Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

Instruction Following on Alpaca (Speedup and Drafting Cost Analysis)

6.27Average Accepted Length

EAGLE-3

1.6634242.8593624.05535.251238May 8, 2026May 9, 2026May 10, 2026May 11, 2026
Updated 21d ago

Evaluation Results

MethodLinks
2026.05
6.273.3531
2026.05
5.172.3630
2026.05
4.743.4318
2026.05
4.723.4715
2026.05
4.663.416
2026.05
4.292.3529
2026.05
4.27454.29-
2026.05
4.252.6617
2026.05
4.232.8626
2026.05
4.222.7215
2026.05
4.141.9429
2026.05
4.092.2824
2026.05
4.022.5714
2026.05
3.941.8425
2026.05
3.82.6219
2026.05
3.782.6417
2026.05
3.732.5814
2026.05
3.722.1419
2026.05
3.692.1918
2026.05
3.67423.82-
2026.05
3.642.1127
2026.05
3.622.515
2026.05
3.612.5113
2026.05
3.592.0614
2026.05
3.57--
2026.05
3.55214
2026.05
3.542.4811
2026.05
3.53342.68-
2026.05
3.532.0613
2026.05
3.49--
2026.05
3.43393.45-
2026.05
3.421.9411
2026.05
3.282.398
2026.05
2.93062.21-
2026.05
2.86442.17-
2026.05
2.851.789
2026.05
2.83482.18-
2026.05
2.8282.94-
2026.05
2.79--
2026.05
2.73--
2026.05
2.71.4838
2026.05
2.61792.63-
2026.05
2.442.016
2026.05
2.39172.41-
2026.05
2.331.0640
2026.05
2.31311.79-
2026.05
2.31041.75-
2026.05
2.071.57
2026.05
2.06232.08-
2026.05
1.85941.9-
2026.05
1.84061.88-