method
active
method:softmax-policy-selection

Softmax policy selection

Selecting policies using a softmax (normalized exponential) function of negative expected free energy.

Neighborhood — ranked by edge-count

Frameworks (1)

framework
  • Active Inference
    implements
    Foundational framework by Karl Friston; the paper extends it to three hierarchical levels for modeling meta-awareness.

Concepts (1)

concept
  • Expected Free Energy
    implements
    Free energy expected under future outcomes; guides policy selection via epistemic and extrinsic value.

Related by similarity (8)

cosine ≥ 0.65 · no typed edge

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

  • Softmax Policy Priorconcept0.845
    Policies assigned probability via softmax of expected free energy; enables self-evidencing behavior.
  • Policy Selectionconcept0.801
    Choosing sequences of actions based on expected free energy; prior probability of policy is softmax of expected free energy
  • Softmax Functionmethod0.746
    Neuronal dynamics computed from free energy gradients; interpreted as average firing rate of neural populations.
  • Soft preference labelsmethod0.741
    Using normalized log-probabilities from the feedback model as soft targets for preference model training.
  • Softmax Inc.institute0.740
    Organisation that hosted the Holistic Intelligence unconference where the paper's ideas originated
  • Softmax Bottleneckconcept0.736
    Failure mode for output-surjectivity: LLMs may lack capacity to predict all tokens due to rank constraints
  • Softmax Activation Function as Neuronal Modelmethod0.726
    Using softmax to translate membrane potentials into firing rates, implementing lateral inhibition.
  • Adaptive Beta Softmax Scalingmethod0.718
    Implementation detail weighting softmax by log(n_memories) to prevent down-weighting of attention values as memory set grows.