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
4,861 results for “keyword:"lm"” in ICLR 2027. 940 withdrawn or desk rejected listed apart. clearsort:relevanceratingnewestmost read
[PDF]When Representations Collapse in Logical Reasoning: Diagnosing and Mitigating Cascade Failures in LLMs▼1.2k3.00
[PDF]RSCB: REGIONAL SEMANTIC CONTEXTUAL BANDIT FOR OPTIMAL SELECTION OVER LLM-GENERATED DYNAMIC CANDIDATES▼1361.33
[PDF]Prompting with Connections: Graph-Guided Attribute Selection for Test-Time Calibration▼4.5k2.67
[PDF]Structure Becomes Signal: Document Hierarchy as a Learned Behavioral Channel in LLM Evaluators▼1.2k3.00
[PDF]When Does Correction Become Repair? Mechanistic Auditing of Internal Interventions in Tool-Using LLMs▼1.2k1.67
[PDF]MARS: Multi-Evidence Adaptive Search with State-Preserving Control for Repository-Level Program Repair▼4.3k2.00
[PDF]Resolving Score-Ordering Reversal in Watermarked Text: An Asymmetric Two-Sided Test for LLM Text Detection▲5103.00
[PDF]What Should LLMs See, and How Often Should They Plan? World-State Encodings for Real-Time Games▼1361.33
[PDF]MICO: VERIFIABLE MILESTONES FOR PROVABLY BOUNDED CREDIT ASSIGNMENT IN AGENTIC REINFORCEMENT LEARNING▲11.00
[PDF]Lies We Can See: Joint Verbal and Non-Verbal Deception by VLM Agents in Embodied Social Interactions▲3.0k2.33
Withdrawn or desk rejected (940)
[PDF]When Is a Policy Violation a Model Failure? Policy Observability and Failure Attribution in Agent-Safety EvaluationDesk rejected
[PDF]-Bench: Benchmarking Proactive Speech Interaction with Full-Duplex Conversational AIDesk rejected