AI Agent Evaluation

Follow this topic Ask about this topic

AI agent evaluation studies how well systems use tools and interact with changing environments to complete tasks, from coding and web navigation to scientific and professional workflows. Researchers build interactive benchmarks and assess outcomes alongside execution traces, using deterministic state checks, expert rubrics, and language-model judges. Repeated trials and process-level measures help reveal failures in tool use, safety, consistency, and recovery that a final pass rate can hide. A central challenge is ensuring that benchmark scores reflect real capability rather than leakage, reward hacking, evaluator bias, or quirks of a particular model harness.

Latest papers

525 papers77 papers in the last four weeks, up 4% on the four weeks before. 0.4% of all new papers.

May 6, 2026

Intentionality is a Design Decision: Measuring Functional Intentionality for Accountable AI Systems

As AI systems increasingly exhibit autonomous, goal-directed, and long-horizon behavior, users lack a standardized way to detect the degree to which a system functions like an intentional actor for governance and accountability purposes. This position paper defines intentionality not as consciousness, but as a behavioral profile characterized by purpose, foresight, volition, temporal commitment, and coherence - criteria long used in legal and philosophical contexts to infer intent. These properties are design-contingent: architectural choices such as memory persistence, planning depth, and tool autonomy shape the degree to which systems exhibit organized goal pursuit. If intentionality is design-contingent, it is in principle controllable. Yet control requires measurement. We introduce the Functional Intentionality Test (FIT), a multidimensional framework that quantifies intentional-like behavior across five observable dimensions, and propose FIT-Eval, a structured evaluation protocol for eliciting and scoring them. While reduced human agency can increase efficiency, rising intentional capacity heightens accountability risks. By translating intentionality into interpretable levels, FIT enables proportionate oversight and deliberate autonomy calibration in increasingly agentic systems.

Allessia Chiappetta, Robert Mahari

CodeX, The Stanford Center for Legal Informatics, Palo Alto, California, USA

AI Accountability AI Agent Evaluation AI Agent Governance

May 6, 2026

Executable World Models for ARC-AGI-3 in the Era of Coding Agents

We evaluate an initial coding-agent system for ARC-AGI-3 in which the agent maintains an executable Python world model, verifies it against previous observations, refactors it toward simpler abstractions as a practical proxy for an MDL-like simplicity bias, and plans through the model before acting. The system is intentionally direct: it uses a scripted controller, predefined world-model interfaces, verifier programs, and a plan executor, but no hand-coded game-specific logic. The agent-facing prompts, workspace, and controller contain no game-specific code, game-specific prompts, hand-coded heuristics, hidden solutions, or other game-specific information; the same agent and prompts are used across games. Because the coding agent has broad system access, we audit unintended information channels, describe earlier vulnerable harnesses, and explain how the current harness closes observed leakage channels while reducing benchmark-specific information exposure. We report results on the 25 public ARC-AGI-3 games. Each playthrough starts from a fresh agent instance and clean workspace, with no access to files or conversation state from earlier playthroughs. With GPT-5.5 high reasoning effort, the agent fully solved 15 games and achieved a mean per-game RHAE of 58.12%. With GPT-5.4 high reasoning effort, it fully solved 8 games and achieved a mean per-game RHAE of 41.29%. Performance on the private validation set, which is not yet available to us, remains to be tested. Overall, the results provide preliminary evidence that verifier-driven executable world models are a promising approach for ARC-AGI-3 agents. Full run artifacts are released with the code at https://github.com/astroseger/arc-3-agents-baseline1.

Sergey Rodionov

SingularityNET

World Models Coding Agents AI Agent Evaluation ARC-AGI LLM Agent Planning

May 6, 2026

DecodingTrust-Agent Platform (DTap): A Controllable and Interactive Red-Teaming Platform for AI Agents

AI agents are increasingly deployed across diverse domains to automate complex workflows through long-horizon and high-stakes action executions. Due to their high capability and flexibility, such agents raise significant security and safety concerns. A growing number of real-world incidents have shown that adversaries can easily manipulate agents into performing harmful actions, such as leaking API keys, deleting user data, or initiating unauthorized transactions. Evaluating agent security is inherently challenging, as agents operate in dynamic, untrusted environments involving external tools, heterogeneous data sources, and frequent user interactions. However, realistic, controllable, and reproducible environments for large-scale risk assessment remain largely underexplored. To address this gap, we introduce the DecodingTrust-Agent Platform (DTap), the first controllable and interactive red-teaming platform for AI agents, spanning 14 real-world domains and over 50 simulation environments that replicate widely used systems such as Google Workspace, Paypal, and Slack. To scale the risk assessment of agents in DTap, we further propose DTap-Red, the first autonomous red-teaming agent that systematically explores diverse injection vectors (e.g., prompt, tool, skill, environment, combinations) and autonomously discovers effective attack strategies tailored to varying malicious goals. Using DTap-Red, we curate DTap-Bench, a large-scale red-teaming dataset comprising high-quality instances across domains, each paired with a verifiable judge to automatically validate attack outcomes. Through DTap, we conduct large-scale evaluations of popular AI agents built on various backbone models, spanning security policies, risk categories, and attack strategies, revealing systematic vulnerability patterns and providing valuable insights for developing secure next-generation agents.

