When Faster Decisions Do Not Make Better Agents: An Empirical Study of Jev in Agent Workflows

Evaluates Jev as a bounded action selector inside three AndroidWorld agents and two text-based ALFWorld agents, separating decision savings from task success, recovery and resources per delivered outcome.

AI-assisted editorial review against a primary source. This is a catalog annotation, not an independent reproduction or endorsement.

What is this resource?

Resource type

Research papers

Environment

android

Topics

efficiency, action-prediction, evaluation, mobile-navigation, self-reflection

Record publication date

2026-10-09

Evidence retrieved

2026-10-20T15:08:54Z

Resource ID

arxiv:2610.13092

Recorded paper version

arXiv v1

Version updated: 2026-10-09T06:45:23Z.

This identifies the retained arXiv metadata snapshot; newly observed changes require separate review.

Authors

Yuchen Ling, Shengcheng Yu, Zhenyu Chen, Chunrong Fang

Cite this paper

Download BibTeX · Download CSL JSON

Citations preserve the selected source edition, its authors and recorded date precision. Unrecorded venues and identifiers are omitted.

Explore related topics

Scope and limitations

The mobile evaluation uses text UI observations and one fixed instance for each of 116 AndroidWorld task types; three repeats share seed 0. Five agent families use different budgets, with candidate-interface and tool-call granularity changes in some variants. Reported costs use fixed experimental prices, and outcome-selected trace labels are not population failure rates. Faster selection does not imply faster or more reliable task completion. Selected sections were source-read; no independent reproduction.

Primary evidence and verification

Source retrieval is separate from experiment reproduction. Use the original publication or publisher page when citing claims about the resource.

1. arxiv api

   Retrieved 2026-10-20T15:06:37Z. Supports: name, authors, published_at, updated_at, identifiers.arxiv, details.arxiv_snapshot.

   Content fingerprint

   a3d753815e82190ee2de6f322cd816c32f2dd391f862845305fc7e57300303a6
2. editorial primary review

   Retrieved 2026-10-20T15:08:54Z. Supports: name, kind, summary, environments, topics, scope, notes.

   Content fingerprint

   3db7474953cbe5a3052fb22793038f2383bdd0e7c775af06494a5e8501646d18

Suggest a correction

Open a source-backed correction. Include this record ID and the original source. Read our editorial policy and coverage boundaries.