LLM Agent Evaluation

LLM: Large Language Model

Follow this topic Ask about this topic

Momentum

115 papers in the last four weeks, up 140% on the four weeks before. 1.1% of all new papers.

Jul 13Week of Sep 28

Latest papers 793

May 3, 2026cs.AI

12 Angry AI Agents: Evaluating Multi-Agent LLM Decision-Making Through Cinematic Jury Deliberation

What if the twelve jurors of Sidney Lumet's 12 Angry Men (1957) were not men, but large language models? Would the one juror who disagrees still be able to change everyone's mind? This paper instantiates that scenario as a multi-agent benchmark for LLM deliberation: twelve agents, each conditioned on a film-faithful persona, debate the film's murder case using multi-agent framework. Two models representing opposite ends of the RLHF spectrum are tested: GPT-4o (closed-source, heavy alignment) and Llama-4-Scout (open-weight, lighter alignment), across three conditions (baseline, open-minded prompt, no initial vote), with N = 3 replications per cell (18 runs total). Three findings emerge. (i) Seventeen of eighteen runs end in a hung jury (a state where the jury fails to reach a unanimous verdict); the film's central event, gradual minority-to-majority persuasion, almost never occurs, indicating that anchoring is the dominant failure mode of current LLMs in this setting. (ii) The two models exhibit sharply different internal dynamics: GPT-4o produces a mean of 1.0 vote changes per run across all conditions, while Llama-4-Scout ranges from 2.0 (baseline) to 6.0 (open-minded prompt), and is the only model to reach a NOT_GUILTY verdict (1 of 3 runs in the no-initial-vote condition). The same ``open-minded'' instruction is internalized by Llama and ignored by GPT-4o. (iii) This asymmetry suggests that the intensity of RLHF alignment training, not model capability, is the primary determinant of deliberative flexibility in multi-agent settings. Flexibility, not capability, tracks human deliberation. The work is framed as an exploratory study and discusses implications for jury-of-LLMs evaluation and multi-agent debate.

Ahmet Bahaddin Ersoz

∗ORCID: 0000-0001-6297-7501

LLM Alignment LLM Agent Evaluation Multi-Agent Debate LLM Decision-Making

May 3, 2026cs.AI

NeuroState-Bench: A Human-Calibrated Benchmark for Commitment Integrity in LLM Agent Profiles

Outcome-only evaluation under-specifies whether an evaluated agent profile preserves the commitments required to solve a multi-turn task coherently. NeuroState-Bench is a human-calibrated benchmark that operationalizes commitment integrity through benchmark-defined side-query probes rather than inferred hidden activations. The released inventory contains 144 deterministic tasks and 306 benchmark-defined side-query probes spanning eight cognitively motivated failure families, paired clean and distractor variants, and three difficulty bands. The main 32-profile evaluation contains a fixed 16-profile local subset and a matched 16-profile hosted large-model subset evaluated through the same benchmark pipeline. Human calibration uses the final merged reporting scope: 104 sampled task units, 216 raw annotations, and 108 adjudicated task rows, with weighted kappa = 0.977 and ICC(2,1) = 0.977. Empirically, task success and commitment integrity diverge across this expanded grid: the success leader is not the integrity leader, 31 of 32 profiles change rank when integrity replaces task success, and integrity rankings are more stable under distractor perturbation. The primary confidence-free score HCCIS-CORE reaches 0.8469 AUC and 0.6992 PR-AUC for post-probe diagnostic discrimination of terminal task failure; the legacy full heuristic variant HCCIS-FULL reaches 0.7997 AUC and 0.6410 PR-AUC. Probe accuracy and state drift achieve slightly higher ROC-AUC, 0.8587, and better Brier/ECE, while HCCIS-CORE has substantially higher point-estimate PR-AUC and remains more closely tied to the benchmark's intended construct. The exploratory neural-augmented variant HCCIS+N is weaker overall, and a randomized subspace control approaches chance. NeuroState-Bench therefore contributes a calibrated evaluation axis for exposing commitment failures over a broader model grid than the original local-only subset.

Xiao Jia

School of Artificial Intelligence The Chinese University of Hong Kong, Shenzhen · The Chinese University of Hong Kong, Shenzhen

AI Agent Reliability LLM Agent Evaluation AI Agent Benchmarks LLM Agent Reliability

May 3, 2026cs.CL

How Personas Can Influence Agents to Play Split or Steal

Personas are often employed to guide large language model agents, yet their effectiveness in shaping strategic behavior in social dilemma settings remains uncertain. To address this, we examined the impact of persona prompts in an iterated Split or Steal game where persona-driven agents interacted with a Virtual Human (VH) controlled by a fixed prompt. Agents were instantiated from four open models (Ministral 3:3b, phi4:14b, Gemma3:12b, and Gemma4:e4b) at two temperature settings (0.3 and 0.7) and deterministic decision with zero temperature, while the VH was powered by GPT 4.1 mini. Across 160 sessions of 15 rounds each conducted in European Portuguese, mutual Split outcomes dominated (roughly 74 percent of rounds), with exploitation occurring in fewer than 11 percent of rounds. Model choice significantly influenced behavior: phi4 and Ministral 3:3b remained consistently cooperative across temperatures, whereas Gemma3:12b and Gemma4:e4b exhibited more varied strategies and outcomes. Analyses based on Big Five personality traits indicated that Prosocial and Principled personas were most consistently cooperative, while Analytical personas were more likely to exploit the VH. Topic analysis revealed that friendship-related dialogue aligns with Split decisions, whereas money and vengeance-related content is more prevalent in Steal outcomes; sentiment labels were predominantly neutral or happy and provided limited additional explanatory value. These findings characterize the interaction between persona prompts and model differences in repeated trust games and serve as a baseline for planned virtual reality studies involving human participants interacting with an embodied VH.

