Problem Distributions as Tasks: Repurposing Meta Learning for Generative Combinatorial Optimization towards Multi-task Pretraining and Adaptation

Wenzheng Pan, Jiale Ma, Nuoyan Chen, Yang Li, Junchi Yan, International Conference on Machine Learning (ICML), 2026. [Code]

We introduce M²GenCO, a meta-generative framework that treats problem distributions as tasks to enable efficient multi-task pretraining, few-shot adaptation, and robust generalization across graph-based combinatorial optimization problems.

NExCO: Native Solution Expansion for Diffusion-based Combinatorial Optimization

Yu Wang, Yang Li, Jiale Ma, Junchi Yan, Yi Chang, International Conference on Learning Representations (ICLR), 2026. [Code]

We propose NExCO, a masked diffusion framework that realizes adaptive solution expansion as a native generative principle for neural combinatorial optimization. Our framework is built on three key components: a CO-specific forward corruption that preserves sparsity and yields semantic partial solutions, a time-agnostic GNN denoiser trained under optimization consistency, and a Native Adaptive Expansion (NAE) inference strategy that progressively selects confident variables under feasibility constraints.

ML4CO-Bench-101: Benchmark Machine Learning for Classic Combinatorial Problems on Graphs

Jiale Ma, Wenzheng Pan, Yang Li, Junchi Yan, Advances in Neural Information Processing Systems (NeurIPS), 2025. [Code]

We establishe ML4CO-Bench-101, a standardized benchmark and modular evaluation framework that systematically categorizes, reproduces, and compares neural solvers across seven mainstream graph-based combinatorial problems.

Generative Modeling Reinvents Supervised Learning: Label Repurposing with Predictive Consistency Learning

Yang Li, Jiale Ma, Yebin Yang, Qitian Wu, Hongyuan Zha, Junchi Yan, International Conference on Machine Learning (ICML), 2025. [Code]

We propose a novel predictive consistency learning framework beyond previous methods with direct prediction, aiming to explore the full potential of label information for supervision during the learning process.

COExpander: Adaptive Solution Expansion for Combinatorial Optimization

Jiale Ma*, Wenzheng Pan*, Yang Li, Junchi Yan, International Conference on Machine Learning (ICML), 2025. [Code]

We introduce COExpander, an adaptive expansion paradigm that bridges global prediction and local construction by progressively determining decision variables with dynamically controlled step sizes for scalable combinatorial optimization.

UniCO: On Unified Combinatorial Optimization via Problem Reduction to Matrix-Encoded General TSP

Wenzheng Pan*, Hao Xiong*, Jiale Ma, Wentao Zhao, Yang Li, Junchi Yan, International Conference on Learning Representations (ICLR), 2025. [Code]

We propose UniCO, a unified neural combinatorial optimization framework that reduces diverse COPs into matrix-encoded general TSP and solves them with tailored matrix-based RL and diffusion solvers: 1) MatPOENet, an RL-based sequential model with pseudo one-hot embedding (POE) scheme and 2) MatDIFFNet, a Diffusion-based generative model with the mix-noised reference mapping scheme.

Unify ML4TSP: Drawing Methodological Principles for TSP and Beyond from Streamlined Design Space of Learning and Search

Yang Li, Jiale Ma, Wenzheng Pan, Runzhong Wang, Haoyu Geng, Nianzu Yang, Junchi Yan, International Conference on Learning Representations (ICLR), 2025. [Code]

We present ML4TSPBench, a modular framework that decomposes learning-based TSP solvers into reusable learning and search components, revealing key design principles for stronger and more principled ML4CO methods.

Pygmtools: A Python Graph Matching Toolkit

Runzhong Wang, Ziao Guo, Wenzheng Pan, Jiale Ma, Yikai Zhang, Nan Yang, Qi Liu, Longxuan Wei, Hanxue Zhang, Chang Liu, Zetian Jiang, Xiaokang Yang, Junchi Yan, Journal of Machine Learning Research (JMLR), 2024. [Code]

We release Pygmtools, an open-source Python toolkit that unifies classical, multi-graph, and learning-based graph matching solvers across multiple numerical backends for research and practical applications.