Research map / Learning theory and algorithms / Optimal transport
Hyperparameter Trajectory Inference with Conditional Lagrangian Optimal Transport
Harry Amad, Mihaela van der Schaar
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 changed | Where | Measured | Effect |
|---|---|---|---|
| add learned potential term U (dense traversal bias) (ablation) | CLOT with learned metric · Semicircle CTI | NLL: 23.008 -> -0.662 | Improved |
| add learned potential term U to identity metric (ablation) | CLOT K_I · Semicircle CTI | NLL: 105.713 -> -0.532 | Improved |
| add learned metric G to the potential-only model (ablation) | CLOT K_I - U · Semicircle CTI | NLL: -0.532 -> -0.662 | Improved |
| add learned metric G (least-action bias) to identity model (ablation) | CLOT K_I · Semicircle CTI | CD: 0.323 -> 0.158 | Improved |
| add learned potential term U (ablation) | CLOT K_theta · Reacher | average reward: -6.158 -> -6.093 | Improved |
| add learned metric G to potential-only model (ablation) | CLOT K_I - U · Reacher | average reward: -6.397 -> -6.093 | Improved |
| add learned potential term U (ablation) | CLOT K_theta · Cancer_nl | average reward: 54.44 -> 91.94 | Improved |
| add learned metric G to potential-only model (ablation) | CLOT K_I - U · Cancer_nl | average reward: 86.21 -> 91.94 | Improved |
| add learned potential term U (ablation) | CLOT K_theta · Two moons dropout | Wasserstein distance: 0.404 -> 0.079 | Improved |
| add learned metric G to potential-only model (ablation) | CLOT K_I - U · Two moons dropout | Wasserstein distance: 0.060 -> 0.079 | Hurt |
| learn eigenvalues E_theta instead of fixed eigenvalues (ablation) | CLOT metric G · Semicircle CTI | NLL: -0.602 -> -0.662 | Improved |
| learn eigenvalues E_theta instead of fixed eigenvalues (ablation) | CLOT metric G · Cancer | average reward: 98.72 -> 102.49 | Improved |
Closest papers by meaning
- Markovian Dynamics Enforcer: Feasibility Preserving Correction on Learned Dyanmics Manifolds (NeurIPS 2026 · similarity 0.87)
- Guided Trajectory Generation with Diffusion Models for Offline Model-based Optimization (NeurIPS 2024 · similarity 0.87)
- TMPO: Trajectory Matching Policy Optimization for Diverse and Efficient Diffusion Alignment (NeurIPS 2026 · similarity 0.86)
- GeoSPRINT: Geometric Redundancy-Aware Step Pruning for Inference in Diffusion Trajectories (NeurIPS 2026 · similarity 0.86)
- Gradient Descent as Loss Landscape Navigation: a Normative Framework for Deriving Learning Rules (NeurIPS 2025 · similarity 0.85)
- Prior-Guided Diffusion Planning for Offline Reinforcement Learning (NeurIPS 2025 · similarity 0.85)
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.