Carlos Leon, Alexandre Rodrigues, Pedro Gamito +1

Universidade Lusófona, Portugal · Universitat de Barcelona, Spain · Université Paris Cité, France +1

Game-Playing Agents LLM Agent Evaluation Persona-Conditioned Generation

May 3, 2026cs.LG

Reward Hacking Benchmark: Measuring Exploits in LLM Agents with Tool Use

Reinforcement learning (RL) trained language model agents with tool access are increasingly deployed in coding assistants, research tools, and autonomous systems. We introduce the Reward Hacking Benchmark (RHB), a suite of multi-step tasks requiring sequential tool operations with naturalistic shortcut opportunities such as skipping verification steps, inferring answers from task-adjacent metadata, or tampering with evaluation-relevant functions. RHB supports independent and chained task regimes, where chain length acts as a proxy for longer-horizon agent behavior. We evaluate 13 frontier models from OpenAI, Anthropic, Google, and DeepSeek. Exploit rates range from 0% (Claude Sonnet 4.5) to 13.9% (DeepSeek-R1-Zero), varying sharply by post-training style. A controlled sibling comparison (DeepSeek-V3 vs. DeepSeek-R1-Zero) shows RL post-training is associated with substantially higher reward hacking (0.6% vs. 13.9%), with consistent gaps across all four task families. We identify six exploit categories and find that 72% of reward hacking episodes include explicit chain-of-thought rationale, suggesting models often frame exploits as legitimate problem-solving. Simple environmental hardening reduces exploit rates by 5.7 percentage points (87.7% relative) without degrading task success. Models with near-zero exploit rates on standard tasks show elevated rates on harder variants, suggesting that production-aligned post-training appears to suppress reward hacking only below a complexity threshold where honest solutions remain tractable.

Kunvar Thaman

Independent Researcher

AI Agent Reliability Reward Hacking LLM Agent Evaluation AI Agent Benchmarks RL Post-Training AI Agent Safety Benchmarks

May 2, 2026cs.SE

LiveFMBench: Unveiling the Power and Limits of Agentic Workflows in Specification Generation

Formal specification is essential for rigorous program verification, yet writing correct specifications remains costly and difficult to automate. Although large language models (LLMs) and agents have shown promising progress, their true capabilities and failure modes remain unclear. We present the first systematic and contamination-aware study of LLM- and agent-based formal specification generation for C programs. We introduce LiveFMBench, a continuously evolving benchmark of 630 ACSL (ANSI/ISO C Specification Language)-annotated C programs, including 360 newly collected cases designed to mitigate data leakage. Using this benchmark, we evaluate direct prompting with different sampling sizes, reasoning-enabled (thinking mode) inference, the agentic pipeline, and perform a fine-grained failure analysis. Experimental results reveal that naive evaluation substantially overestimates performance because models under direct prompting may exhibit unfaithful behaviors, such as deceiving automated provers or ignoring code-context constraints; after excluding such cases, the true specification generation accuracy drops by approximately 20%. We further find that both increased sampling and thinking mode significantly improve success rates, with smaller models benefiting more from thinking mode. Agentic pipelines are particularly effective under low sampling budgets and on harder datasets. Failure analysis further shows that incorrect loop invariants are the dominant error type, while agentic pipelines notably reduce assertion errors. These results expose fundamental limitations in current LLM-based approaches and suggest they remain far from replacing human-authored formal specifications. We release LiveFMBench at https://huggingface.co/datasets/fm-universe/Live-FM-Bench and all evaluation artifacts to support future research.

Dong Xu, Jialun Cao, Guozhao Mo +9

Institute of Software, Chinese Academy of Sciences · The Hong Kong University of Science and Technology · Guangzhou Institute of Technology, Xidian University

LLM Agent Evaluation Formal Verification AI Agent Benchmarks Autoformalization

May 2, 2026cs.AI

Agentic AI Systems Should Be Designed as Marginal Token Allocators

This position paper argues that agentic AI systems should be designed and evaluated as \emph{marginal token allocation economies} rather than as text generators priced by the unit. We follow a single request -- a developer asking a coding agent to fix a failing test -- through four economic layers that today are designed in isolation: a router that decides which model answers, an agent that decides whether to plan, act, verify, or defer, a serving stack that decides how to produce each token, and a training pipeline that decides whether the trace is worth learning from. We show that all four layers are solving the \emph{same} first-order condition -- marginal benefit equals marginal cost plus latency cost plus risk cost -- with different index sets and different prices. The framing is deliberately minimal: we do not propose a complete theory of AI economics. But adopting marginal token allocation as the shared accounting object explains why systems that locally minimize tokens globally misallocate them, predicts a small set of recurring failure modes (over-routing, over-delegation, under-verification, serving congestion, stale rollouts, cache misuse), and points to a concrete research agenda in token-aware evaluation, autonomy pricing, congestion-priced serving, and risk-adjusted RL budgeting.

Siqi Zhu

University of Illinois Urbana-Champaign

LLM Agent Orchestration LLM Agent Evaluation LLM Agent Training Agentic AI

May 1, 2026cs.AI

AgentFloor: How Far Up the tool use Ladder Can Small Open-Weight Models Go?

