Search papers, labs, and topics across Lattice.
The paper introduces CoT2Edit, a novel framework for improving LLMs' ability to edit knowledge by training them to reason through Chain-of-Thought (CoT) examples generated from both structured and unstructured data. This is important because existing knowledge editing methods struggle with generalization and are limited to structured fact triples. CoT2Edit, which combines supervised fine-tuning (SFT), Group Relative Policy Optimization (GRPO), and Retrieval-Augmented Generation (RAG), demonstrates strong generalization across diverse knowledge editing scenarios.
Forget brittle fact memorization – CoT2Edit teaches LLMs to *reason* their way to updated knowledge, unlocking robust generalization in real-world scenarios.
Large language models (LLMs) can effectively handle outdated information through knowledge editing. However, current approaches face two key limitations: (I) Poor generalization: Most approaches rigidly inject new knowledge without ensuring that the model can use it effectively to solve practical problems. (II) Narrow scope: Current methods focus primarily on structured fact triples, overlooking the diverse unstructured forms of factual information (e.g., news, articles) prevalent in real-world contexts. To address these challenges, we propose a new paradigm: teaching LLMs to edit knowledge via Chain of Thoughts (CoTs) reasoning (CoT2Edit). We first leverage language model agents for both structured and unstructured edited data to generate CoTs, building high-quality instruction data. The model is then trained to reason over edited knowledge through supervised fine-tuning (SFT) and Group Relative Policy Optimization (GRPO). At inference time, we integrate Retrieval-Augmented Generation (RAG) to dynamically retrieve relevant edited facts for real-time knowledge editing. Experimental results demonstrate that our method achieves strong generalization across six diverse knowledge editing scenarios with just a single round of training on three open-source language models. The codes are available at https://github.com/FredJDean/CoT2Edit.