Meta-Reinforcement Learning with Self-Reflection for Agentic Search

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Xiao, Teng, Yuan, Yige, Ivison, Hamish, Zhu, Huaisheng, Brahman, Faeze, Lambert, Nathan, Dasigi, Pradeep, Smith, Noah A., Hajishirzi, Hannaneh
Format: Preprint
Published: 2026
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866914404227874816
author Xiao, Teng
Yuan, Yige
Ivison, Hamish
Zhu, Huaisheng
Brahman, Faeze
Lambert, Nathan
Dasigi, Pradeep
Smith, Noah A.
Hajishirzi, Hannaneh
author_facet Xiao, Teng
Yuan, Yige
Ivison, Hamish
Zhu, Huaisheng
Brahman, Faeze
Lambert, Nathan
Dasigi, Pradeep
Smith, Noah A.
Hajishirzi, Hannaneh
contents This paper introduces MR-Search, an in-context meta reinforcement learning (RL) formulation for agentic search with self-reflection. Instead of optimizing a policy within a single independent episode with sparse rewards, MR-Search trains a policy that conditions on past episodes and adapts its search strategy across episodes. MR-Search learns to learn a search strategy with self-reflection, allowing search agents to improve in-context exploration at test-time. Specifically, MR-Search performs cross-episode exploration by generating explicit self-reflections after each episode and leveraging them as additional context to guide subsequent attempts, thereby promoting more effective exploration during test-time. We further introduce a multi-turn RL algorithm that estimates a dense relative advantage at the turn level, enabling fine-grained credit assignment on each episode. Empirical results across various benchmarks demonstrate the advantages of MR-Search over baselines based RL, showing strong generalization and relative improvements of 9.2% to 19.3% across eight benchmarks. Our code and data are available at https://github.com/tengxiao1/MR-Search.
format Preprint
id arxiv_https___arxiv_org_abs_2603_11327
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Meta-Reinforcement Learning with Self-Reflection for Agentic Search
Xiao, Teng
Yuan, Yige
Ivison, Hamish
Zhu, Huaisheng
Brahman, Faeze
Lambert, Nathan
Dasigi, Pradeep
Smith, Noah A.
Hajishirzi, Hannaneh
Machine Learning
Computation and Language
This paper introduces MR-Search, an in-context meta reinforcement learning (RL) formulation for agentic search with self-reflection. Instead of optimizing a policy within a single independent episode with sparse rewards, MR-Search trains a policy that conditions on past episodes and adapts its search strategy across episodes. MR-Search learns to learn a search strategy with self-reflection, allowing search agents to improve in-context exploration at test-time. Specifically, MR-Search performs cross-episode exploration by generating explicit self-reflections after each episode and leveraging them as additional context to guide subsequent attempts, thereby promoting more effective exploration during test-time. We further introduce a multi-turn RL algorithm that estimates a dense relative advantage at the turn level, enabling fine-grained credit assignment on each episode. Empirical results across various benchmarks demonstrate the advantages of MR-Search over baselines based RL, showing strong generalization and relative improvements of 9.2% to 19.3% across eight benchmarks. Our code and data are available at https://github.com/tengxiao1/MR-Search.
title Meta-Reinforcement Learning with Self-Reflection for Agentic Search
topic Machine Learning
Computation and Language
url https://arxiv.org/abs/2603.11327