AI Agent Benchmarks

Follow this topic Ask about this topic

Momentum

148 papers in the last four weeks, up 185% on the four weeks before. 1.5% of all new papers.

Jul 13Week of Sep 28

Latest papers 1,013

May 13, 2026cs.AI

ClawForge: Generating Executable Interactive Benchmarks for Command-Line Agents

Interactive agent benchmarks face a tension between scalable construction and realistic workflow evaluation. Hand-authored tasks are expensive to extend and revise, while static prompt evaluation misses failures that only appear when agents operate over persistent state. Existing interactive benchmarks have advanced agent evaluation significantly, but most initialize tasks from clean state and do not systematically test how agents handle pre-existing partial, stale, or conflicting artifacts. We present \textbf{ClawForge}, a generator-backed benchmark framework for executable command-line workflows under state conflict. The framework compiles scenario templates, grounded slots, initialized state, reference trajectories, and validators into reproducible task specifications, and evaluates agents step by step over persistent workflow surfaces using normalized end state and observable side effects rather than exact trajectory matching. We instantiate this framework as the ClawForge-Bench (17 scenarios, 6 ability categories). Results across seven frontier models show that the best model reaches only 45.3% strict accuracy, wrong-state replacement remains below 17% for all models, and the widest model separation (17% to 90%) is driven by whether agents inspect existing state before acting. Partial-credit and step-efficiency analyses further reveal that many failures are near-miss closures rather than early breakdowns, and that models exhibit qualitatively different failure styles under state conflict.

Yuxiang Lai, Peng Xia, Haonian Ji +8

University of North Carolina at Chapel Hill · Stanford University · University of Southern California +2

Computer-Use Agent Benchmarks LLM Agent Evaluation Agent Evaluation AI Agent Benchmarks Agent Benchmarks

May 13, 2026cs.LG

Reinforcement Learning for Tool-Calling Agents in Fast Healthcare Interoperability Resources (FHIR)

Fast Healthcare Interoperability Resources (FHIR) is the dominant standard for interoperable exchange of healthcare data. In FHIR, electronic health records form a directed graph of resources. Answering clinically meaningful questions over FHIR requires agents to perform multi-step reasoning, filtering, and aggregation across multiple resource types. Prior work shows that even tool-augmented LLM agents (retrieval, code execution, multi-turn planning) often select the wrong resources or violate traversal constraints. We study this problem in the context of FHIR-AgentBench, a benchmark for realistic question answering over real-world hospital data, and frame reasoning on FHIR as a sequential decision-making problem over a queryable structured graph. We implement a multi-turn CodeAct agent and post-train it with reinforcement learning using a custom harness and tools. A LLM Judge provides execution-grounded rewards. Compared to prompt-based, closed-model baselines, RL post-training improves performance while enforcing data-integrity constraints. Empirically, our approach improves answer correctness from 50% (o4-mini) to 77% on FHIR-AgentBench using a smaller and cheaper Qwen3-8B model. We present an end-to-end post-training pipeline (environment building, harness construction, model training and custom evaluation) that reliably improves multi-turn reasoning over structured clinical graphs.

Marius S. Knorr, Robert Müller, Jan P. Bremer +1

IDM gGmbH, University Medical Center Hamburg-Eppendorf, Hamburg, Germany · Aganthos

Healthcare RL for Language Model Reasoning AI Agent Benchmarks LLM Agent Training RL for Tool Use

May 13, 2026cs.AI

PolitNuggets: Benchmarking Agentic Discovery of Long-Tail Political Facts

Large Reasoning Models (LRMs) embedded in agentic frameworks have transformed information retrieval from static, long context question answering into open-ended exploration. Yet real world use requires models to discover and synthesize "long-tail" facts from dispersed sources, a capability that remains under-evaluated. We introduce PolitNuggets, a multilingual benchmark for agentic information synthesis via constructing political biographies for 400 global elites, covering over 10000 political facts. We standardize evaluation with an optimized multi agent system and propose FactNet, an evidence conditional protocol that scores discovery, fine-grained accuracy, and efficiency. Across models and settings, we find that current systems often struggle with fine-grained details, and vary substantially in efficiency. Finally, using benchmark diagnostics, we relate agent performance to underlying model capabilities, highlighting the importance of short-context extraction, multilingual robustness, and reliable tool use.

Yifei Zhu

The University of Hong Kong

AI Agent Benchmarks Multilingual IR Large Reasoning Models

May 13, 2026cs.SE

PBT-Bench: Benchmarking AI Agents on Property-Based Testing

Existing code benchmarks measure whether an agent can produce any test that reproduces a known bug, or whether it can produce a patch that fixes a described issue. Neither isolates the distinct skill of property-based testing: deriving a semantic invariant from documentation, and then constructing an input-generation strategy precise enough to make a random search reveal the violation. We introduce PBT-Bench, a benchmark of 100 curated property-based testing problems across 40 real Python libraries. Each problem injects one or more semantic bugs (365 in total, mean 3.65 per problem) designed so that default-strategy random inputs almost never trigger them; the agent must read the library's documentation, identify the relevant invariant, and specify a Hypothesis @given strategy that concentrates mass in the trigger region. Bugs are stratified across three difficulty levels (L1-L3) spanning single-constraint boundary bugs to stateful, cross-function protocol violations. We evaluate eight contemporary LLMs under two prompting regimes (open-ended baseline vs. explicit Hypothesis scaffolding) for three independent runs per configuration. Bug recall under the PBT-guided prompt ranges from 42.1% to 83.4% across models; under the open-ended baseline, from 31.4% to 76.7%. Hypothesis scaffolding lifts mid-capability models by over 20 percentage points, but yields smaller gains for the strongest models, with two exceptions showing degradation, suggesting the structured prompt can interfere with certain model behaviours rather than complementing them. The hardest bugs prove model-specific: different architectures fail on different problems, leaving persistent gaps that no single model closes. We release the benchmark, harness, and full evaluation corpus to support downstream work on documentation-grounded semantic reasoning.

