跳到正文
研踪PaperTrail
← 研踪 PaperTrail

arXiv:2607.17117 · v3

Persistent Sparse Autoencoders: Learning Feature-Specific Timescales in Language Model Representations

首次提交 2026-07-19 16:07
版本更新 2026-11-19 16:02
雷达观测 2026-11-20 19:58
arXiv ↗PDF ↗

摘要

Sparse autoencoders (SAEs) decompose language model activations into sparse features, yet these models traditionally encode each token independently, failing to expose information that persists across a sequence. We first show that temporal persistence can naturally emerge in standard SAE features: after a feature activates, the hidden state remains aligned with its direction, and past activations reduce the reconstruction error of later hidden states. How long this lasts varies widely across features. We therefore introduce Persistent Sparse Autoencoders (Persistent SAEs), an extension of standard SAEs that learns a persistence coefficient for each feature, allowing feature-specific timescales to emerge from reconstruction. Our experiments show that Persistent SAEs retain competitive reconstruction quality while learning a spectrum of timescales: short-timescale (fast) features remain locally interpretable, whereas long-timescale (slow) features accumulate higher-level contextual information. Moreover, we show in a prompt-injection monitoring case study that slow features retain injection signals over long contexts and causally affect monitor judgements. These results suggest that Persistent SAEs offer new opportunities for interpreting and monitoring language models via persistent features.

AI 论文解读

基于摘要 · 非全文审阅

此版本尚未生成 AI 解读。解读由后台任务生成并按论文版本保存,登录后共享已有结果。

版本差异

本站首次收录为 v3,暂无更早版本可对照。

观测到的版本

相邻记录可能跳过中间版本;首次发现 v2 及以上也会标为版本更新。

v3 · 2026-11-19 16:02 · 最新观测版本

Persistent Sparse Autoencoders: Learning Feature-Specific Timescales in Language Model Representations

Sparse autoencoders (SAEs) decompose language model activations into sparse features, yet these models traditionally encode each token independently, failing to expose information that persists across a sequence. We first show that temporal persistence can naturally emerge in standard SAE features: after a feature activates, the hidden state remains aligned with its direction, and past activations reduce the reconstruction error of later hidden states. How long this lasts varies widely across features. We therefore introduce Persistent Sparse Autoencoders (Persistent SAEs), an extension of standard SAEs that learns a persistence coefficient for each feature, allowing feature-specific timescales to emerge from reconstruction. Our experiments show that Persistent SAEs retain competitive reconstruction quality while learning a spectrum of timescales: short-timescale (fast) features remain locally interpretable, whereas long-timescale (slow) features accumulate higher-level contextual information. Moreover, we show in a prompt-injection monitoring case study that slow features retain injection signals over long contexts and causally affect monitor judgements. These results suggest that Persistent SAEs offer new opportunities for interpreting and monitoring language models via persistent features.

查看此版本解读 →查看此版本原文 ↗