Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Retrieval and Reasoning on RULER (4K-128K Context Sweep)
Loading...
98.62
Retrieval/Reasoning Score (4K Context)
RDKV
75.6048
81.5799
87.555
93.5301
May 8, 2026
Retrieval/Reasoning Score (4K Context)
Retrieval/Reasoning Score (8K Context)
Retrieval/Reasoning Score (16K Context)
Retrieval/Reasoning Score (32K Context)
Retrieval/Reasoning Score (64K Context)
Retrieval/Reasoning Score (128K Context)
Updated 22d ago
Evaluation Results
Method
Method
Links
Retrieval/Reasoning Score (4K Context)
Retrieval/Reasoning Score (8K Context)
Retrieval/Reasoning Score (16K Context)
Retrieval/Reasoning Score (32K Context)
Retrieval/Reasoning Score (64K Context)
Retrieval/Reasoning Score (128K Context)
RDKV
Backbone=LLaMA-3.1-8B-...
2026.05
98.62
98.61
95.65
88.28
80.07
66.95
FullKV
Backbone=LLaMA-3.1-8B-...
2026.05
98.58
98.88
96.98
90.33
88.49
79.91
AdaKV
Backbone=LLaMA-3.1-8B-...
2026.05
92.03
85.49
80.7
75.67
72.47
58.43
ThinK
Backbone=LLaMA-3.1-8B-...
2026.05
91.76
86.44
81.45
75.94
72.44
62.85
SnapKV
Backbone=LLaMA-3.1-8B-...
2026.05
89.48
84.75
80.43
75.82
72.77
62.85
Snap+Zip
Backbone=LLaMA-3.1-8B-...
2026.05
76.49
77.84
76.26
75.05
74.59
62.37
Feedback
Search any
task
Search any
task
