Reinforced Lifelong Editing for Language Models
This addresses the challenge of keeping LLMs accurate over time for users relying on up-to-date information, representing an incremental advance in model editing techniques.
The paper tackles the problem of lifelong editing for large language models (LLMs) to update knowledge without retraining, proposing RLEdit, an RL-based method that improves effectiveness and efficiency, achieving a 59.24% improvement with only 2.11% of the time compared to existing approaches.
Large language models (LLMs) acquire information from pre-training corpora, but their stored knowledge can become inaccurate or outdated over time. Model editing addresses this challenge by modifying model parameters without retraining, and prevalent approaches leverage hypernetworks to generate these parameter updates. However, they face significant challenges in lifelong editing due to their incompatibility with LLM parameters that dynamically change during the editing process. To address this, we observed that hypernetwork-based lifelong editing aligns with reinforcement learning modeling and proposed RLEdit, an RL-based editing method. By treating editing losses as rewards and optimizing hypernetwork parameters at the full knowledge sequence level, we enable it to precisely capture LLM changes and generate appropriate parameter updates. Our extensive empirical evaluation across several LLMs demonstrates that RLEdit outperforms existing methods in lifelong editing with superior effectiveness and efficiency, achieving a 59.24% improvement while requiring only 2.11% of the time compared to most approaches. Our code is available at: https://github.com/zhrli324/RLEdit.