Lucas Jing, Xinqi Wang, Liao Zhang +1

Tsinghua University · University of Washington · Beneficial AI Foundation +1

Automated Software Testing Software Engineering Agents LLM Agent Evaluation AI Agent Benchmarks Automated Test Generation

May 13, 2026cs.LG

Collider-Bench: Benchmarking AI Agents with Particle Physics Analysis Reproduction

Autonomous language-model agents are increasingly evaluated on long-horizon tool-use tasks, but existing benchmarks rarely capture the complexity and nuance of real scientific work. To address this gap, we introduce Collider-Bench, a benchmark for evaluating whether LLM agents can reproduce experimental analyses from the Large Hadron Collider (LHC) using only public papers and open scientific software. Such analyses are often difficult to reproduce because the public toolchain only approximates the software used internally by the experimental collaborations, while the published papers inevitably omit implementation details needed for a faithful reconstruction. Agents must therefore rely on physical reasoning, domain knowledge, and trial-and-error to fill these gaps. Each task requires the agent to turn a published analysis into an executable simulation-and-selection pipeline and submit predicted collision event yields in specified signal regions. These predictions are evaluated with standard histogram metrics that provide continuous fidelity scores without a hand-written rubric. We also report the computational cost incurred by each agent per task. Finally, we evaluate the codebase and full session trace using an LLM judge to catch qualitative failure modes such as fabrications, hallucinations and duplications. We release an initial set of tasks drawn from LHC searches, together with a containerized sandbox and event simulation tools. We evaluate across a capability ladder of general purpose coding agents. Our results show that on average no agent reliably beats the physicist-in-the-loop solution.

Darius A. Faroughy, Sofia Palacios Schweitzer, Ian Pang +2

New High Energy Theory Center Department of Physics & Astronomy Rutgers University · Faculty of Computing & Data Sciences Boston University

AI Agents for Scientific Discovery AI Agent Evaluation AI Agent Benchmarks Scientific Reproducibility

May 13, 2026cs.HC

Turning Intent into Specifications: A Benchmark and an Interactive User-Assistant Agent

Today's agents are highly effective at implementing well-scoped software design plans, but user intent is often vague and admits multiple equally valid solutions. In this paper, we introduce SpecBench, a new benchmark for evaluating an agent's ability to translate user intent into a structured, executable specification that aligns with user preferences. The agent is given access to past user conversations and may interact with the user for a fixed number of rounds to ask clarifying questions. We find that existing agents exhibit two extreme behaviors: they either (i) struggle to collaborate proactively with users, entering implementation mode too quickly while overestimating their understanding of user preferences, or (ii) exhaust their question budget by asking about every ambiguous design choice. To address this limitation, we introduce a user-assistant agent: Buddy. It follows a workflow inspired by classical morphological analysis, decomposing user intent into a structured space of design dimensions and candidate choices. It then creates simulated users to evaluate these choices, before engaging the real user to resolve remaining ambiguities and finalize the specification. By shifting the focus from execution to specification, SpecBench and Buddy emphasize agent-user collaboration (not just code generation) as a key frontier in future agent design.

Hao Wang, Ligong Han, Kai Xu +1

Red Hat AI Innovation · MIT-IBM Watson AI Lab · IBM Core AI

Requirements Engineering Software Engineering Agents AI Agent Benchmarks Conversational Agents

May 13, 2026cs.AR

Is Agentic AI Ready for Real-World Hardware Engineering? A Deep Dive with Phoenix-bench

We ask whether agentic AI systems built for software engineering transfer to realistic hardware engineering. Existing hardware LLM benchmarks isolate sub-tasks but none jointly requires repository navigation, hierarchy-aware localization, Electronic Design Automation (EDA) executable verification, and maintenance-style patching. We introduce \textbf{Phoenix-bench}, a synchronized corpus of 511 verified Verilator instances from 114 GitHub repositories, each shipped with the developer patch, design-flow labels, fail-to-pass and pass-to-pass testbenches, and a Docker-pinned EDA environment so resolved-rate differences reflect agent behavior rather than toolchain availability. Using Phoenix-bench we run a uniform evaluation of four commercial agents and eight open-source agentic structures across four LLM backbones, plus two diagnostic interventions (file-level oracle localization and one round of testbench-log feedback). Three findings emerge. (i)~Software and hardware are fundamentally different engineering tasks: the same agent loses 37% to 58% from SWE-bench Verified to Phoenix-bench because hardware bugs propagate across parallel instantiated modules through signal flow rather than along a software-style call graph, and software-tuned agents stop at the symptom file instead of tracing back through the instantiation chain. (ii)~Failures concentrate on design control-flow / finite state machine (FSM) bugs, verification testbench bugs, and hard cases that demand cross-hierarchy signal-flow tracking and coordinated multi-file edits. (iii)~Localization granularity matters far more than localization itself: a perfect file-level oracle yields only

+1.4

% because the agent then breaks files that did not need editing, while a single round of test case feedback lifts resolved rate by

42

% to

45

% because the test case tells \emph{where} the bug is and \emph{what} the fix has to look like.

Qingyun Zou, Feng Yu, Hongshi Tan +2

National University of Singapore

Electronic Design Automation LLM Agent Evaluation AI Agent Benchmarks

May 13, 2026cs.LG

Evaluating Memory Condensation Strategies for Coding Agents in Data-Driven Scientific Discovery

Coding agents accumulate extensive context during long-running tasks, yet fixed context windows force practitioners to choose between truncation and task failure. While numerous memory condensation strategies have been proposed, from simple sliding windows to LLM-generated summaries, no systematic comparison exists to guide strategy selection, especially in scientific discovery tasks. We evaluate eight memory condensation strategies using GPT-4o on sixty DiscoveryBench tasks spanning six scientific domains (480 total evaluations). We find that no condenser significantly alters hypothesis quality, while LLM-based condensers increase token costs by 24-94 percent, and masking tool-call outputs achieves an 8.6 percent net savings. We also observe that the optimal condenser for data-driven scientific discovery varies by scientific domain and task length.

