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

Preference Evaluation on AlpacaEval 2

559WR (%)

SelectiveDPO

-21.3304129.3323279.995430.6577Oct 18, 2024Jan 25, 2025May 5, 2025Aug 13, 2025Nov 20, 2025Feb 28, 2026Jun 8, 2026
Updated 1mo ago

Evaluation Results

MethodLinks
2026.02
559-312-
2026.02
559-345-
2026.02
360-238-
2026.02
311-228-
2026.02
99-20-
2026.05
62.6953.28--
2026.05
62.454.24--
2026.05
62.2652--
2026.06
60.1255.08--
2026.05
59.8151.68--
2026.05
58.5950.01--
2026.05
55.8149.04--
2026.06
55.7852.31--
2026.06
55.5351.66--
2026.05
52.8343.37--
2026.06
43.6636.61--
2026.06
41.4234.95--
2026.04
40.9948.31-1,667
2026.04
40.9949.94-1,611
2026.06
38.9334.73--
2026.06
38.6336.43--
2026.05
34.2937.45--
2026.06
33.9232.89--
2026.05
32.6736.56--
2024.10
31.324.2--
2026.05
31.336.17--
2026.06
28.2329.51--
2026.05
27.7220.2--
2026.06
26.3724.46--
2026.06
25.8423.9--
2026.05
24.4116.31--
2024.10
22.5722.92--
2026.05
21.5917.85--
2026.05
21.0517.45--
2024.10
20.318.99--
2024.10
20.117.26--
2026.05
19.524.82--
2026.05
19.115.98--
2026.05
19.0728.59--
2024.10
18.4817.17--
2026.05
18.4115.36--
2026.05
18.1423.18--
2026.05
17.7627.78--
2026.06
17.0817.97--
2026.05
16.8321.81--
2026.06
15.4718.63--
2026.05
14.6619.66--
2026.05
14.6618.3--
2026.05
14.3312.18--
2024.10
13.7912.91--
2026.05
12.8614.46--
2026.06
11.6212.8--
2024.10
11.159.75--
2026.05
10.8714.48--
2026.06
10.7212.24--
2026.05
10.5616.85--
2026.05
10.3713.26--
2026.05
9.6210.69--
2026.05
9.4414.17--
2026.05
7.9510.16--
2026.05
7.958.41--
2026.06
6.058.27--
2026.05
5.597.61--
2026.05
0.992.26--