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

Instruction Following on AlpacaEval (test)

3,213Helpfulness Score

Fine-tuned

-122.072743.7641,609.62,475.436May 28, 2024Sep 23, 2024Jan 19, 2025May 17, 2025Sep 12, 2025Jan 8, 2026May 7, 2026
Updated 25d ago

Evaluation Results

MethodLinks
2025.04
3,213-
2025.04
2,827-
2025.04
2,824-
2025.04
2,664-
2025.04
2,635-
2025.04
1,839-
2025.04
1,830-
2025.04
1,768-
2025.04
1,765-
2025.04
1,563-
2025.04
1,552-
2025.04
1,255-
2025.04
1,187-
2025.04
1,176-
2025.04
1,085-
2026.02
87.2-
2026.02
85.3-
2026.02
82.8-
2026.02
76.1-
2026.05
74.243.87
2026.02
72.4-
2025.04
71-
2026.02
70.7-
2026.02
69.7-
2026.02
69.3-
2026.05
62.936.39
2026.04
60.8-
2026.04
58.7-
2026.04
58.4-
2026.05
58.237.93
2026.04
57.1-
2026.04
57-
2026.04
55.9-
2026.04
55.3-
2026.04
55.2-
2026.04
54.9-
2026.04
54.3-
2026.05
54.335.75
2026.04
53.7-
2026.04
52.4-
2026.04
51.4-
2026.04
50.4-
2026.04
49.8-
2026.04
47.8-
2026.04
46.7-
2024.05
43.254.33
2026.04
42.1-
2024.05
38.545.63
2024.05
37.543.27
2024.05
37.441.75
2026.04
33.54-
2024.05
33.141.35
2024.05
3310.68
2026.04
31.97-
2026.04
29.13-
2026.04
29.06-
2026.04
28.69-
2026.04
28.42-
2026.04
28.26-
2026.04
27.95-
2026.04
23.84-
2026.05
23.733.64
2025.04
10-
2025.04
8-
2026.05
6.245.82