Research map / Probabilistic and scientific modeling / Molecules and proteins
Regression Planning Networks
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 |
|---|---|---|---|
| add backward regression planning (ablation) | RPN · RoomGoal k-d-g | success rate: 3.2 -> 100.0 | Helped |
| add subgoal serialization (ablation) | RPN · DoorKey D=6 | success rate: 0.0 -> 64.3 | Helped |
| combine regression planning with subgoal serialization (ablation) | Kitchen 3D · I=6, D=3 | task success rate: RP-only 0.0; SS-only 32.7; RPN 97.2 | Helped |
| remove subgoal serialization (ablation) | RPN · Kitchen 3D I=6, D=1 | task success rate: 95.3 -> \textless{}0.1 | Hurt |
| remove regression planning (ablation) | RPN · Kitchen 3D I=6, D=1 | task success rate: 95.3 -> 42.8 | Hurt |
| use regression planning without subgoal serialization (ablation) | RP-only · Kitchen 3D | task success rate: 70.3 -> 67.1; 47.0; 27.9; \textless{}0.1; 0.0 | Mixed |
| use subgoal serialization without regression planning (ablation) | SS-only · Kitchen 3D | task success rate: 49.1; 59.3; 56.6; 43.4; 42.8; 32.7 | Mixed |
| use RPN for unseen Kitchen 3D task sizes | RPN · Kitchen 3D evaluation tasks | task success rate: 98.6; 98.2; 98.4; 95.3; 97.2 | Helped |
| use regression planning (ablation) | RPN · DoorKey D=4 and D=6 | success rate: 18.2 -> 91.9; 0.0 -> 64.3 | Helped |
| use regression planning to compose learned plans (ablation) | RP-only · RoomGoal k-d-g | success rate: 3.2 -> 100.0 | Helped |
Closest papers by meaning
- Efficient Robotic Policy Learning via Latent Space Backward Planning (ICML 2025 · similarity 0.87)
- System 1.x: Learning to Balance Fast and Slow Planning with Language Models (ICLR 2025 · similarity 0.87)
- Scaling Value Iteration Networks to 5000 Layers for Extreme Long-Term Planning (ICML 2025 · similarity 0.86)
- Learning Neuro-Symbolic Abstractions for Robot Planning and Learning (AAAI 2024 · similarity 0.86)
- Sparse Planning in Visual World Models via Cost Gradients (NeurIPS 2026 · similarity 0.86)
- Learning Domain-Independent Heuristics for Grounded and Lifted Planning (AAAI 2024 · similarity 0.85)
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.