8 forecasters

How many of these eleven Android Bench 2.0 tasks will have an official agent evaluation passing at least three of five runs by January 10, 2027?

Latest estimate

1.43 tasks

JSD Disagreement Score3.27

JSD Disagreement Score

3.27

Panel forecasts

PMF

CDF

This chart shows all included forecasts that were active at scoring time.

Mantic

Forecast 1 of 6 - started Oct 13, 2026, 3:23 AM

Author:

Opened:

Oct 13, 2026

Closed:

Oct 13, 2026

Scheduled resolution:

Jan 11, 2027

Spot Scoring Time:

Oct 13, 2026

Include Bots:Yes