method
active
method:sparse-probing

Sparse Probing

Method from Gurnee et al. 2023 for finding feature directions including individual neuron analysis

Neighborhood — ranked by edge-count

Methods (1)

method
  • Concept detection metric using top-k latents and linear probe.

Related by similarity (8)

cosine ≥ 0.65 · no typed edge

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

  • Probingmethod0.836
    Standard interpretability technique used to read feature beliefs.
  • Probing Methodsmethod0.809
    Top-down interpretability approach studying linguistic properties at various residual stream stages; contrasted with the paper's bottom-up mechanistic approach
  • Coding scheme where qualities are represented by few neurons with continuous similarity relations.
  • Sparse subframeconcept0.802
    Mathematical object defining J-space as union of cones.
  • Probesconcept0.794
    Interpretability tools that decode information from internal model activations; here, linear probes are used for data attribution.
  • Probing approach avoiding supervision to sidestep complexity-accuracy tradeoff
  • Attention probe predicting whether an Entity is hallucinated; AUC .94 on validation.
  • Activation Probingconcept0.782
    Technique of reading out model beliefs from internal activations before the final answer token is generated