framework
active
framework:automated-interpretabilityAutomated Interpretability
Method using large language models (Claude) to generate and test explanations of features at scale
Neighborhood — ranked by edge-count
Papers (2)
paper
Thinkers (1)
thinker
- Steven BillsextendsDeveloped automated interpretability approach using LLMs to explain neuron activations
Methods (1)
method
- Importance Scoringassociated_withWeighted Spearman correlation that corrects for sampling bias in automated interpretability evaluation
Frameworks (1)
framework
- Paper's main proposal: SAE trained on the mean activation across a turn.
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.
- The capability to explain model predictions; a central theme of the paper, with disruption profiles as vehicle.
- Claude-based explanation of features from max-activating examples.
- Socio-philosophical open question.
- Detection-based LLM judging of latent descriptions using gpt-4o-mini.
- Proposed paradigm for evaluating interpretability work through empirical falsifiability rather than benchmarks or user studies
- Second approach: propose concepts and find components corresponding to them.
- Explaining activations without understanding mechanism.
- Using Claude 3 Opus to generate feature explanations and predict held-out activations.