Representational alignment yields generalizable safety in language models

2026-09-03

AI Summary

论文探讨了通过优化语言模型中的潜在表示与人类道德判断的一致性，来提高模型在面对对抗性或不熟悉形式的有害意图时的安全性和鲁棒性。

Abstract

Aligning large language models (LLMs) is essential for their safe deployment. Current alignment methods mainly optimize observable responses, yet models remain vulnerable when the same harmful intent is recast in unfamiliar or adversarial forms that humans can easily recognize. Prototype theory offers an account of this adaptability. Human concepts are represented around central cases, and new instances are categorized according to their graded typicality relative to these prototypes. Here we show that such categorization of moral concepts is weakly preserved in current LLMs. Across 23 LLMs, models often failed to distinguish opposed moral categories or preserve fine-grained typicality within each category. These deficits persist across parameter sizes and alignment stages. We developed representational similarity optimization, which directly aligns the latent representations in LLMs with the categorization expressed in human moral judgements, without supervising generated responses. In matched experiments using the same 251,334 moral annotations, standard behavioral alignment learned the intended moral judgements at the response level while leaving the categorization structure largely unchanged and increasing vulnerability across adversarial evaluations. Reorganizing moral categorization produced more modest gains in explicit judgements but consistently improved adversarial robustness across model scales on diverse benchmarks and attack strategies. Our findings provide functional support for the view that prototype-based categorization contributes to behavioral adaptability. They also show that transferring this representational principle to LLMs yields generalizable safety under adversarial conditions.

Problem

• large language models
• moral categorization
• adversarial conditions
• prototype theory
• alignment

Innovation

• representational similarity optimization
• moral categorization
• adversarial robustness