AI Paper Atlas

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

ICML 2026 · Vision-language and multimodal › Robot action and navigation · approach: Embodied multimodal LLM agents · 0 citations (OpenAlex)

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 changedWhereMeasuredEffect
add SESim (symbolic validation plus real execution feedback, evolving the world) (ablation)SCOPE · ALFRED basic, open-endedClassicalSR (%): 72.3 -> 94.8Improved
add SASMem (predicate-level symbolic memory) (ablation)SCOPE · ALFRED basic, open-endedClassicalSR (%): 80.2 -> 94.8Improved
add both SESim and SASMem (ablation)SCOPE · ALFRED complex, open-endedClassicalSR (%): 45.2 -> 89.0Improved
add SESim (ablation)SCOPE · VirtualHomeSymHalluc (%): 9.88 -> 3.68Improved
add SASMem (ablation)SCOPE · VirtualHomeSymHalluc (%): 5.72 -> 3.68Improved
use both symbolic validation and real execution feedback instead of real execution only (ablation)SESim · VirtualHomeClassicalSR (%): 80.2 -> 92.6Improved
use both feedback sources instead of symbolic validation only (ablation)SESim · VirtualHomeClassicalSR (%): 86.5 -> 92.6Improved
use both feedback sources instead of real execution only (ablation)SESim · VirtualHomeGC (%): 81.6 -> 82.1No clear effect
use SASMem instead of raw-trace memory (ablation)SCOPE · VirtualHome#PlanRefine: 6.51 -> 2.18Improved
use SASMem instead of VLM-summary memory (ablation)SCOPE · VirtualHome#EnvEvolve: 2.02 -> 0.86Improved
use VLM planner with evolving symbolic world instead of classical planner (NESYC)SCOPE vs NESYC · VirtualHome basic, staticSR (%): 90.1 -> 88.2Hurt
use SCOPE instead of NESYC in open-ended settingSCOPE vs NESYC · VirtualHome basic, open-endedSR (%): 56.3 -> 87.6Improved

Closest papers by meaning

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.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy