稀疏自编码表示中的特征竞争：对LLMs中不确定性驱动特征竞争的机理研究
Feature Rivalry in Sparse Autoencoder Representations: A Mechanistic Study of Uncertainty-Driven Feature Competition in LLMs
AI总结：
研究稀疏自编码中特征竞争与模型不确定性之间的关系，发现高熵问题在特定层产生更强的竞争，且竞争方向影响输出，竞争分数预测答案正确性。
摘要：
稀疏自编码（SAEs）将大语言模型表示分解为可解释的特征，但这些特征在不确定性下的交互仍不明确。本文引入特征竞争——负相关的SAE特征对，并在Gemma-2-2B中研究其是否为模型不确定性的机理标志。通过控制实验发现，高熵问题在第0和12层产生更强的竞争，将不确定性局部化到残差流特定阶段。进一步测试发现，沿竞争方向（vec_A - vec_B）的激活引导比随机方向在低引导乘数下产生更多输出变化。最终，基于活跃SAE特征解码器向量的成对余弦相似度计算的每提示竞争分数预测答案正确性（AUROC=0.689），接近但不匹配softmax置信度（AUROC=0.808）。
英文摘要：
Sparse Autoencoders (SAEs) decompose large language model representations into interpretable features, but how these features interact under uncertainty remains poorly understood. We introduce Feature Rivalry -- negatively correlated SAE feature pairs -- and study whether rivalry serves as a mechanistic signature of model uncertainty in Gemma-2-2B using Gemma Scope SAEs. Through a controlled within-domain experiment on PopQA split by response entropy, we find that high-entropy questions produce significantly stronger feature rivalry at layers 0 and 12 relative to low-entropy questions (p=5.3x10^-26 and p=5.8x10^-5 respectively), localizing uncertainty to specific processing stages in the residual stream. We then test whether rivalry is causally upstream of model outputs via activation steering along rivalry axes -- finding that steering along the rivalry direction (vec_A - vec_B) causes more output changes than random directions at low steering multipliers across 15 of 20 rival feature pairs. Finally, a per-prompt rivalry score derived from pairwise cosine similarities of active SAE feature decoder vectors predicts answer correctness (AUROC=0.689), approaching but not matching softmax confidence (AUROC=0.808).
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。
