Research map / Multi-agent systems and games / LLM agents
Opt-Miner: Empowering Information-Seeking Agent with Tree-Guided Data Synthesis for Optimization Modeling
Haoyang Liu, Yuyang Cai, Jie Wang, Xiongwei Han, Minyang Hu, Shuqi LIU, Mingxuan Yuan, Jianye HAO, Feng Wu
Read the paperCheck your idea against 137,217 papers
AI-extracted summary. Written by AI from the paper's full text. Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Proposes Opt-Miner, an information-seeking agent for optimization modeling that uses tree-guided data synthesis to address the knowledge-intensive nature of real-world problems.
Problem
LLM agents for automated optimization modeling struggle with knowledge-intensive real-world problems because static parametric knowledge lacks the domain expertise needed to understand scenarios and apply appropriate mathematical techniques.
Approach
Equips an LLM agent with information-seeking capability (retrieving external domain knowledge) and trains it on tree-guided synthesized data, rather than relying only on parametric knowledge.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add knowledge fogging operator to the synthesis data (ablation) | Opt-Miner (Qwen3-4B-Instruct) · IndustryOR | Pass@1 accuracy (%): 56.3 -> 66.4 | Improved |
| add knowledge fogging operator to the synthesis data (ablation) | Opt-Miner (Qwen3-4B-Instruct) · Opt-Miner-Bench | Pass@1 accuracy (%): 14.8 -> 21.1 | Improved |
| replace 1k extra OptMATH data with 1k union-synthesized Opt-Miner data (ablation) | Qwen3-4B-Instruct trained on OptMATH-1k · Opt-Miner-Bench | Pass@1 accuracy (%): 27.1 -> 34.3 | Improved |
| add 1k transfer-synthesized data to OptMATH-1k (ablation) | Qwen3-4B-Instruct trained on OptMATH-1k · IndustryOR | Pass@1 accuracy (%): 52.1 -> 62.6 | Improved |
| use full Opt-Miner mixed data instead of OptMATH-2k (ablation) | Qwen3-4B-Instruct · IndustryOR | Pass@1 accuracy (%): 54.5 -> 66.4 | Improved |
| add retrieval matching reward to GRPO (R-GRPO) (ablation) | Qwen3-4B-Instruct · IndustryOR | Pass@1 accuracy (%): 57.3 -> 66.4 | Improved |
| add retrieval matching reward to SIRL (R-SIRL) (ablation) | Qwen3-4B-Instruct · OptMATH | Pass@1 accuracy (%): 57.8 -> 62.8 | Improved |
| add retrieval matching reward to SIRL (R-SIRL) (ablation) | Qwen3-4B-Instruct · MAMO ComplexLP | Pass@1 accuracy (%): 91.3 -> 91.9 | Mixed |
| add web research module to base model (ablation) | Qwen3-4B-Instruct · IndustryOR | Pass@1 accuracy (%): 15.2 -> 20.9 | Improved |
| add web research module to base model (ablation) | Qwen3-4B-Instruct · OptMATH | Pass@1 accuracy (%): 40.8 -> 40.5 | No clear effect |
| add web research module to base model (ablation) | DeepSeek-V3 · Opt-Miner-Bench | Pass@1 accuracy (%): 23.4 -> 33.1 | Improved |
| add web research module to trained agent (ablation) | Opt-Miner (Qwen3-4B-Instruct) · Opt-Miner-Bench | Pass@1 accuracy (%): 28.3 -> 35.5 | Improved |
Closest papers by meaning
- Empowering Efficiency and Efficacy in WebAgent via Enabling Info-Rich Seeking (ICLR 2026 · similarity 0.87)
- Search-R2: Enhancing Search-Integrated Reasoning via Actor-Refiner Collaboration (ICML 2026 · similarity 0.86)
- GPTSwarm: Language Agents as Optimizable Graphs (ICML 2024 · similarity 0.86)
- EvoOptiGraph: Weakness-Driven Coevolution via Graph-Based Structural Generation for Optimization Modeling (NeurIPS 2026 · similarity 0.86)
- AgentSwift: Efficient LLM Agent Design via Value-Guided Hierarchical Search (AAAI 2026 · similarity 0.86)
- Language Agent Tree Search Unifies Reasoning, Acting, and Planning in Language Models (ICML 2024 · similarity 0.86)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
