Research map / Reinforcement learning / Visual and representation RL
Domain Adaptation In Reinforcement Learning Via Latent Unified State Representation
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 domain-general latent embeddings alone as the RL state representation | LUSR · CarRacing seen target domains | score: 803.52 vs 834.99; 807.37 vs 819.05; 803.11 vs 713.86; 781.94 vs 711.85 | Helped |
| Use domain-general latent embeddings alone as the RL state representation | LUSR · CarRacing C1 | score: 678.7 vs 715.43 | Hurt |
| Use domain-general latent embeddings alone as the RL state representation | LUSR · CarRacing C2 | score: 800.56 vs 806.76 | Hurt |
| Use domain-general latent embeddings alone as the RL state representation | LUSR · CARLA Clear Noon | score: 1175.61 vs 674.42 | Helped |
| Use domain-general latent embeddings alone as the RL state representation | LUSR · CARLA Hard Rain | score: 1270.32 vs 846.14 | Helped |
| Increase DARLA β to encourage greater disentanglement | DARLA · CARLA Clear Noon and CARLA Hard Rain | score: higher β leads to worse RL training performance | Hurt |
| Include domain-specific features in the latent state representation | VAE-Embedding · CarRacing target domains | score: 616.89; 282.71; 484.57; 223.88; 332.58; 595.42 | Hurt |
| Include domain-specific features in the learned embedding | CURL · CARLA target domains | score: 44.34; 73.63 | Hurt |
Closest papers by meaning
- Cross-Domain Policy Adaptation by Capturing Representation Mismatch (ICML 2024 · similarity 0.89)
- Prompt-based Visual Alignment for Zero-shot Policy Transfer (ICML 2024 · similarity 0.89)
- In-Context Policy Adaptation via Cross-Domain Skill Diffusion (AAAI 2025 · similarity 0.88)
- Foundation Policies with Hilbert Representations (ICML 2024 · similarity 0.88)
- Off-Dynamics Reinforcement Learning via Domain Adaptation and Reward Augmented Imitation (NeurIPS 2024 · similarity 0.87)
- Zero Shot Generalization of Vision-Based RL Without Data Augmentation (ICML 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.