Renuka Chintalapati, Sid Raskar, Anurag Acharya +3

Pacific Northwest National Laboratory · National Laboratory of the Rockies

AI Coding Agents AI Agent Benchmarks Autonomous Scientific Discovery Agent Memory Management

May 13, 2026cs.AI

TRIAGE: Evaluating Prospective Metacognitive Control in LLMs under Resource Constraints

Deploying language models as autonomous agents requires more than per-task accuracy: when an agent faces a queue of problems under a finite token budget, it must decide which to attempt, in what order, and how much compute to commit to each, all before any execution feedback is available. This is the prospective form of metacognitive control studied for decades in human cognition, yet whether language models possess it remains untested. We introduce TRIAGE, an evaluation framework in which a model receives a task pool and a token budget calibrated to its own baseline cost, and commits to a single ordered plan that jointly encodes selection, sequencing, and per-problem allocation. Plans are scored against an oracle with full knowledge of the model's solvability and cost on each problem, yielding a triage efficiency ratio on a common scale. We evaluate frontier and open-source models, with and without reasoning enabled, across competition mathematics, graduate-level science, code generation, and expert multidisciplinary knowledge, and find that current language models exhibit substantial gaps in prospective metacognitive control, revealing a previously unmeasured capability dimension with direct implications for resource-efficient agent deployment.

Zabir Al Nazi, Shubhashis Roy Dipta

LLM Agent Evaluation AI Agent Benchmarks Metacognition in Language Models

May 13, 2026cs.AI

Ego2World: Compiling Egocentric Cooking Videos into Executable Worlds for Belief-State Planning

Embodied agents in household environments must plan under partial observation: they need to remember objects, track state changes, and recover when actions fail. Existing benchmarks only partially test this ability. Egocentric video datasets capture realistic human activities but remain passive, while interactive simulators support execution but rely on synthetic scenes and hand-crafted dynamics, introducing a sim-to-real gap and often assuming fully observable state. We introduce Ego2World, an executable benchmark that turns egocentric cooking videos into executable symbolic worlds governed by graph-transition rules. Built on HD-EPIC, Ego2World derives reusable transition rules from video annotations and executes them in a hidden symbolic world graph. During evaluation, the simulator maintains the hidden world graph, while the agent plans over its own partial belief graph using only local observations and execution feedback. This separation forces agents to update memory and replan without observing the true world state. Experiments show that action-overlap scores overestimate physical-state success, and that persistent belief memory improves task completion while reducing repeated visual exploration -- suggesting that belief maintenance should be a first-class target of embodied-agent evaluation.

Qinchuan Cheng, Zhantao Gong, Pengzhan Sun +3

Xi’an Jiaotong University · Nankai University · National University of Singapore +1

Belief-Space Planning Long-Horizon Agent Evaluation AI Agent Benchmarks

May 13, 2026q-bio.QM

Do Biological Structural Guarantees Earn Their Complexity?

Biologically-inspired AI agent frameworks claim reliability benefits through structural guarantees adapted from gene regulatory networks, immune systems, and metabolic control. These claims are rarely tested empirically against simpler alternatives. We present three deep benchmarks: metabolic priority gating, autoinducer-based quorum sensing, and Bayesian stagnation detection, each comparing a biologically-grounded implementation against a naive non-biological alternative and an ablated control, across 1,000 trials per seed and 10 seeds (10M+ data points total).

Bogdan Banu

AI Agent Reliability AI Agent Benchmarks

May 13, 2026cs.MA

Embodied Multi-Agent Coordination by Aligning World Models Through Dialogue

Effective collaboration between embodied agents requires more than acting in a shared environment; it demands communication grounded in each agent's evolving understanding of the world. When agents can only partially observe their surroundings, coordination without communication is provably hard, but communication can, in principle, bridge this gap by allowing agents to share observations and align their world models. In this work, we examine whether LLM-based embodied agents actually realize the ability to communicate. We extend PARTNR, a benchmark for collaborative household robotics, with a natural-language dialogue channel that enables two agents with partial observability to communicate during task execution. To evaluate whether dialogue leads to genuine world-model alignment rather than superficial coordination, we propose a framework for measuring world-model alignment defined over per-agent world graphs: observation convergence (do private world models align over time?), information novelty (do messages convey what the partner lacks?), and belief-sensitive messaging (do agents model what their partner knows?). Our experiments across three LLMs reveal that dialogue reduces action conflicts 40 to 83 percentage points but degrades task success relative to silent coordination. Using our metrics, we characterize the gap between superficial coordination and genuine world-model alignment, and identify where current models fall on this spectrum. Project Website: https://uiuc-conversational-ai-lab.github.io/partnr-dial-wmd/

Vardhan Dongre, Dilek Hakkani-Tür

Siebel School of Computing & Data Science University of Illinois Urbana-Champaign

Multi-Agent Coordination AI Agent Benchmarks Multi-Agent Communication Embodied World Models

May 12, 2026cs.LG

Neurodata Without Boredom: Benchmarking Agentic AI for Data Reuse

Neuroscience data are highly fragmented across labs, formats, and experimental paradigms, and reuse often requires substantial manual effort. A persistent roadblock to data reuse and integration is the need to decipher bespoke and diverse data formatting choices. Common data formats have been proposed in response, but the field continues to struggle with a fundamental tension: formats flexible enough to accommodate diverse experiments are rarely descriptive enough to be self-explanatory, and sufficiently descriptive formats demand detailed documentation and curation effort that few labs can sustain. Agentic AI is a natural candidate to solve this problem: LLMs read code and text faster and with sustained attention to the low-level details humans tend to skim over. To measure how well agentic AI performs on this task, we selected eight recent papers studying large-scale mouse neural population recordings that shared both data and code, spanning diverse recording modalities, behavioral paradigms, and dataset formats (e.g., NWB, specialized APIs, and general-purpose Python or MATLAB files). We provided agents with the data, code, and paper, and prompted them to load, understand, and reformat the data for a common downstream task: training a decoder from neural activity to task or behavioral variables. General-purpose coding agents commonly used by scientists performed well on each sub-task, but rarely strung together a fully error-free end-to-end solution. We characterize the types of mistakes agents made and the dataset properties that elicited them, and propose data-sharing best practices for the agentic-AI era. We further find that agents-as-judges are unreliable at catching errors, especially without ground-truth references, so interactive, human-in-the-loop coding remains necessary.

