concept
active
concept:predictive-integrationpredictive integration
The mid-to-late layer computational process that converts routed perturbation signals into explicit predictions
Neighborhood — ranked by edge-count
Frameworks (1)
framework
- This paper's proposed mechanistic explanation integrating signal injection, attention routing, predictive integration, and residual recovery
Methods (1)
method
- Logit LensimplementsUnsupervised interpretability technique that projects activations through unembedding matrix; provides comparison point for NLA approach.
Concepts (2)
concept
- residual stream recovery dynamicsassociated_withThe network's tendency to actively attenuate injected perturbations over subsequent layers, erasing the signal before output
- attention-based signal routingassociated_withMechanism by which attention heads detect injected perturbations and route information about them to the final token position
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.
- Key property enabling organized collective action; mechanism by which parts coordinate into functional wholes.
- Computational property emphasized across consciousness theories as necessary for conscious processing
- Related framework emphasizing prediction errors; active inference extends to Markov decision processes.
- Bialek/Ay information-theoretic measure; characterizes agent capacity to predict future from past.
- Theory that perception and cognition involve predictive coding; listed with indicators in Butlin et al. 2023.
- Cognitive framework where top-down predictions shape perception and action.
- Integration of signals across components to compute holistic states, a hallmark of organismic individuality and basal cognition