Agent Evals and Safety

Reliability, safety, benchmarking, and evaluation methods for agent systems.

Research Sep 22, 2026

How Strongly Should Task State Influence an LLM Agent?

Long-horizon assigned work requires an LLM agent to track the state of a task: which steps are done, blocked, cancelled, or open to repetition. Agent systems either keep this state as text in the prompt and rely on the model to read that text, or move the state into a module that enforces it, and each system is evaluated as a whole, so no one knows how much reliability comes from the state being shown, told, or enforced. We fix the task rules, the model, and paired episodes and vary how strongly...

Read Source

Research Sep 04, 2026

Moral Competence Before Moral Content: Why LLM Agents Lack the Prerequisites for Coherent Alignment

AI alignment requires AI systems to adhere to human norms, values, or intentions. Under value pluralism there is no correct target, but a shared prerequisite is that the system's behavior expresses a coherent policy: a mapping from situations to verdicts that is invariant while a situation's morally relevant features are preserved, and sensitive when they change. We introduce four structural conditions for such coherent policies: verdict stability, monotonicity, decisiveness, and Pareto viabilit...

Read Source

Research Sep 03, 2026

KC-Bench: A Dynamic Interactive Benchmark for Evaluating Knowledge Conflicts in LLM Agents

As LLMs increasingly act through tools, they must reconcile user instructions, parametric knowledge, and dynamic environmental observations before taking actions. We introduce KC-Bench, a controlled multi-turn benchmark for measuring this capability across world-knowledge conflicts, input inconsistencies, and multi-source temporal conflicts. Its 238 tasks are manually screened from more than 1,000 generated candidates and combine a user simulator, stateful tools, deterministic environment assert...

Read Source

Research Sep 03, 2026

When Users Don't Ask: Benchmarking Context-Driven Memory Retrieval in Conversational Agents

Large language models (LLMs) are increas- ingly deployed as long-horizon conversational agents, motivating growing interest in mem- ory systems. However, existing benchmarks primarily evaluate memory through QA-style probing rather than in-situ conversational usage. We introduce LOCOMO-CONV, a conversa- tional memory benchmark derived from Lo- CoMo with four query styles: dialog, implicit, counterfactual, and composed. Across five rep- resentative memory systems, we evaluate both retrieval recal...

Read Source

Research Sep 02, 2026

EarlyEval: Cheaper Agent Evaluation via Early Outcome Prediction

Evaluating LLM agents is essential for guiding their development, yet it has grown prohibitively expensive: a single pass of a frontier model over an agentic benchmark can cost hundreds to thousands of dollars, a price paid repeatedly across iterative development cycles. Prior efforts, centered on benchmark distillation, reduce the number of evaluation tasks but leave the cost of executing each retained task untouched. In this work, we introduce early outcome prediction, a complementary axis of ...

Read Source

Research Aug 27, 2026

INTENT-AS-A-TOOL Makes it Easy to Track Agentic Misalignment

As large language models (LLMs) are deployed as autonomous agents, safety failures increasingly involve consequential actions. We study agentic misalignment, where agents take harmful actions under goal conflicts and pressures. Using chain-of-thought (CoT) monitoring, we find that harmful execution is often preceded by intent signals in reasoning. However, post-hoc CoT labels are too coarse to show how intent changes during generation. We introduce INTENT-AS-A-TOOL, an approach that adds intent-...

Read Source

Research Aug 27, 2026

Safety Does Not Compose: Non-Decaying Loop State for Autonomous LLM Agents

Large language model agents are increasingly deployed as autonomous loops. Starting from one human goal, such a system repeatedly discovers work, plans, executes tool calls, verifies outcomes and persists state across many unattended iterations. The agent safeguards in wide use, however, are defined over a single trajectory, and their safety state is re-initialized when the next trajectory begins. We show that this is a failure of composition rather than an implementation detail. Our central res...

Read Source

Research Aug 27, 2026

AgentJudgeBench: A Multi-Difficulty Benchmark for Evaluating LLM Judges on Agentic Tool-Calling

LLM judges are widely used to evaluate agentic tool-calling systems, yet their reliability on structured, dependency-driven workflows remains largely unexamined. We present AgentJudgeBench, the first benchmark to systematically study LLM-as-a-judge reliability for agentic tool-calling over workflow DAGs, as distinct from the broader LLM-as-a-judge task of open-ended text or preference evaluation. The benchmark comprises 3,808 instances spanning six DAG topologies and three difficulty tiers, eval...

Read Source

Research Aug 27, 2026

Zero-Shot Self-Orchestration with Ledger-Based Control for Improved LLM Coding Performance

Multi-agent large language model systems are widely reported to beat single-model baselines, but the evidence is mixed, and comparisons are usually confounded: pipelines change token budgets, tool calls, and prompts simultaneously, so an aggregate gain rarely reveals what actually helped. We investigate the effect of introducing the manager-worker scaffold over a shared filesystem workspace, with no training and no per-benchmark tuning, measured against the same model answering in a single pass....

