The future of AI, with LLMs at the forefront, is promising.

Large Language Models (LLMs) show promise in generating firmware for embedded systems, but often introduce security flaws and fail to meet real-time performance constraints. This paper proposes a three-phase methodology that combines LLM-based firmware generation with automated security validation and iterative refinement in a virtualized environment. Using structured prompts, models like GPT-4 generate firmware for networking and control tasks, deployed on FreeRTOS via QEMU. These implementations are tested using fuzzing, static analysis, and runtime monitoring to detect vulnerabilities such as buffer overflows (CWE-120), race conditions (CWE-362), and denial-of-service threats (CWE-400). Specialized AI agents for Threat Detection, Performance Optimization, and Compliance Verification collaborate to improve detection and remediation. Identified issues are categorized using CWE, then used to prompt targeted LLM-generated patches in an iterative loop. Experiments show a 92.4\% Vulnerability Remediation Rate (37.3\% improvement), 95.8\% Threat Model Compliance, and 0.87 Security Coverage Index. Real-time metrics include 8.6ms worst-case execution time and 195μs jitter. This process enhances firmware security and performance while contributing an open-source dataset for future research.

This paper provides a comprehensive review of the future of cybersecurity through Generative AI and Large Language Models (LLMs). We explore LLM applications across various domains, including hardware design security, intrusion detection, software engineering, design verification, cyber threat intelligence, malware detection, and phishing detection. We present an overview of LLM evolution and its current state, focusing on advancements in models such as GPT-4, GPT-3.5, Mixtral-8x7B, BERT, Falcon2, and LLaMA. Our analysis extends to LLM vulnerabilities, such as prompt injection, insecure output handling, data poisoning, DDoS attacks, and adversarial instructions. We delve into mitigation strategies to protect these models, providing a comprehensive look at potential attack scenarios and prevention techniques. Furthermore, we evaluate the performance of 42 LLM models in cybersecurity knowledge and hardware security, highlighting their strengths and weaknesses. We thoroughly evaluate cybersecurity datasets for LLM training and testing, covering the lifecycle from data creation to usage and identifying gaps for future research. In addition, we review new strategies for leveraging LLMs, including techniques like Half-Quadratic Quantization (HQQ), Reinforcement Learning with Human Feedback (RLHF), Direct Preference Optimization (DPO), Quantized Low-Rank Adapters (QLoRA), and Retrieval-Augmented Generation (RAG). These insights aim to enhance real-time cybersecurity defenses and improve the sophistication of LLM applications in threat detection and response. Our paper provides a foundational understanding and strategic direction for integrating LLMs into future cybersecurity frameworks, emphasizing innovation and robust model deployment to safeguard against evolving cyber threats.

Large Language Models (LLMs) have become extremely potent instruments with exceptional capacities for comprehending and producing human-like text in a wide range of applications. However, the increasing size and complexity of LLMs present significant challenges in both training and deployment, leading to substantial computational and storage costs as well as heightened energy consumption. In this paper, we provide a review of recent advancements and research directions aimed at addressing these challenges and enhancing the efficiency of LLM-based systems. We begin by discussing algorithm-level acceleration techniques focused on optimizing LLM inference speed and resource utilization. We also explore LLM-hardware co-design strategies with a vision to improve system efficiency by tailoring hardware architectures to LLM requirements. Further, we delve into LLM-to-accelerator compilation approaches, which involve customizing hardware accelerators for efficient LLM deployment. Finally, as a case study to leverage LLMs for assisting circuit design, we examine LLM-aided design methodologies for an important task: High-Level Synthesis (HLS) functional verification, by creating a new dataset that contains a large number of buggy and bug-free codes, which can be essential for training LLMs to specialize on HLS verification and debugging. For each aspect mentioned above, we begin with a detailed background study, followed by the presentation of several novel solutions proposed to overcome specific challenges. We then outline future research directions to drive further advancements. Through these efforts, we aim to pave the way for more efficient and scalable deployment of LLMs across a diverse range of applications.

While intelligent virtual assistants like Siri, Alexa, and Google Assistant have become ubiquitous in modern life, they still face limitations in their ability to follow multi-step instructions and accomplish complex goals articulated in natural language. However, recent breakthroughs in large language models (LLMs) show promise for overcoming existing barriers by enhancing natural language processing and reasoning capabilities. Though promising, applying LLMs to create more advanced virtual assistants still faces challenges like ensuring robust performance and handling variability in real-world user commands. This paper proposes a novel LLM-based virtual assistant that can automatically perform multi-step operations within mobile apps based on high-level user requests. The system represents an advance in assistants by providing an end-to-end solution for parsing instructions, reasoning about goals, and executing actions. LLM-based Process Automation (LLMPA) has modules for decomposing instructions, generating descriptions, detecting interface elements, predicting next actions, and error checking. Experiments demonstrate the system completing complex mobile operation tasks in Alipay based on natural language instructions. This showcases how large language models can enable automated assistants to accomplish real-world tasks. The main contributions are the novel LLMPA architecture optimized for app process automation, the methodology for applying LLMs to mobile apps, and demonstrations of multi-step task completion in a real-world environment. Notably, this work represents the first real-world deployment and extensive evaluation of a large language model-based virtual assistant in a widely used mobile application with an enormous user base numbering in the hundreds of millions.

Existing Large Language Models (LLM) can invoke a variety of tools and APIs to complete complex tasks. The computer, as the most powerful and universal tool, could potentially be controlled directly by a trained LLM agent. Powered by the computer, we can hopefully build a more generalized agent to assist humans in various daily digital works. In this paper, we construct an environment for a Vision Language Model (VLM) agent to interact with a real computer screen. Within this environment, the agent can observe screenshots and manipulate the Graphics User Interface (GUI) by outputting mouse and keyboard actions. We also design an automated control pipeline that includes planning, acting, and reflecting phases, guiding the agent to continuously interact with the environment and complete multi-step tasks. Additionally, we construct the ScreenAgent Dataset, which collects screenshots and action sequences when completing a variety of daily computer tasks. Finally, we trained a model, ScreenAgent, which achieved computer control capabilities comparable to GPT-4V and demonstrated more precise UI positioning capabilities. Our attempts could inspire further research on building a generalist LLM agent. The code is available at https://github.com/niuzaisheng/ScreenAgent.

