Log-Driven Semantic Tuning for Mixed Integer Programming Solvers
Abstract
Automated configuration of mixed-integer programming (MIP) solvers largely relies on black-box optimization. Each trial requires an expensive solver run and provides only a scalar outcome, so tuning can require many such evaluations. Human experts take a different approach. They inspect solver logs, identify performance bottlenecks, and adjust the parameters that address them. We propose Log-Driven Semantic Tuning (LDST), which uses a large language model (LLM) to automate this process. LDST extracts compact diagnostics from solver logs, uses the LLM to propose sparse edits to the current best configuration based on parameter meanings and observed solver behavior, and validates each proposal before execution. A regime-adaptive objective selects the tuning signal according to the information available from a time-limited run, while a log-based completion signal estimates completion time from branch-and-bound tree growth recorded in ordinary solver logs, so that solved and unsolved runs share one time scale. On 215 instances from five benchmark sets, LDST reduces geometric-mean loss by 43.9% on easy instances and 16.7% on intermediate instances relative to the best of SMAC3, RBFOpt, and random search. In both regimes, LDST achieves a lower loss with 10 proposals than any of these baselines achieves with 50. Configurations that LDST tunes on a single instance also transfer better than baseline configurations to unseen instances of the same family.
Related
- When Solvers Get Moody: Quantifying and Predicting Performance Variability in Mixed Integer Programming
- Mitigating Performance Variability in Mixed-Integer Programming Solvers via LP State Steering
- Integer linear programming based methods for computing the maximum chordal subgraph of an arbitrary graph
- Hidden Convexity via Symmetric Displacement Covers: Mixed-Integer Quadratic Programming and Integer Cubic Programming in Fixed Dimension
- Learning to Solve Mixed-Integer Nonlinear Programming with Neural Networks and Sequential Quadratic Programming Refinement
BibTeX
@misc{you2026logdriven,
title = {{Log-Driven Semantic Tuning for Mixed Integer Programming Solvers}},
author = {Zhengzhong Ricky You and Kaizhao Sun and Zeping Min and Wotao Yin and Xinshang Wang},
howpublished = {arXiv},
year = {2026},
month = nov,
doi = {10.48550/arxiv.2611.00621},
eprint = {2611.00621},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2611.00621},
}
RIS
TY - UNPB
TI - Log-Driven Semantic Tuning for Mixed Integer Programming Solvers
AU - Zhengzhong Ricky You
AU - Kaizhao Sun
AU - Zeping Min
AU - Wotao Yin
AU - Xinshang Wang
PY - 2026
DA - 2026/11/01
PB - arXiv
DO - 10.48550/arxiv.2611.00621
UR - https://arxiv.org/abs/2611.00621
AB - Automated configuration of mixed-integer programming (MIP) solvers largely relies on black-box optimization. Each trial requires an expensive solver run and provides only a scalar outcome, so tuning can require many such evaluations. Human experts take a different approach. They inspect solver logs, identify performance bottlenecks, and adjust the parameters that address them. We propose Log-Driven Semantic Tuning (LDST), which uses a large language model (LLM) to automate this process. LDST extracts compact diagnostics from solver logs, uses the LLM to propose sparse edits to the current best configuration based on parameter meanings and observed solver behavior, and validates each proposal before execution. A regime-adaptive objective selects the tuning signal according to the information available from a time-limited run, while a log-based completion signal estimates completion time from branch-and-bound tree growth recorded in ordinary solver logs, so that solved and unsolved runs share one time scale. On 215 instances from five benchmark sets, LDST reduces geometric-mean loss by 43.9% on easy instances and 16.7% on intermediate instances relative to the best of SMAC3, RBFOpt, and random search. In both regimes, LDST achieves a lower loss with 10 proposals than any of these baselines achieves with 50. Configurations that LDST tunes on a single instance also transfer better than baseline configurations to unseen instances of the same family.
ER -