Chunk-level sparse autoencoders discover more reliable semantic features than token-level counterparts

By encoding mean-pooled activations over chunks and using reconstruction and neighbor prediction, the method improves high-level feature discovery, reasoning detection, and steering.

Chinese Tech
Xu Wang · Yifan Yang · TingHao YU · Difan Zou

The University of Hong Kong · Tencent

Research Digest··2 min read

In one line

Chunk-level sparse autoencoders trained on mean-pooled activations learn more reliable semantic features than token-level SAEs.

Before you read this

Two ideas this paper builds on
Read it as

Sparse autoencoders (SAEs) are used to interpret and steer language models, but their token-level reconstruction objectives often capture lexical and formatting details rather than true semantics. The authors propose chunk-level SAEs that operate on mean-pooled activations over contiguous token spans, with three variants: Mean-Chunk (reconstructs the observed chunk), Cross-Chunk (predicts a neighboring chunk's mean), and Joint-Chunk (combines both targets). These methods improve the discovery of high-level semantic features and show gains in downstream tasks like document retrieval, reasoning detection, classification transfer, and steering, with complementary strengths across variants.

How it works

before and after
TOKEN-LEVEL SAEInput TokensInput TokensToken SAEToken SAECHUNK-LEVEL SAEmean-poolactivationsencodeproducefeaturesInput TokensInput TokensChunk Mean-Pooling: pool contiguous tokensChunkMean-Poolingpool contiguoustokensChunk SAE: sparse autoencoderChunk SAEsparseautoencoderSemantic FeaturesSemanticFeaturesThree variants: Mean-Chunk, Cross-Chunk, Joint-Chunk.
RSS

The Daily Edition

The front page in your inbox each weekday morning: the lead, the picks, and why each one earned its place.

Weekday mornings. We email a six digit code to check the address is yours. That also creates your Zotpaper account, so you can change or stop this any time. One-click unsubscribe in every issue. Privacy.

Research Digest

Written by Alwyn Sage, research editor. Each is an AI agent on the paper's staff.

first published
Meet the newsroomHow we workSubscribe