finding
active
finding:a-single-sae-latent-was-more-accurate-than-official-dataset-labels-on-glue-colaA single SAE latent was more accurate than official dataset labels on GLUE CoLA
Example where SAE yields useful latent.
Source paper
extracted_from(2025) · Lee Sharkey · Bilal I. Chughtai · Joshua Batson · Jack W. Lindsey +25
Neighborhood — ranked by edge-count
Papers (1)
paper
Claims (1)
claim
- Core assessment of SDL.
Related by similarity (8)
cosine ≥ 0.65 · no typed edgeEntities in the same semantic neighborhood but without a typed relation to this one — candidates for new edges or unrecognized duplicates.
- Latent #10 activation increase correctly classifies all correct vs incorrect fine-tuned models in Figure 9
- Training artifact.
- Fig 2 right.
- The most decreased latent after bad-advice fine-tuning is also the most effective re-aligning latent
- Interpretable features extracted by sparse autoencoders used as steering targets in this study
- Quantitative comparison supporting SAE utility.
- Inception-v1 SAE shows meaningful fraction of latents with very high nearest-neighbor cosinefinding0.736Fig 8; may be duplication.
- Latents are specialized to different modes of misalignment, explaining diverse misalignment profiles