arXiv：cs.CL· Jiayi Zhang, Jianing Yin, Ben Zhou, Li Zhang·· 16 小时前
通过自动形式化实现鲁棒异步规划：LLM 异步规划基准与形式化方法对比研究
Robust Asynchronous Planning via Auto-Formalization
AI 导读
研究将异步规划统一为一种形式化表述，并发布覆盖任务图规模、状态资源约束与执行时事件的三项基准。在 AsyncPlan-XXL 上，PDDL2.1 管线平均规划准确率仅 30.8%，Planner 为 81.7%，CP-SAT Formalizer 达 94.0%；动作数增至 100 时三者分别为 9.0%、44.5%、82.5%。
正文
Abstract:LLMs can plan by generating schedules directly as a Planner, or by translating a task into a formal specification for a solver as a Formalizer. Most real-world tasks are asynchronous, with non-uniform durations, concurrency, and execution-time changes, yet existing benchmarks hardly cover them. We unify asynchronous planning under one formulation and release three benchmarks spanning task-graph scale, grounded state and resource constraints, and execution-time events. Surprisingly, PDDL2.1, the formalism built for temporal planning, is the weakest pipeline. On AsyncPlan-XXL it averages 30.8% plan accuracy, against 81.7% for the Planner and 94.0% for a CP-SAT Formalizer. The gap widens with scale, and at 100 actions the three pipelines reach 9.0%, 44.5%, and 82.5%, respectively. The formalism is not to blame. Compiled from the same model-extracted intermediate representation, PDDL2.1 and CP-SAT reach nearly identical accuracy (81.6% vs. 81.5%), and gold task graphs reach 100% for both. What remains is a bottleneck shared by every pipeline: recovering the task graph from language. Execution-time events further degrade every pipeline, yet updating only event-induced constraints reaches 81.6% versus 41.6% for full regeneration. Robust asynchronous planning thus hinges not on the expressiveness of the target formalism but on how reliably a model can write and edit it, and a well-chosen interface can let a weaker model match a stronger one.
| Comments: | 31 pages, 20 figures, 15 tables |
| Subjects: | Computation and Language (cs.CL) |
| Cite as: | arXiv:2606.00981 [cs.CL] |
| (or arXiv:2606.00981v2 [cs.CL] for this version) | |
| https://doi.org/10.48550/arXiv.2606.00981 arXiv-issued DOI via DataCite |
Submission history
From: Jiayi Zhang [view email]
[v1]
Sun, 31 May 2026 03:28:42 UTC (3,663 KB)
[v2]
Wed, 14 Oct 2026 07:03:15 UTC (2,597 KB)
来源：arXiv：cs.CL · arxiv.org