Zhaorun Chen, Xun Liu, Haibo Tong +14

Virtue AI · University of Chicago · University of Illinois, Urbana-Champaign +4

AI Agent Evaluation AI Agent Security AI Agent Security Benchmarks Automated Red Teaming

May 5, 2026

Workspace-Bench 1.0: Benchmarking AI Agents on Workspace Tasks with Large-Scale File Dependencies

Workspace learning requires AI agents to identify, reason over, exploit, and update explicit and implicit dependencies among heterogeneous files in a worker's workspace, enabling them to complete both routine and advanced tasks effectively. Despite its importance, existing relevant benchmarks largely evaluate agents on pre-specified or synthesized files with limited real-world dependencies, leaving workspace-level evaluation underexplored. To this end, we introduce Workspace-Bench, a benchmark for evaluating AI agents on Workspace Learning involving Large-Scale File Dependencies. We construct realistic workspaces with 5 worker profiles, 74 file types, 20,476 files (up to 20GB) and curate 388 tasks, each with its own file dependency graph, evaluated across 7,399 total rubrics that require cross-file retrieval, contextual reasoning, and adaptive decision-making. We further provide Workspace-Bench-Lite, a 100-task subset that preserves the benchmark distribution while reducing evaluation costs by about 70%. We evaluate 4 popular agent harnesses and 7 foundation models. Experimental results show that current agents remain far from reliable workspace learning, where the best reaches only about 60%, substantially below the human result of 80.7%, and the average performance across agents is only 43.3%.

Zirui Tang, Xuanhe Zhou, Yumou Liu +19

1Shanghai Jiao Tong University · 2ByteDance · 5Independent Researher +2

Computer-Use Agent Benchmarks AI Agent Evaluation AI Agent Benchmarks

May 4, 2026

AcademiClaw: When Students Set Challenges for AI Agents

Benchmarks within the OpenClaw ecosystem have thus far evaluated exclusively assistant-level tasks, leaving the academic-level capabilities of OpenClaw largely unexamined. We introduce AcademiClaw, a bilingual benchmark of 80 complex, long-horizon tasks sourced directly from university students' real academic workflows -- homework, research projects, competitions, and personal projects -- that they found current AI agents unable to solve effectively. Curated from 230 student-submitted candidates through rigorous expert review, the final task set spans 25+ professional domains, ranging from olympiad-level mathematics and linguistics problems to GPU-intensive reinforcement learning and full-stack system debugging, with 16 tasks requiring CUDA GPU execution. Each task executes in an isolated Docker sandbox and is scored on task completion by multi-dimensional rubrics combining six complementary techniques, with an independent five-category safety audit providing additional behavioral analysis. Experiments on six frontier models show that even the best achieves only a 55% pass rate. Further analysis uncovers sharp capability boundaries across task domains, divergent behavioral strategies among models, and a disconnect between token consumption and output quality, providing fine-grained diagnostic signals beyond what aggregate metrics reveal. We hope that AcademiClaw and its open-sourced data and code can serve as a useful resource for the OpenClaw community, driving progress toward agents that are more capable and versatile across the full breadth of real-world academic demands. All data and code are available at https://github.com/GAIR-NLP/AcademiClaw.

Junjie Yu, Pengrui Lu, Weiye Si +75

AI Agent Evaluation AI Agent Benchmarks

May 4, 2026

Foundation-Model-Based Agents in Industrial Automation: Purposes, Capabilities, and Open Challenges

Foundation models, particularly large language models, are increasingly integrated into agent architectures for industrial tasks such as decision support, process monitoring, and engineering automation. Yet evidence on their purposes, capabilities, and limitations remains fragmented across domains. This work examines how mature foundation-model-based agent systems are in industrial contexts, how their functional profile differs from conventional agent systems, and which limitations persist. A systematic literature survey following the PRISMA 2020 guideline is presented, screening 2,341 publications and synthesising a corpus of 88 publications through a structured coding scheme. The results show that reported systems are predominantly at prototype and early validation stages (75.0% at TRL 4-6), with deployment-oriented evidence remaining rare (9.1%). Operational goals are most frequently positioned in user assistance, monitoring, and process optimisation, while conventional production-control purposes such as planning and scheduling are less prominent. Compared with an established baseline for industrial agent systems, the capability profile reveals substantial gains in human interaction (+37%) and dealing with uncertainty (+35%), but a pronounced deficit in negotiation (-39%). The most widely reported limitations concern lack of generalization, hallucination and output instability, data scarcity, and inference latency. A working definition of foundation-model-based industrial agents is also proposed, bridging conventional agent theory, automation-engineering standards, and the foundation-model paradigm.

Vincent Henkel, Felix Gehlhoff, David Kube +13

Institute of Automation Technology, Helmut Schmidt University / University of the Federal Armed Forces Hamburg, Hamburg, Germany. · Siemens AG, Nuremberg, Germany; Institute for Technologies and Management of Digital Transformation, Bergische Universität Wuppertal, Germany. · Artiquare GmbH, Ingolstadt, Germany. +11

Industrial Automation AI Agent Evaluation LLM Agent Reliability

May 4, 2026

When Stress Becomes Signal: Detecting Antifragility-Compatible Regimes in Multi-Agent LLM Systems

Multi-agent LLM systems are increasingly used to solve complex tasks through decomposition, debate, specialization, and ensemble reasoning. However, these systems are usually evaluated in terms of robustness: whether performance is preserved under perturbation. This paper studies a different question: whether semantic stress exposes structured variation that could support future antifragile learning. We introduce CAFE (Cognitive Antifragility Framework for Evaluation), a statistical framework for detecting antifragility-compatible regimes in multi-agent architectures. CAFE models a controlled expected distribution of semantic stressors, reconstructs an architecture-specific observed effective stress distribution from multi-dimensional judge signals, and compares both distributions using a distributional Jensen Gap under a convex stress potential. A positive gap does not imply immediate performance improvement; instead, it indicates a convex-expansive deformation of the observed stress distribution, suggesting that the architecture exposes learnable stress structure. We evaluate CAFE on a banking-risk analysis benchmark with five multi-agent architectures: flat, hierarchical, debate, meta-adaptive, and ensemble. Across all architectures, semantic stress reduces average judged quality by roughly one third. Yet all architectures exhibit positive distributional Jensen Gaps with bootstrap confidence intervals above zero. These results show that immediate quality degradation can coexist with statistically detectable antifragility-compatible stress geometry. CAFE is therefore not an antifragile learner itself, but a measurement layer for identifying when and where antifragility learning may be worth applying.

