DPRM-DLLM / dmpo /matched_step5000_rows.csv
DarkerBu's picture
Add files using upload-large-folder tool
96f01f5 verified
Raw
History Blame Contribute Delete
1.31 kB
host,task,method,variant,metric,direction,value,ci95_low,ci95_high,n,protocol
DMPO,MATH,Progressive DMPO,confidence,mean_pass_at_k,higher,50.433333,,,500,matched_step5000
DMPO,MATH,DMPO-DPRM,dprm_confidence,mean_pass_at_k,higher,52.533333,,,500,matched_step5000
DMPO,MATH,DMPO-DPRM,dprm_confidence,paired_delta,higher,2.100000,-0.333333,4.533333,500,matched_step5000
DMPO,MATH Hard,Progressive DMPO,confidence,mean_pass_at_k,higher,44.270833,,,128,matched_step5000
DMPO,MATH Hard,DMPO-DPRM,dprm_confidence,mean_pass_at_k,higher,47.916667,,,128,matched_step5000
DMPO,MATH Hard,DMPO-DPRM,dprm_confidence,paired_delta,higher,3.645833,-0.911458,8.333333,128,matched_step5000
DMPO,Countdown,Progressive DMPO,confidence,mean_pass_at_k,higher,53.375651,,,5120,matched_step5000
DMPO,Countdown,DMPO-DPRM,dprm_confidence,mean_pass_at_k,higher,55.045573,,,5120,matched_step5000
DMPO,Countdown,DMPO-DPRM,dprm_confidence,paired_delta,higher,1.669922,0.820312,2.548910,5120,matched_step5000
DMPO,Countdown Hard,Progressive DMPO,confidence,mean_pass_at_k,higher,29.638672,,,1024,matched_step5000
DMPO,Countdown Hard,DMPO-DPRM,dprm_confidence,mean_pass_at_k,higher,33.382161,,,1024,matched_step5000
DMPO,Countdown Hard,DMPO-DPRM,dprm_confidence,paired_delta,higher,3.743490,1.464844,6.022135,1024,matched_step5000