framework
active
framework:predictive-concept-decodersPredictive Concept Decoders
Framework learning end-to-end mappings from activations through sparse concept bottlenecks to behavioral predictions
Neighborhood — ranked by edge-count
Papers (1)
paper
Thinkers (1)
thinker
- Huang, V.introducesLead author of Predictive Concept Decoders paper on end-to-end activation-to-language mappings
Hypotheses (1)
hypothesis
- Comparative prediction motivating future work contrasting different approaches to LLM self-knowledge
Related by similarity (8)
cosine ≥ 0.65 · no typed edgeEntities in the same semantic neighborhood but without a typed relation to this one — candidates for new edges or unrecognized duplicates.
- Related framework emphasizing prediction errors; active inference extends to Markov decision processes.
- Theory that perception and cognition involve predictive coding; listed with indicators in Butlin et al. 2023.
- Bialek/Ay information-theoretic measure; characterizes agent capacity to predict future from past.
- How a neural network encodes a semantic concept internally, argued to be better captured by manifolds than by atomic features.
- Apps and Tsakiris's proposal that self-recognition can be modeled probabilistically under the free energy principle
- Cognitive framework where top-down predictions shape perception and action.