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

Instruction Alignment Evaluation on EBench 18K

92.09SRCC

ReasonEdit

1.214824.807448.471.9926May 8, 2026
Updated 23d ago

Evaluation Results

MethodLinks
2026.05
92.0975.5292.91
2026.05
88.370.888.98
2026.05
82.868.9383.9
2026.05
82.0768.0983.87
2026.05
52.4936.7954.52
2026.05
49.6136.8449.94
2026.05
46.6131.748.06
2026.05
36.4125.3638.48
2026.05
36.2724.8232.97
2026.05
34.3226.6137.65
2026.05
28.7519.7831.98
2026.05
28.3918.9826.95
2026.05
22.2214.8929.92
2026.05
22.0814.8526.93
2026.05
21.6614.5226.32
2026.05
21.5214.4922.09
2026.05
20.313.526.7
2026.05
19.3714.2717.2
2026.05
17.0511.3223.01
2026.05
13.669.2315.62
2026.05
11.217.6310.8
2026.05
4.713.0511.96