Research map / Vision-language and multimodal / Robot action and navigation
SCOPE: Evolving Symbolic World for Planning in Open-Ended Environments
Yundaichuan Zhan, Minghe Gao, Zhongqi Yue, Wendong Bu, Wenqiao Zhang, Guoming Wang, Jisheng Dang, Juncheng Li, Siliang Tang, Yueting Zhuang
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 SESim (symbolic validation plus real execution feedback, evolving the world) (ablation) | SCOPE · ALFRED basic, open-ended | ClassicalSR (%): 72.3 -> 94.8 | Improved |
| add SASMem (predicate-level symbolic memory) (ablation) | SCOPE · ALFRED basic, open-ended | ClassicalSR (%): 80.2 -> 94.8 | Improved |
| add both SESim and SASMem (ablation) | SCOPE · ALFRED complex, open-ended | ClassicalSR (%): 45.2 -> 89.0 | Improved |
| add SESim (ablation) | SCOPE · VirtualHome | SymHalluc (%): 9.88 -> 3.68 | Improved |
| add SASMem (ablation) | SCOPE · VirtualHome | SymHalluc (%): 5.72 -> 3.68 | Improved |
| use both symbolic validation and real execution feedback instead of real execution only (ablation) | SESim · VirtualHome | ClassicalSR (%): 80.2 -> 92.6 | Improved |
| use both feedback sources instead of symbolic validation only (ablation) | SESim · VirtualHome | ClassicalSR (%): 86.5 -> 92.6 | Improved |
| use both feedback sources instead of real execution only (ablation) | SESim · VirtualHome | GC (%): 81.6 -> 82.1 | No clear effect |
| use SASMem instead of raw-trace memory (ablation) | SCOPE · VirtualHome | #PlanRefine: 6.51 -> 2.18 | Improved |
| use SASMem instead of VLM-summary memory (ablation) | SCOPE · VirtualHome | #EnvEvolve: 2.02 -> 0.86 | Improved |
| use VLM planner with evolving symbolic world instead of classical planner (NESYC) | SCOPE vs NESYC · VirtualHome basic, static | SR (%): 90.1 -> 88.2 | Hurt |
| use SCOPE instead of NESYC in open-ended setting | SCOPE vs NESYC · VirtualHome basic, open-ended | SR (%): 56.3 -> 87.6 | Improved |
Closest papers by meaning
- World-aware Planning Narratives Enhance Large Vision-Language Model Planner (NeurIPS 2025 · similarity 0.88)
- Towards Reliable Code-as-Policies: A Neuro-Symbolic Framework for Embodied Task Planning (NeurIPS 2025 · similarity 0.87)
- SPIRAL: Symbolic LLM Planning via Grounded and Reflective Search (AAAI 2026 · similarity 0.86)
- Visual Planning: Let's Think Only with Images (ICLR 2026 · similarity 0.86)
- Hierarchical Planning for Complex Tasks with Knowledge Graph-RAG and Symbolic Verification (ICML 2025 · similarity 0.86)
- Planning in the Dark: LLM-Symbolic Planning Pipeline Without Experts (AAAI 2025 · similarity 0.86)
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.