Most prior safety research of large language models (LLMs) has focused on enhancing the alignment of LLMs to better suit the safety requirements of humans. However, internalizing such safeguard features into larger models brought challenges of higher training cost and unintended degradation of helpfulness. To overcome such challenges, a modular approach employing a smaller LLM to detect harmful user queries is regarded as a convenient solution in designing LLM-based system with safety requirements. In this paper, we leverage a smaller LLM for both harmful query detection and safeguard response generation. We introduce our safety requirements and the taxonomy of harmfulness categories, and then propose a multi-task learning mechanism fusing the two tasks into a single model. We demonstrate the effectiveness of our approach, providing on par or surpassing harmful query detection and safeguard response performance compared to the publicly available LLMs.

Large language models (LLMs) show promise in medical diagnosis, but real-world deployment remains challenging due to high-stakes clinical decisions and imperfect reasoning reliability. As a result, careful inspection of model behavior is essential for assessing whether diagnostic reasoning is reliable and clinically grounded. However, debugging medical LLMs remains difficult. First, developers often lack sufficient medical domain expertise to interpret model errors in clinically meaningful terms. Second, models can fail across a large and diverse set of instances involving different input types, tasks, and reasoning steps, making it challenging for developers to prioritize which errors deserve focused inspection. Third, developers struggle to identify recurring error patterns across cases, as existing debugging practices are largely instance-centric and rely on manual inspection of isolated failures. To address these challenges, we present VeriLLMed, a visual analytics system that integrates external biomedical knowledge to audit and debug medical LLM diagnostic reasoning. VeriLLMed transforms model outputs into comparable reasoning paths, constructs knowledge graph-grounded reference paths, and identifies three recurring classes of diagnosis errors: relation errors, branch errors, and missing errors. Case studies and expert evaluation demonstrate that VeriLLMed helps developers identify clinically implausible reasoning and generate actionable insights that can inform the improvement of medical LLMs.

Large Language Models (LLMs) are advanced Artificial Intelligence (AI) systems that have undergone extensive training using large datasets in order to understand and produce language that closely resembles that of humans. These models have reached a level of proficiency where they are capable of successfully completing university exams across several disciplines and generating functional code to handle novel problems. This research investigates the coding proficiency of ChatGPT 3.5, a LLM released by OpenAI in November 2022, which has gained significant recognition for its impressive text generating and code creation capabilities. The skill of the model in creating code snippets is evaluated across 10 various programming languages and 4 different software domains. Based on the findings derived from this research, major unexpected behaviors and limitations of the model have been identified. This study aims to identify potential areas for development and examine the ramifications of automated code generation on the evolution of programming languages and on the tech industry.

Large language model (LLM) scaling laws are empirical formulas that estimate changes in model quality as a result of increasing parameter count and training data. However, these formulas, including the popular DeepMind Chinchilla scaling laws, neglect to include the cost of inference. We modify the Chinchilla scaling laws to calculate the optimal LLM parameter count and pre-training data size to train and deploy a model of a given quality and inference demand. We conduct our analysis both in terms of a compute budget and real-world costs and find that LLM researchers expecting reasonably large inference demand (~1B requests) should train models smaller and longer than Chinchilla-optimal.

Large language model (LLM) agents are evolving from request-response assistants into long-running software actors: they maintain state across model calls, fork subtasks, wait for external events, request human authority, generate tools, and perform side effects that must be resumed and audited. This paper presents Agent libOS, a library-OS-inspired runtime substrate for LLM agents. Agent libOS runs above a conventional host operating system; it does not implement hardware drivers, kernel-mode isolation, or a POSIX-compatible operating system. Instead, it treats an agent as an AgentProcess: a schedulable execution subject with process identity, parent-child lineage, lifecycle state, a tool table derived from an AgentImage, typed Object Memory, explicit capabilities, human queues, checkpoints, events, and audit records. Its central design rule is tools are libc-like wrappers; runtime primitives are the authority boundary. Filesystem access, object access, sleeps, human approval, JIT tool registration, and external side effects are checked at primitive boundaries under explicit capabilities and policy. We describe the design, threat model, Python prototype, and safety-oriented evaluation. The current prototype implements async scheduling, namespace-local Object Memory, runtime-integrated human approval, one-shot permission grants, per-process working directories, shell and image-registration primitives, Deno/TypeScript JIT tools over a libOS syscall broker, filesystem/object bridge tools, an injectable Resource Provider Substrate, deterministic demos, real-model smoke scripts, and 123 regression tests at the time of writing. Rather than improving planner accuracy, Agent libOS demonstrates a runtime substrate in which long-running LLM agents can be scheduled, authorized, resumed, and audited without treating tool dispatch as the trust boundary.

Generating accurate circuit schematics from high-level natural language descriptions remains a persistent challenge in electronics design, as large language models (LLMs) frequently hallucinate in granular details, violate electrical constraints, and produce non-machine-readable outputs. We present CircuitLM, a novel multi-agent LLM-aided circuit design pipeline that translates user prompts into structured, visually interpretable CircuitJSON schematics through five sequential stages: (i) LLM-based component identification, (ii) canonical pinout retrieval, (iii) chain-of-thought reasoning by an electronics expert agent, (iv) JSON schematic synthesis, and (v) force-directed SVG visualization. Anchored by a curated, embedding-powered component knowledge base. While LLMs often violate electrical constraints, CircuitLM bridges this gap by grounding generation in a verified and dynamically extensible component database, initially comprising 50 components. To ensure safety, we incorporate a hybrid evaluation framework, namely Dual-Metric Circuit Validation (DMCV), validated against human-expert assessments, which achieves high fidelity in microcontroller-centric designs. We evaluate the system on 100 diverse embedded-systems prompts across six LLMs and introduce DMCV to assess both structural and electrical validity. This work bridges natural language input to deployable hardware designs, enabling reliable circuit prototyping by non-experts. Our code and data will be made public upon acceptance.

Large language models (LLMs) have revolutionized natural language processing with their exceptional capabilities. However, deploying LLMs on resource-constrained edge devices presents significant challenges due to computational limitations, memory constraints, and edge hardware heterogeneity. This survey summarizes recent developments in edge LLMs across their lifecycle, examining resource-efficient designs from pre-deployment techniques to runtime optimizations. Additionally, it explores on-device LLM applications in personal, enterprise, and industrial scenarios. By synthesizing advancements and identifying future directions, this survey aims to provide a comprehensive understanding of state-of-the-art methods for deploying LLMs on edge devices, bridging the gap between their immense potential and edge computing limitations.

