Training the sparse autoencoder on more data makes features subjectively sharper and more interpretable

Entities in the same semantic neighborhood but without a typed relation to this one — candidates for new edges or unrecognized duplicates.

• Sparse autoencoders produce interpretable features for large models.claim0.875
  Central claim of the paper: the method scales to state-of-the-art transformers.
• Sparse Autoencoders Find Highly Interpretable Features in Language Models (Cunningham et al., 2023)concept0.870
  Core methodology paper for SAE-based interpretable feature extraction
• Sparse Autoencoder Featuresconcept0.856
  Used in Anthropic welfare assessment to identify performative behavior and hidden emotional struggle co-activating features
• Sparse autoencoders are preferable to stronger iterative dictionary learning methods because they cannot recover features the model itself cannot accessclaim0.848
  Rationale for using simpler sparse autoencoders rather than NP-hard compressed sensing algorithms
• Sparse autoencoders don't provide a comprehensive solution because they decode activations, not parametersclaim0.839
  Critique of activation-based interpretability methods.
• Sparse autoencoders extract features that are significantly more monosemantic than neurons, as shown by four independent lines of evidenceclaim0.832
  Central claim of the paper, supported by detailed feature analysis, human evaluation, automated interpretability of activations, and automated interpretability of logit weights
• Sparse Autoencoder for Dictionary Learningframework0.824
  Primary method introduced: trains a one-hidden-layer MLP with L1 sparsity penalty to decompose model activations into overcomplete feature dictionaries
• Sparse Autoencoderframework0.820
  Interpretability framework used to decompose layer-40 activations into sparse feature sets for studying emotional alignment and persistence