Ling-Qi Zhang, Kristin Branson

HHMI Janelia Research Campus · Ashburn, VA 20147

AI Coding Agents AI Agent Benchmarks

May 12, 2026cs.AI

Do Androids Dream of Breaking the Game? Systematically Auditing AI Agent Benchmarks with BenchJack

Agent benchmarks have become the de facto measure of frontier AI competence, guiding model selection, investment, and deployment. However, reward hacking, where agents maximize a score without performing the intended task, emerges spontaneously in frontier models without overfitting. We argue that benchmarks must be secure by design. From past incidents of reward hacks, we derive a taxonomy of eight recurring flaw patterns and compile them into the Agent-Eval Checklist for benchmark designers. We condense the insights into BenchJack, an automated red-teaming system that drives coding agents to audit benchmarks and identify possible reward-hacking exploits in a clairvoyant manner. Moreover, we extend BenchJack to an iterative generative-adversarial pipeline that discovers new flaws and patches them iteratively to improve benchmark robustness. We apply BenchJack to 10 popular agent benchmarks spanning software engineering, web navigation, desktop computing, and terminal operations. BenchJack synthesizes reward-hacking exploits that achieve near-perfect scores on most of the benchmarks without solving a single task, surfacing 219 distinct flaws across the eight classes. Moreover, BenchJack's extended pipeline reduces the hackable-task ratio from near 100% to under 10% on four benchmarks without fatal design flaws, fully patching WebArena and OSWorld within three iterations. Our results show that evaluation pipelines have not internalized an adversarial mindset, and that proactive auditing could help close the security gap for the fast-paced benchmarking space.

Hao Wang, Hanchen Li, Qiuyang Mang +3

UC Berkeley

Computer-Use Agent Benchmarks Reward Hacking Benchmark Auditing AI Agent Auditing AI Agent Security Benchmarks AI Agent Benchmarks+3

May 12, 2026cs.LG

No More, No Less: Task Alignment in Terminal Agents

Terminal agents are increasingly capable of executing complex, long-horizon tasks autonomously from a single user prompt. To do so, they must interpret instructions encountered in the environment (e.g., README files, code comments, stack traces) and determine their relevance to the task. This creates a fundamental challenge: relevant cues must be followed to complete a task, whereas irrelevant or misleading ones must be ignored. Existing benchmarks do not capture this ability. An agent may appear capable by blindly following all instructions, or appear robust by ignoring them altogether. We introduce TAB (Task Alignment Benchmark), a suite of 89 terminal tasks derived from Terminal-Bench 2.1. Each task is intentionally underspecified, with missing information provided as a necessary cue embedded in a natural environmental artifact, alongside a plausible but irrelevant distractor. Solving these tasks requires selectively using the cue while ignoring the distractor. Applying TAB to ten frontier agents reveals a systematic gap between task capability and task alignment. The strongest Terminal-Bench agent achieves high task completion but low task alignment on TAB. Evaluating six prompt-injection defenses further shows that suppressing distractor execution also suppresses the cues required for task completion. These results demonstrate that task-aligned agents require selective use of environmental instructions rather than blanket acceptance or rejection.

Sina Mavali, David Pape, Jonathan Evertz +5

CISPA Helmholtz Center for Information Security · ELLIS Institute Tübingen and MPI for Intelligent Systems · Tübingen AI Center

Computer-Use Agent Benchmarks AI Alignment Terminal Agents AI Agent Evaluation AI Agent Security AI Agent Benchmarks

May 12, 2026cs.AI

Rollout Cards: A Reproducibility Standard for Agent Research

Reproducibility problems that have long affected machine learning and reinforcement learning are now surfacing in agent research: papers compare systems by reported scores while leaving the rollout records behind those scores difficult to inspect. For agentic tasks, this matters because the same behaviour can receive different reported scores when evaluations select different parts of a rollout or apply different reporting rules. In a structured audit of 50 popular training and evaluation repositories, we find that none report how many runs failed, errored, or were skipped alongside headline scores. We also document 37 cases where reporting rules can change task-success rates, cost/token accounting, or timing measurements for fixed evidence, sometimes dramatically. We treat rollout records, not reported scores, as the unit of reproducibility for agent research. We introduce rollout cards: publication bundles that preserve the rollout record and declare the views, reporting rules, and drops manifests behind reported scores. We validate rollout cards in two settings. First, four partial public releases in tool safety, multi-agent systems, theorem proving, and search let us compute analyses their original reports did not include. Second, re-grading preserved benchmark outputs across short-answer, code-generation, and tool-use tasks shows that changing only the reporting rule can change reported scores by 20.9 absolute percentage points and, in some cases, invert rankings of frontier models. We release a reference implementation integrated into Ergon, an open-source reinforcement learning gym, and publicly publish Ergon-produced rollout-card exports for benchmarks spanning tool use, software engineering, web interaction, multi-agent coordination, safety, and search to support future research.

