Research map / Reinforcement learning / Visual and representation RL
First-Order Representation Languages for Goal-Conditioned RL
Simon Ståhlberg, Hector Geffner
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 |
|---|---|---|---|
| relabel goals with propositional subsets instead of full states | R-GNN DQN · 15 planning domains | coverage: 51.0% -> 82.4% | Improved |
| relabel goals with lifted subgoals instead of full states | R-GNN DQN · 15 planning domains | coverage: 51.0% -> 78.6% | Improved |
| lift propositional HER subgoals | R-GNN DQN · 15 planning domains | coverage: 82.4% -> 78.6% | Mixed |
| use lifted HER instead of propositional | R-GNN DQN · Maze | coverage: 0 relabeled -> 100% coverage | Improved |
| use lifted HER instead of state HER | R-GNN DQN · Miconic-L | instances solved: 0 -> 100 | Improved |
| use lifted HER instead of propositional HER | R-GNN DQN · Childsnack | instances solved: 100 -> 56 | Hurt |
| use any HER variant | R-GNN DQN · Hiking | instances solved: 0 -> 0 | No clear effect |
| use any HER variant on Delivery | R-GNN DQN · Delivery | instances solved: State 0; lifted 12; propositional 11 | Hurt |
| use learned greedy policy instead of LAMA | R-GNN DQN propositional HER · 15 planning domains | coverage: 80.8% -> 82.4% | Improved |
Closest papers by meaning
- Horizon Generalization in Reinforcement Learning (ICLR 2025 · similarity 0.89)
- Flattening Hierarchies with Policy Bootstrapping (NeurIPS 2025 · similarity 0.88)
- Goal-Conditioned On-Policy Reinforcement Learning (NeurIPS 2024 · similarity 0.88)
- PcLast: Discovering Plannable Continuous Latent States (ICML 2024 · similarity 0.88)
- The Synergy of LLMs & RL Unlocks Offline Learning of Generalizable Language-Conditioned Policies with Low-fidelity Data (ICML 2025 · similarity 0.88)
- Are Expressive Models Truly Necessary for Offline RL? (AAAI 2025 · 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.