Conventional evaluation methods for multimodal LLMs (MLLMs) lack interpretability and are often insufficient to fully disclose significant capability gaps across models. To address this, we introduce AuditDM, an automated framework that actively discovers and rectifies MLLM failure modes by auditing their divergence. AuditDM fine-tunes an MLLM as an auditor via reinforcement learning to generate challenging questions and counterfactual images that maximize disagreement among target models. Once trained, the auditor uncovers diverse, interpretable exemplars that reveal model weaknesses and serve as annotation-free data for rectification. When applied to SoTA models like Gemma-3 and PaliGemma-2, AuditDM discovers more than 20 distinct failure types. Fine-tuning on these discoveries consistently improves all models across 16 benchmarks, and enables a 3B model to surpass its 28B counterpart. Our results suggest that as data scaling hits diminishing returns, targeted model auditing offers an effective path to model diagnosis and improvement.

Recent advancements in large language models (LLMs) have prompted interest in deploying these models on mobile devices to enable new applications without relying on cloud connectivity. However, the efficiency constraints of deploying LLMs on resource-limited devices present significant challenges. In this paper, we conduct a comprehensive measurement study to evaluate the efficiency tradeoffs between mobile-based, edge-based, and cloud-based deployments for LLM applications. We implement AutoLife-Lite, a simplified LLM-based application that analyzes smartphone sensor data to infer user location and activity contexts. Our experiments reveal that: (1) Only small-size LLMs (<4B parameters) can run successfully on powerful mobile devices, though they exhibit quality limitations compared to larger models; (2) Model compression is effective in lower the hardware requirement, but may lead to significant performance degradation; (3) The latency to run LLMs on mobile devices with meaningful output is significant (>30 seconds), while cloud services demonstrate better time efficiency (<10 seconds); (4) Edge deployments offer intermediate tradeoffs between latency and model capabilities, with different results on CPU-based and GPU-based settings. These findings provide valuable insights for system designers on the current limitations and future directions for on-device LLM applications.

Large Language Models (LLMs) have gained popularity in time series forecasting, but their potential for anomaly detection remains largely unexplored. Our study investigates whether LLMs can understand and detect anomalies in time series data, focusing on zero-shot and few-shot scenarios. Inspired by conjectures about LLMs' behavior from time series forecasting research, we formulate key hypotheses about LLMs' capabilities in time series anomaly detection. We design and conduct principled experiments to test each of these hypotheses. Our investigation reveals several surprising findings about LLMs for time series: 1. LLMs understand time series better as images rather than as text 2. LLMs did not demonstrate enhanced performance when prompted to engage in explicit reasoning about time series analysis 3. Contrary to common beliefs, LLM's understanding of time series do not stem from their repetition biases or arithmetic abilities 4. LLMs' behaviors and performance in time series analysis vary significantly across different model architectures This study provides the first comprehensive analysis of contemporary LLM capabilities in time series anomaly detection. Our results suggest that while LLMs can understand time series anomalies, many common conjectures based on their reasoning capabilities do not hold. Our code and data are available at `https://github.com/Rose-STL-Lab/AnomLLM/`.

System log anomaly detection is critical for maintaining the reliability of large-scale software systems, yet traditional methods struggle with the heterogeneous and evolving nature of modern log data. Recent advances in Large Language Models (LLMs) offer promising new approaches to log understanding, but a systematic comparison of LLM-based methods against established techniques remains lacking. In this paper, we present a comprehensive benchmark study evaluating both LLM-based and traditional approaches for log anomaly detection across four widely-used public datasets: HDFS, BGL, Thunderbird, and Spirit. We evaluate three categories of methods: (1) classical log parsers (Drain, Spell, AEL) combined with machine learning classifiers, (2) fine-tuned transformer models (BERT, RoBERTa), and (3) prompt-based LLM approaches (GPT-3.5, GPT-4, LLaMA-3) in zero-shot and few-shot settings. Our experiments reveal that while fine-tuned transformers achieve the highest F1-scores (0.96-0.99), prompt-based LLMs demonstrate remarkablezero-shot capabilities (F1: 0.82-0.91) without requiring any labeled training data -- a significant advantage for real-world deployment where labeled anomalies are scarce. We further analyze the cost-accuracy trade-offs, latency characteristics, and failure modes of each approach. Our findings provide actionable guidelines for practitioners choosing log anomaly detection methods based on their specific constraints regarding accuracy, latency, cost, and label availability. All code and experimental configurations are publicly available to facilitate reproducibility.

Large language models (LLMs) are susceptible to a variety of risks, from non-faithful output to biased and toxic generations. Due to several limiting factors surrounding LLMs (training cost, API access, data availability, etc.), it may not always be feasible to impose direct safety constraints on a deployed model. Therefore, an efficient and reliable alternative is required. To this end, we present our ongoing efforts to create and deploy a library of detectors: compact and easy-to-build classification models that provide labels for various harms. In addition to the detectors themselves, we discuss a wide range of uses for these detector models - from acting as guardrails to enabling effective AI governance. We also deep dive into inherent challenges in their development and discuss future work aimed at making the detectors more reliable and broadening their scope.

Recently, researchers have attempted to investigate the capability of LLMs in handling videos and proposed several video LLM models. However, the ability of LLMs to handle video grounding (VG), which is an important time-related video task requiring the model to precisely locate the start and end timestamps of temporal moments in videos that match the given textual queries, still remains unclear and unexplored in literature. To fill the gap, in this paper, we propose the LLM4VG benchmark, which systematically evaluates the performance of different LLMs on video grounding tasks. Based on our proposed LLM4VG, we design extensive experiments to examine two groups of video LLM models on video grounding: (i) the video LLMs trained on the text-video pairs (denoted as VidLLM), and (ii) the LLMs combined with pretrained visual description models such as the video/image captioning model. We propose prompt methods to integrate the instruction of VG and description from different kinds of generators, including caption-based generators for direct visual description and VQA-based generators for information enhancement. We also provide comprehensive comparisons of various VidLLMs and explore the influence of different choices of visual models, LLMs, prompt designs, etc, as well. Our experimental evaluations lead to two conclusions: (i) the existing VidLLMs are still far away from achieving satisfactory video grounding performance, and more time-related video tasks should be included to further fine-tune these models, and (ii) the combination of LLMs and visual models shows preliminary abilities for video grounding with considerable potential for improvement by resorting to more reliable models and further guidance of prompt instructions.

LLM agents are evolving from conversational chatbots to operational tools in real-world workspaces. In local agentic harnesses, an LLM can read and write files, call tools, and reuse workspace state across sessions. While such capabilities enhance utility, they also expose a new attack surface for attackers. Attackers can embed a prompt injection within a file or tool output. Agents may read this hidden instruction, store it, and execute it later. In this multi-step trojan attack paradigm, no individual step appears malicious on its own, but these steps can collectively turn untrusted text into persistent control content. However, existing defenses often inspect each step in isolation. As a result, they can block a clear harmful action, but fail to detect the earlier write operation that plants the backdoor. To reveal this threat, we introduce ClawTrojan, a benchmark designed to identify multi-step trojan attacks in local agentic harnesses. In an OpenClaw-style simulated workspace with GPT-5.4, ClawTrojan reaches a 95.5% attack success rate (ASR), while existing single-turn prompt-injection attacks produce near-zero ASR on the same model. To address this threat, we propose DASGuard, which scans control-like text in sensitive local files, traces its origin, and removes control content that does not originate from a trusted source. Our results show that DASGuard achieves strong dynamic defense by combining runtime attack blocking with sanitized commits to the workspace.

This paper provides a comprehensive review of the integration of Large Language Models (LLMs) with visual analytics, addressing their foundational concepts, capabilities, and wide-ranging applications. It begins by outlining the theoretical underpinnings of visual analytics and the transformative potential of LLMs, specifically focusing on their roles in natural language understanding, natural language generation, dialogue systems, and text-to-media transformations. The review further investigates how the synergy between LLMs and visual analytics enhances data interpretation, visualization techniques, and interactive exploration capabilities. Key tools and platforms including LIDA, Chat2VIS, Julius AI, and Zoho Analytics, along with specialized multimodal models such as ChartLlama and CharXIV, are critically evaluated. The paper discusses their functionalities, strengths, and limitations in supporting data exploration, visualization enhancement, automated reporting, and insight extraction. The taxonomy of LLM tasks, ranging from natural language understanding (NLU), natural language generation (NLG), to dialogue systems and text-to-media transformations, is systematically explored. This review provides a SWOT analysis of integrating Large Language Models (LLMs) with visual analytics, highlighting strengths like accessibility and flexibility, weaknesses such as computational demands and biases, opportunities in multimodal integration and user collaboration, and threats including privacy concerns and skill degradation. It emphasizes addressing ethical considerations and methodological improvements for effective integration.

LLM applications (i.e., LLM apps) leverage the powerful capabilities of LLMs to provide users with customized services, revolutionizing traditional application development. While the increasing prevalence of LLM-powered applications provides users with unprecedented convenience, it also brings forth new security challenges. For such an emerging ecosystem, the security community lacks sufficient understanding of the LLM application ecosystem, especially regarding the capability boundaries of the applications themselves. In this paper, we systematically analyzed the new development paradigm and defined the concept of the LLM app capability space. We also uncovered potential new risks beyond jailbreak that arise from ambiguous capability boundaries in real-world scenarios, namely, capability downgrade and upgrade. To evaluate the impact of these risks, we designed and implemented an LLM app capability evaluation framework, LLMApp-Eval. First, we collected application metadata across 4 platforms and conducted a cross-platform ecosystem analysis. Then, we evaluated the risks for 199 popular applications among 4 platforms and 6 open-source LLMs. We identified that 178 (89.45%) potentially affected applications, which can perform tasks from more than 15 scenarios or be malicious. We even found 17 applications in our study that executed malicious tasks directly, without applying any adversarial rewriting. Furthermore, our experiments also reveal a positive correlation between the quality of prompt design and application robustness. We found that well-designed prompts enhance security, while poorly designed ones can facilitate abuse. We hope our work inspires the community to focus on the real-world risks of LLM applications and foster the development of a more robust LLM application ecosystem.

Automatic disease diagnosis has become increasingly valuable in clinical practice. The advent of large language models (LLMs) has catalyzed a paradigm shift in artificial intelligence, with growing evidence supporting the efficacy of LLMs in diagnostic tasks. Despite the increasing attention in this field, a holistic view is still lacking. Many critical aspects remain unclear, such as the diseases and clinical data to which LLMs have been applied, the LLM techniques employed, and the evaluation methods used. In this article, we perform a comprehensive review of LLM-based methods for disease diagnosis. Our review examines the existing literature across various dimensions, including disease types and associated clinical specialties, clinical data, LLM techniques, and evaluation methods. Additionally, we offer recommendations for applying and evaluating LLMs for diagnostic tasks. Furthermore, we assess the limitations of current research and discuss future directions. To our knowledge, this is the first comprehensive review for LLM-based disease diagnosis.

With the introduction of ChatGPT, Large Language Models (LLMs) have received enormous attention in healthcare. Despite their potential benefits, researchers have underscored various ethical implications. While individual instances have drawn much attention, the debate lacks a systematic overview of practical applications currently researched and ethical issues connected to them. Against this background, this work aims to map the ethical landscape surrounding the current stage of deployment of LLMs in medicine and healthcare. Electronic databases and preprint servers were queried using a comprehensive search strategy. Studies were screened and extracted following a modified rapid review approach. Methodological quality was assessed using a hybrid approach. For 53 records, a meta-aggregative synthesis was performed. Four fields of applications emerged and testify to a vivid exploration phase. Advantages of using LLMs are attributed to their capacity in data analysis, personalized information provisioning, support in decision-making, mitigating information loss and enhancing information accessibility. However, we also identifies recurrent ethical concerns connected to fairness, bias, non-maleficence, transparency, and privacy. A distinctive concern is the tendency to produce harmful misinformation or convincingly but inaccurate content. A recurrent plea for ethical guidance and human oversight is evident. Given the variety of use cases, it is suggested that the ethical guidance debate be reframed to focus on defining what constitutes acceptable human oversight across the spectrum of applications. This involves considering diverse settings, varying potentials for harm, and different acceptable thresholds for performance and certainty in healthcare. In addition, a critical inquiry is necessary to determine the extent to which the current experimental use of LLMs is necessary and justified.

As Large Language Models (LLMs) are rapidly growing in popularity, LLM inference services must be able to serve requests from thousands of users while satisfying performance requirements. The performance of an LLM inference service is largely determined by the hardware onto which it is deployed, but understanding of which hardware will deliver on performance requirements remains challenging. In this work we present LLM-Pilot - a first-of-its-kind system for characterizing and predicting performance of LLM inference services. LLM-Pilot performs benchmarking of LLM inference services, under a realistic workload, across a variety of GPUs, and optimizes the service configuration for each considered GPU to maximize performance. Finally, using this characterization data, LLM-Pilot learns a predictive model, which can be used to recommend the most cost-effective hardware for a previously unseen LLM. Compared to existing methods, LLM-Pilot can deliver on performance requirements 33% more frequently, whilst reducing costs by 60% on average.