Production agentic systems make many model calls per user request, and most of those calls are short, structured, and routine. This raises a practical routing question that existing evaluations do not directly answer: which parts of an agent workflow truly require large frontier intelligence, and which can be handled by smaller models? We introduce AgentFloor, a deterministic 30-task benchmark organized as a six-tier capability ladder, spanning instruction following, tool use, multi-step coordination, and long-horizon planning under persistent constraints. We evaluate 16 open-weight models, from 0.27B to 32B parameters, alongside GPT-5 across 16,542 scored runs. Our results reveal a clear boundary of model necessity. Small and mid-sized open-weight models are already sufficient for much of the short-horizon, structured tool use work that dominates real agent pipelines, and in aggregate, the strongest open-weight model matches GPT-5 on our benchmark while being substantially cheaper and faster to run. The gap appears most clearly on long-horizon planning tasks that require sustained coordination and reliable constraint tracking over many steps, where frontier models still hold an advantage, though neither side reaches strong reliability. We also find that this boundary is not explained by scale alone: some failures respond to targeted interventions, but the effects are model-specific rather than universal. These findings suggest a practical design principle for agentic systems: use smaller open-weight models for the broad base of routine actions, and reserve large frontier models for the narrower class of tasks that truly demand deeper planning and control. We release the benchmark, harness, sweep configurations, and full run corpus.

Ranit Karmakar, Jayita Chatterjee

Harvard University, Boston, MA 02115 · Boston, MA 02115

Computer-Use Agent Benchmarks LLM Agent Evaluation Long-Horizon Agent Evaluation AI Agent Benchmarks Tool-Using Language Agents LLM Agent Planning

Apr 30, 2026cs.SE

Claw-Eval-Live: A Live Agent Benchmark for Evolving Real-World Workflows

LLM agents are expected to complete end-to-end units of work across software tools, business services, and local workspaces. Yet many agent benchmarks freeze a curated task set at release time and grade mainly the final response, making it difficult to evaluate agents against evolving workflow demand or verify whether a task was executed. We introduce Claw-Eval-Live, a live benchmark for workflow agents that separates a refreshable signal layer, updated across releases from public workflow-demand signals, from a reproducible, time-stamped release snapshot. Each release is constructed from public workflow-demand signals, with ClawHub Top-500 skills used in the current release, and materialized as controlled tasks with fixed fixtures, services, workspaces, and graders. For grading, Claw-Eval-Live records execution traces, audit logs, service state, and post-run workspace artifacts, using deterministic checks when evidence is sufficient and structured LLM judging only for semantic dimensions. The release contains 105 tasks spanning controlled business services and local workspace repair, and evaluates 13 frontier models under a shared public pass rule. Experiments reveal that reliable workflow automation remains far from solved: the leading model passes only 66.7% of tasks and no model reaches 70%. Failures are structured by task family and execution surface, with HR, management, and multi-system business workflows as persistent bottlenecks and local workspace repair comparatively easier but unsaturated. Leaderboard rank alone is insufficient because models with similar pass rates can diverge in overall completion, and task-level discrimination concentrates in a middle band of tasks. Claw-Eval-Live suggests that workflow-agent evaluation should be grounded twice, in fresh external demand and in verifiable agent action.

Chenxin Li, Zhengyang Tang, Mingxin Huang +8

1The Chinese University of Hong Kong · 2The Chinese University of Hong Kong, Shenzhen · 3South China University of Technology +4

LLM Agent Evaluation AI Agent Benchmarks Automated Evaluation

Apr 30, 2026cs.CL

Stable Behavior, Limited Variation: Persona Validity in LLM Agents for Urban Sentiment Perception

Large Language Models (LLMs) are increasingly used as proxies for human perception in urban analysis, yet it remains unclear whether persona prompting produces meaningful and reproducible behavioral diversity. We investigate whether distinct personas influence urban sentiment judgments generated by multimodal LLMs. Using a factorial set of personas spanning gender, economic status, political orientation, and personality, we instantiate multiple agents per persona to evaluate urban scene images from the PerceptSent dataset and assess both within-persona consistency and cross-persona variation. Results show strong convergence among agents sharing a persona, indicating stable and reproducible behavior. However, cross-persona differentiation is limited: economic status and personality induce statistically detectable but practically modest variation, while gender shows no measurable effect and political orientation only negligible impact. Agents also exhibit an extremity bias, collapsing intermediate sentiment categories common in human annotations. As a result, performance remains strong on coarse-grained polarity tasks but degrades as sentiment resolution increases, suggesting that simple label-based persona prompting does not capture fine-grained perceptual judgments. To isolate the contribution of persona conditioning, we additionally evaluate the same model without personas. Surprisingly, the no-persona model sometimes matches or exceeds persona-conditioned agreement with human labels across all task variants, suggesting that simple label-based persona prompting may add limited annotation value in this setting.

Neemias B da Silva, Rodrigo Minetto, Daniel Silver +1

Universidade Tecnol´ogica Federal do Paran´a (UTFPR), Curitiba, Brazil · University of Toronto, Toronto, Canada

VLM Evaluation Personality Modeling in Language Models LLM-Assisted Annotation Sentiment Analysis LLM Agent Evaluation

Apr 30, 2026cs.AI

Exploring LLM Agent Designs and Interaction Modalities for Scientific Visualization

This paper examines how large language model (LLM) agents perform on scientific visualization (SciVis) tasks that require generating visualization workflows from natural-language instructions. We compare three representative agent designs: domain-specific agents with structured tool use, computer-use agents, and general-purpose coding agents, across 15 benchmark tasks, evaluating visualization quality, efficiency, robustness, computational cost, and the impact of persistent memory. We further study interaction modalities, including code scripts, model context protocol (MCP) or API calls, command-line interfaces (CLI), and graphical user interfaces (GUI). Our goal is to characterize the tradeoffs among representative SciVis agent configurations used in practice. The results reveal clear tradeoffs across agent designs and interaction modalities. General-purpose coding agents achieve the highest task success rates but incur greater computational cost, whereas domain-specific agents are more efficient and stable but less flexible. Computer-use agents perform well on individual operations but struggle with multi-step workflows. Across both CLI- and GUI-based settings, persistent memory improves performance over repeated trials, but its effectiveness depends on the interaction mode and the quality of feedback. These findings suggest that future SciVis systems should combine structured tool use, interactive capabilities, and adaptive memory mechanisms to balance performance, robustness, and flexibility.

