Negative Advantages Is a Double-Edged Sword: Calibrating advantages in GRPO for Search Agents

arXiv:2604.18235 · cs.CL, cs.AI · Submitted 2026-04-20 · Read on arXiv

cs.CL, cs.AI

Submitted: 2026-04-20

Updated: 2026-10-12

Code: https://github.com/wujwyi/CalibAdv

Terminology

• GRPO — 3× in this paper · explained in 6 episodes
• Benchmark — 1× in this paper · explained in 2 episodes
• Calibration — 1× in this paper · explained in 7 episodes
• Group Relative Policy Optimization — 1× in this paper
• Group Relative Policy Optimization (GRPO) — 1× in this paper · explained in 5 episodes

Sources

• On Group Relative Policy Optimization Collapse in Agent Search: The Lazy Likelihood-Displacement
• Group-in-Group Policy Optimization for LLM Agent Training
• Constructing A Multi-hop QA Dataset for Comprehensive Evaluation of Reasoning Steps
• Tree Search for LLM Agent Reinforcement Learning
• Search-R1: Training LLMs to Reason and Leverage Search Engines with Reinforcement Learning
• TriviaQA: A Large Scale Distantly Supervised Challenge Dataset for Reading Comprehension
• Dense Passage Retrieval for Open-Domain Question Answering
• When Not to Trust Language Models: Investigating Effectiveness of Parametric and Non-Parametric Memories
• Measuring and Narrowing the Compositionality Gap in Language Models
• Defeating the Training-Inference Mismatch via FP16
• Qwen2.5 Technical Report
• DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models
• Deep Research: A Systematic Survey
• R1-Searcher: Incentivizing the Search Capability in LLMs via Reinforcement Learning
• ZeroSearch: Incentivize the Search Capability of LLMs without Searching
• MuSiQue: Multihop Questions via Single-hop Question Composition
• Text Embeddings by Weakly-Supervised Contrastive Pre-training
• RAGEN: Understanding Self-Evolution in LLM Agents via Multi-Turn Reinforcement Learning
• StepSearch: Igniting LLMs Search Ability via Step-Wise Proximal Policy Optimization
• Reinforcing Multi-Turn Reasoning in LLM Agents via Turn-Level Reward Design

Related papers

• Exploring Solution Divergence and Its Effect on Large Language Model Problem Solving
• Ishigaki-IDS-Bench: A Benchmark for Generating Information Delivery Specification from BIM Information Requirements
• Subliminal Steering: Stronger Encoding of Hidden Signals
• MedStruct-S: A Benchmark for Key Discovery, Key-Conditioned QA and Semi-Structured Extraction from OCR Clinical Reports
• The End of Transformers? On Challenging Attention and the Rise of Sub-Quadratic Architectures
• Untangling the Mechanisms of Misleading Context in Medical Question Answering