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
273 results for “keyword:"tool use"” in ICLR 2027. 61 withdrawn or desk rejected listed apart. clearsort:relevanceratingnewestmost read
[PDF]Learning to Zoom at the Right Difficulty: Dynamic Curricula for Visual Tool-Use Reinforcement Learning▲5103.00
[PDF]When Top-K Misses the Decision: Tool-Call Drift in Multi-Teacher On-Policy Distillation▼1361.33
[PDF]Track, Rank, Crack: Epistemic Working Memory Scales Multi-Hop Reasoning in Language Agents▼1361.33
[PDF]EarthVerse: Benchmarking Scientific Agents across Dynamic Earth Systems and Natural Hazards▼4.3k2.00
[PDF]Interface-Induced Trajectory Censoring: When Tool-Call Interfaces Distort Evaluation and RL Experience▼5.3k1.33
[PDF]Causal Experience Instantiation: Grounding Procedural Memory with Temporal Graphs for Tool-Use Agents▲5103.00
[PDF]Trajectory2Gym: Decision-Grounded Compositional Interaction Synthesis for Tool-Use Agents▼2.0k1.00
Withdrawn or desk rejected (61)
[PDF]When Is a Policy Violation a Model Failure? Policy Observability and Failure Attribution in Agent-Safety EvaluationDesk rejected
[PDF]Auditing Reward-Model Selection on Tool Trajectories: Identification, Policy Utility, and Natural TransferWithdrawn1.67
[PDF]Policy-Verified Self-Training: Learning Agent Compliance from Verifiable Policy RewardsWithdrawn1.67
[PDF]From Loading Skills to Knowing Them: Scaffolding-Removed Fine-Tuning for Agent Skill InternalizationWithdrawn