Similar Items: Senses Wide Shut: A Representation-Action Gap in Omnimodal LLMs
- LLMs Improving LLMs: Agentic Discovery for Test-Time Scaling
- Misaligned by Reward: Socially Undesirable Preferences in LLMs
- Beyond Benchmarks: MathArena as an Evaluation Platform for Mathematics with LLMs
- Beyond Confidence: Rethinking Self-Assessments for Performance Prediction in LLMs
- SERE: Structural Example Retrieval for Enhancing LLMs in Event Causality Identification
- Can RL Teach Long-Horizon Reasoning to LLMs? Expressiveness Is Key