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

Alignment on MT-Bench

9.12MT-Bench Score

Qwen2-72B-Instruct

3.62885.05446.487.9056Jul 15, 2024Nov 1, 2024Feb 18, 2025Jun 7, 2025Sep 24, 2025Jan 11, 2026Apr 30, 2026
Updated 1mo ago

Evaluation Results

MethodLinks
2024.07
9.12---
2024.07
8.95---
2024.07
8.88---
2024.07
8.66---
2024.07
8.61---
2026.04
8.16---
2026.04
8.15---
2026.04
8.11---
2026.04
8.07---
2026.04
8.01---
2026.04
7.96---
2026.04
7.89---
2026.04
7.87---
2026.04
7.84---
2026.04
7.83---
2026.04
7.64---
2026.04
7.62---
2026.04
7.6---
2026.04
7.58---
2026.04
7.53---
2026.04
7.45---
2026.04
7.44---
2026.04
7.42---
2026.04
7.34---
2026.04
7.29---
2025.05
5.51--49.4
2025.05
5.42--47.5
2025.05
5.28--47.5
2025.05
5.21--43.1
2025.05
5.19--48.1
2025.05
5.15--48.8
2025.05
5.15--49.7
2025.05
5.13--50.9
2025.05
5.01--45
2025.05
4.98--45
2025.05
4.96--45.6
2025.05
4.8--38.4
2025.05
4.61--27.8
2025.05
4.61--27.8
2025.05
4.61--33.1
2025.05
4.54--33.1
2025.05
4.51--34.1
2025.05
4.49--30
2025.05
4.46--34.7
2025.05
4.4--23.1
2025.05
4.04--21.6
2025.05
4.04--21.6
2025.05
3.88--19.4
2025.05
3.84--20.3
2025.08
-7.645.71-
2025.08
-7.916-
2025.08
-7.415.84-
2025.08
-7.636.74-
2025.08
-6.834.86-
2025.08
-8.516.95-