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

Long-context Reasoning on BABILong 8k

34.7Accuracy

Full (exact)

7.24414.37221.528.628Apr 7, 2026
Updated 10d ago

Evaluation Results

MethodLinks
2026.04
34.7
2026.04
33
2026.04
33
2026.04
32.5
2026.04
32.5
2026.04
32.3
2026.04
32.2
2026.04
32.1
2026.04
31.9
2026.04
31.6
2026.04
31.3
2026.04
31.3
2026.04
30.6
2026.04
30.3
2026.04
30.1
2026.04
29.9
2026.04
29.5
2026.04
28.9
2026.04
28.9
2026.04
28.8
2026.04
28.5
2026.04
28.4
2026.04
28.1
2026.04
27.9
2026.04
27.9
2026.04
27.7
2026.04
27.5
2026.04
27.3
2026.04
27.3
2026.04
26.9
2026.04
26.8
2026.04
26.8
2026.04
26.7
2026.04
26.7
2026.04
26.5
2026.04
26.5
2026.04
26.5
2026.04
26.4
2026.04
26.3
2026.04
25.7
2026.04
25.4
2026.04
25.4
2026.04
25.2
2026.04
25.2
2026.04
25.2
2026.04
24.6
2026.04
24.5
2026.04
24.5
2026.04
24.4
2026.04
24.3
2026.04
22.8
2026.04
22.8
2026.04
22.2
2026.04
20.2
2026.04
19.4
2026.04
18.7
2026.04
15
2026.04
14.3
2026.04
13.2
2026.04
12.6
2026.04
11.9
2026.04
11.4
2026.04
11.3
2026.04
9.8
2026.04
8.3