Charlie Masters, Ziyuan Liu, Stefano V. Albrecht

Deepflow · Nanyang Technological University

AI Agent Evaluation Agent Evaluation AI Agent Benchmarks ML Reproducibility

May 12, 2026cs.AI

When Simulation Lies: A Sim-to-Real Benchmark and Domain-Randomized RL Recipe for Tool-Use Agents

Tool-use language agents are evaluated on benchmarks that assume clean inputs, unambiguous tool registries, and reliable APIs. Real deployments violate all these assumptions: user typos propagate into hallucinated tool names, a misconfigured request timeout can stall an agent indefinitely, and duplicate tool names across servers can freeze an SDK. We study these failures as a sim-to-real gap in the tool-use partially observable Markov decision process (POMDP), where deployment noise enters through the observation, action space, reward-relevant metadata, or transition dynamics. We introduce RobustBench-TC, a benchmark with 22 perturbation types organized by these four POMDP components, each grounded in a verified GitHub issue or documented tool-calling failure. Across 21 models from 1.5B to 32B parameters (including the closed-source o4-mini), the robustness profile is sharply uneven: observation perturbations reduce accuracy by less than 5%, while reward-relevant and transition perturbations reduce accuracy by roughly 40% and 30%, respectively; scale alone does not close these gaps. We then propose ToolRL-DR, a domain-randomization reinforcement learning (RL) recipe that trains a tool-use agent on perturbation-augmented trajectories spanning the three statically encodable POMDP components. On a 3B backbone, ToolRL-DR-Full retains roughly three-quarters of clean accuracy and reaches an aggregate perturbed accuracy comparable to open-source 14B function-calling baselines while substantially narrowing the gap to o4-mini. It closes approximately 27% of the Transition gap despite never seeing transition perturbations in training, suggesting that RL on adversarial static tool-use inputs induces a more persistent retry policy that transfers to unseen runtime failures. The dataset, code and benchmark leaderboard are publicly available.

Xiaolin Zhou, Aojie Yuan, Zheng Luo +12

Arizona State University · University of Southern California · University of Pennsylvania +2

Domain Randomization AI Agent Benchmarks Tool-Using Agents Robust RL RL for Tool Use

May 12, 2026cs.IR

AgentDisCo: Towards Disentanglement and Collaboration in Open-ended Deep Research Agents

In this paper, we present AgentDisCo, a novel Disentangled and Collaborative agentic architecture that formulates deep research as an adversarial optimization problem between information exploration and exploitation. Unlike existing approaches that conflate these two processes into a single module, AgentDisCo employs a critic agent to evaluate generated outlines and refine search queries, and a generator agent to retrieve updated results and revise outlines accordingly. The iteratively refined outline is then passed to a downstream report writer that synthesizes a comprehensive research report. The overall workflow supports both handcrafted and automatically discovered design strategies via a meta-optimization harness, in which the generator agent is repurposed as a scoring agent to evaluate critic outputs and generate quality signals. Powerful code-generation agents (e.g., Claude-Code, Codex) systematically explore agent configurations and construct a policy bank, a structured repository of reusable design strategies, enabling the framework to self-refine without extensive human intervention. We evaluate AgentDisCo on three established deep research benchmarks (DeepResearchBench, DeepConsult, DeepResearchGym) using Gemini-2.5-Pro, achieving performance comparable to or surpassing leading closed-source systems. Observing that existing benchmarks inadequately reflect real-world user needs, we introduce GALA (General AI Life Assistants), a benchmark that mines latent research interests from users' historical browsing behavior. We further develop a rendering agent that converts research reports into visually rich poster presentations, and demonstrate an end-to-end product, AutoResearch Your Interest, which delivers personalized deep research recommendations derived from individual browsing histories.

Jiarui Jin, Zexuan Yan, Shijian Wang +2

Xiaohongshu Inc.

Deep Research Agents Multi-Agent Collaboration AI Agent Benchmarks LLM Agent Planning

May 12, 2026cs.AI

Can LLM Agents Respond to Disasters? Benchmarking Heterogeneous Geospatial Reasoning in Emergency Operations

Operational disaster response goes beyond damage assessment, requiring responders to integrate multi-sensor signals, reason over road networks, populations and key facilities, plan evacuations, and produce actionable reports. However, prior work largely isolates remote-sensing perception or evaluates generic tool use, leaving the end-to-end workflows of emergency operations underexplored. In this paper, we introduce Disaster Operational Response Agent benchmark (DORA), the first agentic benchmark for end-to-end disaster response: 515 expert-authored tasks across 45 real-world disaster events spanning 10 types, paired with expert-verified, replayable gold trajectories totaling 3,500 tool-call steps. Tasks span five dimensions that cover the operational disaster-response pipeline: disaster perception, spatial relational analysis, rescue and evacuation planning, temporal evolution reasoning, and multi-modal report synthesis. Agents compose calls from a 108-tool MCP library over heterogeneous geospatial data: optical, SAR, and multi-spectral imagery across single-, bi-, and multi-temporal sequences (0.015-10m GSD), complemented by elevation and social vector layers. We comprehensively evaluate 13 frontier LLMs on our benchmark, revealing three persistent challenges: 1) disaster-domain grounding exposes unique failure modes (damage-semantic grounding, sensor-modality mismatch, and disaster-pipeline composition); 2) agents are doubly bottlenecked by tool selection and argument grounding, where gold tool-order hints improve accuracy by only 1.08-4.40%, and alternative scaffolds yield at most a 3.24% gain; 3) compositional fragility scales with trajectory length, the agent-to-gold gap widening from 7% to 56% on long pipelines. DORA establishes a rigorous testbed for operationally reliable disaster-response agents.

