NeurIPS 2026 reviews are up: reviews, scores, rebuttals and meta-reviews for 7,877 papers. See the highest rated
acceptodds: the papers of ICLR 2027
446 results for “keyword:"benchmarks"” in ICLR 2027. 117 withdrawn or desk rejected listed apart. clearsort:relevanceratingnewestmost read
[PDF]LLM Verifiers Are Wrong Together: Confident Consensus Errors Set the Floor for Failure Prediction▼1361.33
[PDF]EviVQA: A Dataset and Benchmark for Evidence-Grounded Evaluation of E-Commerce Visual Question Answering▼4.3k2.00
[PDF]An Agent Is More Than Its Model: Controlled Safety Measurement Across CLI Model–Harness Configurations▲3.0k2.33
[PDF]Neuro-Symbolic Anchored Governance for Alleviating Action-Level Hallucinations in Large Language Model Agents▲3.0k2.33
[PDF]FIELD: A Prospective Cross-Sport Benchmark for Language-Model Forecasting and Information Use▲6.1k2.00
[PDF]CurriculumWorldBench: Benchmarking Autonomous Reconstruction of Executable University Curricula from the Web▲3.0k2.33
[PDF]StoreBench: A Live-Commerce Environment for Evaluating and Training Autonomous Operator Agents▲3.0k2.33
[PDF]Perturb-Path: Benchmarking H&E Pathology Foundation Models with in vivo Genetic Perturbations in a Lung Cancer Model▲5.4k3.00
[PDF]CLASHarness: Benchmarking Multi-Agent Conflict in Computer Use via a Controllable and Verifiable Generation Pipeline▲1.9k3.67
[PDF]AhaBench: Do Agents Turn Experience into Reusable Insights? A Long-Horizon Benchmark for Continual Learning▼1.2k1.67
[PDF]The Missing Complement: State-Conditioned Minimal Sufficient Evidence for Coding Agents▼5.8k2.33
[PDF]DARPinBench: A Dataset and Library-Screening Benchmark for Designed Ankyrin Repeat Proteins▼4.3k2.00
[PDF]IMPACT: Measuring Conditional Workflow Responses to Single-Span Model Substitution in Stateful LLM Agents▼1361.33
[PDF]TRUSTCONS: EXACTLY VERIFYING CONSISTENCY, COMPLETENESS , AND CONDITIONED RE-ENUMERATION IN LANGUAGE MODELS▼1361.33
Withdrawn or desk rejected (117)
[PDF]Counterfactual Auditing of Object Presence and Quantity Judgments in Vision-Language ModelsWithdrawn1.67
[PDF]Correlation Does Not Certify a Policy Choice: What Benchmarks Should Publish InsteadWithdrawn2.33
[PDF]EarthArena: Benchmarking Agents on Real-world, Long-Horizon Scientific Tasks in Earth ObservationWithdrawn1.67
[PDF]AskInfer-Bench: Profiles Are Evidence, Not Intent Labels, in Personalized Deep ResearchWithdrawn1.00
[PDF]CommerceAgentBench: Benchmarking Long-Horizon Agents in High-Fidelity, Stateful, and Reproducible Replicas of Real Online ServicesDesk rejected2.00