method
active
method:circuit-tracingCircuit tracing
Anthropic technique revealing computational pathways in Claude.
Neighborhood — ranked by edge-count
Papers (1)
paper
Institutes (1)
institute
- AnthropicstudiesLab behind Claude models and Constitutional AI training approach; represents highest baseline scores and lowest prompt lift.
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.
- Interpretability technique for identifying functional sub-circuits in neural networks, supported by pyvene
- Mechanistic interpretability technique for locating factual associations, mentioned as future work direction.
- Fine-grained approach to identifying specific network components responsible for reflection, mentioned as future direction.
- An open scientific collaboration hosted on Distill slack studying the inner workings of neural networks via zoomed-in mechanistic investigation
- Advantage of DiffLogic CA over NCA — learned rules are pure binary logic circuits that can be visualized and analyzed
- Interactive tool for visualizing and inspecting learned binary logic circuits using modified DigitalJS library
- A recurring, abstract pattern found in circuits (e.g., equivariance, unioning over cases), inspired by circuit motifs in systems biology
- Circuits active by default, e.g. can't-answer, suppressed by contrary evidence.