arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2502.20952cs.CRcs.LG
收藏

通过冻结训练高效越狱大型模型:较低层对有害内容表现出更高的敏感性

Efficient Jailbreaking of Large Models by Freeze Training: Lower Layers Exhibit Greater Sensitivity to Harmful Content

Peking University(北京大学);Ant Group(蚂蚁集团)

Hongyuan Shen, Min Zheng, Jincheng Wang, Yang Zhao

AI总结:

本研究通过统计分析发现大型模型较低层对有害内容更敏感,据此提出仅微调较低层的冻结训练方法,显著降低训练开销并保持高越狱成功率,且可泛化至其他模型。

摘要:

随着大型语言模型在各领域的广泛应用,其安全问题日益受到学术界和工业界的广泛关注。本研究对大型语言模型的参数进行采样和归一化处理,生成参数分布的可视化表示和热力图,揭示了隐藏层中某些层参数分布的显著差异。进一步分析中,我们计算每一层的统计指标,并基于这些指标计算综合敏感性得分,该得分识别出较低层对有害内容生成特别敏感。基于这一发现,我们采用冻结训练策略,仅对较低层进行有监督微调。实验结果表明,该方法显著减少了训练时间和GPU内存消耗,同时保持了较高的越狱成功率和较高的危害得分,优于在所有层应用LoRA方法进行有监督微调所取得的结果。此外,该方法已成功扩展到其他开源大型模型,验证了其在不同模型架构上的通用性和有效性。进一步地,我们将我们的方法与其他越狱方法进行比较,展示了我们方法的优越性能。通过创新性地提出一种逐层统计分析并比较大型模型参数的方法,本研究为大型模型的可解释性提供了新的见解。这些发现强调了在快速发展的大型语言模型领域进行持续研究和实施自适应安全措施的必要性,以防止潜在的越狱攻击风险,从而促进更稳健、更安全的大型语言模型的发展。

英文摘要:

With the widespread application of Large Language Models across various domains, their security issues have increasingly garnered significant attention from both academic and industrial communities. This study conducts sampling and normalization of the parameters of the LLM to generate visual representations and heatmaps of parameter distributions, revealing notable discrepancies in parameter distributions among certain layers within the hidden layers. Further analysis involves calculating statistical metrics for each layer, followed by the computation of a Comprehensive Sensitivity Score based on these metrics, which identifies the lower layers as being particularly sensitive to the generation of harmful content. Based on this finding, we employ a Freeze training strategy, selectively performing Supervised Fine-Tuning only on the lower layers. Experimental results demonstrate that this method significantly reduces training duration and GPU memory consumption while maintaining a high jailbreak success rate and a high harm score, outperforming the results achieved by applying the LoRA method for SFT across all layers. Additionally, the method has been successfully extended to other open-source large models, validating its generality and effectiveness across different model architectures. Furthermore, we compare our method with ohter jailbreak method, demonstrating the superior performance of our approach. By innovatively proposing a method to statistically analyze and compare large model parameters layer by layer, this study provides new insights into the interpretability of large models. These discoveries emphasize the necessity of continuous research and the implementation of adaptive security measures in the rapidly evolving field of LLMs to prevent potential jailbreak attack risks, thereby promoting the development of more robust and secure LLMs.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