method
active
method:gradient-based-data-attribution

Gradient-based data attribution

Baseline method against which probe-based ranking is compared; more computationally expensive.

Neighborhood — ranked by edge-count

Findings (1)

finding

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.

  • Gradient Descentmethod0.836
    Used for updating hidden state expectations; provides dynamical process theory testable against neuronal data
  • Data Attributionconcept0.834
    The task of attributing model behaviors to specific training datapoints.
  • Gradientsconcept0.825
    The property that qualities vary slowly, subtly, gradually across the extent of each living thing; gradients arise as natural responses to changing circumstances and create field-like character that points toward and establishes centers
  • Linear classifier approach applied to model activations to identify which training datapoints caused undesired behaviors in post-training.
  • Gradient that tells a cell its correct position; stress arises from deviation from this gradient.
  • Gradient methodmethod0.803
    Optimization technique that computes weight changes by following the gradient of an error function; contrasted with evolutionary stochastic search.
  • Computing attribution as the dot product of the output logit gradient with the SAE decoder weight, multiplied by feature activation.