paper-with-me

Papers

To bootstrap or to rollout? An optimal and adaptive interpolation

2024-11-14 · Wenlong Mou, Jian Qian

Bootstrapping and rollout are two fundamental principles for value function estimation in reinforcement learning (RL). We introduce a novel class of Bellman operators, called subgraph Bellman operators, that interpolate between bootstrapping and rollout methods. Our estimator, derived by solving the fixed point of the empirical subgraph Bellman operator, combines the strengths of the bootstrapping-based temporal difference (TD) estimator and the rollout-based Monte Carlo (MC) methods. Specifically, the error upper bound of our estimator approaches the optimal variance achieved by TD, with an additional term depending on the exit probability of a selected subset of the state space. At the same time, the estimator exhibits the finite-sample adaptivity of MC, with sample complexity depending only on the occupancy measure of this subset. We complement the upper bound with an information-theoretic lower bound, showing that the additional term is unavoidable given a reasonable sample size. Together, these results establish subgraph Bellman estimators as an optimal and adaptive framework for reconciling TD and MC methods in policy evaluation.

📄 PDF Abstract BibTeX arXiv:2411.09731

Code (0)

등록된 구현이 없습니다.

Tasks

Reinforcement Learning (RL)

Similar Papers 제목 키워드 기반

Fine-Grained Gap-Dependent Bounds for Tabular MDPs via Adaptive Multi-Step Bootstrap

2021-02-09 · Haike Xu, Tengyu Ma, Simon S. Du

This paper presents a new model-free algorithm for episodic finite-horizon Markov Decision Processes (MDP), Adaptive Multi-step Bootstrap (AMB), which enjoys a stronger gap-dependent regret bound. The first innovation is…

Multi-Armed Bandits

Adaptive N-step Bootstrapping with Off-policy Data

2021-01-01 · Guan Wang, Dong Yan, Hang Su, Jun Zhu

The definition of the update target is a crucial design choice in reinforcement learning. Due to the low computation cost and empirical high performance, n-step returns with off-policy data is a widely used…

Atari GamesMuJoCo

CADENCE: Closing the Reasoning Gap via Coverage-Adaptive On-Policy Distillation

2026-07-18 · Satyam Kumar, Saurabh Jha hf

On-policy knowledge distillation transfers reasoning from large teachers to compact students, but existing approaches suffer three compounding failure modes: (i) cold-start collapse, where a fresh student assigns near-ze…

Knowledge Distillation

Contextual Rollout Bandits for Reinforcement Learning with Verifiable Rewards

2026-02-09 · Xiaodong Lu, Xiaohan Wang, Jiajun Chai, Guojun Yin 외 arxiv

Reinforcement Learning with Verifiable Rewards (RLVR) is an effective paradigm for improving the reasoning capabilities of large language models. However, existing RLVR methods utilize rollouts in an indiscriminate and s…

Reinforcement LearningMathematical Reasoning

Rollout Algorithms and Approximate Dynamic Programming for Bayesian Optimization and Sequential Estimation

2022-12-15 · Dimitri Bertsekas

We provide a unifying approximate dynamic programming framework that applies to a broad variety of problems involving sequential estimation. We consider first the construction of surrogate cost functions for the purposes…

Bayesian Optimization