From Off-Policy to On-Policy: Enhancing GUI Agents via Bi-level Expert-to-Policy Assimilation
Year: 2026
Type: conference-paper
Abstract: Vision-language models are increasingly deployed as computer-use agents (CUAs) that operate desktops and browsers.Top-performing CUAs are framework-based systems that decompose planning and execution, while endto-end screenshot-to-action policies are easier to deploy but lag behind on benchmarks such as OSWorld-Verified.GUI datasets like OSWorld pose two bottlenecks: they expose only a few hundred interactive, verifiable tasks and environments, and expert trajectories must be gathered by interacting with these environments, making such data hard to scale.We therefore ask how reinforcement learning from verifiable rewards (RLVR) can best exploit a small pool of exist expert trajectories to train end-to-end policies.Naïvely mixing these offpolicy traces into on-policy RLVR is brittle: even after format conversion, expert trajectories exhibit structural mismatch and distribution shift from the learner.We propose BEPA (Bi-Level Expert-to-Policy Assimilation), which turns static expert traces into policy-aligned guidance via self-rolled reachable trajectories under the base policy (LEVEL-1) and a pertask, dynamically updated cache used in RLVR (LEVEL-2).On OSWorld-Verified, BEPA improves UITARS1.5-7Bsuccess from 22.87% to 32.13% and raises a held-out split from 5.74% to 10.30%, with consistent gains on MMBench-GUI and Online-Mind2Web.Our code and data are available at
Language: English
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: hybrid
OpenAlex ID: W7166854919