Article

Regret Is Weighted Forgetting

2026-07-06

Abstract excerpt

How much of an agent's regret comes from a bad representation, and how much from a bad policy? Here I give an exact answer. Fix a representation $M$, a function that compresses histories into states, and a finite evaluation distribution over pairs of a history and a binary diagnostic test. Sort the pairs into cells, one cell for each combination of representation state and test, so that a policy reading only the r...

Topics

Open a Topic to create a Post that cites this publication.

Identifiers and source

Literature Corpus work: aeb2106f-63d5-507e-bab4-9e2b6a89acd7
DOI: 10.20944/preprints202603.1546.v2

Open publication

Related research

Semantic proximity does not establish scientific evidence.

Click a neighbor to travelStep 1 · 12 closest

Regret Is Weighted ForgettingDOI 10.20944/preprints202603.1546.v2Read

Regret Is Weighted ForgettingSemantic proximity Read

The Value of AbstractionSemantic proximity Read

Reward-predictive representations generalize across tasks in reinforcement learningSemantic proximity Read

A Research Programme for Continual and Neurosymbolic Learning in Epistemic Artificial IntelligenceSemantic proximity Read

Multi-Agent based Dynamic Anchors for Interpretation of Deep Learning ClassifiersSemantic proximity Read

Why Compression Creates Intelligence: The Architecture of Experience in Large ModelsSemantic proximity Read

Learning Safe Behaviour via Justified Human Preferences and Hypothetical QueriesSemantic proximity Read

Learning Safe Behaviour via Justified Human Preferences and Hypothetical QueriesSemantic proximity Read

Computation noise promotes cognitive resilience to adverse conditions during decision-makingSemantic proximity Read

Predictions and outcomes independently shape the subjective experience of regretSemantic proximity Read

Hierarchical clustering optimizes the tradeoff between compositionality and expressivity of task structures for flexible reinforcement learningSemantic proximity Read

Connecting Exploration, Generalization, and Planning in Correlated TreesSemantic proximity Read