List your company

Logical Reasoning on BBH-7

100Boolean Expressions AccuracySPEAR
65.05674.12883.292.272May 25, 2026

Evaluation results

MethodLinks
SPEARBootstrap resampling=N=10002026.05100——76.698.498.495.21008893.897.391.2
TextGradBootstrap resampling=N=10002026.0599.2——72.30.85.619.299.242.448.48.578.3
GEPABootstrap resampling=N=10002026.0597.6——71.312.82.472.898.48462.829.387.9
SeedBootstrap resampling=N=10002026.0566.4——63.802.4099.2033.10.857.4