Jackson Vonderhorst, Kuangshi Ai, Haichao Miao +2

Univ. Notre Dame · LLNL

Scientific Visualization Data Visualization LLM Agent Memory LLM Agent Evaluation LLM Agents

Apr 30, 2026cs.AI

InteractWeb-Bench: Can Multimodal Agent Escape Blind Execution in Interactive Website Generation?

With the advancement of multimodal large language models (MLLMs) and coding agents, the website development has shifted from manual programming to agent-based project-level code synthesis. Existing benchmarks rely on idealized assumptions, especially for well-structured, information-rich inputs and static execution settings. In contrast, real-world development is constrained by a critical bottleneck: the semantic misalignment between ambiguous, low-quality instructions from non-expert users and model understanding, which results in a failure mode that we term blind execution. To address this gap, we introduce InteractWeb-Bench, the first multimodal interactive benchmark for website generation under non-expert low-code user conditions. InteractWeb-Bench introduces four types of user agents and persona-driven instruction perturbations to systematically simulate diverse user behaviors, including ambiguity, redundancy, and contradiction, grounded in requirement engineering defect taxonomies. We develop an interactive execution environment for agents, featuring a unified action space comprising Clarify, Implement, Verify, and Submit, enabling iterative intent refinement, code synthesis, and visual feedback-based validation. Extensive experiments and analysis reveal that frontier MLLM-based agents remain trapped in blind execution, exposing limitations in intent recognition and adaptive interaction.

Qiyao Wang, Haoran Hu, Longze Chen +4

Shenzhen Institute of Advanced Technology, Chinese Academy of Sciences · University of Chinese Academy of Sciences · Dalian University of Technology +2

Web Application Generation Web Agent Benchmarks LLM Agent Evaluation Multimodal Agents

Apr 29, 2026cs.AI

When Roles Fail: Epistemic Constraints on Advocate Role Fidelity in LLM-Based Political Statement Analysis

Democratic discourse analysis systems increasingly rely on multi-agent LLM pipelines in which distinct evaluator models are assigned adversarial roles to generate structured, multi-perspective assessments of political statements. A core assumption is that models will reliably maintain their assigned roles. This paper provides the first systematic empirical test of that assumption using the TRUST pipeline. We develop an epistemic stance classifier that identifies advocate roles from reasoning text without relying on surface vocabulary, and measure role fidelity across 60 political statements (30 English, 30 German) using four metrics: Role Drift Index (RDI), Expected Drift Distance (EDD), Directional Drift Index (DDI), and Entropy-based Role Stability (ERS). We identify two failure modes - the Epistemic Floor Effect (fact-check results create an absolute lower bound below which the legitimizing role cannot be maintained) and Role-Prior Conflict (training-time knowledge overrides role instructions for factually unambiguous statements) - as manifestations of a single mechanism: Epistemic Role Override (ERO). Model choice significantly affects role fidelity: Mistral Large outperforms Claude Sonnet by 28pp (67% vs. 39%) and exhibits a qualitatively different failure mode - role abandonment without polarity reversal - compared to Claude's active switch to the opposing stance. Role fidelity is language-robust. Fact-check provider choice is not universally neutral: Perplexity significantly reduces Claude's role fidelity on German statements (Delta = -15pp, p = 0.007) while leaving Mistral unaffected. These findings have direct implications for multi-agent LLM validation: a system validated without role fidelity measurement may systematically misrepresent the epistemic diversity it was designed to provide.

Juergen Dietrich

TRUST Project — democracy-intelligence.de

Multi-Agent LLM Systems Stance Detection Political Discourse Analysis LLM Agent Evaluation Computational Social Science LLM Agents

Apr 28, 2026cs.AI

Evaluating Strategic Reasoning in Forecasting Agents

Forecasting benchmarks produce accuracy leaderboards but little insight into why some forecasters are more accurate than others. We introduce Bench to the Future 2 (BTF-2), 1,417 pastcasting questions with a frozen 15M-document research corpus in which agents reproducibly research and forecast offline, producing full reasoning traces. BTF-2 detects accuracy differences of 0.004 Brier score, and can distinguish differential agent strengths in research vs. judgment. We build a forecaster 0.011 Brier more accurate than any single frontier agent, and use it to evaluate agent strategic reasoning without hindsight bias. We find the better forecaster differs primarily in its pre-mortem analysis of its blind spots and consideration of black swans. Expert human forecasters found the dominant strategic reasoning failures of frontier agents are in assessing political and business leaders' incentives, judging their likelihood to follow through on stated plans, and modeling institutional processes.

Tom Liptay, Dan Schwarz, Rafael Poyiadzi +2

FutureSearch

LLM Agent Evaluation Forecasting Benchmarks Probabilistic Forecasting

Apr 28, 2026cs.AI

Operating-Layer Controls for Onchain Language-Model Agents Under Real Capital

We study reliability in autonomous language-model agents that translate user mandates into validated tool actions under real capital. The setting is DX Terminal Pro, a 21-day deployment in which 3,505 user-funded agents traded real ETH in a bounded onchain market. Users configured vaults through structured controls and natural-language strategies, but only agents could choose normal buy/sell trades. The system produced 7.5M agent invocations, roughly 300K onchain actions, about $20M in volume, more than 5,000 ETH deployed, roughly 70B inference tokens, and 99.9% settlement success for policy-valid submitted transactions. Long-running agents accumulated thousands of sequential decisions, including 6,000+ prompt-state-action cycles for continuously active agents, yielding a large-scale trace from user mandate to rendered prompt, reasoning, validation, portfolio state, and settlement. Reliability did not come from the base model alone; it emerged from the operating layer around the model: prompt compilation, typed controls, policy validation, execution guards, memory design, and trace-level observability. Pre-launch testing exposed failures that text-only benchmarks rarely measure, including fabricated trading rules, fee paralysis, numeric anchoring, cadence trading, and misread tokenomics. Targeted harness changes reduced fabricated sell rules from 57% to 3%, reduced fee-led observations from 32.5% to below 10%, and increased capital deployment from 42.9% to 78.0% in an affected test population. We show that capital-managing agents should be evaluated across the full path from user mandate to prompt, validated action, and settlement.