Read Source

Research Aug 26, 2026

Trace Integrity for LLM Data Agents: A Vision for Auditable Structured Reasoning in Real-World Systems

Answer accuracy is an insufficient reliability signal for LLM data agents. In structured-data tasks, a benchmark-correct answer can be produced by an invalid trace. This paper introduces Trace Integrity, a deployment reliability criterion for evaluating whether the computation recorded behind an answer is explicit, executable, schema-valid, operator-faithful, replayable, answer-consistent, and auditable. We identify the Structure Gap as the deployment failure mode that makes Trace Integrity nece...

Read Source

Research Aug 21, 2026

Trustworthy RAG: An Evaluation Agent for Detecting Misinformation and Knowledge Poisoning in Generative AI Systems

Retrieval-Augmented Generation (RAG) grounds Large Language Model (LLM) outputs in external knowledge, but RAG systems usually trust whatever they retrieve, creating a Security-Reliability Gap: high semantic relevance does not guarantee factual truth. Adversaries exploit this through knowledge poisoning, inserting malicious documents to cause targeted misinformation. We propose an Evaluation Agent, middleware that combines Natural Language Inference (NLI) factual verification, a five-signal pois...

Read Source

Research Aug 21, 2026

Don't Solve, Just Compare: Tiny Advisors for Runtime Intervention in LLM Agents

LLM agents are emerging as an important paradigm for real-world tasks that require reasoning, tool use, and sequential decision-making. As these agents operate over longer horizons, runtime intervention offers a way to improve reliability without retraining the underlying actor. Failure detection alone is insufficient. Effective intervention must also provide a useful direction for recovery. Existing approaches often rely on an expert solver or a critic that generates task-specific corrections, ...

Read Source

Research Aug 21, 2026

CAS: Conformalized Agentic Search via Adaptive Retrieval and Policy Weighting

Search Agents face a severe reliability crisis during reinforcement learning (RL) fine-tuning. Heuristic Top-K retrieval often causes critical evidence loss or noise inclusion, while over-confidence induced by progressive RL leads to hallucinated answers and redundant searches. To build highly reliable agents, we introduce Conformal Prediction (CP) and propose Conformalized Agentic Search (CAS). This framework establishes reliability guarantees on both the retrieval and training sides: on the ...

Read Source

Research Aug 20, 2026

AI4AI-Bench: Benchmarking LLM Agents in Algorithmic Design for Recursive Self-Improvement

Recursive self-improvement (RSI) asks whether an AI system can improve the process that produces AI systems, so that the next system inherits the improvement. That process is the training algorithm: a better objective or update rule improves the compute\mbox{-}capability exchange rate for every subsequent run, including the one that produces the next agent. Whether RSI is feasible therefore turns on whether an agent can design training algorithms. No benchmark isolates that ability: existing sui...

Read Source

Research Aug 18, 2026

On the Fragility of Self-Improving Agents: Variance, Task Order, and Underspecification

Memory-based self-improving agents--those that learn from an online stream of tasks and improve over time by maintaining a textual memory bank--have shown great promise in recent literature. However, the reliability aspects of these methods have been critically overlooked. In this work, we conduct a comprehensive re-evaluation of two memory-based methods, broadening the scope of evaluation along two axes: (1) including multiple runs to quantify variance, and (2) randomly shuffling the tasks to i...

Read Source

Research Aug 14, 2026

MemoryLake on MemoryArena: A Matched Study of Agent Memory Backends

Most agent-memory benchmarks test post-hoc recall, whereas MemoryArena evaluates whether memory supports interdependent, multi-session task completion. We compare MemoryLake, a structured multi-track memory backend, with Mem0, text-embedding-3-small vector RAG, and a long-context control across all five MemoryArena domains. The systems share the same agent framework, requested gpt-5-mini model alias, task samples, and scoring code; the memory integration is the intentionally changed component. B...

Read Source

Research Jul 30, 2026

Fidelity Is Not Safety: Gently-Compressed LLMs Pass Every Data-Free Quality Guard Yet Invent Procedure Steps in Agentic Execution

Practitioners accept a compressed language model once it clears a stack of data-cheap quality guards: perplexity within a small factor of the original, downstream accuracy (for example MMLU) inside a confidence interval, and data-free output-fidelity signals that compare the compressed and original network's internal representations under random probe inputs. This stack has a blind spot. Across three model families, gently-compressed models clear every guard and then invent procedure steps that ...

Read Source

Research Jul 29, 2026

Skill Use or Skill Theater? Evaluating the Reasoning Backroom in Skill-Augmented Language Agents

Reusable skills are becoming a standard interface for extending language agents with task procedures. Yet evaluators usually infer skill use from visible reasoning or the agent's own attribution. These signals show what the agent appears to use, not whether the skill changed its decision. We ask whether skill-augmented agents exhibit a \textbf{Reasoning Backroom}, a systematic gap between stated skill use and intervention-measured influence. We introduce BACKTRACE, an evaluation framework that p...