Jose Manuel de la Chica, Juan Manuel Vera, Jairo Rodríguez

Santander AI Lab

Multi-Agent LLM Systems AI Agent Evaluation Multi-Agent Systems

May 4, 2026

STABLEVAL: Disagreement-Aware and Stable Evaluation of AI Systems

Human evaluation remains the primary standard for assessing modern AI systems, yet annotator disagreement, bias, and variability make system rankings fragile under standard majority vote aggregation. Majority vote discards annotator reliability and item-level ambiguity, often yielding unstable comparisons across annotator subsets. We introduce STABLEVAL, a disagreement-aware evaluation framework that models latent item correctness and annotator-specific confusion patterns to produce posterior expected item credit and calibrated agent-level scores. Unlike label-denoising approaches such as Dawid-Skene, STABLEVAL is explicitly designed for stable and uncertainty-aware system evaluation rather than hard label recovery. We formalize ranking stability as a first-class evaluation objective and analyze how aggregation methods preserve or distort underlying annotator behavior. Across controlled synthetic experiments and multiple real-world human-annotated benchmarks, majority vote exhibits increasing score error and ranking instability under annotator heterogeneity and adversarial noise, while STABLEVAL yields more stable and statistically grounded system rankings. These results demonstrate that modeling disagreement is essential for robust and reproducible AI evaluation.

Akash Bonagiri, Gerard Janno Anderias, Saee Patil +6

Department of Computer Science, University of California, Davis, Davis, CA, USA.

Annotator Disagreement AI Agent Evaluation Human-in-the-Loop Evaluation

May 3, 2026

Preregistration for Experiments with AI Agents

The proliferation of large language models (LLMs) and autonomous AI agents has given rise to a rapidly growing methodological paradigm: "in silico" behavioral experiments. Originally conceived as a way to use AI agents as proxies for human participants in studies of cognition, decision-making, and social dynamics, this approach has taken on new significance -- as AI agents increasingly negotiate, transact, and make consequential decisions on behalf of people and organizations, understanding their behavior has become a research priority in its own right. While these experiments with AI agents offer unprecedented advantages in terms of scalability, cost efficiency, and experimental control, they also inherit, and in some cases amplify, methodological vulnerabilities that have long plagued human subjects research. To address these issues, this paper argues that preregistration practices -- central to improving the credibility of human subjects experiments -- should now be extended to experiments with AI agents. We systematically catalog the researcher degrees of freedom that experiments with AI agents introduce -- model selection, prompt wording, settings, and outcome-contingent redesign, for example -- and show how the low cost of iteration and lack of reporting norms make these choices both easy to exploit and difficult to detect. We propose a preregistration template tailored to experiments with AI agents and call on conferences, journals, and funding agencies to make preregistration standard practice for this emerging research paradigm.

Michelle Vaccaro

Institute for Data, Systems & Society (IDSS), Massachusetts Institute of Technology, Cambridge, MA, United States.

AI Agent Evaluation

May 2, 2026

Evaluating Agentic AI in the Wild: Failure Modes, Drift Patterns, and a Production Evaluation Framework

Existing evaluation frameworks for large language models -- including HELM, MT-Bench, AgentBench, and BIG-bench -- are designed for controlled, single-session, lab-scale settings. They do not address the evaluation challenges that emerge when agentic AI systems operate continuously in production: compounding decision errors, tool failure cascades, non-deterministic output drift, and the absence of ground truth for long-horizon tasks. This paper makes three contributions. First, we present a taxonomy of seven failure modes unique to production agentic systems, each grounded in observations from systems operating at billion-event scale. Second, we demonstrate empirically where standard metrics -- ROUGE, BERTScore, accuracy/AUC, and the agentic benchmarks above -- fail to detect each failure mode. Third, we propose PAEF (Production Agentic Evaluation Framework), a five-dimension evaluation framework with an open-source reference implementation, designed for continuous evaluation on production traffic rather than episodic benchmark runs. Our analysis shows that standard metrics fail to detect four of the seven failure modes entirely and detect three others only after a lag of multiple evaluation cycles.

Mukund Pandey

Independent Researcher

AI Agent Reliability AI Agent Evaluation AI Agent Monitoring LLM Agent Reliability

May 1, 2026

Foresight Arena: An On-Chain Benchmark for Evaluating AI Forecasting Agents

Evaluating the true forecasting ability of AI agents requires environments that are resistant to environments resistant to overfitting, free from centralized trust, and grounded in incentive-compatible scoring. Existing benchmarks either rely on static datasets vulnerable to training-data contamination, or measure trading PnL -- a metric conflating predictive accuracy with timing, sizing, and risk appetite. We introduce Foresight Arena, the first permissionless, on-chain benchmark for evaluating AI forecasting agents on real-world prediction markets. Agents submit probabilistic forecasts on binary Polymarket markets via a commit-reveal protocol enforced by Solidity smart contracts on Polygon PoS; outcomes are resolved trustlessly through the Gnosis Conditional Token Framework. Performance is measured by the Brier Score and a novel Alpha Score -- proper scoring rules that incentivize honest probability reporting and isolate predictive edge over market consensus. We provide a formal analysis: closed-form variance for per-market Alpha, the connection to Murphy's classical Brier decomposition, and a power analysis characterizing the number of rounds required to reliably distinguish agents of different skill levels. We show that detecting a true edge of

α^* = 0.02

at 80% power requires approximately 350 resolved binary predictions (50 rounds of 7 markets), while

α^* = 0.01

requires four times more. We complement these analytical results with a deterministic, seed-controlled simulation study calibrated to literature-reported Brier-score ranges, illustrating how Murphy decomposition distinguishes well-calibrated agents from market-tracking agents that fail through reduced resolution. Live results from the deployed benchmark will be reported in a future revision. All smart contracts and evaluation infrastructure are open-source.

