Year: 2026
Type: conference-paper
Abstract: Large-language-model (LLM) agents are becoming competent at straightforward web tasks, such as opening an item page or submitting a form, but still struggle with objectives that require long-horizon navigation, large-scale information extraction, and reasoning under constraints.We present WEBDART, a general framework that enables a single LLM to handle such complex chores.WEBDART (i) dynamically decomposes each objective into three focused subtasks-navigation, information extraction, and execution-so the model concentrates on one skill at a time, and (ii) continuously re-plans the decomposition as new webpages are revealed, taking advantage of newly discovered filters or shortcuts and avoiding redundant exploration.Evaluated on WebChore-Arena, WEBDART lifts end-to-end success rates by up to 13.7 percentage points over previous state-of-the-art agents, while matching their performance on the easier WebArena suite and completing tasks with up to 14.7 fewer navigation steps.
Authors: Jingbo Yang, Bairu Hou, Wei Wei, Shiyu Chang, Yujia Bao
Institutions: Accenture (Switzerland), Accenture (United States)
Language: English
Topic: Web Data Mining and Analysis
Subfield: Information Systems
Field: Computer Science
Domain: Physical Sciences
Open Access status: hybrid
Funder: National Science Foundation
Awards: CAREER: Understanding, Correcting, and Adapting Large Language Models from a Knowledge-Oriented Perspective in NLP Applications and Beyond, Building A Teacher-AI Collaborative System for Personalized Instruction and Assessment of Comprehension Skills, Collaborative Research: RI: Small: Advancing Theory and Practice of Trustworthy Machine Learning via Bi-Level Optimization
OpenAlex ID: W4415315941