Transcoders

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

• Autoencoderconcept0.756
  Neural network architecture that learns compressed representations; SOHMs are functionally equivalent.
• Deep Auto Encoderframework0.749
• Masked Autoencodersconcept0.735
  Self-supervised learning method that optimizes reconstruction tasks; included in the paper's analysis as a multi-task objective
• Deep Autoencoderframework0.735
• Transformer decoder architectureframework0.707
  Base architecture of reasoning LLMs studied, with attention and MLP blocks per layer
• Signal Transformerconcept0.707
  Core abstraction in Fruit: pure function mapping signals to signals; enables compositional GUI definitions.
• Pre-Encoder Biasconcept0.702
  Architectural modification subtracting a learned bias from autoencoder inputs before encoding; initialized to geometric median of dataset; improves autoencoder performance
• Precision Encodingconcept0.701
  Quantifies certainty in beliefs; modulates policy selection; related to dopamine encoding.