Large Language Models (LLMs) have recently emerged as a focal point of research and application, driven by their unprecedented ability to understand and generate text with human-like quality. Even more recently, LLMs have been extended into multi-modal large language models (MM-LLMs) which extends their capabilities to deal with image, video and audio information, in addition to text. This opens up applications like text-to-video generation, image captioning, text-to-speech, and more and is achieved either by retro-fitting an LLM with multi-modal capabilities, or building a MM-LLM from scratch. This paper provides an extensive review of the current state of those LLMs with multi-modal capabilities as well as the very recent MM-LLMs. It covers the historical development of LLMs especially the advances enabled by transformer-based architectures like OpenAI's GPT series and Google's BERT, as well as the role of attention mechanisms in enhancing model performance. The paper includes coverage of the major and most important of the LLMs and MM-LLMs and also covers the techniques of model tuning, including fine-tuning and prompt engineering, which tailor pre-trained models to specific tasks or domains. Ethical considerations and challenges, such as data bias and model misuse, are also analysed to underscore the importance of responsible AI development and deployment. Finally, we discuss the implications of open-source versus proprietary models in AI research. Through this review, we provide insights into the transformative potential of MM-LLMs in various applications.

Recent advances such as LLaVA and Mini-GPT4 have successfully integrated visual information into LLMs, yielding inspiring outcomes and giving rise to a new generation of multi-modal LLMs, or MLLMs. Nevertheless, these methods struggle with hallucinations and the mutual interference between tasks. To tackle these problems, we propose an efficient and accurate approach to adapt to downstream tasks by utilizing LLM as a bridge to connect multiple expert models, namely u-LLaVA. Firstly, we incorporate the modality alignment module and multi-task modules into LLM. Then, we reorganize or rebuild multi-type public datasets to enable efficient modality alignment and instruction following. Finally, task-specific information is extracted from the trained LLM and provided to different modules for solving downstream tasks. The overall framework is simple, effective, and achieves state-of-the-art performance across multiple benchmarks. We also release our model, the generated data, and the code base publicly available.

Large language models (LLMs) are leading significant progress in code generation. Beyond one-pass code generation, recent works further integrate unit tests and program verifiers into LLMs to iteratively refine the generated programs. However, these works consider the generated programs as an indivisible entity, which falls short for LLMs in debugging the programs, especially when the programs contain complex logic flows and data operations. In contrast, when human developers debug programs, they typically set breakpoints and selectively examine runtime execution information. The execution flow and the intermediate variables play a crucial role in the debugging process, yet they are underutilized in the existing literature on code generation. In this study, we introduce Large Language Model Debugger (LDB), a novel debugging framework that enables LLMs to refine their generated programs with the runtime execution information. Specifically, LDB segments the programs into basic blocks and tracks the values of intermediate variables after each block throughout the runtime execution. This allows LLMs to concentrate on simpler code units within the overall execution flow, verify their correctness against the task description block by block, and efficiently pinpoint any potential errors. Experiments demonstrate that LDB consistently enhances the baseline performance by up to 9.8% across the HumanEval, MBPP, and TransCoder benchmarks, archiving new state-of-the-art performance in code debugging for various LLM selections.

We explore the integration of large language models (LLMs) into visual analytics (VA) systems to transform their capabilities through intuitive natural language interactions. We survey current research directions in this emerging field, examining how LLMs are integrated into data management, language interaction, visualisation generation, and language generation processes. We highlight the new possibilities that LLMs bring to VA, especially how they can change VA processes beyond the usual use cases. We especially highlight building new visualisation-language models, allowing access of a breadth of domain knowledge, multimodal interaction, and opportunities with guidance. Finally, we carefully consider the prominent challenges of using current LLMs in VA tasks. Our discussions in this paper aim to guide future researchers working on LLM-assisted VA systems and help them navigate common obstacles when developing these systems.

The continuous flow of data collected by Internet of Things (IoT) devices, has revolutionised our ability to understand and interact with the world across various applications. However, this data must be prepared and transformed into event data before analysis can begin. In this paper, we shed light on the potential of leveraging Large Language Models (LLMs) in event abstraction and integration. Our approach aims to create event records from raw sensor readings and merge the logs from multiple IoT sources into a single event log suitable for further Process Mining applications. We demonstrate the capabilities of LLMs in event abstraction considering a case study for IoT application in elderly care and longitudinal health monitoring. The results, showing on average an accuracy of 90% in detecting high-level activities. These results highlight LLMs' promising potential in addressing event abstraction and integration challenges, effectively bridging the existing gap.

Large language models (LLMs) have been applied to tasks in healthcare, ranging from medical exam questions to responding to patient questions. With increasing institutional partnerships between companies producing LLMs and healthcare systems, real world clinical application is coming closer to reality. As these models gain traction, it is essential for healthcare practitioners to understand what LLMs are, their development, their current and potential applications, and the associated pitfalls when utilized in medicine. This review and accompanying tutorial aim to give an overview of these topics to aid healthcare practitioners in understanding the rapidly changing landscape of LLMs as applied to medicine.

Large language models (LLMs) have proven their remarkable versatility in handling a comprehensive range of language-centric applications. To expand LLMs' capabilities to a broader spectrum of modal inputs, multimodal large language models (MLLMs) have attracted growing interest. This work delves into enabling LLMs to tackle more vision-language-related tasks, particularly image captioning, visual question answering (VQA,) and visual grounding. To this end, we implemented a three-stage training scheme: starting with lightweight alignment pretraining, then moderate-weight multitask hybrid training, and finally, LLM fine-tuning to improve instruction following capability. Throughout the training process, the requirements on GPU memory gradually increase. To effectively manage the number of visual embeddings passed to the LLM while preserving their positional information, we introduce a straightforward visual adapter module dubbed pool-adapter. Our experiments demonstrate that preserving the positional information of visual embeddings through the pool-adapter is particularly beneficial for tasks like visual grounding. We name our proposed approach InfMLLM and have evaluated it extensively on various benchmark datasets. Our results demonstrate that InfMLLM achieves either state-of-the-art (SOTA) performance or performance comparable to recent MLLMs. The code and model will be made open-source at: https://github.com/mightyzau/InfMLLM.

