List your company

General Multitask Language Understanding on MMLU-Pro (Pass@1, # Tokens, Saved %)

989.8Avg TokensTRACES
929.3161,337.5831,745.852,154.117Apr 22, 2026

Evaluation results

MethodLinks
TRACESthreshold=0.9, Backbone=DS-Qwen14B, Seed=422026.04989.863.460.44
TRACESthreshold=0.8, Backbone=DS-Qwen14B, Seed=422026.041,286.96548.56
TRACESthreshold=0.7, Backbone=DS-Qwen14B, Seed=422026.041,614.765.235.46
Basel. P^{(1)}_{system}Backbone=DS-Qwen14B, Seed=422026.041,705.663.731.83
TRACESthreshold=0.6, Backbone=DS-Qwen14B, Seed=422026.041,766.366.529.4
TRACESthreshold=0.5, Backbone=DS-Qwen14B, Seed=422026.041,880.666.424.83
TRACESthreshold=0.4, Backbone=DS-Qwen14B, Seed=422026.041,923.166.923.13
Basel. P^{(3)}_{system}Backbone=DS-Qwen14B, Seed=422026.041,935.261.322.65
Basel. P^{(0)}_{user}Backbone=DS-Qwen14B, Seed=422026.042,194.166.512.3
Basel. P^{(0)}_{system}Backbone=DS-Qwen14B, Seed=422026.042,292.967.58.35
StandardBackbone=DS-Qwen14B, Seed=422026.042,501.965.2—