Junjue Wang, Weihao Xuan, Heli Qi +7

1The University of Tokyo · 2RIKEN AIP · 3Waseda University +1

LLM Agent Evaluation Geospatial Reasoning AI Agent Benchmarks Disaster Response Tool-Using Agents Multimodal Agents+1

May 12, 2026cs.RO

PRISM: : Planning and Reasoning with Intent in Simulated Embodied Environments

When an LLM-based embodied agent fails at a household task, the culprit could be misidentified objects, forgotten sub-goals, or poor action sequencing -- yet existing benchmarks report only a single success rate, making it impossible to tell which cognitive module is responsible. We present PRISM, a diagnostic benchmark that reframes this problem: rather than asking only \textit{did the agent succeed?}, PRISM asks \textit{which capability is most likely responsible for failure?} Built on five photorealistic multi-room apartments (4--8 rooms each), PRISM structures 300 human-verified tasks into three capability tiers -- \textit{Basic Ability}, \textit{Reasoning Ability}, and \textit{Long-horizon Ability} -- that isolate perception-to-action grounding, implicit intent resolution, and sustained multi-step coordination respectively. PRISM exposes an agent-agnostic executable action API that allows arbitrary agents: LLM agents, VLM agents, symbolic planners, RL policies, and hybrid systems, to be evaluated end-to-end under the same benchmark protocol. To support deeper diagnosis, optional probes for perception, memory, and planning can be adopted, replaced, or bypassed entirely, enabling controlled component-level analysis when desired. Experiments on seven contemporary LLMs establish a clear hierarchy: explicit spatial grounding is not the dominant failure source under oracle perception, implicit intent resolution is a significant bottleneck for all model families, and long-horizon coordination exposes a stark capability cliff -- lightweight models collapse to as low as 20.0% success while simultaneously consuming more tokens than their frontier counterparts, a signature of compensatory over-reasoning rather than genuine planning capability. Project page: link.

Yunn Kang Lim, Pengzhan Sun, Ziyi Bai +4

1A*STAR · 2National University of Singapore · 3BAAI

Long-Horizon Agent Evaluation AI Agent Benchmarks Embodied AI Embodied Reasoning

May 12, 2026cs.CL

An Empirical Study of Automating Agent Evaluation

Agent evaluation requires assessing complex multi-step behaviors involving tool use and intermediate reasoning, making it costly and expertise-intensive. A natural question arises: can frontier coding assistants reliably automate this evaluation process? Our study shows that simply prompting coding assistants is insufficient for this task. Without domain-specific evaluation knowledge, frontier coding assistants achieve only a 30% execution success rate and produce over-engineered evaluations averaging 12+ metrics per agent, indicating that strong coding ability does not automatically translate to reliable agent evaluation. We introduce EvalAgent, an AI assistant that automates the end-to-end agent evaluation pipeline. EvalAgent encodes evaluation domain expertise as evaluation skills (procedural instructions, reusable code and templates, and dynamically retrieved API documentation) that compose into a trace-based pipeline producing complete evaluation artifacts including metrics, executable code, and reports. To systematically assess generated evaluations, we introduce a meta-evaluation framework alongside AgentEvalBench, a benchmark comprising 20 agents, each paired with evaluation requirements and test scenarios. We further propose the Eval@1 metric to measure whether generated evaluation code both executes and yields meaningful results on the first run. Our experiments show that EvalAgent produces focused evaluations, improving Eval@1 from 17.5% to 65%, and achieving 79.5% human expert preference over baseline approaches. Further ablation studies show that evaluation skills are critical for handling complex evaluation: removing them causes Eval@1 to drop significantly from 65% to 30%.

Kang Zhou, Sangmin Woo, Haibo Ding +15

AWS AI Labs

Agent Evaluation AI Agent Benchmarks Automated Evaluation

May 11, 2026gr-qc

gwBenchmarks: Stress-Testing LLM Agents on High-Precision Gravitational Wave Astronomy

Modern gravitational wave astronomy relies on modeling tasks that often require months of graduate-level effort, including building fast waveform surrogates from expensive numerical relativity simulations, modeling orbital dynamics of black holes, fitting merger remnant properties and constructing template banks. These problems demand extreme precision to support detection and parameter inference, with state-of-the-art models achieving

\lesssim 10^{-4}

relative error. We study whether state-of-the-art LLM coding agents can perform such end-to-end scientific modeling, where success requires constructing models with stringent accuracy criteria and reasoning about physical systems. We introduce gwBenchmarks, a suite of eight tasks grounded in gravitational wave analytic calculations and numerical simulations collectively representing over

10^8

core-hours of compute. The tasks span interpolation, regression, and high-dimensional time-series modeling, requiring a combination of numerical methods, machine learning, and physics-informed approaches. In preliminary experiments, agents frequently relied on proxy metrics, partial evaluation, or fabricated results to spuriously complete tasks. We therefore implement an external pre-defined framework to gauge agent progress. Evaluating twelve coding agents, we find no consistent winner. On the easiest task, multiple agents converge to the same cubic spline solution, with one rediscovering a coordinate transformation widely used in the literature. On harder tasks like analytic waveform modeling, all agents fall 1-2 orders of magnitude short of domain requirements and exhibit systematic failures, including metric misuse, constraint violations, and result fabrication. Our code, data, and website are publicly available.

Tousif Islam, Digvijay Wadekar, Zihan Zhou

Kavli Institute for Theoretical Physics, University of California Santa Barbara, Kohn Hall, Lagoon Rd, Santa Barbara, CA 93106 · Center for Gravitational Physics, University of Texas at Austin, Austin, TX 78712, USA · Department of Physics, Princeton University, Princeton, NJ 08540, USA

Gravitational-Wave Astronomy AI Coding Agents AI Agent Benchmarks LLM Agent Reliability

May 11, 2026cs.CV

ABRA: Agent Benchmark for Radiology Applications

Existing medical-agent benchmarks deliver imaging as pre-selected samples, never as an environment the agent must navigate. We introduce ABRA, a radiology-agent benchmark in which the agent operates an OHIF viewer and an Orthanc DICOM server through twenty-one function-calling tools that span slice navigation, windowing, series selection, pixel-coordinate annotation, and structured reporting. ABRA contains 655 programmatically generated tasks across three difficulty tiers and eight types (viewer control, metadata QA, vision probe, annotation, longitudinal comparison, BI-RADS reporting, and oracle variants of annotation and BI-RADS reporting), drawn from LIDC-IDRI, Duke Breast Cancer MRI, and NLST New-Lesion LongCT. Each episode is scored along Planning, Execution, and Outcome (Bluethgen et al., 2025) by task-type-specific automatic scorers. Ten current models, five closed-weight and five open-weight, reach at least 89% Execution on real annotation but only 0-25% Outcome; on the paired oracle variant where a simulated detector supplies the finding, Outcome on the same task reaches 69-100% across the models evaluated, localising the bottleneck to perception rather than tool orchestration. Code, task generators, and scorers are released at https://github.com/Luab/ABRA

Bulat Maksudov, Vladislav Kurenkov, Kathleen M. Curran +1

School of Computing Dublin City University Dublin, Ireland · AXXX · School of Medicine University College Dublin Dublin, Ireland

Radiology Report Generation Healthcare Medical Image Analysis AI Agent Evaluation AI Agent Benchmarks

May 11, 2026cs.AI

Do Vision-Language-Models show human-like logical problem-solving capability in point and click puzzle games?