Recent advancements indicate that scaling up Multimodal Large Language Models (MLLMs) effectively enhances performance on downstream multimodal tasks. The prevailing MLLM paradigm, e.g., LLaVA, transforms visual features into text-like tokens using a static vision-language mapper, thereby enabling static LLMs to develop the capability to comprehend visual information through visual instruction tuning. Although promising, the static tuning strategy~The static tuning refers to the trained model with static parameters. that shares the same parameters may constrain performance across different downstream multimodal tasks. In light of this, we introduce HyperLLaVA, which involves adaptive tuning of the projector and LLM parameters, in conjunction with a dynamic visual expert and language expert, respectively. These experts are derived from HyperNetworks, which generates adaptive parameter shifts through visual and language guidance, enabling dynamic projector and LLM modeling in two-stage training. Our experiments demonstrate that our solution significantly surpasses LLaVA on existing MLLM benchmarks, including MME, MMBench, SEED-Bench, and LLaVA-Bench. ~Our project is available on the link https://github.com/DCDmllm/HyperLLaVA.

The powerful ability to understand, follow, and generate complex language emerging from large language models (LLMs) makes LLM-generated text flood many areas of our daily lives at an incredible speed and is widely accepted by humans. As LLMs continue to expand, there is an imperative need to develop detectors that can detect LLM-generated text. This is crucial to mitigate potential misuse of LLMs and safeguard realms like artistic expression and social networks from harmful influence of LLM-generated content. The LLM-generated text detection aims to discern if a piece of text was produced by an LLM, which is essentially a binary classification task. The detector techniques have witnessed notable advancements recently, propelled by innovations in watermarking techniques, zero-shot methods, fine-turning LMs methods, adversarial learning methods, LLMs as detectors, and human-assisted methods. In this survey, we collate recent research breakthroughs in this area and underscore the pressing need to bolster detector research. We also delve into prevalent datasets, elucidating their limitations and developmental requirements. Furthermore, we analyze various LLM-generated text detection paradigms, shedding light on challenges like out-of-distribution problems, potential attacks, and data ambiguity. Conclusively, we highlight interesting directions for future research in LLM-generated text detection to advance the implementation of responsible artificial intelligence (AI). Our aim with this survey is to provide a clear and comprehensive introduction for newcomers while also offering seasoned researchers a valuable update in the field of LLM-generated text detection. The useful resources are publicly available at: https://github.com/NLP2CT/LLM-generated-Text-Detection.

Monitoring autonomous large language model (LLM) agents for covert malicious behavior is challenging due to delayed, context-dependent, and long-horizon attack patterns. Agents may pursue hidden objectives while maintaining superficially benign behavior, making detection difficult even with full trajectory access. Prior monitoring approaches improve scaffolding or ensemble aggregation, but treat each trajectory independently and do not learn from prior monitoring experience. Moreover, standard reasoning methods explain observed behavior without explicitly reasoning about agent beliefs, intentions, and goal alignment required to distinguish benign task execution from covert deviation. We propose Agent-ToM, a learning-to-monitor framework grounded in Theory-of-Mind (ToM) reasoning for security analysis of autonomous agents. Agent-ToM performs structured full-trajectory analysis by inferring beliefs, intent hypotheses with calibrated confidence, expected actions, and deviations from task-consistent behavioral baselines. At inference time, it employs a Reason-Verify-Refine pipeline to construct and validate monitoring decisions. At training time, Agent-ToM distills critique signals into a persistent semantic guardrail memory, enabling reusable belief- and intent-conditioned constraints across episodes. We evaluate Agent-ToM on adversarial agent monitoring benchmarks (SHADE-Arena and CUA-SHADE-Arena). Agent-ToM achieves strong precision-recall balance and outperforms state-of-the-art monitoring baselines, including ensemble methods, while using a coherent two-call reasoning pipeline. These results demonstrate that learning at the monitoring layer, combined with structured ToM reasoning and verification, provides an effective and deployable foundation for securing autonomous LLM agents.

Large language models (LLMs) are being rapidly integrated into decision-support tools, automation workflows, and AI-enabled software systems. However, their behavior in production environments remains poorly understood, and their failure patterns differ fundamentally from those of traditional machine learning models. This paper presents a system-level taxonomy of fifteen hidden failure modes that arise in real-world LLM applications, including multi-step reasoning drift, latent inconsistency, context-boundary degradation, incorrect tool invocation, version drift, and cost-driven performance collapse. Using this taxonomy, we analyze the growing gap in evaluation and monitoring practices: existing benchmarks measure knowledge or reasoning but provide little insight into stability, reproducibility, drift, or workflow integration. We further examine the production challenges associated with deploying LLMs - including observability limitations, cost constraints, and update-induced regressions - and outline high-level design principles for building reliable, maintainable, and cost-aware LLM systems. Finally, we outline high-level design principles for building reliable, maintainable, and cost-aware LLM-based systems. By framing LLM reliability as a system-engineering problem rather than a purely model-centric one, this work provides an analytical foundation for future research on evaluation methodology, AI system robustness, and dependable LLM deployment.

Large language model (LLM) systems are increasingly used to support high-stakes decision-making, but they typically perform worse when the available evidence is internally inconsistent. Such a scenario exists in real-world healthcare settings, with patient-reported symptoms contradicting medical signs. To study this problem, we introduce MIMIC-DOS, a dataset for short-horizon organ dysfunction worsening prediction in the intensive care unit (ICU) setting. We derive this dataset from the widely recognized MIMIC-IV, a publicly available electronic health record dataset, and construct it exclusively from cases in which discordance between signs and symptoms exists. This setting poses a substantial challenge for existing LLM-based approaches, with single-pass LLMs and agentic pipelines often struggling to reconcile such conflicting signals. To address this problem, we propose CARE: a multi-stage privacy-compliant agentic reasoning framework in which a remote LLM provides guidance by generating structured categories and transitions without accessing sensitive patient data, while a local LLM uses these categories and transitions to support evidence acquisition and final decision-making. Empirically, CARE achieves stronger performance across all key metrics compared to multiple baseline settings, showing that CARE can more robustly handle conflicting clinical evidence while preserving privacy.

Large Language Models (LLMs) have gained significant attention in the medical domain for their human-level capabilities, leading to increased efforts to explore their potential in various healthcare applications. However, despite such a promising future, there are multiple challenges and obstacles that remain for their real-world uses in practical settings. This work discusses key challenges for LLMs in medical applications from four unique aspects: operational vulnerabilities, ethical and social considerations, performance and assessment difficulties, and legal and regulatory compliance. Addressing these challenges is crucial for leveraging LLMs to their full potential and ensuring their responsible integration into healthcare.

