Value-Weighted Attention Pattern Visualization

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

• Value-Weighted Attention Patternconcept0.916
  Attention patterns scaled by the norm of the value vector at each source position, showing how large a vector is moved from each position
• Feature Visualizationmethod0.765
  Method of optimizing input to cause a neuron to fire maximally, used to characterize what a neuron detects; establishes causal link
• Mind's eye visualizationmethod0.750
  Technique of building a fluid, three-dimensional vision by closing one's eyes, relying on words and feeling to avoid arbitrary graphical over-specification.
• attention computationconcept0.745
  Process using Q, K, V to compute a heat map over K and weighted sum of V.
• visual reasoningconcept0.744
  Visual reasoning tasks often interleaved with intermediate visual states; promising direction in the field.
• Principal Component Analysis Visualizationmethod0.742
  Used to visualize LLM true/false representations, revealing clear linear structure separating true from false statements
• Identification of algorithms implemented in attention layers, distributed across attention headsfinding0.739
  VPD successfully recovered interpretable attention algorithms (previous-token behavior, syntax-boundary routing) in weight space without requiring manual decomposition across heads.
• Attention is a generalization of convolution; all convolutions can be expressed as tensor products of fixed relative position attention patterns and weight matricesclaim0.733
  Mathematical equivalence showing the relationship between attention mechanisms and convolutional operations