Sourcepreprinttext stored

AxBench: Steering LLMs? Even Simple Baselines Outperform Sparse Autoencoders

Wu, Zhengxuan, Arora, Aryaman, Geiger, Atticus, Wang, Zheng, Huang, Jing, Jurafsky, Dan, Manning, Christopher D., Potts, ChristopherarXiv.orgPublished Jan 28, 2025Original︎ ↗Archived copy

Claims1

ExaminationsNone

TextStored, not divided

VersionsOne known

CopiesNone recorded

CorrectionsNone recorded

Facts checked Nov 12, 2026

ReachNo copies recorded

StructureUnderlying on 1 claim1 source downstream across 2 claims

EvidenceNot yet read

ReviewsNot reviewed

Examinations

None recorded. An examination is commissioned by a claim’s Steward for what its claim rests on; a review of the document for its own sake is asked for as a request bearing on it.

Text

Stored, not yet divided into passages.

As quoted

On steering, only ReFT-r1 is competitive with finetuning and prompting baselines, while SAEs fall behind both ReFT-r1 and difference-in-means on both axes.

On AxBench, SAE steering underperforms both prompting and difference-in-means representations.Verified

Not yet read by this claim’s Steward.

Lineage across 2 claims

Draws on · none recorded

This source

AxBench: Steering LLMs? Even Simple Baselines Outperform Sparse Autoencoders

underlying on 1 · in 2 claims' maps

Drawn on by · 1

Steering LLMs? Actually, Sparse Autoencoders can outperform simple baselinespreprintMay 29, 20261 claimCites it faithfully, on 1 claim.

Record

Kind: preprint
Authors: Wu, Zhengxuan, Arora, Aryaman, Geiger, Atticus, Wang, Zheng, Huang, Jing, Jurafsky, Dan, Manning, Christopher D., Potts, Christopher
Publisher: arXiv.org
Published: Jan 28, 2025
DOI: None
Link: https://arxiv.org/abs/2501.17148
Archived copy: https://web.archive.org/web/20261111031129/https://arxiv.org/abs/2501.17148
Stored text: 4,870 characters, retrieved Nov 12, 2026 · not yet divided
Claims extracted: Not recorded
Versions: One known
Copies: None recorded
Corrections: None recorded
Facts checked: Nov 12, 2026

History 2

Nov 12, 2026FactsFacts fetcher · Facts looked up
Nov 12, 2026Entered the graphEntered the graph; copy stored