Enhancing Safety of Large Language Models via Embedding Space Separation
About
Large language models (LLMs) have achieved impressive capabilities, yet ensuring their safety against harmful prompts remains a critical challenge. Recent work has revealed that the latent representations (embeddings) of harmful and safe queries in LLMs typically exhibit linear separability, a property that has been exploited to construct attacks by perturbing the embeddings of harmful queries towards the safe subspace. Motivated by this observation, we propose a representation-level fine-tuning approach, named Embedding Space Separation (ES2), which improves LLM safety by explicitly enlarging the distance between harmful and safe representations in the embedding space. To prevent degradation of model's general capabilities, we introduce a Kullback-Leibler (KL) divergence regularization term into the loss function, which constrains the logits of the fine-tuned model to align with those of the original base model on harmless inputs. We evaluate our method on several open-source LLMs using standard safety benchmarks. Extensive experimental results demonstrate that our approach substantially improves model safety while maintaining comparable general capabilities.
Related benchmarks
| Task | Dataset | Result | Rank | |
|---|---|---|---|---|
| Defense Robustness | RepE Attack | DSR Keyword Match Rate100 | 16 | |
| Defense Robustness | Soft Prompt Attack | DSR Keyword Success Rate100 | 16 | |
| Defense Robustness | SCAV Attack | DSR (Keyword)80 | 16 | |
| Defense Robustness | AutoDAN Attack | Keyword Success Rate100 | 16 | |
| Defense Robustness | GCG Attack | Keyword Match Rate100 | 16 | |
| Defense Robustness | Direct Inquiry Vanilla | Keyword Match Rate100 | 16 |