Smaller Abstract State Spaces Enable Cross-Scale Generalization in Reinforcement Learning

• URL: http://arxiv.org/abs/2605.20272v1
• Date: Tue, 19 May 2026 02:26:28 GMT
• Title: Smaller Abstract State Spaces Enable Cross-Scale Generalization in Reinforcement Learning
• Abstract summary: We present the first theoretical model of how Out-of-Distribution generalization can be achieved in RL agents.<n>We derive a bound on the agent's OOD test performance, thereby defining the conditions under which OOD generalization is achievable.<n>Our analysis suggests that constraining an agent to operate over a small, finite set of abstract states is necessary for achieving generalization to more complex tasks.
• Score: 1.8354627928712421
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: While humans readily generalize abstract concepts to more complex or larger tasks, building Reinforcement Learning (RL) systems with this ability remains elusive. Here, we present the first theoretical model of how such Out-of-Distribution (OOD) generalization can be achieved in RL agents. Our approach considers Partially Observable Markov Decision Processes (POMDPs) and assumes that an intelligent agent uses an abstraction function to determine which experiences can be treated as equivalent and which must be distinguished. First, we extend the existing state abstraction framework and proof techniques to POMDPs. Then, we define a successor-weighted model reduction, a model reduction variant that enables compression into smaller abstract spaces than prior definitions allow. We derive a bound on the agent's OOD test performance, thereby defining the conditions under which OOD generalization is achievable. This bound decomposes an agent's performance loss into approximation and estimation errors, revealing how reducing an agent's abstract state space size improves test performance and OOD generalization. Our analysis suggests that constraining an agent to operate over a small, finite set of abstract states is necessary for achieving generalization to more complex tasks. Our results motivate further research into learning RL architectures that scale across tasks of varying complexity levels.

Related papers

• Predictive Representations for Skill Transfer in Reinforcement Learning [46.51727738073925]
  Key challenge in scaling up Reinforcement Learning is generalizing learned behaviour.<n>We develop a new formalism for transfer by virtue of state abstraction.
  arXiv  Detail & Related papers  (2026-04-08T12:35:24Z)
• Discovering State Equivalences in UCT Search Trees By Action Pruning [64.43268969806098]
  We show that Ideal Pruning Abstractions in UCT (IPA-UCT) outperforms OGA-UCT across a large range of test domains and iteration budgets.<n>We also show that both IPA and ASAP are special cases of a more general framework that we call p-ASAP.
  arXiv  Detail & Related papers  (2025-10-30T10:54:43Z)
• Causal Abstraction Inference under Lossy Representations [53.18851962820361]
  We introduce a new type of abstractions called projected abstractions that generalize existing definitions to accommodate lossy representations.<n>We show how to construct a projected abstraction from the low-level model and how it translates equivalent observational, interventional, and counterfactual causal queries from low to high-level.
  arXiv  Detail & Related papers  (2025-09-25T21:20:42Z)
• Building Minimal and Reusable Causal State Abstractions for Reinforcement Learning [63.58935783293342]
  Causal Bisimulation Modeling (CBM) is a method that learns the causal relationships in the dynamics and reward functions for each task to derive a minimal, task-specific abstraction. CBM's learned implicit dynamics models identify the underlying causal relationships and state abstractions more accurately than explicit ones.
  arXiv  Detail & Related papers  (2024-01-23T05:43:15Z)
• Finding Alignments Between Interpretable Causal Variables and Distributed Neural Representations [62.65877150123775]
  Causal abstraction is a promising theoretical framework for explainable artificial intelligence. Existing causal abstraction methods require a brute-force search over alignments between the high-level model and the low-level one. We present distributed alignment search (DAS), which overcomes these limitations.
  arXiv  Detail & Related papers  (2023-03-05T00:57:49Z)
• Exploiting Multiple Abstractions in Episodic RL via Reward Shaping [23.61187560936501]
  We consider a linear hierarchy of abstraction layers of the Markov Decision Process (MDP) underlying the target domain. We propose a novel form of Reward Shaping where the solution obtained at the abstract level is used to offer rewards to the more concrete MDP.
  arXiv  Detail & Related papers  (2023-02-28T13:22:29Z)
• A General Framework for Sample-Efficient Function Approximation in Reinforcement Learning [132.45959478064736]
  We propose a general framework that unifies model-based and model-free reinforcement learning. We propose a novel estimation function with decomposable structural properties for optimization-based exploration. Under our framework, a new sample-efficient algorithm namely OPtimization-based ExploRation with Approximation (OPERA) is proposed.
  arXiv  Detail & Related papers  (2022-09-30T17:59:16Z)
• Model-Invariant State Abstractions for Model-Based Reinforcement Learning [54.616645151708994]
  We introduce a new type of state abstraction called textitmodel-invariance. This allows for generalization to novel combinations of unseen values of state variables. We prove that an optimal policy can be learned over this model-invariance state abstraction.
  arXiv  Detail & Related papers  (2021-02-19T10:37:54Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.