NeurPRISE models and data

Models and inputs for Learning Scenario Reduction for Two-Stage Robust Optimization with Discrete Uncertainty (NeurIPS 2026). Code: https://github.com/lintianjue/NeurPRISE

checkpoints/<experiment>/<tag>[_<variant>].pt   NeurPRISE models used in the paper
data/<problem>/inst_*.pkl                       test instances
data/<problem>/label_prise_*.pkl                PRISE labels and exact objectives V(Ξ), x*
eval_regret/full_objs_cache.json                V(Ξ) of every test instance

Download with scripts/reproduce.sh download from the code repository, then rerun the paper's evaluations with scripts/reproduce.sh eval. See doc/REPRODUCE.md there. Optimization requires a Gurobi license. Rerun results can vary with hardware, solver version and settings.

Downloads last month: -; Downloads are not tracked for this model. How to track

Inference Providers NEW

This model isn't deployed by any Inference Provider. 🙋 Ask for provider support