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

Multi-task Reasoning on Average Out-of-domain

49.57Accuracy (OOD)

DFT

4.361216.098127.83539.5719Oct 31, 2025
Updated 1mo ago

Evaluation Results

MethodLinks
2025.10
49.57
2025.10
45.93
2025.10
44.83
2025.10
43.78
2025.10
42.64
2025.10
42.52
2025.10
37.29
2025.10
35.34
2025.10
35.26
2025.10
35.26
2025.10
35.07
2025.10
32.87
2025.10
32.49
2025.10
30.62
2025.10
30.43
2025.10
30.34
2025.10
29.58
2025.10
26.12
2025.10
20.18
2025.10
9.59
2025.10
9.07
2025.10
7.71
2025.10
7.17
2025.10
6.1