Research map / Reinforcement learning / Visual and representation RL
Value Function Spaces: Skill-Centric State Abstractions for Long-Horizon Reasoning
Dhruv Shah, Peng Xu, Yao Lu, Ted Xiao, Alexander T Toshev, Sergey Levine, brian ichter
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 VFS instead of the strongest listed baseline | model-free HRL · MultiRoom-10 | success rate: 0.43 -> 0.77 | Improved |
| Use VFS instead of the strongest listed baseline | model-free HRL · KeyCorridor-7 | success rate: 0.54 -> 0.68 | Improved |
| Use VFS instead of CURL | model-free HRL · KeyCorridor-3 | success rate: 0.83 -> 0.82 | No clear effect |
| Use VFS instead of a learned image representation | model-based manipulation · O10 | task success rate: 2/20 -> 80% | Improved |
| Combine learned representations with VFS by concatenating their embeddings (ablation) | model-free HRL · MultiRoom-4 (SPL) | SPL: 0.84 -> 0.90 | Improved |
| Combine learned representations with VFS by concatenating their embeddings (ablation) | model-free HRL · KeyCorridor-7 (SPL) | SPL: 0.59 -> 0.65 | Improved |
| Increase CURL embedding size from 1x to 5x (ablation) | CURL · MultiRoom-10 | success rate: 0.43 -> 0.48 | Improved |
| Increase CURL embedding size from 1x to 5x (ablation) | CURL · KeyCorridor-7 | success rate: 0.54 -> 0.56 | Improved |
| Increase VAE embedding size from 1x to 5x (ablation) | VAE · MultiRoom-10 | success rate: 0.58 -> 0.62 | Improved |
| Increase VAE embedding size from 1x to 5x (ablation) | VAE · KeyCorridor-7 | success rate: 0.50 -> 0.53 | Improved |
| Increase AE embedding size from 1x to 2x (ablation) | AE · MultiRoom-10 | success rate: 0.34 -> 0.33 | Hurt |
| Increase AE embedding size from 1x to 5x (ablation) | AE · KeyCorridor-7 | success rate: 0.33 -> 0.40 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- CURL — CURL: Contrastive Unsupervised Representations for Reinforcement Learning (ICML 2020) · better on 5 of 6 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- When Do Skills Help Reinforcement Learning? A Theoretical Analysis of Temporal Abstractions (ICML 2024 · similarity 0.88)
- Option-aware Temporally Abstracted Value for Offline Goal-Conditioned Reinforcement Learning (NeurIPS 2025 · similarity 0.88)
- Reconciling Spatial and Temporal Abstractions for Goal Representation (ICLR 2024 · similarity 0.87)
- Goal-Oriented Skill Abstraction for Offline Multi-Task Reinforcement Learning (ICML 2025 · similarity 0.87)
- Efficient Language-instructed Skill Acquisition via Reward-Policy Co-Evolution (AAAI 2025 · similarity 0.87)
- Your Language Model is Its Own Critic: Reinforcement Learning with Value Estimation from Actor’s Internal States (NeurIPS 2026 · 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.
