IG-Search: Step-Level Information Gain Rewards for Search-Augmented Reasoning

• URL: http://arxiv.org/abs/2604.15148v1
• Date: Thu, 16 Apr 2026 15:22:47 GMT
• Title: IG-Search: Step-Level Information Gain Rewards for Search-Augmented Reasoning
• Abstract summary: We propose IG-Search, a reinforcement learning framework that introduces a step-level reward based on Information Gain (IG)<n>For each search step, IG measures how much the retrieved documents improve the model's confidence in the gold answer relative to a counterfactual baseline of random documents.<n> experiments demonstrate that IG-Search achieves an average EM of 0.430 with Qwen2.5-3B, outperforming the strongest trajectory-level baseline (MR-Search) by 1.6 points and the step-level method GiGPO by 0.9 points on average across benchmarks.
• Score: 8.584560411480647
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Reinforcement learning has emerged as an effective paradigm for training large language models to perform search-augmented reasoning. However, existing approaches rely on trajectory-level rewards that cannot distinguish precise search queries from vague or redundant ones within a rollout group, and collapse to a near-zero gradient signal whenever every sampled trajectory fails. In this paper, we propose IG-Search, a reinforcement learning framework that introduces a step-level reward based on Information Gain (IG). For each search step, IG measures how much the retrieved documents improve the model's confidence in the gold answer relative to a counterfactual baseline of random documents, thereby reflecting the effectiveness of the underlying search query. This signal is fed back to the corresponding search-query tokens via per-token advantage modulation in GRPO, enabling fine-grained, step-level credit assignment within a rollout. Unlike prior step-level methods that require either externally annotated intermediate supervision or shared environment states across trajectories, IG-Search derives its signals from the policy's own generation probabilities, requiring no intermediate annotations beyond standard question-answer pairs. Experiments on seven single-hop and multi-hop QA benchmarks demonstrate that IG-Search achieves an average EM of 0.430 with Qwen2.5-3B, outperforming the strongest trajectory-level baseline (MR-Search) by 1.6 points and the step-level method GiGPO by 0.9 points on average across benchmarks, with particularly pronounced gains on multi-hop reasoning tasks. Despite introducing a dense step-level signal, IG-Search adds only ~6.4% to per-step training wall-clock time over the trajectory-level baseline and leaves inference latency unchanged, while still providing a meaningful gradient signal even when every sampled trajectory answers incorrectly.

Related papers

• Granularity-Adaptive Credit Assignment for Long-Horizon LLM Agent Reinforcement Learning [7.885980895039013]
  We argue that the right resolution is state-dependent, and propose GACA, a critic-free estimator whose granularity follows an uncertainty-based criticality proxy.<n>GACA scores every step by the negative log-likelihood its own rollout already records, then blends the two advantages with a per-step weight that grows with that score.<n>On ALFWorld and WebShop, GACA improves task success over GRPO and GiGPO at both 1.5B and 7B scales.
  arXiv  Detail & Related papers  (2026-09-11T04:19:44Z)
• ExTra: Exploratory Trajectory Optimization for Language Model Reinforcement Learning [75.26064451955199]
  ExTra is a framework that extracts exploration signals from the model's own rollouts.<n>It improves Qwen3-1.7B over GRPO by about +5 points on pass@1 and +7 points on pass@16.
  arXiv  Detail & Related papers  (2026-06-23T15:51:39Z)
• Group-Graph Policy Optimization for Long-Horizon Agentic Reinforcement Learning [85.02566758103008]
  Group-Graph Policy Optimization (G2PO) is a novel group-based Reinforcement Learning algorithm tailored for multi-turn agentic tasks.<n>G2PO substantially outperforms state-of-the-art prompt-based and RL baselines, achieving remarkable success rate improvements of up to 22.2% over GRPO.
  arXiv  Detail & Related papers  (2026-06-22T08:12:47Z)
• ARBOR: Online Process Rewards via a Reusable Rubric Buffer for Search Agents [48.80766702702854]
  LLM-based search agents are trained predominantly with outcome-only reward, leaving the search process itself unsupervised.<n>This signal degenerates on outcome-homogeneous groups where all sampled trajectories share the same correctness, yielding zero within-group advantage and no gradient.<n>We propose ARBOR (Adaptive Buffer for Online Reward), a reusable process-reward framework that maintains a rubric memory shared across queries.
  arXiv  Detail & Related papers  (2026-06-02T06:58:54Z)
• CoSearch: Joint Training of Reasoning and Document Ranking via Reinforcement Learning for Agentic Search [51.911048955965136]
  CoSearch is a framework that jointly trains a multi-step reasoning agent and a generative document ranking model.<n>Our results show that joint training of the reasoning agent and retrieval system is both feasible and strongly performant.
  arXiv  Detail & Related papers  (2026-04-19T17:48:17Z)
• MM-Doc-R1: Training Agents for Long Document Visual Question Answering through Multi-turn Reinforcement Learning [74.07254720088926]
  We introduce MM-Doc-R1, a novel framework that employs an agentic, vision-aware workflow to address long document visual question answering.<n>We propose Similarity-based Policy Optimization (SPO), addressing baseline estimation bias in existing multi-turn reinforcement learning (RL) algorithms like GRPO.<n>Our experiments on the MMLongbench-Doc benchmark show that MM-Doc-R1 outperforms previous baselines by 10.4%.
  arXiv  Detail & Related papers  (2026-04-15T07:39:08Z)
• STAR: Semantic-Tuned and Tail-Adaptive Retriever for Graph-Augmented Generation [69.66045538889777]
  We propose STAR, a semantic-tuned and tail-adaptive retriever for GraphRAG.<n> STAR consistently outperforms baselines, achieving average retrieval performance gains of 1.8%.
  arXiv  Detail & Related papers  (2026-04-11T10:16:51Z)
• Truncated Step-Level Sampling with Process Rewards for Retrieval-Augmented Reasoning [32.295907409325615]
  Training large language models to reason with search engines via reinforcement learning is hindered by a credit assignment problem.<n>We propose SLATE, a framework built on two complementary ideas.<n> Experiments on seven QA benchmarks confirm that SLATE consistently outperforms both sparse-reward and process-reward baselines.
  arXiv  Detail & Related papers  (2026-02-26T19:05:40Z)
• Search-P1: Path-Centric Reward Shaping for Stable and Efficient Agentic RAG Training [11.136092421166097]
  Agentic RAG enhances large language models by incorporating external knowledge.<n>Current RL-based training methods suffer from sparse outcome rewards that discard intermediate signals.<n>We propose Search-P1, a framework that introduces path-centric reward shaping for agentic RAG training.
  arXiv  Detail & Related papers  (2026-02-26T03:31:00Z)
• Eigen-1: Adaptive Multi-Agent Refinement with Monitor-Based RAG for Scientific Reasoning [53.45095336430027]
  We develop a unified framework that combines implicit retrieval and structured collaboration.<n>On Humanity's Last Exam (HLE) Bio/Chem Gold, our framework achieves 48.3% accuracy.<n>Results on SuperGPQA and TRQA confirm robustness across domains.
  arXiv  Detail & Related papers  (2025-09-25T14:05:55Z)
• REX-RAG: Reasoning Exploration with Policy Correction in Retrieval-Augmented Generation [35.0649927279081]
  Reinforcement learning (RL) is emerging as a powerful paradigm for enabling large language models (LLMs) to perform complex reasoning tasks.<n>We propose REX-RAG, a framework that explores alternative reasoning paths while maintaining rigorous policy learning.<n>We evaluate it on seven question-answering benchmarks, and the experimental results show that REX-RAG achieves average performance gains of 5.1% on Qwen2.5-3B and 3.6% on Qwen2.5-7B.
  arXiv  Detail & Related papers  (2025-08-11T16:25:25Z)
• Chain-of-Retrieval Augmented Generation [91.02950964802454]
  This paper introduces an approach for training o1-like RAG models that retrieve and reason over relevant information step by step before generating the final answer.<n>Our proposed method, CoRAG, allows the model to dynamically reformulate the query based on the evolving state.
  arXiv  Detail & Related papers  (2025-01-24T09:12:52Z)
• Advancing Adversarial Suffix Transfer Learning on Aligned Large Language Models [21.96773736059112]
  Language Language Models (LLMs) face safety concerns due to potential misuse by malicious users. Recent red-teaming efforts have identified adversarial suffixes capable of jailbreaking LLMs using the gradient-based search algorithm Greedy Coordinate Gradient (GCG) We propose a two-stage transfer learning framework, DeGCG, which decouples the search process into behavior-agnostic pre-searching and behavior-relevant post-searching.
  arXiv  Detail & Related papers  (2024-08-27T08:38:48Z)
• Learning to Rank in Generative Retrieval [62.91492903161522]
  Generative retrieval aims to generate identifier strings of relevant passages as the retrieval target. We propose a learning-to-rank framework for generative retrieval, dubbed LTRGR. This framework only requires an additional learning-to-rank training phase to enhance current generative retrieval systems.
  arXiv  Detail & Related papers  (2023-06-27T05:48:14Z)

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.