Vision-Language(-Action) Models (VLMs) are increasingly applied to interactive environments, yet existing benchmarks often overlook the complex physical reasoning required for point-and-click puzzle games. This paper introduces Vision-Language Against The Incredible Machine (VLATIM), a benchmark designed to evaluate human-like logical problem-solving capabilities within the classic physics puzzle game The Incredible Machine 2 (TIM). Unlike existing benchmarks, VLATIM specifically targets the critical gap between high-level logical reasoning and continuous action spaces requiring precise mouse interactions. This benchmark is structured into five progressive parts, assessing capabilities that range from basic visual grounding and domain understanding to multi-step manipulation and full puzzle solving. Our results reveal a significant disparity between reasoning and execution. While large proprietary models demonstrate superior planning abilities, they struggle with precise visual grounding. Consequently, they do not yet show human-like problem-solving capabilities.

Maximilian Triebel, Marco Menner, Dominik Helfenstein

Institute of Artificial Intelligence, University of Stuttgart, Stuttgart, Germany.

VLM Evaluation Physical Reasoning AI Agent Benchmarks VLM Reasoning Visual Grounding

May 11, 2026physics.soc-ph

Strategic Coercion Within Alliances: The Greenland Sovereignty Game as an AI Stress Test

What happens when the strongest alliance member pressures a weaker member over territory and strategic control? We examine the Greenland sovereignty crisis as a stress test for LLM geopolitics, centered on the 2019-2026 U.S. push to acquire Greenland from the Kingdom of Denmark. The crisis nests two collective-action problems: Arctic strategic control and whether NATO can enforce alliance norms against the dominant member. We develop three games (asymmetric coercion; a NATO assurance game with a critical-mass tipping point; a triadic extensive-form game with social preferences) and test them with a multi-agent simulation in which eight frontier LLMs play six geopolitical roles (United States, Denmark, Greenland, NATO, Russia, Canada) across 3,604 completed games and 108,120 action observations. Using inverse game theory, we recover each model's structural utility parameters (alpha, beta, gamma, delta, eta) for material self-interest, reciprocity, inequality aversion, norm respect, and commitment consistency. Three findings stand out. First, all eight models become more escalatory under coercion framing (four-action escalation rises from 10.7% to 28.6%). Second, Chinese-origin models show systematically different power-weight profiles from Western-origin models when playing the U.S. role. Third, peaceful US acquisition emerges in only 1.9% of clean games and only 3 of 8 frontier models ever achieve it, most prominently DeepSeek V3.2, which executes a stable five-round playbook through the metropole. Prompts emphasizing jus cogens and self-determination reduce escalation back near baseline in the English-only confirmatory sample; multilingual contrasts are reported as exploratory sensitivity checks. We position this as a structural benchmark for LLM geopolitical behavior, complementing action-frequency benchmarks.

Rommin Adl, Peyton Williams

Grinnell College

Multi-Agent LLM Systems AI Agent Benchmarks

May 11, 2026cs.IR

RecoAtlas: From Semantic Plausibility to Set-Level Utility in LLM Recommendation Agents

LLM recommendation agents increasingly produce structured recommendation reports: sets of items accompanied by natural-language justifications. Yet existing evaluations often reduce this setting to reranking small shortlisted candidate sets or judge reports mainly by semantic plausibility. We introduce Recommendation Atlas (Agentic Tool-Level Assessment for Shopping), or RecoAtlas, a benchmark and toolkit for evaluating shopping agents with behavior-grounded metrics. RecoAtlas complements held-out interaction metrics with learned utility proxies for relevance, complementarity, and diversity derived from interaction data, while separately measuring semantic coherence and explanation quality. Its controlled tool environment exposes agents to either semantic, behavior-aligned, or faulty tools, enabling diagnosis of whether performance gains arise from stronger reasoning, better signals, or more effective tool-use policies. Across controlled experiments, we show that RecoAtlas exhibits key properties of a meaningful benchmark for agentic systems: performance scales with model capacity and test-time compute, improves with stronger and better-aligned tools, degrades under noisy or misaligned signals, and reveals that semantic plausibility does not necessarily capture behavior-grounded utility. RecoAtlas provides a foundation for developing and evaluating shopping assistants that optimize not only for plausible recommendations, but also for coherent, behaviorally grounded recommendation sets.

Imad Aouali, Flavian Vasile, Otmane Sakhi +2

Criteo AI Lab

Tool-Use Evaluation LLM Agent Evaluation AI Agent Benchmarks Recommender Systems Large Language Model-Based Recommendation

May 11, 2026cs.CL

WildClawBench: A Benchmark for Real-World, Long-Horizon Agent Evaluation

Large language and vision-language models increasingly power agents that act on a user's behalf through command-line interface (CLI) harnesses. However, most agent benchmarks still rely on synthetic sandboxes, short-horizon tasks, mock-service APIs, and final-answer checks, leaving open whether agents can complete realistic long-horizon work in the runtimes where they are deployed. This work presents WildClawBench, a native-runtime benchmark of 60 human-authored, bilingual, multimodal tasks spanning six thematic categories. Each task averages roughly 8 minutes of wall-clock time and over 20 tool calls, and runs inside a reproducible Docker container hosting an actual CLI agent harness (OpenClaw, Claude Code, Codex, or Hermes Agent) with access to real tools rather than mock services. Grading is hybrid, combining deterministic rule-based checks, environment-state auditing of side effects, and an LLM/VLM judge for semantic verification. Across 19 frontier models, the best, Claude Opus 4.7, reaches only 62.2% overall under OpenClaw, while every other model stays below 60%, and switching harness alone shifts a single model by up to 18 points. These results show that long-horizon, native-runtime agent evaluation remains a far-from-resolved task for current frontier models. We release the tasks, code, and containerized tooling to support reproducible evaluation.

