Bi-level Agentic Reinforcement Learning for Web Decision-Making under Multi-Cost and Failure Risk Constraints
American Journal Of Big Data · 25.03.2026 · 7 (1) · 193-201
DOI: 10.71465/ajbd3654 ↗Kurz gefasst
Fachartikel in der Zeitschrift American Journal Of Big Data, online erschienen am 25.03.2026. Verfasst von James Thompson und 2 weiteren Autor:innen. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).
Abstract (Originalsprache)
Web agent decision-making naturally involves nested objectives, where high-level task completion depends on lower-level interaction strategies that incur heterogeneous costs and failure risks. This study formulates agentic web reinforcement learning as a bi-level optimization problem. The upper-level objective optimizes task success and global failure tolerance, while the lower-level problem minimizes cumulative operational costs subject to interaction feasibility. The two levels are coupled through shared policy parameters and budget signals. A differentiable bi-level training scheme is introduced to approximate the lower-level optimal response during policy updates. The proposed model is evaluated on 1,000–1,800 web tasks spanning transactional, informational, and configuration studyflows. Empirical results demonstrate improved cost efficiency and reduced catastrophic failure frequency compared with single-level constrained reinforcement learning, particularly in tasks requiring long sequences of dependent actions.
Zitieren
APA
Thompson, J., Carter, E., & Wright, D. (2026). Bi-level Agentic Reinforcement Learning for Web Decision-Making under Multi-Cost and Failure Risk Constraints. American Journal Of Big Data, 7(1), 193-201. https://doi.org/10.71465/ajbd3654
Vancouver
Thompson J, Carter E, Wright D. Bi-level Agentic Reinforcement Learning for Web Decision-Making under Multi-Cost and Failure Risk Constraints. American Journal Of Big Data. 2026;7(1):193-201. doi:10.71465/ajbd3654
BibTeX
@article{Thompson2026,
author = {Thompson, James and Carter, Emily and Wright, Daniel},
title = {Bi-level Agentic Reinforcement Learning for Web Decision-Making under Multi-Cost and Failure Risk Constraints},
journal = {American Journal Of Big Data},
year = {2026},
volume = {7},
number = {1},
pages = {193--201},
doi = {10.71465/ajbd3654}
}
EUVIMED