Research map / Reinforcement learning / Visual and representation RL
Saliency-Guided Representation with Consistency Policy Learning for Visual Unsupervised Reinforcement Learning
Jingbo Sun, Qichao Zhang, Songjun Tu, Xing Fang, Yupeng Zheng, Haoran Li, Ke Chen, Dongbin Zhao
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 saliency-guided dynamics representation learning (decouple encoder from SR objective) (ablation) | HILP + consistency policy · ExORL RND, Walker | return: 345 -> 439 | Improved |
| add saliency-guided dynamics representation learning (ablation) | HILP + consistency policy · ExORL RND, Quadruped | return: 352 -> 485 | Improved |
| add consistency policy (ablation) | SRCP with saliency representation · ExORL RND, Walker | return: 396 -> 439 | Improved |
| add consistency policy (ablation) | SRCP with saliency representation · ExORL RND, Quadruped | return: 406 -> 485 | Improved |
| add saliency representation and consistency policy (ablation) | HILP · ExORL RND, Cheetah | return: 599 -> 602 | No clear effect |
| add SRCP components (ablation) | HILP · ExORL RND, Jaco | return: 34 -> 50 | Improved |
| use classifier-free guidance (omega=3 vs 0) (ablation) | SRCP · ExORL RND, Quadruped | return: 234 -> 485 | Improved |
| increase guidance weight omega from 3 to 6 (ablation) | SRCP · ExORL RND, Quadruped | return: 485 -> 455 | Hurt |
| add saliency loss weight beta=0.5 (vs 0) (ablation) | SRCP · ExORL RND, Walker | return: 341 -> 439 | Improved |
| increase beta from 0.5 to 1 (ablation) | SRCP · ExORL RND, Walker | return: 439 -> 396 | Hurt |
| plug SRCP into forward-backward (FB) learner | FB · ExORL RND, Walker and Quadruped (8 tasks) | return: +23% to +490% across tasks | Improved |
| replace HILP encoder objective with SRCP vs TACO/Premier-TACO contrastive representations (ablation) | HILP · ExORL visual URL | return | Mixed |
Closest papers by meaning
- VIGOR: Zero-Shot Visual Generalization via Latent-Space Consistency in Model-Based Reinforcement Learning (NeurIPS 2026 · similarity 0.89)
- Focus On What Matters: Separated Models For Visual-Based RL Generalization (NeurIPS 2024 · similarity 0.88)
- A Simple Framework for Generalization in Visual RL under Dynamic Scene Perturbations (NeurIPS 2024 · similarity 0.88)
- Generalizing Consistency Policy to Visual RL with Prioritized Proximal Experience Regularization (NeurIPS 2024 · similarity 0.88)
- Learning Successor Features the Simple Way (NeurIPS 2024 · similarity 0.88)
- What Effects the Generalization in Visual Reinforcement Learning: Policy Consistency with Truncated Return Prediction (AAAI 2024 · similarity 0.88)
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.