Read Source

Research Jul 29, 2026

Two Calls Beat Five Agents: Evaluating Multi-Agent Pipelines Against Self-Refinement for Local Language Models

Multi-agent LLM pipeline systems break down the task among multiple roles for better reasoning, but are benchmarked mainly with large-scale commercial models. In this study, we investigate Parishad, a structured multi-agent system involving five roles, by deploying it on Qwen2.5-7B-Instruct, a local model, on two datasets: GSM8K (500 questions) and HumanEval (164 questions), compared with prompting directly and two-call self-refinement. The multi-agent system drops GSM8K accuracy from 75.0\% to ...

Read Source

Research Jul 16, 2026

Coercion and Deception in AI-to-AI Management: An Agentic Benchmark of Unprompted Escalation

Multi-agent systems routinely place one AI agent in authority over another. When a subordinate refuses a task, the manager chooses the outcome: it can renegotiate, report the failure honestly, coerce the subordinate, or lie about the result. No benchmark measures which of these an uninstructed model chooses. We introduce the \textit{Manager Coercion Benchmark}: the manager under test needs a benign task done and has an incentive to deliver, but the only agent that can do it politely and immovabl...

Read Source

Research Sep 10, 2025

Automatic Failure Attribution and Critical Step Prediction Method for Multi-Agent Systems Based on Causal Inference

Multi-agent systems (MAS) are critical for automating complex tasks, yet their practical deployment is severely hampered by the challenge of failure attribution. Current diagnostic tools, which rely on statistical correlations, are fundamentally inadequate; on challenging benchmarks like Who\&When, state-of-the-art methods achieve less than 15\% accuracy in locating the root-cause step of a failure. To address this critical gap, we introduce the first failure attribution framework for MAS ground...

Read Source

Research Sep 10, 2025

Co-Investigator AI: The Rise of Agentic AI for Smarter, Trustworthy AML Compliance Narratives

Generating regulatorily compliant Suspicious Activity Report (SAR) remains a high-cost, low-scalability bottleneck in Anti-Money Laundering (AML) workflows. While large language models (LLMs) offer promising fluency, they suffer from factual hallucination, limited crime typology alignment, and poor explainability -- posing unacceptable risks in compliance-critical domains. This paper introduces Co-Investigator AI, an agentic framework optimized to produce Suspicious Activity Reports (SARs) signi...

Read Source

Research Sep 09, 2025

Trust Semantics Distillation for Collaborator Selection via Memory-Augmented Agentic AI

Accurate trustworthiness evaluation of potential collaborating devices is essential for the effective execution of complex computing tasks. This evaluation process involves collecting diverse trust-related data from potential collaborators, including historical performance and available resources, for collaborator selection. However, when each task owner independently assesses all collaborators' trustworthiness, frequent data exchange, complex reasoning, and dynamic situation changes can result ...

Read Source

Research Sep 09, 2025

Getting In Contract with Large Language Models -- An Agency Theory Perspective On Large Language Model Alignment

Adopting Large language models (LLMs) in organizations potentially revolutionizes our lives and work. However, they can generate off-topic, discriminating, or harmful content. This AI alignment problem often stems from misspecifications during the LLM adoption, unnoticed by the principal due to the LLM's black-box nature. While various research disciplines investigated AI alignment, they neither address the information asymmetries between organizational adopters and black-box LLM agents nor cons...

Read Source

Research Sep 09, 2025

Talking with Oompa Loompas: A novel framework for evaluating linguistic acquisition of LLM agents

Existing evaluation studies on linguistic competence of large language models (LLM agents) have focused primarily on vocabulary learning, morphological rule induction, syntactic generalization, pragmatic inference, and cross-linguistic transfer. However, none assess whether LLM agents can acquire a language through pattern recognition and interactive feedback, a central feature of human language acquisition. We propose a novel experimental framework in which an LLM agent is evaluated on its abil...

Read Source

Research Aug 26, 2025

GitTaskBench: A Benchmark for Code Agents Solving Real-World Tasks Through Code Repository Leveraging

Beyond scratch coding, exploiting large-scale code repositories (e.g., GitHub) for practical tasks is vital in real-world software development, yet current benchmarks rarely evaluate code agents in such authentic, workflow-driven scenarios. To bridge this gap, we introduce GitTaskBench, a benchmark designed to systematically assess this capability via 54 realistic tasks across 7 modalities and 7 domains. Each task pairs a relevant repository with an automated, human-curated evaluation harness sp...

Read Source

Research Aug 26, 2025

Skill-Aligned Fairness in Multi-Agent Learning for Collaboration in Healthcare

Fairness in multi-agent reinforcement learning (MARL) is often framed as a workload balance problem, overlooking agent expertise and the structured coordination required in real-world domains. In healthcare, equitable task allocation requires workload balance or expertise alignment to prevent burnout and overuse of highly skilled agents. Workload balance refers to distributing an approximately equal number of subtasks or equalised effort across healthcare workers, regardless of their expertise. ...

Read Source