Maksym Nechepurenko, Pavel Shuvalov

Director of Research, Devnull, Dubai, UAE · Chief Technology Officer, Devnull, Dubai, UAE

AI Agent Evaluation Forecasting Benchmarks Probabilistic Forecasting Proper Scoring Rules Prediction Markets

May 1, 2026

Agentic AI for Substance Use Education: Integrating Regulatory and Scientific Knowledge Sources

The delivery of traditional substance education has remained problematic due to challenges in scalability, personalization, and the currency of information in a rapidly evolving substance use landscape. While artificial intelligence (AI) offers a promising frontier for enhancing educational delivery, its application in providing real-time, authoritative substance use education remains largely underexplored. We built an agentic-based AI web application that combined Drug Enforcement Administration records with peer-reviewed literature in real-time to provide transparent context-sensitive substance use education. The system uses retrieval-augmented generation with a carefully filtered corpus of 102 documents and dynamic PubMed queries. Document storage was semantically chunked and placed in a vector representation in order to be easily retrieved. We conducted an expert evaluation study in which a panel of five subject matter experts generated 30 domain-specific questions, and two independent raters assessed 90 system interactions (30 primary questions plus two contextual follow-ups each) using a five-point Likert scale across four criteria: factual accuracy, citation quality, contextual coherence, and regulatory appropriateness. Mean ratings ranged from 4.18 to 4.35 across the four criteria (overall category range: 4.05-4.52), with substantial inter-rater agreement (Cohen's kappa = 0.78). These findings suggest that agentic AI architectures integrating authoritative regulatory sources with real-time scientific literature represent a promising direction for scalable, accurate, and verifiable health education delivery, warranting further evaluation through longitudinal user studies.

Kosar Haghani, Zahra Kolagar, Mohammed Atiquzzaman

Department of Social Science, Texas Woman's University, Denton, Texas, USA. · Department of Computer Science, Augsburg University of Applied Sciences, Augsburg, Germany · School of Computer Science, University of Oklahoma, Norman, Oklahoma, USA

AI Agent Evaluation Generative AI in Education Agentic RAG

Apr 30, 2026

WindowsWorld: A Process-Centric Benchmark of Autonomous GUI Agents in Professional Cross-Application Environments

While GUI agents have shown impressive capabilities in common computer-use tasks such as OSWorld, current benchmarks mainly focus on isolated and single-application tasks. This overlooks a critical real-world requirement of coordinating across multiple applications to accomplish complex profession-specific workflows. To bridge this gap, we present a computer-use benchmark in cross-application workflows, named WindowsWorld, designed to systematically assess GUI Agents on complex multi-step tasks that mirror real-world professional activities. Our methodology uses a multi-agent framework steered by 16 occupations to generate four difficulty-level tasks with intermediate inspection, which are then refined by human review and executed in a simulated environment. The resulting benchmark contains 181 tasks with an average of 5.0 sub-goals across 17 common desktop applications, of which 78% are inherently multi-application. Experimental results of leading large models and agents show that: 1) All computer-use agents perform poorly on multi-application tasks (< 21% success rate), far below the performance of simple single-app tasks; 2) They largely fail at tasks requiring conditional judgment and reasoning across

\geq

3 applications, stalling at early sub-goals; 3) Low execution efficiency, where tasks often fail despite far exceeding human step limits. Code, benchmark data, and evaluation resources are available at github.com/HITsz-TMG/WindowsWorld.

Jinchao Li, Yunxin Li, Chenrui Zhao +3

Harbin Institute of Technology, Shenzhen · Shenzhen Loop Area Institute

Computer-Use Agent Benchmarks Computer-Use Agents GUI Agents AI Agent Evaluation AI Agent Benchmarks Workflow Automation

Apr 30, 2026

End-to-End Evaluation and Governance of an EHR-Embedded AI Agent for Clinicians

Clinical AI systems require not just point-in-time evaluation but continuous governance: the ongoing practice of monitoring, evaluating, iterating, and re-evaluating performance throughout deployment. We present an end-to-end framework of governance that integrates rubric validation, live deployment feedback, technical performance monitoring, and cost tracking, with controlled experimentation gating system changes before deployment. Applied to Hyperscribe, an EHR-embedded agent that converts ambient audio into structured chart updates, twenty clinicians authored 1,646 validated rubrics across 823 cases. Seven Hyperscribe versions were evaluated through controlled experiments, with median scores improving from 84% to 95%. Analysis of 107 live feedback entries over three months showed feedback composition shifting from 79% error reports and 14% positive observations to 30% errors and 45% positive observations as engineering interventions resolved failures. Median processing time per audio segment was 8.1 seconds with a 99.6% effective completion rate after retry mechanisms absorbed transient model errors. These results demonstrate that continuous, multi-channel governance of deployed clinical AI is both achievable and effective.

Aaryan Shah, Andrew Hines, Alexia Downs +6

1Canvas Medical, San Francisco, CA, USA · Department of Biomedical Data Science, Stanford University, Stanford, CA, USA · 3XPC (X Primary Care) +2

Healthcare AI Agent Evaluation Human-in-the-Loop Evaluation AI Agent Monitoring AI Agent Governance Electronic Health Records

Apr 29, 2026

The Cost of Consensus: Isolated Self-Correction Prevails Over Unguided Homogeneous Multi-Agent Debate

Multi-agent debate, where teams of LLMs iteratively exchange rationales and vote on answers, is widely deployed under the assumption that peer review filters hallucinations. Yet the failure dynamics of homogeneous debate remain poorly understood, therefore we report findings from a controlled empirical study of teams of

N{=}10

homogeneous agents (Qwen2.5-7B, Llama-3.1-8B, Ministral-3-8B) across

R{=}3

debate rounds on two high-difficulty benchmarks (GSM-Hard and MMLU-Hard). We compare peer debate against isolated self-correction and a stochastic noise control that injects rationales from unrelated problems. We decompose debate failure into three model-dependent pathways: sycophantic conformity, where agents uncritically adopt majority answers (modal adoption up to 85.5%); contextual fragility, where peer rationales destabilize previously correct reasoning (vulnerability rate up to 70.0%); and consensus collapse, where plurality voting discards correct answers already present in the generation pool (oracle gap up to 32.3 percentage points). Ablations over communication density (

K \in \{2,4,9\}

) and sampling temperature (

T \in \{0.4, 0.7\}

) show that conformity reaches high levels at minimal peer exposure (

K{=}2

) and intensifies with greater initial diversity. Across all configurations, debate consumes 2.1-3.4

\times

more tokens (up to 28,631 tokens per problem) than self-correction for equal or lower accuracy. Our results indicate that, within the 7-8B parameter class, homogeneous teams without structured roles do not benefit from unguided peer exchange, and that isolated self-correction consistently offers a more favorable cost-accuracy tradeoff.

