CRISP: Critical Step Perception for Training Efficient Deep Search Agents

arXiv:2608.01867 · cs.CL · Submitted 2026-08-03 · Read on arXiv

Haosi Mo, Zihao Yan, Ruiqing Zhang, Zhongli Li, Hexuan Deng, Xuebo Liu, Min Zhang

cs.CL

Submitted: 2026-08-03

Comments: 15 pages, 6 figures

License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Terminology

• Accuracy — 35× in this paper · explained in 1 episode
• GRPO — 11× in this paper · explained in 6 episodes
• Benchmark — 10× in this paper · explained in 2 episodes
• Redundancy — 10× in this paper · explained in 2 episodes
• Reinforcement Learning — 10× in this paper · explained in 4 episodes
• Distillation — 7× in this paper · explained in 2 episodes
• Trace — 4× in this paper · explained in 2 episodes
• Language Models (LLMs) — 3× in this paper · explained in 2 episodes
• Large Language Models (LLMs) — 3× in this paper · explained in 9 episodes
• Pruning — 3× in this paper · explained in 4 episodes
• Fine-Tuning — 2× in this paper · explained in 5 episodes
• Foundation Model — 2× in this paper · explained in 3 episodes
• Foundation Models — 2× in this paper · explained in 8 episodes
• Hallucination — 2× in this paper · explained in 4 episodes
• Scaffold — 2× in this paper · explained in 2 episodes
• Absorb — 1× in this paper · explained in 1 episode
• Benchmarking — 1× in this paper · explained in 4 episodes
• Calibration — 1× in this paper · explained in 9 episodes
• Composition — 1× in this paper · explained in 2 episodes
• Credit assignment — 1× in this paper · explained in 3 episodes
• Evaluation Protocol — 1× in this paper · explained in 2 episodes
• Failure Modes — 1× in this paper · explained in 3 episodes
• Hyperparameters — 1× in this paper · explained in 2 episodes
• LLM judge — 1× in this paper · explained in 2 episodes
• Multi-Agent Framework — 1× in this paper · explained in 3 episodes
• Papers — 1× in this paper · explained in 3 episodes
• Verifiable Rewards — 1× in this paper · explained in 2 episodes
• Verification — 1× in this paper · explained in 2 episodes
• Weight Decay — 1× in this paper · explained in 2 episodes
• World Model — 1× in this paper · explained in 8 episodes

Sources

• gpt-oss-120b & gpt-oss-20b Model Card
• Efficient Agentic Reinforcement Learning with On-Policy Intrinsic Knowledge Boundary Enhancement
• ReSearch: Learning to Reason with Search for LLMs via Reinforcement Learning
• ET-Agent: Incentivizing Effective Tool-Integrated Reasoning Agent via Behavior Calibration
• Tool-Star: Empowering LLM-Brained Multi-Tool Reasoner via Reinforcement Learning
• S1-DeepResearch: Beyond Search, Toward Real-World Long-Horizon Research Agents
• DeepSearchQA: Bridging the Comprehensiveness Gap for Deep Research Agents
• Deep Research Agents: A Systematic Examination And Roadmap
• Reinforced Internal-External Knowledge Synergistic Reasoning for Efficient Adaptive Search Agent
• DeepWideSearch: Benchmarking Depth and Width in Agentic Information Seeking
• ToolPRMBench: Evaluating and Advancing Process Reward Models for Tool-using Agents
• WebSailor-V2: Bridging the Chasm to Proprietary Agents via Synthetic Data and Scalable Reinforcement Learning
• Verified Critical Step Optimization for LLM Agents
• AdaSearch: Balancing Parametric Knowledge and Search in Large Language Models via Reinforcement Learning
• SFR-DeepResearch: Towards Effective Reinforcement Learning for Autonomously Reasoning Single Agents
• Humanity's Last Exam
• Flash-Searcher: Fast and Effective Web Agents via DAG-Based Parallel Execution
• Rewarding the Scientific Process: Process-Level Reward Modeling for Agentic Data Analysis
• MatchTIR: Fine-Grained Supervision for Tool-Integrated Reasoning via Bipartite Matching
• DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models

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