Research map / Reinforcement learning / RL from feedback
Behavioral Mode Discovery for Fine-tuning Multimodal Generative Policies
Alberta Longhini, David Emukpere, Jean-Michel Renders, Seungsu Kim
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 BMD mutual-information regularization | DPPO · Gaussian mixture G1 | SR_M: 0.58 -> 1.00 | Improved |
| add BMD mutual-information regularization | DPPO · Gaussian mixture G2 | SR_M: 0.42 -> 0.75 | Improved |
| add BMD mutual-information regularization | RES · Gaussian mixture G2 | SR_M: 0.50 -> 1.00 | Improved |
| add BMD mutual-information regularization | DSRL · Gaussian mixture G1 | SR_M: 0.25 -> 0.33 | Improved |
| add BMD mutual-information regularization | DSRL · Gaussian mixture G2 | SR_M: 0.08 -> 0.08 | No clear effect |
| add BMD mutual-information regularization | RES · ManiSkill Lift | SR_M: 0.50 -> 0.99 | Improved |
| add BMD mutual-information regularization | DSRL · ManiSkill Lift | SR_M: 0.78 -> 0.88 | Improved |
| add BMD mutual-information regularization | DPPO · ManiSkill Lift | SR_M: 0.57 -> 0.55 | No clear effect |
| add BMD mutual-information regularization | RES · ManiSkill Avoid | SR_M: 0.04 -> 0.30 | Improved |
| add BMD mutual-information regularization | DSRL · ManiSkill Avoid | SR_M: 0.09 -> 0.42 | Improved |
| add BMD mutual-information regularization | DSRL · ANYmal locomotion | SR_M: 0.25 -> 0.89 | Improved |
| add BMD mutual-information regularization | DSRL · Franka Kitchen | SR_M: 0.04 -> 0.10 | Improved |
Closest papers by meaning
- Learning Multimodal Behaviors from Scratch with Diffusion Policy Gradient (NeurIPS 2024 · similarity 0.88)
- Text-Aware Diffusion for Policy Learning (NeurIPS 2024 · similarity 0.88)
- MSRL: Scaling Generative Multimodal Reward Modeling via Multi-Stage Reinforcement Learning (CVPR 2026 · similarity 0.87)
- Beyond the Dirac Delta: Mitigating Diversity Collapse in Reinforcement Fine-Tuning for Image Generation (NeurIPS 2026 · similarity 0.87)
- Score Regularized Policy Optimization through Diffusion Behavior (ICLR 2024 · similarity 0.87)
- Improving Generative Behavior Cloning via Self-Guidance and Adaptive Chunking (NeurIPS 2025 · similarity 0.87)
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.
