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

Semantic Needle-In-A-Haystack on S-NIAH

52.4Accuracy

RLM

0.81614.20827.640.992Mar 20, 2026
Updated 26d ago

Evaluation Results

MethodLinks
2026.03
52.4
2026.03
51.3
2026.03
49.8
2026.03
46.8
2026.03
44.2
2026.03
41.8
2026.03
40.2
2026.03
38.6
2026.03
35.2
2026.03
31.7
2026.03
31.5
2026.03
29.7
2026.03
28.3
2026.03
26.4
2026.03
24.6
2026.03
22.6
2026.03
22.1
2026.03
19.1
2026.03
18.5
2026.03
18.4
2026.03
14.3
2026.03
10.2
2026.03
8.4
2026.03
6.1
2026.03
4.1
2026.03
3.2
2026.03
2.8