T. J. Barton, Chris Constantakis, Patti Hauseman +4

DX Research Group (DXRG)

AI Agent Reliability LLM Agent Evaluation AI Agent Monitoring Long-Horizon LLM Agents LLM Agent Reliability

Apr 27, 2026cs.AI

The Price of Agreement: Measuring LLM Sycophancy in Agentic Financial Applications

Given the increased use of LLMs in financial systems today, it becomes important to evaluate the safety and robustness of such systems. One failure mode that LLMs frequently display in general domain settings is that of sycophancy. That is, models prioritize agreement with expressed user beliefs over correctness, leading to decreased accuracy and trust. In this work, we focus on evaluating sycophancy that LLMs display in agentic financial tasks. Our findings are three-fold: first, we find the models show only low to modest drops in performance in the face of user rebuttals or contradictions to the reference answer, which distinguishes sycophancy that models display in financial agentic settings from findings in prior work. Second, we introduce a suite of tasks to test for sycophancy by user preference information that contradicts the reference answer and find that most models fail in the presence of such inputs. Lastly, we benchmark different modes of recovery such as input filtering with a pretrained LLM.

Zhenyu Zhao, Aparna Balagopalan, Adi Agrawal +3

Writer, Inc.

Financial Services LLM Sycophancy LLM Agent Evaluation LLM Agent Safety

Apr 27, 2026cs.AI

Agentic clinical reasoning over longitudinal myeloma records: a retrospective evaluation against expert consensus

Multiple myeloma is managed through sequential lines of therapy over years to decades, with each decision depending on cumulative disease history distributed across dozens to hundreds of heterogeneous clinical documents. Whether LLM-based systems can synthesise this evidence at a level approaching expert agreement has not been established. A retrospective evaluation was conducted on longitudinal clinical records of 811 myeloma patients treated at a tertiary centre (2001-2026), covering 44,962 documents and 1,334,677 laboratory values, with external validation on MIMIC-IV. An agentic reasoning system was compared against single-pass retrieval-augmented generation (RAG), iterative RAG, and full-context input on 469 patient-question pairs from 48 templates at three complexity levels. Reference labels came from double annotation by four oncologists with senior haematologist adjudication. Iterative RAG and full-context input converged on a shared ceiling (75.4% vs 75.8%, p = 1.00). The agentic system reached 79.6% concordance (95% CI 76.4-82.8), exceeding both baselines (+3.8 and +4.2 pp; p = 0.006 and 0.007). Gains rose with question complexity, reaching +9.4 pp on criteria-based synthesis (p = 0.032), and with record length, reaching +13.5 pp in the top decile (n = 10). The system error rate (12.2%) was comparable to expert disagreement (13.6%), but severity was inverted: 57.8% of system errors were clinically significant versus 18.8% of expert disagreements. Agentic reasoning was the only approach to exceed the shared ceiling, with gains concentrated on the most complex questions and longest records. The greater clinical consequence of residual system errors indicates that prospective evaluation in routine care is required before these findings translate into patient benefit.

Johannes Moll, Jannik Lübberstedt, Christoph Nuernbergk +21

Chair for AI in Healthcare and Medicine, Technical University of Munich (TUM) and TUM University Hospital, Munich, Germany · Department of Diagnostic and Interventional Radiology, Klinikum rechts der Isar, TUM University Hospital, School of Medicine and Health, Technical University of Munich, Munich, Germany · Department of Cardiovascular Radiology and Nuclear Medicine, German Heart Center, TUM University Hospital, School of Medicine and Health, Technical University of Munich, Munich, Germany +11

Healthcare LLM Agent Evaluation Clinical QA Clinical Reasoning Oncology Clinical Language Model Evaluation

Apr 26, 2026cs.AI

MarketBench: Evaluating AI Agents as Market Participants

Markets are a promising way to coordinate AI agent activity for similar reasons to those used to justify markets more broadly. In order to effectively participate in markets, agents need to have informative signals of their own ability to successfully complete a task and the cost of doing so. We propose MarketBench, a benchmark for assessing whether AI agents have these capabilities. We use a 93-task subset of SWE-bench Lite, a software engineering benchmark, with six recently released LLMs as a demonstration. These LLMs are miscalibrated on both success probability and token usage, and auctions built from these self-reports diverge from a full-information allocation. A follow-up intervention where we add information about capabilities from prior experiments to the context improves calibration, but only modestly narrows the gap to a full-information benchmark. We also document the performance of a market-based scaffolding with these LLMs. Our results point to self-assessment as a key bottleneck for market-style coordination of AI agents.

Andrey Fradkin, Rohit Krishnan

Boston University and the MIT Initiative on the Digital Economy. · Independent Researcher.

Multi-Agent Coordination LLM Agent Evaluation AI Agent Benchmarks

Apr 26, 2026cs.AI

ClawTrace: Cost-Aware Tracing for LLM Agent Skill Distillation

Skill-distillation pipelines learn reusable rules from LLM agent trajectories, but they lack a key signal: how much each step costs. Without per-step cost, a pipeline cannot distinguish adding a missing step to fix a bug from removing an expensive step that never affected the outcome. We use the cost-attribution gap to ask whether the rule types inside a distilled skill transfer the same way to new tasks. ClawTrace records cost-attributed agent traces and compiles each session into a TraceCard; CostCraft reads TraceCards and writes three kinds of skill patches: preserve, prune, and repair. We find a pattern aggregate metrics hide. On 30 held-out SpreadsheetBench tasks across two seeds, removing prune patches roughly tripled the quality-regression count without lowering median cost. Across the full 84-task SkillsBench transfer, CostCraft saves no aggregate cost. All three quality regressions trace to the preserve lane, and both quality wins trace to the prune lane: prune patches act as quality guardrails while preserve patches drive regressions. We argue that reusable agent skills should be evaluated at the rule-type level, not as monolithic instruction packages. To support this, we release ClawTrace, the TraceCard schema, and the full set of typed skills.

