ResearchBenchmarksGeneral Capability Evaluation on Capability BenchmarksGeneral Capability Evaluation on Capability BenchmarksFollow74.32ScoreGCWM38.367247.701157.03566.3689May 10, 2026Evaluation resultsMethodMethodLinksScoreGainWin CountGCWMScale=14BScale=14B2026.0574.324.284GCWMScale=8BScale=8B2026.0572.542.644GCWMScale=4BScale=4B2026.0570.275.716TIESScale=14BScale=14B2026.0570.04——TIESScale=8BScale=8B2026.0569.9——TIESScale=4BScale=4B2026.0564.56——GCWMScale=1.7BScale=1.7B2026.0558.263.245TIESScale=1.7BScale=1.7B2026.0555.02——GCWMScale=0.6BScale=0.6B2026.0539.960.212TIESScale=0.6BScale=0.6B2026.0539.75——