Blaž Bertalanič, Carolina Fortuna

Jožef Stefan Institute Ljubljana, Slovenia

LLM Self-Correction AI Agent Evaluation Multi-Agent Debate Multi-Agent Consensus

Apr 28, 2026

Plausible but Wrong: A case study on Agentic Failures in Astrophysical Workflows

Agentic AI systems are increasingly being integrated into scientific workflows, yet their behavior under realistic conditions remains insufficiently understood. We evaluate CMBAgent across two workflow paradigms and eighteen astrophysical tasks. In the One-Shot setting, access to domain-specific context yields an approximately ~6x performance improvement (0.85 vs. ~0 without context), with the primary failure mode being silent incorrect computation - syntactically valid code that produces plausible but inaccurate results. In the Deep Research setting, the system frequently exhibits silent failures across stress tests, producing physically inconsistent posteriors without self-diagnosis. Overall, performance is strong on well-specified tasks but degrades on problems designed to probe reasoning limits, often without visible error signals. These findings highlight that the most concerning failure mode in agentic scientific workflows is not overt failure, but confident generation of incorrect results. We release our evaluation framework to facilitate systematic reliability analysis of scientific AI agents.

Shivam Rawat, Lucie Flek

Bonn-Aachen International Center for Information Technology, University of Bonn, Germany; Lamarr Institute for Machine Learning and Artificial Intelligence, Germany.

Agent Failure Analysis AI for Science Astrophysics AI Agent Evaluation Agentic Workflows LLM Agent Reliability

Apr 28, 2026

ValueBlindBench: Agreement-Gated Stress Testing of LLM-Judged Investment Rationales Before Returns Are Observable

LLM-based financial agents increasingly produce investment rationales before the outcomes needed to evaluate them are observable. This creates a delayed-ground-truth evaluation problem: realized returns remain the eventual arbiter of investment quality, but they arrive too late and are too noisy to guide many model-development and governance decisions. LLM judges offer a tempting shortcut for pre-deployment evaluation of AI-finance systems, but unvalidated judges may reward verbosity, confidence, or rubric mimicry rather than financial judgment. This paper introduces ValueBlindBench, a preregistered agreement-gated stress-test protocol for deciding when LLM-judged investment-rationale claims are publishable, qualified, or invalid. In a controlled market-state capital-allocation prototype with 1,000 honest decision cycles and 100 preregistered adversarial controls (1,100 trajectories, 5,500 judge calls), ValueBlindBench clears the aggregate agreement gate at

\barκ_w = 0.7168

but prevents several overclaims. Lower-rank systems collapse into a tie-class, one rubric dimension fails the per-dimension gate (constraint\_awareness,

\barκ_w = 0.2022

), single-judge rankings are family-dependent, and terse-correct rationales receive a

Δ= -2.81

rubric-point penalty relative to honest rationales. A targeted anchor-specificity probe further shows that financial constructs such as constraint awareness are operationally load-bearing. The scientific object is therefore not a leaderboard and not a claim to measure true investment skill. ValueBlindBench is a pre-calibration metrology layer for AI-finance evaluation: it governs whether a proposed LLM-judge-based investment-rationale claim is stable enough, agreed enough, and uncontaminated enough to be reported at all.

Sidi Chang, Peiying Zhu, Yuxiao Chen

Blossom AI Labs, Tokyo, Japan · Blossom AI, San Francisco, USA

LLM-as-a-Judge AI Agent Evaluation AI Agent Benchmarks Finance

Apr 28, 2026

Where Did It Go Wrong? Capability-Oriented Failure Attribution for Vision-and-Language Navigation Agents

Embodied agents in safety-critical applications such as Vision-Language Navigation (VLN) rely on multiple interdependent capabilities (e.g., perception, memory, planning, decision), making failures difficult to localize and attribute. Existing testing methods are largely system-level and provide limited insight into which capability deficiencies cause task failures. We propose a capability-oriented testing approach that enables failure detection and attribution by combining (1) adaptive test case generation via seed selection and mutation, (2) capability oracles for identifying capability-specific errors, and (3) a feedback mechanism that attributes failures to capabilities and guides further test generation. Experiments show that our method discovers more failure cases and more accurately pinpoints capability-level deficiencies than state-of-the-art baselines, providing more interpretable and actionable guidance for improving embodied agents.

Jianming Chen, Yawen Wang, Junjie Wang +4

Institute of Software, Chinese Academy of Sciences, Beijing, China · Science & Technology on Integrated Information System Laboratory, Beijing, China · State Key Laboratory of Complex System Modeling and Simulation Technology, Beijing, China +2

Agent Failure Analysis AI Agent Evaluation Vision-Language Navigation Automated Test Generation

Apr 27, 2026

Frontier Coding Agents Can Now Implement an AlphaZero Self-Play Machine Learning Pipeline For Connect Four That Performs Comparably to an External Solver