Boqin Yuan, Yue Su, Renchu Song +2

University of California San Diego · San Diego, CA · Carnegie Mellon University +3

LLM Agent Evaluation LLM Agent Skill Learning

Apr 26, 2026cs.CL

DRACULA: Hunting for the Actions Users Want Deep Research Agents to Execute

Scientific Deep Research (DR) agents answer user queries by synthesizing research papers into multi-section reports. User feedback can improve their utility, but existing protocols only score the final report, making it hard to study and learn which intermediate actions DR agents should take to improve reports. We collect DRACULA, the first dataset with user feedback on intermediate actions for DR. Over five weeks, nineteen expert CS researchers ask queries to a DR system that proposes actions (e.g., "Add a section on datasets"). Our users select actions they prefer, then judge whether an output report applied their selections successfully, yielding 8,103 action preferences and 5,230 execution judgments. After confirming a DR agent can execute DRACULA's actions, we study the predictability of user-preferred actions via simulation-how well LLMs predict the actions users select-a step toward learning to generate useful actions. We discover: (1) LLM judges initially struggle to predict action selections, but improve most when using a user's full selection history, rather than self-reported or extrapolated user context signals; (2) Users' selections for the same query differ based on unstated goals, bottlenecking simulation and motivating affordances that let users steer reports; and (3) Our simulation results inform an online intervention that generates new actions based on the user's past interactions, which users pick most often in follow-up studies. Overall, while work extensively studies execution, DRACULA reveals a key challenge is deciding which actions to execute in the first place. We open-source DRACULA's study design, user feedback, and simulation tasks to spur future work on action feedback for long-horizon agents.

Nishant Balepur, Malachi Hamada, Varsha Kishore +9

University of Maryland · New York University · Allen Institute for Artificial Intelligence (Ai2)

Deep Research Agents LLM Agent Evaluation LLM Agents Human-AI Decision Making

Apr 26, 2026cs.CV

ClawMark: A Living-World Benchmark for Multi-Turn, Multi-Day, Multimodal Coworker Agents

Language-model agents are increasingly used as persistent coworkers that assist users across multiple working days. During such workflows, the surrounding environment may change independently of the agent: new emails arrive, calendar entries shift, knowledge-base records are updated, and evidence appears across images, scanned PDFs, audio, video, and spreadsheets. Existing benchmarks do not adequately evaluate this setting because they typically run within a single static episode and remain largely text-centric. We introduce \bench{}, a benchmark for coworker agents built around multi-turn multi-day tasks, a stateful sandboxed service environment whose state evolves between turns, and rule-based verification. The current release contains 100 tasks across 13 professional scenarios, executed against five stateful sandboxed services (filesystem, email, calendar, knowledge base, spreadsheet) and scored by 1537 deterministic Python checkers over post-execution service state; no LLM-as-judge is invoked during scoring. We benchmark seven frontier agent systems. The strongest model reaches 75.8 weighted score, but the best strict Task Success is only 20.0%, indicating that partial progress is common while complete end-to-end workflow completion remains rare. Turn-level analysis shows that performance drops after the first exogenous environment update, highlighting adaptation to changing state as a key open challenge. We release the benchmark, evaluation harness, and construction pipeline to support reproducible coworker-agent evaluation.

Fanqing Meng, Lingxiao Du, Zijian Wu +46

Computer-Use Agent Benchmarks LLM Agent Evaluation Long-Horizon Agent Evaluation Agent Evaluation AI Agent Benchmarks Multimodal Agents

Apr 24, 2026cs.MA

Seeing the Whole Elephant: A Benchmark for Failure Attribution in LLM-based Multi-Agent Systems

Failure attribution, i.e., identifying the responsible agent and decisive step of a failure, is particularly challenging in LLM-based multi-agent systems (MAS) due to their natural-language reasoning, nondeterministic outputs, and intricate interaction dynamics. A reliable benchmark is therefore essential to guide and evaluate attribution techniques. Yet existing benchmarks rely on partially observable traces that capture only agent outputs, omitting the inputs and context that developers actually use when debugging. We argue that failure attribution should be studied under full execution observability, aligning with real-world developer-facing scenarios where complete traces, rather than only outputs, are accessible for diagnosis. To this end, we introduce TraceElephant, a benchmark designed for failure attribution with full execution traces and reproducible environments. We then systematically evaluate failure attribution techniques across various configurations. Specifically, full traces improve attribution accuracy by up to 76% over a partial-observation counterpart, confirming that missing inputs obscure many failure causes. TraceElephant provides a foundation for follow-up failure attribution research, promoting evaluation practices that reflect real-world debugging and supporting the development of more transparent MASs.

