Our new X account is live! Follow @wizwand_team for updates
Home
/
Benchmarks
Needle-In-A-Haystack Retrieval on RULER (test)
Loading...
2,800
Multi-Query Success Rate
DroPE transformer
-112
644
1,400
2,156
Dec 13, 2025
Multi-Query Success Rate
Multi-Key Success Rate
Multi-Value Success Rate
Updated 4d ago
Evaluation Results
Method
Method
Links
Multi-Query Success Rate
Multi-Key Success Rate
Multi-Value Success Rate
DroPE transformer
mode=Zero-shot, contex...
2025.12
2,800
4,160
2,330
RoPE transformer + RoPE-NTK
mode=Zero-shot, contex...
2025.12
2,110
1,940
1,650
RoPE transformer + YaRN
mode=Zero-shot, contex...
2025.12
1,780
50
1,460
NoPE transformer
mode=Zero-shot, contex...
2025.12
920
3,620
2,140
ALiBi transformer
mode=Zero-shot, contex...
2025.12
520
0
110
RNoPE-SWA transformer
mode=Zero-shot, contex...
2025.12
520
2,560
2,060
RoPE transformer
mode=Zero-shot, contex...
2025.12
0
0
0
RoPE transformer + PI
mode=Zero-shot, contex...
2025.12
0
0
0
Feedback
Search any
task
Search any
task