Get Confused Cautiously: Textual Sequence Memorization Erasure with Selective Entropy Maximization
This addresses privacy and copyright issues for users of large language models by improving the ability to forget memorized text without harming model performance, though it is incremental as it builds on existing erasure methods.
The paper tackles the problem of textual sequence memorization (TSM) in large language models, which raises privacy and copyright concerns, by proposing a new framework called Entropy Maximization with Selective Optimization (EMSO) that achieves a better trade-off between erasing memorized text and preserving model utility, with extensive experiments showing effectiveness across three model scales.
Large Language Models (LLMs) have been found to memorize and recite some of the textual sequences from their training set verbatim, raising broad concerns about privacy and copyright issues when using LLMs. This Textual Sequence Memorization (TSM) phenomenon leads to a high demand to regulate LLM output to prevent it from generating certain memorized text to meet user requirements. However, our empirical study reveals that existing methods for TSM erasure fail to forget massive memorized samples without substantially jeopardizing the model utility. To achieve a better trade-off between the effectiveness of TSM erasure and model utility in LLMs, our paper proposes a new framework based on Entropy Maximization with Selective Optimization (EMSO), where the updated weights are chosen with a novel contrastive gradient metric without any participation of additional model or data. Our analysis shows that training with the entropy maximization loss has a more stable optimization process and better keeps model utility than existing methods. The contrastive gradient metric localizes the most influential weight for TSM erasure by taking both the gradient magnitude and direction into consideration. Extensive experiments across three model scales demonstrate that our method excels in handling large-scale forgetting requests while preserving model ability in language generation and reasoning.