Activation-based interpretability does not immediately explain the computations that gave rise to activations; understanding parameters is necessary for deeper insight

{
  "doc": "paper.md",
  "context": "Motivates shift from studying model activations ('thoughts') to understanding parameters ('the computations themselves').",
  "category": "ai",
  "norm_label": "Activation-based interpretability does not immediately explain the computations that gave rise to activations; understanding parameters is necessary for deeper insight",
  "graphify_id": "activation_vs_parameter_interpretability",
  "source_file": "paper.md",
  "imported_from": "/Users/antonborzov/Documents/Research.nosync/papers/extract_typed_out/paper/graph.json",
  "extracted_type": "claim",
  "source_location": "intro",
  "graphify_file_type": "claim"
}