一作ACL 2026Findings· Findings of ACL
Reason- KE ++: Aligning the Process, Not Just the Outcome, for Faithful LLM Knowledge Editing
Yuchen Wu · Liang Ding · Li Shen · Dacheng Tao
摘要
Aligning Large Language Models (LLMs) to be faithful to new knowledge in complex, multi-hop reasoning tasks is a critical, yet unsolved, challenge. We find that SFT-based methods, e.g., Reason-KE, while state-of-the-art, suffer from a “faithfulness gap”: they optimize for format mimicry rather th...