The widespread availability of large language models (LLMs), such as ChatGPT, has significantly impacted education, raising both opportunities and challenges. Students can frequently interact with LLM-powered, interactive learning tools, but their usage patterns need to be analyzed to ensure ethical usage of these tools. To better understand how students interact with LLMs in an academic setting, we introduce StudyChat, a publicly available dataset capturing real-world student interactions with an LLM-powered tutoring chatbot in a semester-long, university-level artificial intelligence (AI) course. We deploy a web application that replicates ChatGPT's core functionalities, and use it to log student interactions with the LLM while working on programming assignments. We collect 1,197 conversations, which we annotate using a dialogue act labeling schema inspired by observed interaction patterns and prior research. Additionally, we analyze these interactions, highlight behavioral trends, and analyze how specific usage patterns relate to course outcomes. StudyChat provides a rich resource for the learning sciences and AI in education communities, enabling further research into the evolving role of LLMs in education.

Visual grounding is an essential tool that links user-provided text queries with query-specific regions within an image. Despite advancements in visual grounding models, their ability to comprehend complex queries remains limited. To overcome this limitation, we introduce LLM-Optic, an innovative method that utilizes Large Language Models (LLMs) as an optical lens to enhance existing visual grounding models in comprehending complex text queries involving intricate text structures, multiple objects, or object spatial relationships, situations that current models struggle with. LLM-Optic first employs an LLM as a Text Grounder to interpret complex text queries and accurately identify objects the user intends to locate. Then a pre-trained visual grounding model is used to generate candidate bounding boxes given the refined query by the Text Grounder. After that, LLM-Optic annotates the candidate bounding boxes with numerical marks to establish a connection between text and specific image regions, thereby linking two distinct modalities. Finally, it employs a Large Multimodal Model (LMM) as a Visual Grounder to select the marked candidate objects that best correspond to the original text query. Through LLM-Optic, we have achieved universal visual grounding, which allows for the detection of arbitrary objects specified by arbitrary human language input. Importantly, our method achieves this enhancement without requiring additional training or fine-tuning. Extensive experiments across various challenging benchmarks demonstrate that LLM-Optic achieves state-of-the-art zero-shot visual grounding capabilities. Project Page: https://haoyu-zhao.github.io/LLM-Optic.github.io/.

Large language models (LLMs) have become increasingly sophisticated, leading to widespread deployment in sensitive applications where safety and reliability are paramount. However, LLMs have inherent risks accompanying them, including bias, potential for unsafe actions, dataset poisoning, lack of explainability, hallucinations, and non-reproducibility. These risks necessitate the development of "guardrails" to align LLMs with desired behaviors and mitigate potential harm. This work explores the risks associated with deploying LLMs and evaluates current approaches to implementing guardrails and model alignment techniques. We examine intrinsic and extrinsic bias evaluation methods and discuss the importance of fairness metrics for responsible AI development. The safety and reliability of agentic LLMs (those capable of real-world actions) are explored, emphasizing the need for testability, fail-safes, and situational awareness. Technical strategies for securing LLMs are presented, including a layered protection model operating at external, secondary, and internal levels. System prompts, Retrieval-Augmented Generation (RAG) architectures, and techniques to minimize bias and protect privacy are highlighted. Effective guardrail design requires a deep understanding of the LLM's intended use case, relevant regulations, and ethical considerations. Striking a balance between competing requirements, such as accuracy and privacy, remains an ongoing challenge. This work underscores the importance of continuous research and development to ensure the safe and responsible use of LLMs in real-world applications.

Guard models are widely used to detect harmful content in user prompts and LLM responses. However, state-of-the-art guard models rely solely on terminal-layer representations and overlook the rich safety-relevant features distributed across internal layers. We present SIREN, a lightweight guard model that harnesses these internal features. By identifying safety neurons via linear probing and combining them through an adaptive layer-weighted strategy, SIREN builds a harmfulness detector from LLM internals without modifying the underlying model. Our comprehensive evaluation shows that SIREN substantially outperforms state-of-the-art open-source guard models across multiple benchmarks while using 250 times fewer trainable parameters. Moreover, SIREN exhibits superior generalization to unseen benchmarks, naturally enables real-time streaming detection, and significantly improves inference efficiency compared to generative guard models. Overall, our results highlight LLM internal states as a promising foundation for practical, high-performance harmfulness detection.

With the success of large language models (LLMs), integrating the vision model into LLMs to build vision-language foundation models has gained much more interest recently. However, existing LLM-based large multimodal models (e.g., Video-LLaMA, VideoChat) can only take in a limited number of frames for short video understanding. In this study, we mainly focus on designing an efficient and effective model for long-term video understanding. Instead of trying to process more frames simultaneously like most existing work, we propose to process videos in an online manner and store past video information in a memory bank. This allows our model to reference historical video content for long-term analysis without exceeding LLMs' context length constraints or GPU memory limits. Our memory bank can be seamlessly integrated into current multimodal LLMs in an off-the-shelf manner. We conduct extensive experiments on various video understanding tasks, such as long-video understanding, video question answering, and video captioning, and our model can achieve state-of-the-art performances across multiple datasets. Code available at https://boheumd.github.io/MA-LMM/.

Large Language Models (LLMs) are highly proficient in language-based tasks. Their language capabilities have positioned them at the forefront of the future AGI (Artificial General Intelligence) race. However, on closer inspection, Valmeekam et al. (2024); Zecevic et al. (2023); Wu et al. (2024) highlight a significant gap between their language proficiency and reasoning abilities. Reasoning in LLMs and Vision Language Models (VLMs) aims to bridge this gap by enabling these models to think and re-evaluate their actions and responses. Reasoning is an essential capability for complex problem-solving and a necessary step toward establishing trust in Artificial Intelligence (AI). This will make AI suitable for deployment in sensitive domains, such as healthcare, banking, law, defense, security etc. In recent times, with the advent of powerful reasoning models like OpenAI O1 and DeepSeek R1, reasoning endowment has become a critical research topic in LLMs. In this paper, we provide a detailed overview and comparison of existing reasoning techniques and present a systematic survey of reasoning-imbued language models. We also study current challenges and present our findings.

