一作ACL 2026Poster· Natural Language Processing
K now RL : Exploring Knowledgeable Reinforcement Learning for Factuality
Baochang Ren · Shuofei Qiao · Ningyu Zhang · Da Zheng · Huajun Chen
摘要
Slow-thinking Large Language Models (LLMs) have demonstrated strong reasoning capabilities but often suffer from severe hallucinations due to an inability to recognize their knowledge boundaries. Existing Reinforcement Learning (RL) approaches typically rely on outcome-oriented rewards, which can...