# DualWorldBench: Can Agents Plan Deliveries across Symbolic and Grounded Worlds?

DevFeed: [DualWorldBench: Can Agents Plan Deliveries across Symbolic and Grounded Worlds?](<https://devfeed.tech/articles/dualworldbench-can-agents-plan-deliveries-across-symbolic-and-grounded-worlds-117445.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2611.04228>)

Author: Yang Chen, Bin Wen, Danyang Peng, Hong-Jie You, Zi-Qiao Shang, Lan-Zhe Guo

Published: 2026-11-04T05:00:00Z

Content type: article

Language: en

Sources: [arXiv: Artificial Intelligence](<https://devfeed.tech/sources/arxiv-artificial-intelligence.md>)

Topics: [AI Agent](<https://devfeed.tech/topics/ai-agent.md>), [Combinatorial optimization](<https://devfeed.tech/topics/combinatorial-optimization.md>), [Machine Perception](<https://devfeed.tech/topics/machine-perception.md>), [Robot Navigation](<https://devfeed.tech/topics/robot-navigation.md>)

Tags: [agents](<https://devfeed.tech/tags/agents.md>), [benchmark](<https://devfeed.tech/tags/benchmark.md>), [components](<https://devfeed.tech/tags/components.md>), [cs-ai](<https://devfeed.tech/tags/cs-ai.md>), [datasets](<https://devfeed.tech/tags/datasets.md>), [gap](<https://devfeed.tech/tags/gap.md>), [multimodal](<https://devfeed.tech/tags/multimodal.md>), [path-planning](<https://devfeed.tech/tags/path-planning.md>), [planning](<https://devfeed.tech/tags/planning.md>), [temporal](<https://devfeed.tech/tags/temporal.md>), [testbed](<https://devfeed.tech/tags/testbed.md>), [topology](<https://devfeed.tech/tags/topology.md>)

## AI overview

DualWorldBench tests whether agents can turn symbolic or grounded evidence into globally feasible multi-goal delivery plans. Its shared pickup-delivery interface spans TextWorld, which varies task composition and release dynamics; SimWorld, which tests path planning across topologies using text, image, or multimodal inputs; and DualWorld-VQA, which diagnoses perception, scene understanding, and planning. The reported experiments find that agents struggle to sequence tasks optimally under interacting temporal constraints and do not reliably convert grounded observations into feasible actions. The diagnostic results likewise show that perception and topology understanding do not ensure constrained planning. The abstract identifies integration of evidence into globally feasible and efficient plans as the central challenge.

## Source excerpt

arXiv:2611.04228v1 Announce Type: new Abstract: Recent multimodal agents have shown promising progress in embodied planning and web interaction. However, existing evaluations often rely on simplified tasks with limited combinatorial structure, leaving it unclear whether current agents can robustly integrate perception and reasoning for multi-goal planning under temporal, physical, and multimodal constraints. We introduce DualWorldBench, a unified benchmark for evaluating agent planning under controlled symbolic--temporal and grounded--topological complexity. Built around a shared pickup--delivery interface, DualWorldBench consists of three complementary components: TextWorld, which isolates symbolic planning under varying task compositions and release dynamics; SimWorld, which evaluates grounded path planning under controlled topological variation with text, image, or multimodal inputs; and DualWorld-VQA, a diagnostic suite spanning perception, scene understanding, and planning. Experiments reveal a common gap between local understanding and globally consistent planning. In TextWorld, agents struggle to derive optimal task sequences under interacting temporal constraints; in SimWorld, grounded observations are not reliably converted into feasible action sequences. DualWorld-VQA further confirms that perception and topology understanding do not necessarily translate into constrained planning. Together, these results identify the central challenge as composing symbolic or grounded evidence into globally feasible and efficient multi-goal plans. These findings establish DualWorldBench as a challenging testbed for agents with tighter perception--reasoning--planning integration. All datasets and code are publicly available at our project website https://dualworld.netlify.app/.