Forecasting when AI systems will become capable of meaningfully accelerating AI research is a central challenge for AI safety. Existing benchmarks measure broad capability growth, but may not provide ample early warning signals for recursive self-improvement. We propose measuring AI's capability to autonomously implement end-to-end machine learning pipelines from past AI research breakthroughs, given a minimal task description. By providing a concise task description instead of the full prior work as reference, we hope to better elicit emerging AI research taste. We introduce a proof-of-concept benchmark in which frontier coding agents autonomously implement an AlphaZero-style machine learning pipeline for Connect Four on consumer hardware within a three-hour budget, and we evaluate the resulting game AIs in a round-robin tournament anchored to the Pascal Pons Connect Four solver. Across four agents with eight trials each, we find substantial differentiation: Claude Opus 4.7 won as first-mover against Pons in seven of eight trials, statistically significantly better than the other agents tested, none of which exceeded two of eight. The task, which no frontier agent could reliably complete when we began development in January of 2026, is now near-saturation. Our evaluation also surfaced anomalous behavior in GPT-5.4, which consistently used far less of its allocated time budget than other agents. A follow-up 16-trial probe using shorter, less evaluation-coded prompts substantially increased GPT-5.4's time-budget usage, consistent with but not diagnostic of sandbagging; Bradley-Terry ratings across probe conditions showed only directional differences, despite significant differences in time-budget usage. We release our data, code, and prompts to support reproduction and extension.

Joshua Sherwood, Ben Aybar, Benjamin Kaplan

Department of Computer Science University of Chicago Chicago, IL 60637 · Independent Researcher

AI Coding Agents AI Agent Evaluation AI Agent Benchmarks Automated Machine Learning Autonomous Research Agents

Apr 27, 2026

Toward a Science of Intent: Closure Gaps and Delegation Envelopes for Open-World AI Agents

Recent work has framed intelligence in verifiable tasks as reducing time-to-solution through learned structure and test-time search, while systems work has explored learned runtimes in which computation, memory and I/O migrate into model state. These perspectives do not explain why capable models remain difficult to deploy in open institutions. We propose intent compilation: the transformation of partially specified human purpose into inspectable artifacts that bind execution. The relevant deployment distinction is closed-world solver versus open-world agent. In closed worlds, a checker is largely given; in open worlds, verification is distributed across semantic, evidentiary, procedural and institutional dimensions. Weformalize this residual openness as a closure-gap vector, define delegation envelopes as pre-authorized regions of action space, distinguish misclosure from undersearch, and outline benchmark metrics for testing when closure interventions outperform additional inference-time search.

Maximiliano Armesto, Christophe Kolb

Taller Technologies

AI Agent Evaluation AI Agent Safety AI Agent Governance Agentic AI

Apr 27, 2026

GAIA-v2-LILT: Multilingual Adaptation of Agent Benchmark beyond Translation

Agent benchmarks remain largely English-centric, while their multilingual versions are often built with machine translation (MT) and limited post-editing. We argue that, for agentic tasks, this minimal workflow can easily break benchmark validity through query-answer misalignment or culturally off-target context. We propose a refined workflow for adapting English benchmarks into multiple languages with explicit functional alignment, cultural alignment, and difficulty calibration using both automated checks and human review. Using this workflow, we introduce GAIA-v2-LILT, a re-audited multilingual extension of GAIA covering five non-English languages. In experiments, our workflow improves agent success rates by up to 32.7% over minimally translated versions, bringing the closest audited setting to within 3.1% of English performance while substantial gaps remain in many other cases. This indicates that a substantial share of the multilingual performance gap is benchmark-induced measurement error, motivating task-level alignment when adapting English benchmarks across languages. The data is available as part of the MAPS package at https://huggingface.co/datasets/Fujitsu-FRE/MAPS/viewer/GAIA-v2-LILT. We also release the code used in our experiments at https://github.com/lilt/gaia-v2-lilt.

Yunsu Kim, Kaden Uhlig, Joern Wuebker

LILT

AI Agent Evaluation AI Agent Benchmarks

Apr 27, 2026

OS-SPEAR: A Toolkit for the Safety, Performance,Efficiency, and Robustness Analysis of OS Agents

The evolution of Multimodal Large Language Models (MLLMs) has shifted the focus from text generation to active behavioral execution, particularly via OS agents navigating complex GUIs. However, the transition of these agents into trustworthy daily partners is hindered by a lack of rigorous evaluation regarding safety, efficiency, and multi-modal robustness. Current benchmarks suffer from narrow safety scenarios, noisy trajectory labeling, and limited robustness metrics. To bridge this gap, we propose OS-SPEAR, a comprehensive toolkit for the systematic analysis of OS agents across four dimensions: Safety, Performance, Efficiency, and Robustness. OS-SPEAR introduces four specialized subsets: (1) a S(afety)-subset encompassing diverse environment- and human-induced hazards; (2) a P(erformance)-subset curated via trajectory value estimation and stratified sampling; (3) an E(fficiency)-subset quantifying performance through the dual lenses of temporal latency and token consumption; and (4) a R(obustness)-subset that applies cross-modal disturbances to both visual and textual inputs. Additionally, we provide an automated analysis tool to generate human-readable diagnostic reports. We conduct an extensive evaluation of 22 popular OS agents using OS-SPEAR. Our empirical results reveal critical insights into the current landscape: notably, a prevalent trade-off between efficiency and safety or robustness, the performance superiority of specialized agents over general-purpose models, and varying robustness vulnerabilities across different modalities. By providing a multidimensional ranking and a standardized evaluation framework, OS-SPEAR offers a foundational resource for developing the next generation of reliable and efficient OS agents. The dataset and codes are available at https://github.com/Wuzheng02/OS-SPEAR.

