List your company

Multitask Language Understanding on MMLU (Accuracy and Token Usage)

65.1AccuracyLaser
37.74844.84951.9559.051May 8, 2026

Evaluation results

MethodLinks
LaserBackbone=DeepSeek-R1-Distill-Qwen-7B2026.0565.1781.1
ExpThinkBackbone=DeepSeek-R1-Distill-Qwen-7B2026.0565.1448.5
JETBackbone=DeepSeek-R1-Distill-Qwen-7B2026.0564.61,036
VanillaBackbone=DeepSeek-R1-Distill-Qwen-7B2026.0563.91,170.8
LC-R1Backbone=DeepSeek-R1-Distill-Qwen-7B2026.0563.4649.6
AdaptThinkBackbone=DeepSeek-R1-Distill-Qwen-7B2026.0563.2902.6
AutoThinkBackbone=DeepSeek-R1-Distill-Qwen-7B2026.0557.4861
JETBackbone=DeepSeek-R1-Distill-Qwen-1.5B2026.0546.71,027.1
AdaptThinkBackbone=DeepSeek-R1-Distill-Qwen-1.5B2026.0544.7995.2
ExpThinkBackbone=DeepSeek-R1-Distill-Qwen-1.5B2026.0544.2622.1
LaserBackbone=DeepSeek-R1-Distill-Qwen-1.5B2026.0542.5943.5
LC-R1Backbone=DeepSeek-R1-Distill-Qwen-1.5B2026.0541927
AutoThinkBackbone=DeepSeek-R1-Distill-Qwen-1.5B2026.0540.51,068.1
VanillaBackbone=DeepSeek-R1-Distill-Qwen-1.5B2026.0538.81,680.3