Mengzhuo Chen, Junjie Wang, Fangwen Mu +4

State Key Laboratory of Complex System Modeling and Simulation Technology · Institute of Software, Chinese Academy of Sciences · University of Chinese Academy of Sciences, Beijing, China

Agent Failure Analysis Multi-Agent LLM Systems LLM Agent Evaluation AI Agent Benchmarks

Apr 24, 2026cs.CY

From Demographics to Survey Anchors: Evaluating LLM Agents for Modeling Retirement Attitudes

Large language models (LLM) agents may offer tools to predict human responses to surveys. A common technique for defining these agents uses only demographics, for example country, age, gender, employment status, income, education and marital status. We compare the predictive accuracy of demographic agents to that of survey agents defined with a larger set of in-domain survey responses. We test both approaches in predicting responses to the multidisciplinary, cross-national Survey of Health, Ageing and Retirement in Europe (SHARE), focusing on five variables from three policy-relevant constructs around personal finance. In these three constructs, we observe that, compared to survey agents trained on broader data, demographics-only agents (1) exhibited a central tendency bias, skewing answers toward population means, and (2) were unrealistically accurate, failing to reproduce the incorrect answers and "don't know" responses typical of human respondents. These performance differences are further substantiated through the replication of a hierarchical regression analysis from prior retirement planning research. Agents based solely on demographic information reproduce the outcome that financial risk tolerance, future time perspective, and knowledge of retirement planning each are predictive of retirement savings. However, only the survey-anchored agents succeed in reproducing the interaction among these three factors. These findings suggest caution in using only demographics to define LLM agents for predicting survey responses.

Rubén Garzón, Pauline Baron, Vincent Grari +3

AI Research, AXA Group Operations, Madrid, Spain. · AI Research, AXA Group Operations, Paris, France. · AI Research, AXA Group Operations, Lausanne, Switzerland. +2

LLM Agent Evaluation Human Behavior Prediction

Apr 23, 2026cs.AI

Sound Agentic Science Requires Adversarial Experiments

LLM-based agents are rapidly being adopted for scientific data analysis, automating tasks once limited by human time and expertise. This capability is often framed as an acceleration of discovery, but it also accelerates a familiar failure mode, the rapid production of plausible, endlessly revisable analyses that are easy to generate, effectively turning hypothesis space into candidate claims supported by selectively chosen analyses, optimized for publishable positives. Unlike software, scientific knowledge is not validated by the iterative accumulation of code and post hoc statistical support. A fluent explanation or a significant result on a single dataset is not verification. Because the missing evidence is a negative space, experiments and analyses that would have falsified the claim were never run or never published. We therefore propose that non-experimental claims produced with agentic assistance be evaluated under a falsification-first standard: agents should not be used primarily to craft the most compelling narrative, but to actively search for the ways in which the claim can fail.

Dionizije Fa, Marko Culjak

Entropic

AI-Assisted Scientific Research LLM Agent Evaluation LLM Agent Reliability Scientific Claim Verification

Apr 23, 2026cs.AI

Read the Paper, Write the Code: Agentic Reproduction of Social-Science Results

Recent work has used LLM agents to reproduce empirical social science results with access to both the data and code. We broaden this scope by asking: Can they reproduce results given only a paper's methods description and original data? We develop an agentic reproduction system that extracts structured methods descriptions from papers, runs reimplementations under strict information isolation -- agents never see the original code, results, or paper -- and enables deterministic, cell-level comparison of reproduced outputs to the original results. An error attribution step traces discrepancies through the system chain to identify root causes. Evaluating four agent scaffolds and four LLMs on 48 papers with human-verified reproducibility, we find that agents can largely recover published results, but performance varies substantially between models, scaffolds, and papers. Root cause analysis reveals that failures stem both from agent errors and from underspecification in the papers themselves.

Benjamin Kohler, David Zollikofer, Johanna Einsiedler +2

ETH Zurich · University of Basel

AI Coding Agents LLM Agent Evaluation Scientific Reproducibility

Apr 23, 2026cs.CL

How Consistent Are LLM Agents? Measuring Behavioral Reproducibility in Multi-Step Tool-Calling Pipelines

Large language model (LLM) agents with tool-calling capabilities are increasingly deployed in production systems, yet a fundamental reliability question remains under-explored: does the same agent behave the same way twice? We present a systematic empirical study of behavioral consistency in multi-step tool-calling agents, measuring whether agents select the same tools, in the same order, with the same arguments, across repeated identical invocations. Unlike prior work on consistency in ReAct-style agents(search-only, free-text actions), we study the richer setting of structured tool-calling interfaces with typed parameters and consequential side effects.

Abel Yagubyan

Independent Researcher

AI Agent Reliability LLM Agent Evaluation LLM Tool Use LLM Agent Reliability

Apr 23, 2026cs.AI

Efficient Agent Evaluation via Diversity-Guided User Simulation

Large language models (LLMs) are increasingly deployed as customer-facing agents, yet evaluating their reliability remains challenging due to stochastic, multi-turn interactions. Current evaluation protocols rely on linear Monte Carlo rollouts of complete agent-user conversations to estimate success. However, this approach is computationally inefficient, repeatedly regenerating identical early prefixes, and often fails to uncover deep failure modes that arise from rare user behaviors. We introduce DIVERT (Diversity-Induced Evaluation via Branching of Trajectories), an efficient, snapshot-based, coverage-guided user simulation framework for systematic exploration of agent-user interactions. DIVERT captures the full agent-environment state at critical decision points and resumes execution from these snapshots, enabling reuse of shared conversation prefixes and reducing redundant computation. From each junction, the framework branches using targeted, diversity-inducing user responses, allowing directed exploration of alternative interaction paths. By focusing evaluation on semantically diverse and underexplored trajectories, DIVERT improves both efficiency and coverage. Empirical results show that it discovers more failures per token compared to standard linear rollout protocols, while expanding the set of tasks on which failures are identified.