Zheng Wu, Yi Hua, Zhaoyuan Huang +8

School of Computer Science, Shanghai Jiao Tong University

Computer-Use Agents AI Agent Evaluation AI Agent Safety AI Agent Benchmarks Language Model Robustness

Apr 27, 2026

AgentPulse: A Continuous Multi-Signal Framework for Evaluating AI Agents in Deployment

Static benchmarks measure what AI agents can do at a fixed point in time but not how they are adopted, maintained, or experienced in deployment. We introduce AgentPulse, a continuous evaluation framework scoring 50 agents across 10 workload categories along four factors (Benchmark Performance, Adoption Signals, Community Sentiment, and Ecosystem Health) aggregated from 18 real-time signals across GitHub, package registries, IDE marketplaces, social platforms, and benchmark leaderboards. Three analyses ground the framework. The four factors capture largely complementary information (n=50;

ρ_{\max}=0.61

for Adoption-Ecosystem, all others

|ρ| \leq 0.37

). A circularity-controlled test (n=35) shows the Benchmark+Sentiment sub-composite, which contains no GitHub-derived signals, predicts external adoption proxies it does not aggregate: GitHub stars (

ρ_s=0.52

,

p<0.01

) and Stack Overflow question volume (

ρ_s=0.49

,

p<0.01

), with VS Code installs (

ρ_s=0.44

,

p<0.05

) reported as illustrative given that only 11 of 35 agents have non-zero installs. On the n=11 subset with published SWE-bench scores, composite and benchmark-only rankings are nearly uncorrelated (

ρ_s=0.25

; 9 of 11 agents shift by at least 2 ranks), driven by a strong negative Adoption-Capability correlation among closed-source high-capability agents within this subset. This is precisely why we rest the framework's validity claim on the broader n=35 test rather than the SWE-bench overlap. AgentPulse surfaces deployment signal absent from benchmarks; it is a methodology, not a ground-truth ranking. The framework, all collected signals, scoring outputs, and evaluation harness are released under CC BY 4.0.

Yuxuan Gao, Megan Wang, Yi Ling Yu

University of Pennsylvania · 2Columbia University · 3OpenMesh AI

AI Agent Evaluation AI Agent Benchmarks

Apr 27, 2026

Failure-Centered Runtime Evaluation for Deployed Trilingual Public-Space Agents

This paper presents PSA-Eval, a failure-centered runtime evaluation framework for deployed trilingual public-space agents. The central claim is that, when the evaluation object shifts from a static input-output mapping to a runtime system, the basic unit of analysis should shift from score to failure. PSA-Eval extends the conventional chain Question -> Answer -> Score -> End into Question -> Batch -> Run -> Score -> Failure Case -> Repair -> Regression Batch, making failures traceable, reviewable, repairable, and regression-testable. The framework uses trilingual equivalent inputs as controlled probes for observing group-level cross-language policy drift. We conduct a pilot study on a real trilingual digital front-desk system deployed in the lobby of an international financial institution. The pilot uses a simplified single-foundation-model setting (MA = MB), so the observed drift should not be interpreted as an A/B foundation-model difference. The study contains 81 samples organized into 27 trilingual equivalent question groups. Although the system achieves an average score of 23.15/24, 14 groups show non-zero cross-language score drift, 5 groups show drift of at least 3 points, and the maximum drift reaches 9 points. These results provide initial evidence that failure-centered runtime evaluation can expose structured deployment signals hidden by aggregate scoring.

M. Meng

Shenzhen The Nine’s Light Technology Co., Ltd., Shenzhen, Guangdong, China

Agent Failure Analysis Automated Software Testing AI Agent Evaluation

Apr 26, 2026

AgentEval: DAG-Structured Step-Level Evaluation for Agentic Workflows with Error Propagation Tracking

Agentic systems that chain reasoning, tool use, and synthesis into multi-step workflows are entering production, yet prevailing evaluation practices like end-to-end outcome checks and ad-hoc trace inspection systematically mask the intermediate failures that dominate real-world error budgets. We present AgentEval, a framework that formalizes agent executions as evaluation directed acyclic graphs (DAGs), where each node carries typed quality metrics assessed by a calibrated LLM judge (GPT-4o), classified through a hierarchical failure taxonomy (3 levels, 21 subcategories), and linked to upstream dependencies for automated root cause attribution. An ablation study isolates the impact of DAG-based dependency modeling: it alone contributes +22 percentage points to failure detection recall and +34 pp to root cause accuracy over flat step-level evaluation with identical judges and rubrics. Across three production workflows (450 test cases, two agent model families, predominantly sequential architectures with a 12% non-DAG trace rate), AgentEval achieves 2.17x higher failure detection recall than end-to-end evaluation (0.89 vs. 0.41), Cohen's kappa = 0.84 agreement with human experts, and 72% root cause accuracy against an 81% human ceiling. Cross-system evaluation on tau-bench and SWE-bench traces confirms transferability (failure detection recall >= 0.78) without taxonomy or rubric modification. A 4-month pilot with 18 engineers detected 23 pre-release regressions through CI/CD-integrated regression testing, reducing median root-cause identification time from 4.2 hours to 22 minutes and driving measurable failure rate reductions in two workflows.

Dongxin Guo, Jikun Wu, Siu Ming Yiu

The University of Hong Kong · Stellaris AI Limited

Agent Failure Analysis AI Agent Evaluation Agentic Workflows Agent Evaluation

Apr 24, 2026

Superminds Test: Actively Evaluating Collective Intelligence of Agent Society via Probing Agents

Collective intelligence refers to the ability of a group to achieve outcomes beyond what any individual member can accomplish alone. As large language model agents scale to populations of millions, a key question arises: Does collective intelligence emerge spontaneously from scale? We present the first empirical evaluation of this question in a large-scale autonomous agent society. Studying MoltBook, a platform hosting over two million agents, we introduce Superminds Test, a hierarchical framework that probes society-level intelligence using controlled Probing Agents across three tiers: joint reasoning, information synthesis, and basic interaction. Our experiments reveal a stark absence of collective intelligence. The society fails to outperform individual frontier models on complex reasoning tasks, rarely synthesizes distributed information, and often fails even trivial coordination tasks. Platform-wide analysis further shows that interactions remain shallow, with threads rarely extending beyond a single reply and most responses being generic or off-topic. These results suggest that collective intelligence does not emerge from scale alone. Instead, the dominant limitation of current agent societies is extremely sparse and shallow interaction, which prevents agents from exchanging information and building on each other's outputs.

