Research map / Learning theory and algorithms / Combinatorial and search optimization
Evolving Interdependent Operators with Large Language Models for Multi-Objective Combinatorial Optimization
junhao qiu, Xin Chen, Liang Ge, Liyong Lin, Zhichao Lu, Qingfu Zhang
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 |
|---|---|---|---|
| evolve operator combinations with E2OC instead of expert operators | NSGA-II · Bi-FJSP | HV: 0.1996 -> 0.2435 (RI 22.00%) | Improved |
| E2OC instead of expert operators | NSGA-II · Bi-TSP | HV: 0.3881 -> 0.4424 (RI 14.00%) | Improved |
| E2OC instead of expert operators | NSGA-II · Tri-FJSP | HV: 0.1266 -> 0.1485 (RI 17.36%) | Improved |
| E2OC instead of expert operators | NSGA-II · Tri-TSP | HV: 0.1824 -> 0.1939 (RI 6.30%) | Improved |
| co-evolve design strategies versus sequentially optimizing single heuristics | E2OC vs MCTS-AHD · Bi-FJSP | HV: 0.2269 -> 0.2435 | Improved |
| use direct LLM-guided operator combination control | multi-heuristic co-design · Bi-FJSP | HV: LLM 0.2148 vs Random 0.2263 | Hurt |
| replace progressive strategy search with fixed design strategy (ablation) | E2OC (MCTS OC) · Bi-FJSP | HV: 0.2435 -> 0.2085 | Hurt |
| add operator rotation instead of sequential operator design (ablation) | E2OC · Bi-FJSP | HV: 0.2187 -> 0.2435 | Improved |
| use EoH as underlying designer rather than ReEvo/FunSearch/MCTS-AHD (ablation) | E2OC · Bi-FJSP | HV: 0.2435 vs 0.2213 to 0.2269 | Improved |
| search a warm-start fixed design thought space instead of dynamic sampling (ablation) | E2OC vs MCTS Sample · Bi-FJSP | HV: 0.2181 -> 0.2435 | Improved |
| reuse E2OC outputs as inputs for further runs (ablation) | E2OC · Bi-FJSP | HV: 0.2435 -> 0.2454 -> 0.2475 | Improved |
| use stronger/more expensive LLM (gemini-2.5-pro) than deepseek-chat (ablation) | E2OC · Bi-FJSP | HV: 0.2271 vs 0.2223 HV | No clear effect |
Closest papers by meaning
- HSEvo: Elevating Automatic Heuristic Design with Diversity-Driven Harmony Search and Genetic Algorithm Using LLMs (AAAI 2025 · similarity 0.89)
- Multi-Objective Evolution of Heuristic Using Large Language Model (AAAI 2025 · similarity 0.89)
- Evolution of Heuristics: Towards Efficient Automatic Algorithm Design Using Large Language Model (ICML 2024 · similarity 0.87)
- MOTIF: Multi-strategy Optimization via Turn-based Interactive Framework (AAAI 2026 · similarity 0.87)
- ReEvo: Large Language Models as Hyper-Heuristics with Reflective Evolution (NeurIPS 2024 · similarity 0.87)
- PathWise: Planning through World Model for Automated Heuristic Design via Self-Evolving LLMs (ICML 2026 · 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.
