Back to AI News
arxiv

CrossEnv: Synthesizing Verifiable Cross-Environment Tasks for Long-Horizon Agent Training

CrossEnv is a framework for generating executable, verifiable training tasks that span multiple environments, using an entity-tool graph to identify connections between services and align task-specific data. It also introduces CrossEnv-Curriculum, which moves from single-environment tool use to coordination across more environments and business domains before adding user interaction. Across four external benchmarks, combining cross-environment training data with this curriculum achieved competitive performance with fewer training examples, with analyses indicating transfer to new environment combinations, less similar external tasks, and longer workflows.

AI Brief

It offers a concrete method for training long-horizon agents on verifiable cross-service tasks, addressing a gap left by single-environment synthetic task generation.

arXiv:2611.22777v1 Announce Type: new Abstract: Executable synthetic environments provide scalable training tasks and verifiable feedback for long-horizon agents. Existing environment synthesis methods primarily generate tasks within individual environments. However, these tasks do not adequately capture real-world interactions across services, where actions in one service depend on information obtained from another. Synthesizing such tasks is challenging because data generated across environments must remain consistent to support executable and verifiable interactions. To address this challenge, we introduce CrossEnv, a framework that uses an entity-tool graph to identify connections between environments and aligns task-specific data to synthesize executable, verifiable cross-environment tasks. We further propose CrossEnv-Curriculum, which progresses from single-environment tool use to coordination across more environments and business domains, before introducing user interaction. Across four external benchmarks, combining cross-environment training data with this curriculum achieves competitive performance with fewer training examples. Further analyses support the benefits of cross-environment data and curriculum training, and provide evidence of transfer to new environment combinations, external tasks less similar to the training data, and longer workflows.
Be the first to rate this article.

Comments & Questions (0)

Captcha
Please be respectful — let's keep the conversation friendly.

No comments yet

Be the first to comment!