Research map / Probabilistic and scientific modeling / Uncertainty and Gaussian processes
ReBaPL: Repulsive Bayesian Prompt Learning
Yassir Bendou, Omar Ezzahir, Eduardo Montesuma, Gabriel Mahuas, Victoria Shevchenko, Mike Gartrell
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 repulsion force (MMD) to ReBaPL sampling (ablation) | MaPLe + ReBaPL · Base-to-novel, 11-dataset average | HM: 78.93 -> 79.52 | Improved |
| add repulsion force (MMD) to ReBaPL sampling (ablation) | MaPLe + ReBaPL · Base-to-novel, 11-dataset average | Novel accuracy: 75.47 -> 76.08 | Improved |
| use Wasserstein instead of MMD for repulsion (ablation) | MaPLe + ReBaPL · Base-to-novel, 11-dataset average | HM: 79.52 -> 79.44 | No clear effect |
| add Bayesian posterior sampling (cyclical SGHMC) without repulsion (ablation) | MaPLe · Base-to-novel, 11-dataset average | HM: 78.37 -> 78.93 | Improved |
| add ReBaPL to MMRL | MMRL · Base-to-novel, 11-dataset average | Novel accuracy: 76.52 -> 77.44 | Improved |
| add ReBaPL to MMRL | MMRL · Base-to-novel, 11-dataset average | Base accuracy: 85.54 -> 85.74 | Improved |
| add ReBaPL to MaPLe | MaPLe · Base-to-novel, 11-dataset average | HM: 78.37 -> 79.52 | Improved |
| add ReBaPL to MaPLe | MaPLe · Base-to-novel FGVCAircraft | Novel accuracy: 34.90 -> 34.03 | Hurt |
| add ReBaPL to MMRL | MMRL · Base-to-novel EuroSAT | HM: 85.50 -> 89.71 | Improved |
| add ReBaPL to MMRL | MMRL · Cross-dataset transfer | Average accuracy: 66.87 -> 67.62 | Improved |
| add ReBaPL to MaPLe | MaPLe · Cross-dataset transfer | Average accuracy: 65.63 -> 66.77 | Improved |
| add ReBaPL to MaPLe | MaPLe · Domain generalization ImageNet-R | accuracy: 75.33 -> 75.40 | No clear effect |
Closest papers by meaning
- Bayesian Prompt Learning for Image-Language Model Generalization (ICCV 2023 · similarity 0.89)
- Make Prompts Adaptable: Bayesian Modeling for Vision-Language Prompt Learning with Data-Dependent Prior (AAAI 2024 · similarity 0.87)
- Understanding Prompt Tuning and In-Context Learning via Meta-Learning (NeurIPS 2025 · similarity 0.86)
- Scalable Bayesian Learning with posteriors (ICLR 2025 · similarity 0.85)
- Large Language Models to Enhance Bayesian Optimization (ICLR 2024 · similarity 0.85)
- BayesPrompt: Prompting Large-Scale Pre-Trained Language Models on Few-shot Inference via Debiased Domain Abstraction (ICLR 2024 · 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.