Large Language Models (LLMs) have showcased impressive capabilities in text comprehension and generation, prompting research efforts towards video LLMs to facilitate human-AI interaction at the video level. However, how to effectively encode and understand videos in video-based dialogue systems remains to be solved. In this paper, we investigate a straightforward yet unexplored question: Can we feed all spatial-temporal tokens into the LLM, thus delegating the task of video sequence modeling to the LLMs? Surprisingly, this simple approach yields significant improvements in video understanding. Based upon this, we propose ST-LLM, an effective video-LLM baseline with Spatial-Temporal sequence modeling inside LLM. Furthermore, to address the overhead and stability issues introduced by uncompressed video tokens within LLMs, we develop a dynamic masking strategy with tailor-made training objectives. For particularly long videos, we have also designed a global-local input module to balance efficiency and effectiveness. Consequently, we harness LLM for proficient spatial-temporal modeling, while upholding efficiency and stability. Extensive experimental results attest to the effectiveness of our method. Through a more concise model and training pipeline, ST-LLM establishes a new state-of-the-art result on VideoChatGPT-Bench and MVBench. Codes have been available at https://github.com/TencentARC/ST-LLM.

VLMs (Vision-Language Models) extend the capabilities of LLMs (Large Language Models) to accept multimodal inputs. Since it has been verified that LLMs can be induced to generate harmful or inaccurate content through specific test cases (termed as Red Teaming), how VLMs perform in similar scenarios, especially with their combination of textual and visual inputs, remains a question. To explore this problem, we present a novel red teaming dataset RTVLM, which encompasses 10 subtasks (e.g., image misleading, multi-modal jail-breaking, face fairness, etc) under 4 primary aspects (faithfulness, privacy, safety, fairness). Our RTVLM is the first red-teaming dataset to benchmark current VLMs in terms of these 4 different aspects. Detailed analysis shows that 10 prominent open-sourced VLMs struggle with the red teaming in different degrees and have up to 31% performance gap with GPT-4V. Additionally, we simply apply red teaming alignment to LLaVA-v1.5 with Supervised Fine-tuning (SFT) using RTVLM, and this bolsters the models' performance with 10% in RTVLM test set, 13% in MM-Hal, and without noticeable decline in MM-Bench, overpassing other LLaVA-based models with regular alignment data. This reveals that current open-sourced VLMs still lack red teaming alignment. Our code and datasets will be open-source.

Large language models (LLMs) have been increasingly used in time series analysis. However, the potential of multimodal LLMs (MLLMs), particularly vision-language models, for time series remains largely under-explored. One natural way for humans to detect time series anomalies is through visualization and textual description. Motivated by this, we raise a critical and practical research question: Can multimodal LLMs perform time series anomaly detection? To answer this, we propose VisualTimeAnomaly benchmark to evaluate MLLMs in time series anomaly detection (TSAD). Our approach transforms time series numerical data into the image format and feed these images into various MLLMs, including proprietary models (GPT-4o and Gemini-1.5) and open-source models (LLaVA-NeXT and Qwen2-VL), each with one larger and one smaller variant. In total, VisualTimeAnomaly contains 12.4k time series images spanning 3 scenarios and 3 anomaly granularities with 9 anomaly types across 8 MLLMs. Starting with the univariate case (point- and range-wise anomalies), we extend our evaluation to more practical scenarios, including multivariate and irregular time series scenarios, and variate-wise anomalies. Our study reveals several key insights: 1) MLLMs detect range- and variate-wise anomalies more effectively than point-wise anomalies. 2) MLLMs are highly robust to irregular time series, even with 25% of the data missing. 3) Open-source MLLMs perform comparably to proprietary models in TSAD. While open-source MLLMs excel on univariate time series, proprietary MLLMs demonstrate superior effectiveness on multivariate time series. To the best of our knowledge, this is the first work to comprehensively investigate MLLMs for TSAD, particularly for multivariate and irregular time series scenarios. We release our dataset and code at https://github.com/mllm-ts/VisualTimeAnomaly to support future research.

With advancements in Large Language Models (LLMs), a major use case that has emerged is querying databases in plain English, translating user questions into executable database queries, which has improved significantly. However, real-world datasets often feature a vast array of attributes and complex values, complicating the LLMs task of accurately identifying relevant columns or values from natural language queries. Traditional methods cannot fully relay the datasets size and complexity to the LLM. To address these challenges, we propose a novel framework that leverages Full-Text Search (FTS) on the input table. This approach not only enables precise detection of specific values and columns but also narrows the search space for language models, thereby enhancing query accuracy. Additionally, it supports a custom auto-complete feature that suggests queries based on the data in the table. This integration significantly refines the interaction between the user and complex datasets, offering a sophisticated solution to the limitations faced by current table querying capabilities. This work is accompanied by an application for both Mac and Windows platforms, which readers can try out themselves on their own data.

Large language models (LLMs) are increasingly integrated into applications ranging from review summarization to medical diagnosis support, where they affect human decisions. Even though LLMs perform well in many tasks, they may also inherit societal or cognitive biases, which can inadvertently transfer to humans. We investigate when and how LLMs expose users to biased content and quantify its severity. Specifically, we assess three LLM families in summarization and news fact-checking tasks, evaluating how much LLMs stay consistent with their context and/or hallucinate. Our findings show that LLMs expose users to content that changes the sentiment of the context in 21.86% of the cases, hallucinates on post-knowledge-cutoff data questions in 57.33% of the cases, and primacy bias in 5.94% of the cases. We evaluate 18 distinct mitigation methods across three LLM families and find that targeted interventions can be effective. Given the prevalent use of LLMs in high-stakes domains, such as healthcare or legal analysis, our results highlight the need for robust technical safeguards and for developing user-centered interventions that address LLM limitations.

Production LLM systems often rely on separate models for safety and other classification-heavy steps, increasing latency, VRAM footprint, and operational complexity. We instead reuse computation already paid for by the serving LLM: we train lightweight probes on its hidden states and predict labels in the same forward pass used for generation. We frame classification as representation selection over the full token-layer hidden-state tensor, rather than committing to a fixed token or fixed layer (e.g., first-token logits or final-layer pooling). To implement this, we introduce a two-stage aggregator that (i) summarizes tokens within each layer and (ii) aggregates across layer summaries to form a single representation for classification. We instantiate this template with direct pooling, a 100K-parameter scoring-attention gate, and a downcast multi-head self-attention (MHA) probe with up to 35M trainable parameters. Across safety and sentiment benchmarks our probes improve over logit-only reuse (e.g., MULI) and are competitive with substantially larger task-specific baselines, while preserving near-serving latency and avoiding the VRAM and latency costs of a separate guard-model pipeline.