AutoMedBench: Towards Medical AutoResearch with Agentic AI Models

• URL: http://arxiv.org/abs/2606.01961v2
• Date: Wed, 03 Jun 2026 05:43:27 GMT
• Title: AutoMedBench: Towards Medical AutoResearch with Agentic AI Models
• Abstract summary: AutoMedBench is a workflow-aware benchmark for autonomous medical-AI research.<n>It comprises long-horizon tasks with each run averaging 33 agent turns.
• Score: 31.335309699875264
• License: http://creativecommons.org/licenses/by-nc-sa/4.0/
• Abstract: Autonomous agents are increasingly expected to support end-to-end medical-AI research workflows, moving beyond isolated prediction tasks or short-form clinical question answering. However, existing medical agent benchmarks primarily evaluate final outputs, providing limited visibility into agent behavior within the research process. To address this gap, we present AutoMedBench, a workflow-aware benchmark for autonomous medical-AI research across diverse medical imaging and multimodal inference tasks, organizing agent execution into a unified five-stage workflow (S1-S5): Plan, Setup, Validate, Inference, and Submit. It comprises long-horizon tasks with each run averaging 33 agent turns, spanning five research tracks: segmentation, image enhancement, visual question answering (VQA), report generation, and lesion detection. Each task is evaluated under two difficulty tiers, Lite and Standard, which use the same data and metrics but differ in the amount of task-brief scaffolding, and each run is scored using both final task performance and S1-S5 stage scores, enabling stage-level analysis from the initial task brief to the final submitted artifact. Across thousands of recorded runs, stage-level scoring reveals that Validate is the weakest workflow stage on average, whereas Setup is the strongest, suggesting that current agents are better at making pipelines executable than at verifying their reliability. Post-run error analysis further shows that verification and submission failures dominate tagged errors, accounting for 37.7% and 38.1% of fired codes respectively, whereas task-understanding errors are rare at 0.9%, and runs with one fired error code have a 48% lower overall score than runs with no error code on average.