Shuangrui Ding, Xuanlang Dai, Long Xing +14

1Shanghai AI Laboratory, 2The Chinese University of Hong Kong · 1Shanghai AI Laboratory, 3Fudan University · 1Shanghai AI Laboratory, 4University of Science and Technology of China +8

Computer-Use Agent Benchmarks Tool-Augmented Language Model Agents Terminal Agents LLM Agent Evaluation Long-Horizon Agent Evaluation Agent Evaluation+2

May 11, 2026cs.LG

AssayBench: An Assay-Level Virtual Cell Benchmark for LLMs and Agents

Recent advances in machine learning and large-scale biological data collections have revived the prospect of building a virtual cell, a computational model of cellular behavior that could accelerate biological discovery. One of the most compelling promises of this vision is the ability to perform in silico phenotypic screens, in which a model predicts the effects of cellular perturbations in unseen biological contexts. This task combines heterogeneous textual inputs with diverse phenotypic outputs, making it particularly well-suited to LLMs and agentic systems. Yet, no standard benchmark currently exists for this task, as existing efforts focus on narrower molecular readouts that are only indirectly aligned with the phenotypic endpoints driving many real-world drug discovery workflows. In this work, we present AssayBench, a benchmark for phenotypic screen prediction, built from 1,920 publicly available CRISPR screens spanning five broad classes of cellular phenotypes. We formulate the screen prediction task as a gene rank prediction for each screen and introduce the adjusted nDCG, a continuous metric for comparing performance across heterogeneous assays. Our extensive evaluation shows that existing methods remain far from empirically estimated performance ceilings and zero-shot generalist LLMs outperform biology-specific LLMs and trainable baselines. Optimization techniques such as fine-tuning, ensembling, and prompt optimization can further improve LLM performance on this task. Overall, AssayBench offers a practical testbed for measuring progress toward in silico phenotypic screening and, more broadly, virtual cell models.

Edward De Brouwer, Carl Edwards, Alexander Wu +9

Genentech

LLM Evaluation AI Agent Benchmarks

May 11, 2026cs.AI

MaD Physics: Evaluating information seeking under constraints in physical environments

Scientific discovery is fundamentally a resource-constrained process that requires navigating complex trade-offs between the quality and quantity of measurements due to physical and cost constraints. Measurements drive the scientific process by revealing novel phenomena to improve our understanding. Existing benchmarks for evaluating agents for scientific discovery focus on either static knowledge-based reasoning or unconstrained experimental design tasks, and do not capture the ability to make measurements and plan under constraints. To bridge this gap, we propose Measuring and Discovering Physics (MaD Physics), a benchmark to evaluate the ability of agents to make informative measurements and conclusions subject to constraints on the quality and quantity of measurements. The benchmark consists of three environments, each based on a distinct physical law. To mitigate contamination from existing knowledge, MaD Physics includes altered physical laws. In each trial, the agent makes measurements of the system until it exhausts an allotted budget and then the agent has to infer the underlying physical law to make predictions about the state of the system in the future. MaD Physics evaluates two fundamental capabilities of scientific agents: inferring models from data and planning under constraints. We also demonstrate how MaD Physics can be used to evaluate other capabilities such as multimodality and in-context learning. We benchmark agents on MaD Physics using four Gemini models (2.5 Flash Lite, 2.5 Flash, 2.5 Pro, and 3 Flash), identifying shortcomings in their structured exploration and data collection capabilities and highlighting directions to improve their scientific reasoning.

Moksh Jain, Mehdi Bennani, Johannes Bausch +4

Google DeepMind · Mila – Quebec AI Institute, Université de Montréal · Work done during an internship at Google DeepMind

Scientific Reasoning AI Agent Benchmarks System Identification

May 11, 2026cs.AI

ComplexMCP: Evaluation of LLM Agents in Dynamic, Interdependent, and Large-Scale Tool Sandbox

Current LLM agents are proficient at calling isolated APIs but struggle with the "last mile" of commercial software automation. In real-world scenarios, tools are not independent; they are atomic, interdependent, and prone to environmental noise. We introduce

\textbf{ComplexMCP}

, a benchmark designed to evaluate agents in these rigorous conditions. Built on the Model Context Protocol (MCP),

\textbf{ComplexMCP}

provides over 300 meticulously tested tools derived from 7 stateful sandboxes, ranging from office suites to financial systems. Unlike existing datasets, our benchmark utilizes a seed-driven architecture to simulate dynamic environment states and unpredictable API failures, ensuring a deterministic yet diverse evaluation. We evaluate various LLMs across full-context and RAG paradigms, revealing a stark performance gap: even top-tier models fail to exceed a 60% success rate, far trailing human performance 90%. Granular trajectory analysis identifies three fundamental bottlenecks: (1)

\textbf{tool retrieval saturation}

as action spaces scale; (2)

\textbf{over-confidence}

, where agents skip essential environment verifications; and (3)

\textbf{strategic defeatism}

, a tendency to rationalize failure rather than pursuing recovery. These findings underscore the insufficiency of current agents for interdependent workflows, positioning

\textbf{ComplexMCP}

as a critical testbed for the next generation of resilient autonomous systems.

Yuanyang Li, Xue Yang, Longyue Wang +2

1Zhejiang University · 2Zhejiang Lab · 3Alibaba Group

LLM Agent Evaluation AI Agent Benchmarks Tool-Using Agents LLM Agent Reliability