Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Long-context retrieval on RULER-NIAH 64k
Loading...
98.8
Accuracy
BF16
-3.952
22.724
49.4
76.076
May 18, 2026
Accuracy
Updated 14d ago
Evaluation Results
Method
Method
Links
Accuracy
BF16
Backbone=GLM-4.7-FP8†
2026.05
98.8
QuaRot-INT2
Backbone=GLM-4.7-FP8†
2026.05
98.8
OSCAR
Backbone=GLM-4.7-FP8†
2026.05
98.8
BF16
Backbone=Qwen3-4B-Thin...
2026.05
85.3
BF16
Backbone=Qwen3-8B
2026.05
79.2
OSCAR
Backbone=Qwen3-4B-Thin...
2026.05
61.9
OSCAR
Backbone=Qwen3-8B
2026.05
61.9
QuaRot-INT2
Backbone=Qwen3-4B-Thin...
2026.05
15.6
QuaRot-INT2
Backbone=Qwen3-8B
2026.05
0
Feedback
Search any
task
Search any
task