研究显示上下文压缩会削弱模型的拒绝对齐行为
Does Context Compression Preserve Refusal Alignment?
把长上下文压缩后模型可能更容易回答不该答的问题，这篇给出机制解释和一个不用重训练的修复方法，做 RAG 或长文本系统的人该看看。
一项 arXiv 研究发现，基于编码器的上下文压缩会在保持高重建保真度的同时，系统性削弱指令微调语言模型的拒绝行为，该效应跨多个模型家族和压缩架构持续存在。机制分析显示，压缩会削弱解码器已学习的拒绝方向上的激活。研究者提出 Memory Steering，一种轻量级推理时干预方法，可在不重训练的情况下将拒绝率恢复到接近基线水平，且完全在压缩表示空间内运行。结果表明语义保真不等于行为保真，压缩感知系统需要显式保留对齐相关特征。
Does Context Compression Preserve Refusal Alignment?
Context compression reduces inference cost by encoding inputs into compact representations while preserving semantic content. An open question is whether semantic preservation alone is sufficient to maintain downstream behaviours such as refusal alignment. We investigate this question and find that encoder-based compression systematically weakens refusal behaviour in instruction-tuned language models, despite high reconstruction fidelity. This effect persists across model families and compression architectures. Mechanistic analysis shows that compression attenuates activation along the decoder's learned refusal direction. We further explore Memory Steering, a lightweight inference-time intervention that restores refusal rates to near-baseline levels without retraining and operates entirely in compressed representation space. These results demonstrate that semantic preservation does not guarantee behavioural preservation under compression, highlighting the need to explicitly preserve alignment-relevant features in compression-aware systems.