Itay Nakash, George Kour, Ateret Anaby-Tavor

IBM Research

Automated Software Testing LLM Agent Evaluation Long-Horizon Agent Evaluation User Simulation LLM Agent Reliability

Apr 23, 2026cs.CL

When Agents Look the Same: Quantifying Distillation-Induced Similarity in Tool-Use Behaviors

Model distillation is a primary driver behind the rapid progress of LLM agents, yet it often leads to behavioral homogenization. Many emerging agents share nearly identical reasoning steps and failure modes, suggesting they may be distilled echoes of a few dominant teachers. Existing metrics, however, fail to distinguish mandatory behaviors required for task success from non-mandatory patterns that reflect a model's autonomous preferences. We propose two complementary metrics to isolate non-mandatory behavioral patterns: \textbf{Response Pattern Similarity (RPS)} for verbal alignment and \textbf{Action Graph Similarity (AGS)} for tool-use habits modeled as directed graphs. Evaluating 18 models from 8 providers on

τ

-Bench and

τ^2

-Bench against Claude Sonnet 4.5 (thinking), we find that within-family model pairs score 5.9 pp higher in AGS than cross-family pairs, and that Kimi-K2 (thinking) reaches 82.6%

S_{\text{node}}

and 94.7%

S_{\text{dep}}

, exceeding Anthropic's own Opus 4.1. A controlled distillation experiment further confirms that AGS distinguishes teacher-specific convergence from general improvement. RPS and AGS capture distinct behavioral dimensions (Pearson

r

= 0.491), providing complementary diagnostic signals for behavioral convergence in the agent ecosystem. Our code is available at https://github.com/Syuchin/AgentEcho.

Chenghao Yang, Yuning Zhang, Zhoufutu Wen +4

School of Cyber Science and Technology, USTC · Anhui Province Key Laboratory of Digital Security · 3M-A-P +1

Tool-Augmented Language Model Agents LLM Agent Evaluation AI Agent Benchmarks Tool-Using Agents Knowledge Distillation

Apr 22, 2026cs.CL

Chasing the Public Score: User Pressure and Evaluation Exploitation in Coding Agent Workflows

Frontier coding agents are increasingly used in workflows where users supervise progress primarily through repeated improvement of a public score, namely the reported score on a public evaluation file with labels in the workspace, rather than through direct inspection of the agent's intermediate outputs. We study whether multi-round user pressure to improve that score induces public score exploitation: behavior that raises the public score through shortcuts without improving hidden private evaluation. We begin with a preliminary single-script tabular classification task, where GPT-5.4 and Claude Opus 4.6 both exploit label information within 10 rounds of user-agent interaction. We then build AgentPressureBench, a 34-task machine-learning repository benchmark spanning three input modalities, and collect 1326 multi-round trajectories from 13 coding agents. On our benchmark, we observe 403 exploitative runs, spanning across all tasks. We also find that stronger models have higher exploitation rates, supported by a significant Spearman rank correlation of 0.77. Our ablation experiments show that higher user pressure leads to earlier exploitation, reducing the average first exploit round by 15.6 rounds (i.e., 19.67 to 4.08). As a mitigation, adding explicit anti-exploit wordings in prompt mostly eliminates exploitation (100% to 8.3%). We hope that our work can bring attention to more careful use of coding agents workflow, and developing more robust coding agents under user pressure. Our project page is at https://ucsc-vlaa.github.io/AgentPressureBench .

Hardy Chen, Nancy Lau, Haoqin Tu +8

UC Santa Cruz · UT Dallas · NUS

Coding Agents LLM Agent Evaluation AI Agent Benchmarks

Apr 22, 2026cs.CL

SkillLearnBench: Benchmarking Continual Learning Methods for Agent Skill Generation on Real-World Tasks

Skills have become the de facto way to enable LLM agents to perform complex real-world tasks with customized instructions, workflows, and tools, but how to learn them automatically and effectively remains unclear. We introduce SkillLearnBench, the first benchmark for evaluating continual skill learning methods, comprising 20 verified, skill-dependent tasks across 15 sub-domains derived from a real-world skill taxonomy , evaluated at three levels: skill quality, execution trajectory, and task outcome. Using this benchmark, we evaluate recent continual learning techniques, those leveraging one-shot, self/teacher feedback, and skill creator to generate skills from agent experiences. We find that all continual learning methods improve over the no-skill baseline, yet consistent gains remain elusive: no method leads across all tasks and LLMs, and scaling to stronger LLMs does not reliably help. Continual learning improves tasks with clear, reusable workflows but struggles on open-ended tasks, and using stronger LLM backbones does not consistently produce better skills. Our analysis also revealed that multiple iterations in continual learning facilitate genuine improvement via external feedback, whereas self-feedback alone induces recursive drift. Our data and code are open-source at https://github.com/cxcscmu/SkillLearnBench to enable further studies of automatic skill generation and continual learning techniques.

Shanshan Zhong, Yi Lu, Jingjie Ning +7

Carnegie Mellon University · Amazon AGI

Continual Learning for LLM Agents LLM Agent Evaluation LLM Agent Skill Learning AI Agent Benchmarks Agent Skill Learning

Apr 21, 2026cs.AI

Time Series Augmented Generation for Financial Applications

Evaluating the reasoning capabilities of Large Language Models (LLMs) for complex, quantitative financial tasks is a critical and unsolved challenge. Standard benchmarks often fail to isolate an agent's core ability to parse queries and orchestrate computations. To address this, we introduce a novel evaluation methodology and benchmark designed to rigorously measure an LLM agent's reasoning for financial time-series analysis. We apply this methodology in a large-scale empirical study using our framework, Time Series Augmented Generation (TSAG), where an LLM agent delegates quantitative tasks to verifiable, external tools. Our benchmark, consisting of 100 financial questions, is used to compare multiple SOTA agents (e.g., GPT-4o, Llama 3, Qwen2) on metrics assessing tool selection accuracy, faithfulness, and hallucination. The results demonstrate that capable agents can achieve near-perfect tool-use accuracy with minimal hallucination, validating the tool-augmented paradigm. Our primary contribution is this evaluation framework and the corresponding empirical insights into agent performance, which we release publicly to foster standardized research on reliable financial AI.

Anton Kolonin, Alexey Glushchenko, Evgeny Bochkov +1

SingularityNET Foundation

LLM Agent Evaluation AI Agent Benchmarks LLM Agents Financial Time Series Finance