4 papers
Preserving Reasoning Diversity in Reinforcement Learning with Verifiable Rewards via Adaptive Replay
Wen-Zhi Li, Sainyam Galhotra, Claire Cardie
Reinforcement Learning with Verifiable Rewards (RLVR) improves the reasoning capabilities of large language models, but often reduces the diversity of their reasoning behaviors. We…
Beyond Correctness: Resolving Underspecification in Agentic Text-to-SQL
Wen-Zhi Li, Yue Gong, Konstantinos Kanellis +1
Agentic Text-to-SQL systems can interact with users to clarify underspecified queries before generating SQL. However, a correct execution result does not necessarily imply that the…
Dual-Layer Agentic Memory with Fast Write Routing and Slow Consolidation
Wenzhi Li, Dong Nie, Rui Lan +7
Large language model (LLM) agents operate in dynamic environments where knowledge continuously evolves. Existing memory systems typically treat external memory as a monotonically g…
ChainPrune: Evaluating and Reducing Redundancy in Long Chain-of-Thought Reasoning
Weihang Pan, Zhengxu Yu, Yuxiang Zhang +5
Chain-of-Thought (CoT) reasoning has significantly enhanced the multi-step problem-solving capabilities of large language models (LLMs) by introducing explicit intermediate reasoni…