Plan-and-Act: Improving Planning of Agents for Long-Horizon Tasks
Fuente:
arXiv
Salvato in:
| Autori principali: | , , , , , , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2025
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866911087712010240 |
|---|---|
| author | Erdogan, Lutfi Eren Lee, Nicholas Kim, Sehoon Moon, Suhong Furuta, Hiroki Anumanchipalli, Gopala Keutzer, Kurt Gholami, Amir |
| author_facet | Erdogan, Lutfi Eren Lee, Nicholas Kim, Sehoon Moon, Suhong Furuta, Hiroki Anumanchipalli, Gopala Keutzer, Kurt Gholami, Amir |
| contents | Large language models (LLMs) have shown remarkable advancements in enabling language agents to tackle simple tasks. However, applying them for complex, multi-step, long-horizon tasks remains a challenge. Recent work have found success by separating high-level planning from low-level execution, which enables the model to effectively balance high-level planning objectives and low-level execution details. However, generating accurate plans remains difficult since LLMs are not inherently trained for this task. To address this, we propose Plan-and-Act, a novel framework that incorporates explicit planning into LLM-based agents and introduces a scalable method to enhance plan generation through a novel synthetic data generation method. Plan-and-Act consists of a Planner model which generates structured, high-level plans to achieve user goals, and an Executor model that translates these plans into environment-specific actions. To train the Planner effectively, we introduce a synthetic data generation method that annotates ground-truth trajectories with feasible plans, augmented with diverse and extensive examples to enhance generalization. We evaluate Plan-and-Act using web navigation as a representative long-horizon planning environment, demonstrating a state-of-the-art 57.58% success rate on the WebArena-Lite benchmark as well as a text-only state-of-the-art 81.36% success rate on WebVoyager. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2503_09572 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Plan-and-Act: Improving Planning of Agents for Long-Horizon Tasks Erdogan, Lutfi Eren Lee, Nicholas Kim, Sehoon Moon, Suhong Furuta, Hiroki Anumanchipalli, Gopala Keutzer, Kurt Gholami, Amir Computation and Language Large language models (LLMs) have shown remarkable advancements in enabling language agents to tackle simple tasks. However, applying them for complex, multi-step, long-horizon tasks remains a challenge. Recent work have found success by separating high-level planning from low-level execution, which enables the model to effectively balance high-level planning objectives and low-level execution details. However, generating accurate plans remains difficult since LLMs are not inherently trained for this task. To address this, we propose Plan-and-Act, a novel framework that incorporates explicit planning into LLM-based agents and introduces a scalable method to enhance plan generation through a novel synthetic data generation method. Plan-and-Act consists of a Planner model which generates structured, high-level plans to achieve user goals, and an Executor model that translates these plans into environment-specific actions. To train the Planner effectively, we introduce a synthetic data generation method that annotates ground-truth trajectories with feasible plans, augmented with diverse and extensive examples to enhance generalization. We evaluate Plan-and-Act using web navigation as a representative long-horizon planning environment, demonstrating a state-of-the-art 57.58% success rate on the WebArena-Lite benchmark as well as a text-only state-of-the-art 81.36% success rate on WebVoyager. |
| title | Plan-and-Act: Improving Planning of Agents for Long-Horizon Tasks |
| topic | Computation and Language |
| url | https://arxiv.org/abs/2503.09572 |