ADeptS-Bench

What is ADeptS-Bench?

Computer Use Agents (CUAs) are increasingly deployed to navigate mobile and desktop applications on behalf of users, yet no benchmark comprehensively evaluates…

Released

2026-08-25

Evaluates

Agents & Tool Use, Manufacturing & Process Control, cs.CR

Openness

unknown

Importer

Claire Radar

Review status

rule-auto-admitted

Reported scores

0

Source provenance

• Original evidence https://arxiv.org/abs/2608.26204

ADeptS-Bench paper

• Paper https://arxiv.org/abs/2608.26204

Which sources cite ADeptS-Bench?

• ADeptS-Bench

Related Agents & Tool Use, Manufacturing & Process Control, cs.CR benchmarks

• $\tau^\tau$-Bench: An Environment for End-To-End, Realistic Agent Construction
• A3S-Bench
• AdaPlanBench
• Adaptive Adversaries
• AeroCopilotBench
• AgentHijack
• Agentic Evaluation in 2D Mazes
• AgentLens