Long-horizon tasks
AssessTechniques
Evaluation tasks that require completing multiple dependent steps over a longer workflow.
Why it's here
Placed in Assess: 1 article(s) of evidence from 1 source(s), led by framework updates, with 1 in the last 30 days. Confidence 24%. Low accumulated evidence, so it defaults conservatively pending more signal.
Evidence (1)
- 6InfoQ·10/9/2026framework_updateAndroid Bench 2 Adds Long-Horizon Tasks and Agentic Evaluation
Google has released Android Bench 2.0, an updated benchmark framework for evaluating AI models and agents on Android development tasks. The new version adds long-horizon tasks, agent-based evaluation, and continuous scoring to better measure performance on complex multi-step workflows.