SIGHT: Reinforcement Learning with Self-Evidence and Information-Gain Diverse Branching for Search Agent

Fuente: arXiv

Saved in:

Bibliographic Details
Main Authors:	Zhong, Wenlin, Yang, Jinluan, Wu, Yiquan, Liu, Yi, Yao, Jianhang, Kuang, Kun
Format:	Preprint
Published:	2026
Subjects:	Computation and Language
Online Access:	Acceder al recurso
Tags:	Add Tag No Tags, Be the first to tag this record!

_version_	1866914324405026816
author	Zhong, Wenlin Yang, Jinluan Wu, Yiquan Liu, Yi Yao, Jianhang Kuang, Kun
author_facet	Zhong, Wenlin Yang, Jinluan Wu, Yiquan Liu, Yi Yao, Jianhang Kuang, Kun
contents	Reinforcement Learning (RL) has empowered Large Language Models (LLMs) to master autonomous search for complex question answering. However, particularly within multi-turn search scenarios, this interaction introduces a critical challenge: search results often suffer from high redundancy and low signal-to-noise ratios. Consequently, agents easily fall into "Tunnel Vision," where the forced interpretation of early noisy retrievals leads to irreversible error accumulation. To address these challenges, we propose SIGHT, a framework that enhances search-based reasoning through Self-Evidence Support (SES) and Information-Gain Driven Diverse Branching. SIGHT distills search results into high-fidelity evidence via SES and calculates an Information Gain score to pinpoint pivotal states where observations maximally reduce uncertainty. This score guides Dynamic Prompting Interventions - including de-duplication, reflection, or adaptive branching - to spawn new branches with SES. Finally, by integrating SES and correctness rewards via Group Relative Policy Optimization, SIGHT internalizes robust exploration strategies without external verifiers. Experiments on single-hop and multi-hop QA benchmarks demonstrate that SIGHT significantly outperforms existing approaches, particularly in complex reasoning scenarios, using fewer search steps.
format	Preprint
id	arxiv_https___arxiv_org_abs_2602_11551
institution	arXiv
publishDate	2026
record_format	arxiv
spellingShingle	SIGHT: Reinforcement Learning with Self-Evidence and Information-Gain Diverse Branching for Search Agent Zhong, Wenlin Yang, Jinluan Wu, Yiquan Liu, Yi Yao, Jianhang Kuang, Kun Computation and Language Reinforcement Learning (RL) has empowered Large Language Models (LLMs) to master autonomous search for complex question answering. However, particularly within multi-turn search scenarios, this interaction introduces a critical challenge: search results often suffer from high redundancy and low signal-to-noise ratios. Consequently, agents easily fall into "Tunnel Vision," where the forced interpretation of early noisy retrievals leads to irreversible error accumulation. To address these challenges, we propose SIGHT, a framework that enhances search-based reasoning through Self-Evidence Support (SES) and Information-Gain Driven Diverse Branching. SIGHT distills search results into high-fidelity evidence via SES and calculates an Information Gain score to pinpoint pivotal states where observations maximally reduce uncertainty. This score guides Dynamic Prompting Interventions - including de-duplication, reflection, or adaptive branching - to spawn new branches with SES. Finally, by integrating SES and correctness rewards via Group Relative Policy Optimization, SIGHT internalizes robust exploration strategies without external verifiers. Experiments on single-hop and multi-hop QA benchmarks demonstrate that SIGHT significantly outperforms existing approaches, particularly in complex reasoning scenarios, using fewer search steps.
title	SIGHT: Reinforcement Learning with Self-Evidence and Information-Gain Diverse Branching for Search Agent
topic	Computation and Language
url	https://arxiv.org/abs/2602.11551