Metraipolis › archive › dc_7da0408c0850
Do LLM Agents Execute the Plans They Declare? From Planning-Mode Declaration to Pattern-Specific Execution — arXiv:2609.38108
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2609.38108
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-07T16:24:29.754Z
- SHA-256 of the archived text
461f7f8bfc2247a6f6ea666167925d5982ebb26e53982382174b08ff8230e4ba- Stable URL of this page
https://metraipolis.com/documento/dc_7da0408c0850
How to cite
Do LLM Agents Execute the Plans They Declare? From Planning-Mode Declaration to Pattern-Specific Execution — arXiv:2609.38108 — arXiv (en), https://arxiv.org/abs/2609.38108 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_7da0408c0850 (Metraipolis archive dc_7da0408c0850, sha256 461f7f8bfc2247a6f6ea666167925d5982ebb26e53982382174b08ff8230e4ba)
BibTeX
@misc{dc7da0408c0850,
title = {Do LLM Agents Execute the Plans They Declare? From Planning-Mode Declaration to Pattern-Specific Execution — arXiv:2609.38108},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2609.38108},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_7da0408c0850; sha256 461f7f8bfc2247a6f6ea666167925d5982ebb26e53982382174b08ff8230e4ba},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.38108. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Large language models (LLMs) enable agents to solve long-horizon tasks by generating a plan and then executing it in an environment. However, successful planning requires two distinct capabilities: selecting an appropriate plan for the task and executing it faithfully. Existing planner--executor systems can fail at either stage, while final task success alone cannot distinguish selection from execution failures. We therefore study the Plan Declaration--Execution Gap and introduce Planning-as-Routing, where an LLM declares one of four planning modes: Predefined, Sequential, Hierarchical, or Search, and a deterministic router dispatches the task to the corresponding pattern-specific executor. Across four benchmarks and three LLMs, we find three consistent patterns. First, generic Plan+ReAct often fails to preserve declared planning structure, especially for longer plans: across three benchmarks, only (22)--(45%) of trajectories preserve it, whereas pattern-specific executors enforce the intended structure. Second, planning-mode effectiveness varies across environments and models: Search performs best on ALFWorld, Hierarchical on SWE-bench, and the strongest pattern can vary across models within the same benchmark. Third, the largest gains come from execution: pattern-specific executors improve task success from (0.48) to (0.92) on ALFWorld and from (0.36) to (0.44) on SWE-bench Verified over Plan+ReAct. Current LLMs, however, do not reliably select the strongest mode for each task, although few-shot examples improve selection in some benchmark--model combinations. Overall, reliable agent planning requires both effective mode selection and faithful execution: routing substantially closes the execution gap, while task-specific mode selection remains open. Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-30 in cs.AI). Categorie arXiv: Artificial Intelligence (cs.AI) ; Machine Learning (cs.LG). Pagina del lavoro: https://arxiv.org/abs/2609.38108 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito). Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina. Letto il 2026-10-07T15:54:53.690Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-09-30?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).