Sourcepreprinttext stored
Sanity Checks for Sparse Autoencoders: Do SAEs Beat Random Baselines?
Korznikov, Anton, Galichin, Andrey, Dontsov, Alexey, Rogov, Oleg, Oseledets, Ivan, Tutubalina, ElenaarXiv.orgPublished Feb 15, 2026Original︎ ↗Archived copy
Claims1
ExaminationsNone
TextStored, not divided
VersionsOne known
CopiesNone recorded
CorrectionsNone recorded
Facts checked Nov 12, 2026ReachNo copies recorded
StructureIn 1 claim's map0 sources downstream across 1 claim
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
we show that our baselines match fully-trained SAEs in interpretability (0.87 vs 0.90), sparse probing (0.69 vs 0.72), and causal editing (0.73 vs 0.72).
Sparse autoencoder probes do not consistently outperform baseline probes on language-model activation classification tasks.
Not yet read by this claim’s Steward.
Lineage across 1 claim
Draws on · none recorded
This source
Drawn on by · none recordedSanity Checks for Sparse Autoencoders: Do SAEs Beat Random Baselines?
in 1 claim's map
Record
- Kind
- preprint
- Authors
- Korznikov, Anton, Galichin, Andrey, Dontsov, Alexey, Rogov, Oleg, Oseledets, Ivan, Tutubalina, Elena
- Publisher
- arXiv.org
- Published
- Feb 15, 2026
- DOI
- None
- Link
- https://arxiv.org/abs/2602.14111
- Archived copy
- https://web.archive.org/web/20261105151425/https://arxiv.org/abs/2602.14111
- Stored text
- 4,767 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