Related papers

• Towards Autonomous and Auditable Medical Imaging Model Development [53.326510777116766]
  We introduce AMID, an autonomous multi-agent framework for medical imaging model development.<n>AMID first proposes Data-Conditioned Method Planning, which refines coarse task-level search spaces into executable, parallelizable method lanes.<n>It then develops Verification-Guided Two-Stage Optimization, moving from broad early exploration of diverse method lanes to selective exploitation of promising candidates.
  arXiv  Detail & Related papers  (2026-07-12T00:54:06Z)
• DRFLOW: A Deep Research Benchmark for Personalized Workflow Prediction [44.59825034567626]
  Many enterprise tasks require an agent to identify concrete which is a sequence of action-steps.<n>We introduce DRFLOW, a benchmark for evaluating personalized predictions from heterogeneous sources.<n>DRFLOW contains 100 tasks across five domains, with 1,246 reference workflow steps grounded in more than 3,900 sources.
  arXiv  Detail & Related papers  (2026-06-16T17:22:07Z)
• EnactToM: An Evolving Benchmark for Functional Theory of Mind in Embodied Agents [75.01735520608075]
  Existing benchmarks mostly test literal Theory of Mind (ToM) by asking direct belief questions.<n>We introduce EnactToM, an evolving benchmark of 300 embodied multi-agent tasks set in a 3D household.<n>On the hard split, all seven evaluated frontier models score 0.0% Pass3 on functional task completion, while averaging 45.0% on literal belief probes.
  arXiv  Detail & Related papers  (2026-05-11T00:04:19Z)
• Unified Ultrasound Intelligence Toward an End-to-End Agentic System [8.884663182013282]
  USTri is a tri-stage ultrasound intelligence pipeline for unified multi-organ, multi-task analysis.<n>Stage I trains a universal generalist USGen on different domains to learn broad, transferable priors.<n>Stage II builds USpec keeping USGen frozen and finetuning heads.<n>Stage III introduces USAgent, which mimics clinician specialists for multi-step inference and structured reports.
  arXiv  Detail & Related papers  (2026-04-18T08:46:58Z)
• OccuBench: Evaluating AI Agents on Real-World Professional Tasks via Language Environment Simulation [57.505743202759646]
  OccuBench is a benchmark covering 100 real-world professional task scenarios across 10 industry categories and 65 specialized domains.<n>Our multi-agent synthesis pipeline automatically produces evaluation instances with guaranteed solvability, calibrated difficulty, and document-grounded diversity.
  arXiv  Detail & Related papers  (2026-04-13T00:27:32Z)
• GUIDE: Interpretable GUI Agent Evaluation via Hierarchical Diagnosis [17.279226216630065]
  We introduce GUIDE (GUI Understanding and Interpretable Diagnostic Evaluation), a framework that decomposes trajectory assessment into three sequential stages.<n>Subtask Diagnosis evaluates each unit in context, assigning a completion verdict and generating a structured error analysis with corrective recommendations.<n>We validate GUIDE on three benchmarks: an industrial e-commerce dataset of 932 trajectories, AGENTREWARDBENCH spanning five web agent tasks with 1302 trajectories, and AndroidBench for mobile device control.
  arXiv  Detail & Related papers  (2026-04-06T03:58:43Z)
• Agentic-MME: What Agentic Capability Really Brings to Multimodal Intelligence? [35.30497528897595]
  Agentic-MME is a process-verified benchmark for Multimodal Agentic Capabilities.<n>It contains 418 real-world tasks across 6 domains and 3 difficulty levels.<n>It features over 2,000 stepwise checkpoints that average 10+ person-hours of manual annotation per task.
  arXiv  Detail & Related papers  (2026-04-03T13:02:01Z)
• Human-Guided Agentic AI for Multimodal Clinical Prediction: Lessons from the AgentDS Healthcare Benchmark [0.5066646435185324]
  We investigate how human guidance of agentic AI can improve multimodal clinical prediction.<n>We present our approach to three benchmark challenges: 30-day hospital prediction, emergency department cost forecasting, and discharge readiness assessment.<n>Our approach ranked 5th overall in the healthcare domain, with a 3rd-place finish on the discharge readiness task.
  arXiv  Detail & Related papers  (2026-02-23T04:37:45Z)
• ResearchGym: Evaluating Language Model Agents on Real-World AI Research [48.46915933681714]
  We introduce ResearchGym, a benchmark and execution environment for evaluating AI agents on end-to-end research.<n>To instantiate this, we repurpose five oral and spotlight papers from ICML, ICLR, and ACL.<n>In a controlled evaluation of an agent powered by GPT-5, we observe a sharp capability--reliability gap.
  arXiv  Detail & Related papers  (2026-02-16T19:00:03Z)
• AgentIF-OneDay: A Task-level Instruction-Following Benchmark for General AI Agents in Daily Scenarios [49.90735676070039]
  The capacity of AI agents to effectively handle tasks of increasing duration and complexity continues to grow.<n>We argue that current evaluations prioritize increasing task difficulty without sufficiently addressing the diversity of agentic tasks.<n>We propose AgentIF-OneDay, aimed at determining whether general users can utilize natural language instructions and AI agents to complete a diverse array of daily tasks.
  arXiv  Detail & Related papers  (2026-01-28T13:49:18Z)
• ART: Action-based Reasoning Task Benchmarking for Medical AI Agents [0.0]
  We introduce Action-based Reasoning clinical Task benchmark for medical AI agents.<n>We identify three dominant error categories: retrieval failures, aggregation errors, and conditional logic misjudgments.<n>Our four-stage pipeline produces diverse, clinically validated tasks grounded in real patient data.
  arXiv  Detail & Related papers  (2026-01-13T21:26:11Z)
• Eigen-1: Adaptive Multi-Agent Refinement with Monitor-Based RAG for Scientific Reasoning [53.45095336430027]
  We develop a unified framework that combines implicit retrieval and structured collaboration.<n>On Humanity's Last Exam (HLE) Bio/Chem Gold, our framework achieves 48.3% accuracy.<n>Results on SuperGPQA and TRQA confirm robustness across domains.
  arXiv  Detail & Related papers  (2025-09-25T14:05:55Z)
• Which Agent Causes Task Failures and When? On Automated Failure Attribution of LLM Multi-Agent Systems [50.29939179830491]
  Failure attribution in LLM multi-agent systems remains underexplored and labor-intensive.<n>We develop and evaluate three automated failure attribution methods, summarizing their corresponding pros and cons.<n>The best method achieves 53.5% accuracy in identifying failure-responsible agents but only 14.2% in pinpointing failure steps.
  arXiv  Detail & Related papers  (2025-04-30T23:09:44Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.