Does DAS scale with large foundation models?

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

• will these methods work for large models?question0.762
  Motivating question for the paper, addressed by scaling SAEs to Claude 3 Sonnet.
• Foundation Modelsconcept0.750
  Large pretrained models used as backbones across tasks; their universality motivates the convergence hypothesis
• Bigger models are more likely to converge to a shared representation than smaller modelshypothesis0.728
  Selective pressure toward convergence via model capacity
• Representation engineering for large-language models: Survey and research challenges (Bartoszcze et al., 2025)concept0.725
  Survey of representation engineering methods cited as related work
• Foundation models trained on different data converge on similar latent representations, suggesting a Platonic form.claim0.721
• As LLMs scale, they develop increasingly general abstractions, with large models linearly representing abstract concepts like truth that capture shared properties of diverse inputsclaim0.720
  Interpretive claim connecting scale to abstraction level in LLM representations
• as the subject model scales, how does the ideal expansion factor and required training data for dictionary learning change?question0.718
  Scaling laws for dictionary learning are unknown and needed to assess feasibility on frontier models
• Features can be used to steer large models.claim0.717
  Clamping feature activations causally alters model behavior in interpretable ways.