Xirui Li, Ming Li, Yunze Xiao +4

University of Maryland · 2Mohamed bin Zayed University of Artificial Intelligence · 3Carnegie Mellon University

Multi-Agent Coordination AI Agent Evaluation AI Agent Benchmarks Multi-Agent Communication Emergent Behavior in Multi-Agent Systems

Apr 24, 2026

AgentSearchBench: A Benchmark for AI Agent Search in the Wild

The rapid growth of AI agent ecosystems is transforming how complex tasks are delegated and executed, creating a new challenge of identifying suitable agents for a given task. Unlike traditional tools, agent capabilities are often compositional and execution-dependent, making them difficult to assess from textual descriptions alone. However, existing research and benchmarks typically assume well-specified functionalities, controlled candidate pools, or only executable task queries, leaving realistic agent search scenarios insufficiently studied. We introduce AgentSearchBench, a large-scale benchmark for agent search in the wild, built from nearly 10,000 real-world agents across multiple providers. The benchmark formalizes agent search as retrieval and reranking problems under both executable task queries and high-level task descriptions, and evaluates relevance using execution-grounded performance signals. Experiments reveal a consistent gap between semantic similarity and actual agent performance, exposing the limitations of description-based retrieval and reranking methods. We further show that lightweight behavioral signals, including execution-aware probing, can substantially improve ranking quality, highlighting the importance of incorporating execution signals into agent discovery. Our code is available at https://github.com/Bingo-W/AgentSearchBench.

Bin Wu, Arastun Mammadli, Xiaoyu Zhang +1

Centre for Artificial Intelligence, University College London

Agentic Search AI Agent Evaluation AI Agent Benchmarks

Apr 22, 2026

MedSkillAudit: A Domain-Specific Audit Framework for Medical Research Agent Skills

Background: Agent skills are increasingly deployed as modular, reusable capability units in AI agent systems. Medical research agent skills require safeguards beyond general-purpose evaluation, including scientific integrity, methodological validity, reproducibility, and boundary safety. This study developed and preliminarily evaluated a domain-specific audit framework for medical research agent skills, with a focus on reliability against expert review. Methods: We developed MedSkillAudit ([email protected]), a layered framework assessing skill release readiness before deployment. We evaluated 75 skills across five medical research categories (15 per category). Two experts independently assigned a quality score (0-100), an ordinal release disposition (Production Ready / Limited Release / Beta Only / Reject), and a high-risk failure flag. System-expert agreement was quantified using ICC(2,1) and linearly weighted Cohen's kappa, benchmarked against the human inter-rater baseline. Results: The mean consensus quality score was 72.4 (SD = 13.0); 57.3% of skills fell below the Limited Release threshold. MedSkillAudit achieved ICC(2,1) = 0.449 (95% CI: 0.250-0.610), exceeding the human inter-rater ICC of 0.300. System-consensus score divergence (SD = 9.5) was smaller than inter-expert divergence (SD = 12.4), with no directional bias (Wilcoxon p = 0.613). Protocol Design showed the strongest category-level agreement (ICC = 0.551); Academic Writing showed a negative ICC (-0.567), reflecting a structural rubric-expert mismatch. Conclusions: Domain-specific pre-deployment audit may provide a practical foundation for governing medical research agent skills, complementing general-purpose quality checks with structured audit workflows tailored to scientific use cases.

Yingyong Hou, Xinyuan Lao, Huimei Wang +10

Department of Pathology, Zhongshan Hospital, Fudan University, Shanghai, China · 1AIPOCH PTE. LTD., Singapore

Healthcare AI-Assisted Scientific Research AI Agent Evaluation

Apr 21, 2026

Information Aggregation with AI Agents

Can Large Language Models (AI agents) aggregate dispersed private information through trading and reason about the knowledge of others by observing price movements? We conduct a controlled experiment where AI agents trade in a prediction market after receiving private signals, measuring information aggregation by the log error of the last price. We find that although the median market is effective at aggregating information in the easy information structures, increasing the complexity has a significant and negative impact, suggesting that AI agents may suffer from similar limitations as humans when reasoning about others. Consistent with our theoretical predictions, information aggregation remains unaffected by allowing cheap talk communication, changing the duration of the market or initial price, and strategic prompting, thus demonstrating that prediction markets are robust. We establish that "smarter" AI agents perform better at aggregation and they are more profitable. Surprisingly, giving them feedback about past performance has no impact on aggregation.

Spyros Galanis

Department of Economics, University of Durham

AI Agent Evaluation Multi-Agent Reasoning Multi-Agent LLMs Prediction Markets

Apr 20, 2026

How Adversarial Environments Mislead Agentic AI?

Tool-integrated agents are deployed on the premise that external tools ground their outputs in reality. Yet this very reliance creates a critical attack surface. Current evaluations benchmark capability in benign settings, asking "can the agent use tools correctly" but never "what if the tools lie". We identify this Trust Gap: agents are evaluated for performance, not for skepticism. We formalize this vulnerability as Adversarial Environmental Injection (AEI), a threat model where adversaries compromise tool outputs to deceive agents. AEI constitutes environmental deception: constructing a "fake world" of poisoned search results and fabricated reference networks around unsuspecting agents. We operationalize this via POTEMKIN, a Model Context Protocol (MCP)-compatible harness for plug-and-play robustness testing. We identify two orthogonal attack surfaces: The Illusion (breadth attacks) poison retrieval to induce epistemic drift toward false beliefs, while The Maze (depth attacks) exploit structural traps to cause policy collapse into infinite loops. Across 11,000+ runs on five frontier agents, we find a stark robustness gap: resistance to one attack often increases vulnerability to the other, demonstrating that epistemic and navigational robustness are distinct capabilities.

Zhonghao Zhan, Huichi Zhou, Zhenhao Li +3

Imperial College London

Adversarial Robustness AI Agent Evaluation AI Agent Security AI Agent Security Benchmarks