Research map / Learning theory and algorithms / Optimal transport
Supervised Training of Conditional Monge Maps
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 changed | Where | Measured | Effect |
|---|---|---|---|
| add context conditioning (CondOT vs context-free ICNN OT) | ICNN OT · Givinostat dosage in-sample | MMD: 0.0365 -> 0.0111 | Helped |
| add context conditioning (CondOT vs ICNN OT) | ICNN OT · Givinostat dosage out-of-sample | MMD: 0.0466 -> 0.0374 | Helped |
| CondOT identity init vs CPA | CPA · Givinostat dosage in-sample | MMD: 0.1502 -> 0.0111 | Helped |
| CondOT vs CPA conditioned on cell line | CPA · Givinostat cell line in-sample | MMD: 0.2551 -> 0.0148 | Helped |
| Gaussian instead of identity initialization (ablation) | CondOT · Givinostat dosage in-sample | MMD: 0.0111 -> 0.0128 | No clear effect |
| Gaussian instead of identity initialization (ablation) | CondOT · Givinostat dosage out-of-sample | MMD: 0.0374 -> 0.0325 | Helped |
| Gaussian instead of identity initialization (ablation) | CondOT · Givinostat/Trametinib 10,000 nM | Wasserstein loss W_eps: 20.97 -> 21.30 | No clear effect |
| proposed initialization vs vanilla random initialization (ablation) | ICNN potentials · synthetic 2D mapping | convergence speed | Helped |
| multi-task CondOT vs per-condition single-task ICNN OT (ablation) | CondOT · Givinostat/Trametinib 10 nM | Wasserstein loss W_eps: 10.98 -> 10.54 | No clear effect |
| multi-task CondOT vs single ICNN OT trained on all conditions | ICNN OT · Givinostat/Trametinib 10,000 nM | Wasserstein loss W_eps: 31.02 -> 20.97 | Helped |
Closest papers by meaning
- Unbalancedness in Neural Monge Maps Improves Unpaired Domain Translation (ICLR 2024 · similarity 0.88)
- GENOT: Entropic (Gromov) Wasserstein Flow Matching with Applications to Single-Cell Genomics (NeurIPS 2024 · similarity 0.87)
- Hierarchical Refinement: Optimal Transport to Infinity and Beyond (ICML 2025 · similarity 0.87)
- Estimation of Stochastic Optimal Transport Maps (NeurIPS 2025 · similarity 0.86)
- Overcoming Spurious Solutions in Semi-Dual Neural Optimal Transport: A Smoothing Approach for Learning the Optimal Transport Plan (ICML 2025 · similarity 0.86)
- Progressive Entropic Optimal Transport Solvers (NeurIPS 2024 · similarity 0.86)
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.