Curve detectors found across AlexNet, InceptionV1, VGG19, ResNetV2-50 and models trained on Places365

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

• High-low frequency detectors found across AlexNet, InceptionV1, VGG19, and ResNetV2-50finding0.861
  Second low-level feature type demonstrating cross-architecture universality
• Weights between early and full curve detectors in InceptionV1 form a curve of positive weights at tangent positions, with opposing orientations inhibitoryfinding0.813
  Demonstrates that meaningful algorithms can be read directly off floating-point weights in a neural network
• Curve detecting neurons found in every non-trivial vision model carefully examinedfinding0.780
  Empirical basis for treating curve detectors as a canonical example of meaningful, understandable features
• InceptionV1 spreads car feature from a pure car detector in mixed4c across dog detector neurons in the next layerfinding0.766
  Circuit-level evidence that polysemantic neurons arise deliberately through superposition rather than entangled computation
• We hypothesize that high-low frequency detectors, if predicted by artificial neural network universality, might be found in biological neural networks.hypothesis0.743
  Specific cross-domain prediction mentioned by neuroscientists in conversation with the authors
• The gradual reconvergence of aT and aF activation clusters in final layers reflects the model's internal conflict or moral dilemma during deceptive behavior generationclaim0.730
  Interpretive claim attributing representational pattern to internal model state during threat-based deception
• InceptionV1 implements a four-layer circuit for pose-invariant dog head detection with mirrored left/right pathways that inhibit each other then unite, exhibiting XOR-like propertiesfinding0.722
  Evidence that neural networks learn sophisticated invariance mechanisms through structured circuits rather than loose feature aggregation
• All 32 attention heads at layer 3 achieve 100% localization accuracy for injections at layer 2 (5-way classification, 20% chance)finding0.720
  Striking mechanistic finding that injection creates universally detectable perturbation in residual stream immediately downstream