Research map / Reinforcement learning / Visual and representation RL
For Pre-Trained Vision Models in Motor Control, Not All Policy Learning Methods are Created Equal
Yingdong Hu, Renhao Wang, Li Erran Li, Yang Gao
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 |
|---|---|---|---|
| Use visual reward function (VRF) instead of RL for evaluation | pre-trained vision models for motor control · Meta-World, Robosuite, Franka-Kitchen | consistency of model rankings across runs and environments | Improved |
| Use RL (DrQ-v2) as downstream policy learning method | pre-trained vision models · Meta-World vs Robosuite | model ranking consistency | Hurt |
| Increase seeds from 3 to 40+ runs for RL comparisons (ablation) | RL evaluation of vision models · Meta-World (5 tasks) | stability of probability of improvement | Improved |
| Use linear probing loss as a proxy for BC performance (lower loss) | pre-trained vision models with BC · Franka-Kitchen (5 tasks) | correlation with BC success rate: Pearson r = -0.78 | Improved |
| Use ImageNet k-NN accuracy as predictor of VRF performance | ImageNet-pretrained vision models with VRF · all environments | correlation with VRF performance: r = 0.91 | Improved |
| Use MAE pre-trained features directly (masked image modeling) as frozen representations for VRF | MAE with VRF · all environments | IQM success rate | Hurt |
| Use joint embedding architectures (contrastive/self-distillation) rather than masked image modeling | pre-trained vision models with VRF · all environments | VRF success rate | Improved |
| Use iBOT (MIM plus self-distillation) over other ViTs | ViT-B/16 models with BC · BC across environments | BC performance | Improved |
| Use ResNet-50 backbone instead of ViT-B/16 | pre-trained vision models with BC · BC across environments | BC performance | Improved |
Closest papers by meaning
- Latent Policy Barrier: Learning Robust Visuomotor Policies by Staying In-Distribution (NeurIPS 2025 · similarity 0.90)
- World Models as Reference Trajectories for Rapid Motor Adaptation (NeurIPS 2025 · similarity 0.88)
- DMC-VB: A Benchmark for Representation Learning for Control with Visual Distractors (NeurIPS 2024 · similarity 0.88)
- DynaMo: In-Domain Dynamics Pretraining for Visuo-Motor Control (NeurIPS 2024 · similarity 0.88)
- Perception-R1: Pioneering Perception Policy with Reinforcement Learning (NeurIPS 2025 · similarity 0.87)
- When Should We Prefer State-to-Visual DAgger over Visual Reinforcement Learning? (AAAI 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.
