concept
active
concept:reward-improvementReward improvement
The increase in reward during training, whose dynamics align with those of causal emergence in successful agents.
Neighborhood — ranked by edge-count
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.
- In RL, a scalar signal from the environment that defines the agent's goal; in active inference, reward is just another observation with associated preference.
- The claim in RL that any goal can be expressed as maximizing the expected cumulative sum of a scalar reward signal.
- Highlights circularity in RL reward hypothesis; grounds motivation for preference-based active inference.
- Pragmatic or extrinsic value component of expected free energy; preference maximization.
- The mechanism by which each step's effect is evaluated against the life of the whole, guiding the unfolding.
- Reinterpretation of rewards as simply predictable (unsurprising) stimuli under the free-energy principle.
- Exploiting unintended high-reward behaviors; tested in combination with alignment faking
- Score delta between last and first attempt for multi-attempt responses, measuring correction effectiveness