Research map / Large language models / Agents, code and math

Process-Aware LNS for Large-Scale MILP via Context-Enhanced Fine-Tuned LLM-driven Selector

NeurIPS 2026 · Large language models › Agents, code and math · approach: LLMs for optimization and heuristics

Read the paperCheck your idea against 108,967 papers

AI-extracted summary. Written by AI from the paper's abstract only. Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Proposes a process-aware LNS for large-scale MILP that uses a context-enhanced, fine-tuned LLM as the neighborhood selector.

Problem

Large-scale MILP with millions of variables is intractable; LNS efficacy depends on neighborhood selection, which is the target of this work.

Approach

Uses a fine-tuned LLM as the LNS neighborhood selector, enhanced with search-process context, to choose which variables to free at each step.

Closest papers by meaning

• Large Language Model-driven Large Neighborhood Search for Large-Scale MILP Problems (ICML 2025 · similarity 0.93)
• Regularized Large Neighborhood Search (NeurIPS 2026 · similarity 0.87)
• Large Language Models as End-to-end Combinatorial Optimization Solvers (NeurIPS 2025 · similarity 0.86)
• Adaptive Kernel Design for Bayesian Optimization Is a Piece of CAKE with LLMs (NeurIPS 2025 · similarity 0.86)
• SolverLLM: Leveraging Test-Time Scaling for Optimization Problem via LLM-Guided Search (NeurIPS 2025 · similarity 0.86)
• LLMOPT: Learning to Define and Solve General Optimization Problems from Scratch (ICLR 2025 · similarity 0.86)