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

Dialogue Evaluation on Arena-Hard

3.29Accepted Length (τ)

DSpark

2.44762.66632.8853.1037Jul 6, 2026
Updated 18d ago

Evaluation Results

MethodLinks
2026.07
3.29
2026.07
3.21
2026.07
3.13
2026.07
2.92
2026.07
2.83
2026.07
2.81
2026.07
2.72
2026.07
2.72
2026.07
2.59
2026.07
2.55
2026.07
2.54
2026.07
2.48