concept
active
concept:surprise-minimizationSurprise Minimization
Core principle: acting to maximize value is equivalent to minimizing surprise by sampling environment to conform to expectations.
Neighborhood — ranked by edge-count
Frameworks (1)
framework
- Active InferenceimplementsFoundational framework by Karl Friston; the paper extends it to three hierarchical levels for modeling meta-awareness.
Claims (1)
claim
- Reinterprets classical reward/value concepts through free energy lens.
Concepts (2)
concept
- free energyassociated_withimplementsThermodynamic potential ΔF = ΔE − TΔS; domain walls form if ΔF < 0
- Epistemic Explorationassociated_withBayes-optimal exploration driven by uncertainty minimization; natural behavior in active inference without handcrafted mechanisms.
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.
- The principle that agents must minimise prediction error (surprisal) to persist.
- The progressive reduction of error (stress) as cells move toward their target positions.
- The core imperative under the Free Energy Principle; systems must reduce the difference between predicted and actual sensory states.
- The drive to reduce expected ambiguity about outcomes given states, leading to seeking well-lit, informative environments.
- Negative log probability of an outcome under the generative model; minimized in active inference.
- The negative log probability of sensory samples; minimized by free energy.
- Minimizing expected free energy for planning, decision-making, and action selection.