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

Commonsense Reasoning on OBQA

89.2Accuracy

HydraLoRA

11.40831.60451.871.996Oct 17, 2022May 16, 2023Dec 14, 2023Jul 13, 2024Feb 9, 2025Sep 9, 2025Apr 9, 2026
Updated 8d ago

Evaluation Results

MethodLinks
2026.02
89.2--
2026.02
89--
2026.02
87.8--
2026.02
87--
2024.06
82.738.360.56
2024.06
82.128.70.52
2024.06
82.0611.380.66
2024.06
81.793.40.53
2024.06
81.7213.10.77
2024.06
81.5212.550.73
2024.06
81.523.770.5
2024.06
81.3813.260.75
2024.06
81.3815.341.06
2026.03
80.5--
2026.03
78.4--
2025.12
78--
2025.12
76--
2025.12
75--
2026.03
74--
2022.10
72--
2025.12
71--
2025.12
69--
2026.03
68.2--
2025.12
68--
2022.10
67.8--
2026.03
67.8--
2025.12
67--
2022.10
66.9--
2022.10
64.9--
2025.12
64--
2023.06
61.58--
2023.06
60.9--
2023.02
60.2--
2023.06
58.9--
2023.02
58.6--
2023.06
58.6--
2023.06
58.5--
2023.06
58.43--
2023.06
58.2--
2023.06
57.97--
2023.02
57.6--
2023.06
57.6--
2023.02
57.2--
2023.06
56.95--
2023.06
56.54--
2023.02
56.4--
2023.06
56.17--
2023.02
53.4--
2023.06
51.2--
2023.02
50.4--
2025.02
44.2--
2025.02
42.7--
2025.12
42--
2025.02
42--
2025.02
41.7--
2025.02
41.4--
2025.02
41.3--
2025.02
40.2--
2025.02
39.6--
2025.02
39.3--
2025.02
39.1--
2025.12
38--
2025.02
37.5--
2025.02
37.4--
2025.12
37--
2025.02
37--
2025.02
36.1--
2025.02
35.2--
2025.07
34.6--
2026.03
34--
2026.03
33.6--
2026.03
33.4--
2026.03
33.4--
2026.03
33.4--
2026.03
33.2--
2026.02
32.4--
2025.12
31--
2026.03
29.4--
2025.12
28--
2026.04
27.8--
2025.07
27.6--
2026.04
27.3--
2026.04
27.3--
2026.04
27.3--
2026.04
27.1--
2026.04
26.9--
2025.07
26.8--
2026.04
26.4--
2025.12
26--
2025.07
26--
2025.07
26--
2025.12
23--
2025.12
21--
2026.03
17.8--
2026.03
17.4--
2026.03
17.2--
2026.03
17--
2026.04
14.8--
2026.04
14.6--
2026.04
14.4--
Showing 100 of 117 rows