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

Question Answering on OpenBookQA (Accuracy and Performance Gain)

84.4Accuracy

Qwen3 8B Base

23.0438.9754.970.83Mar 16, 2024Jul 15, 2024Nov 13, 2024Mar 14, 2025Jul 13, 2025Nov 11, 2025Mar 13, 2026
Updated 1mo ago

Evaluation Results

MethodLinks
2026.01
84.4-0.6
2026.01
84.2-0.8
2026.02
79.35.8
2026.01
76.61
2026.01
76.61
2026.01
73.40.2
2026.01
73-0.2
2026.01
68.80.6
2026.01
68.40.2
2026.01
530.8
2026.01
52.20
2026.01
47.43.8
2025.07
45.6-
2025.07
44.8-
2025.07
44.4-
2025.07
44.2-
2026.01
440.4
2026.03
44-
2026.03
43.8-
2026.03
43.8-
2025.07
43.6-
2026.03
43.6-
2026.03
43.6-
2026.03
43.6-
2026.03
43.6-
2026.03
43.4-
2026.03
43-
2026.03
43-
2024.03
42.8-
2025.07
42.4-
2026.03
42.4-
2026.03
42.4-
2026.03
42.2-
2026.03
42.2-
2026.03
42.2-
2026.03
42.16-
2025.07
42-
2026.03
42-
2025.07
41.6-
2025.07
41.6-
2026.03
41.4-
2026.03
41.2-
2025.07
40.8-
2025.07
40.8-
2026.03
40.6-
2026.03
40.2-
2026.03
40-
2024.03
39.6-
2025.07
39.4-
2026.03
39.4-
2025.07
38.8-
2024.03
38.4-
2025.07
38.2-
2025.07
38-
2024.03
37.8-
2025.07
37.6-
2026.03
37.4-
2025.07
37.2-
2025.07
36.8-
2025.07
36-
2025.07
35.6-
2025.07
35.2-
2025.07
34.4-
2025.07
34-
2025.07
32.4-
2025.07
32.4-
2025.07
31.4-
2025.10
31.4-
2025.10
31.4-
2025.10
30.8-
2025.10
30.6-
2025.07
29.8-
2025.10
29.8-
2025.07
29.6-
2025.10
29.6-
2025.10
29.4-
2025.07
29.4-
2025.07
29.4-
2025.07
29.2-
2025.07
28.8-
2025.07
28.8-
2025.07
28.2-
2025.07
27.4-
2025.10
27.2-
2025.07
27.2-
2025.07
26.8-
2025.10
26.8-
2025.07
26.8-
2025.07
26.6-
2025.10
26.6-
2025.07
26.4-
2025.07
26.2-
2025.07
26-
2025.07
26-
2025.10
26-
2025.10
26-
2025.10
25.8-
2025.10
25.8-
2025.10
25.8-
2025.07
25.4-
Showing 100 of 126 rows