Research map / Large language models / Agents, code and math
Navigating the Alpha Jungle: An LLM-Powered MCTS Framework for Formulaic Alpha Factor Mining
Yu Shi, Yitong Duan, Jian Li
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add Diversity feedback to MCTS (Effectiveness only -> Eff.+Div.) (ablation) | MCTS, LightGBM · Alpha mining, CSI stock pools | IC: 0.0409 -> 0.0501 | Improved |
| add Turnover feedback (ablation) | MCTS, LightGBM · Alpha mining, CSI stock pools | IC: 0.0501 -> 0.0492 | Hurt |
| add Turnover feedback (ablation) | MCTS, LightGBM · Alpha mining, CSI stock pools | IR: 1.0106 -> 1.1062 | Improved |
| add Turnover feedback (ablation) | MCTS, MLP · Alpha mining, CSI stock pools | IR: 1.0462 -> 1.2556 | Improved |
| add Stability feedback (on top of Eff.+Div.+Turn.) (ablation) | MCTS, LightGBM · Alpha mining, CSI stock pools | IR: 1.1062 -> 1.0331 | Hurt |
| add Stability feedback (on top of Eff.+Div.+Turn.) (ablation) | MCTS, MLP · Alpha mining, CSI stock pools | IR: 1.2556 -> 1.1773 | Hurt |
| add Overfitting Risk feedback (on top of Eff.+Div.+Turn.+Stab.) (ablation) | MCTS, LightGBM · Alpha mining, CSI stock pools | IC: 0.0495 -> 0.0515 | Improved |
| add Frequent Subtree Avoidance (ablation) | MCTS, LightGBM · Alpha mining, CSI stock pools | IC: 0.0515 -> 0.0549 | Improved |
| add Frequent Subtree Avoidance (ablation) | MCTS, MLP · Alpha mining, CSI stock pools | IR: 1.2127 -> 1.2712 | Improved |
| replace CoT with MCTS search (same Eff.+Div. feedback) (ablation) | alpha mining with GPT-4.1, LightGBM · Alpha mining, CSI stock pools | IC: 0.0434 -> 0.0501 | Improved |
| replace ToT with MCTS search (same Eff.+Div. feedback) (ablation) | alpha mining with GPT-4.1, LightGBM · Alpha mining, CSI stock pools | IC: 0.0459 -> 0.0501 | Improved |
| MCTS with Effectiveness feedback only vs CoT (ablation) | alpha mining with GPT-4.1, LightGBM · Alpha mining, CSI stock pools | IC: 0.0434 -> 0.0409 (IC); 0.0707 -> 0.0941 (AER) | Mixed |
Closest papers by meaning
- AlphaPROBE: Alpha Mining via principled retriveval and on-graph biased evolution (NeurIPS 2026 · similarity 0.92)
- AlphaForge: A Framework to Mine and Dynamically Combine Formulaic Alpha Factors (AAAI 2025 · similarity 0.92)
- AlphaQCM: Alpha Discovery in Finance with Distributional Reinforcement Learning (ICML 2025 · similarity 0.89)
- AlphaPareto: Formulaic Alpha Discovery with LLM-Guided Multi-Objective Reinforcement Learning (NeurIPS 2026 · similarity 0.88)
- R&D-Agent-Quant: A Multi-Agent Framework for Data-Centric Factors and Model Joint Optimization (NeurIPS 2025 · similarity 0.86)
- FinBen: A Holistic Financial Benchmark for Large Language Models (NeurIPS 2024 · similarity 0.85)
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.
