List your company

Instruct-tuning performance on Leaderboard

0.4159BBH ScoreTaskPGM (PMI)
0.342580.3616150.380650.399685Jul 16, 2025

Evaluation results

MethodLinks
TaskPGM (PMI)Backbone=Llama2-7B, Sample Size=50K2025.070.41590.27940.35770.01890.19670.4511
LESSBackbone=Llama2-7B, Sample Size=50K2025.070.41330.24750.34070.01360.19060.4226
TaskPGM (PMI)Backbone=Llama2-7B, Sample Size=25K2025.070.40950.27180.35610.01590.19240.4298
LESSBackbone=Llama2-7B, Sample Size=25K2025.070.40590.24120.34170.00010.18160.4157
TaskPGM (JSD)Backbone=Llama2-7B, Sample Size=50K2025.070.36240.28020.39250.00980.19270.4246
EPMBackbone=Llama2-7B, Sample Size=25K2025.070.35930.26010.34050.01510.18360.4286
RandomBackbone=Llama2-7B, Sample Size=50K2025.070.35650.26680.35810.01340.18110.377
EPMBackbone=Llama2-7B, Sample Size=50K2025.070.35320.26340.35070.01280.17990.4206
UniformBackbone=Llama2-7B, Sample Size=25K2025.070.35010.27010.35010.01510.17680.4027
RandomBackbone=Llama2-7B, Sample Size=25K2025.070.34820.26260.34650.00980.18770.3677
UniformBackbone=Llama2-7B, Sample Size=50K2025.070.3480.27850.39010.01610.18950.4057
TaskPGM (JSD)Backbone=Llama2-7B, Sample Size=25K2025.070.34540.27850.37290.01510.1790.4021