AI Paper Atlas

Research map / Label-efficient and robust learning / Recognition and long-tail

Dynamic Stratified Contrastive Learning with Upstream Augmentation for MILP Branching

Tongkai Lu, Shuai Ma, Chongyang Tao

ICML 2026 · Label-efficient and robust learning › Recognition and long-tail · approach: Dense features and foundation priors · 0 citations (OpenAlex)

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 changedWhereMeasuredEffect
add upstream-augmented MILP derivation (ablation)SC-MILP · Set Coveringacc@1 (%): 64.2 -> 68.9Improved
add equivalent MILP derivation (linear transformation, redundant constraint) (ablation)SC-MILP · Set Coveringacc@1 (%): 64.8 -> 68.9Improved
add perturbed MILP derivation (ablation)SC-MILP · Set Coveringacc@1 (%): 68.1 -> 68.9Improved
add dynamic stratified contrastive learning (ablation)SC-MILP · Set Covering Hardsolving time (s): 1137.06 -> 953.24Improved
add dynamic stratified weights to contrastive loss (ablation)SC-MILP · Set Coveringacc@1 (%): 66.4 -> 68.9Improved
add hybrid branching strategy (top-k then strong branching at upstream nodes) (ablation)SC-MILP · Set Covering Hardsolving time (s): 1011.35 -> 953.24Improved
add hybrid branching strategy (ablation)SC-MILP · Set Covering Easysolving time (s): 5.95 -> 5.99No clear effect
train GCNN on upstream-augmented data instead of equal-sized original data (ablation)GCNN · Set Coveringacc@1 (%): 62.1 -> 65.4Improved
train MILP-Evolve on upstream-augmented data instead of equal-sized original data (ablation)MILP-Evolve · Set Coveringacc@1 (%): 61.7 -> 63.6Improved
add dynamic stratified contrastive learning on top of upstream-augmented data (ablation)GCNN · Set Covering Hardsolving time (s): 1154.33 -> 953.24Improved
increase number of strata m (ablation)SC-MILP · four generated benchmarkssolving timeMixed
increase hybrid ratio rho (ablation)SC-MILP · four generated benchmarksaccuracy and timeMixed

Closest papers by meaning

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.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy