Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Long-context Question Answering on Clip 128K (Acc)
Loading...
52
Accuracy
RECONTEXT
16.64
25.82
35
44.18
Jul 2, 2026
Accuracy
Updated 22d ago
Evaluation Results
Method
Method
Links
Accuracy
RECONTEXT
Backbone=Qwen3-4B
2026.07
52
A-MEM
Backbone=Qwen3-4B
2026.07
48
DySCO
Backbone=Qwen3-4B
2026.07
44
AttnSharp
Backbone=Qwen3-4B
2026.07
42
RECONTEXT
Backbone=Llama3-8B
2026.07
40
Vanilla
Backbone=Qwen3-4B
2026.07
38
DySCO
Backbone=Llama3-8B
2026.07
38
DySCO
Backbone=Qwen3-8B
2026.07
34
RECONTEXT
Backbone=Qwen3-8B
2026.07
33
AttnSharp
Backbone=Qwen3-8B
2026.07
32
Vanilla
Backbone=Llama3-8B
2026.07
32
A-MEM
Backbone=Llama3-8B
2026.07
32
Vanilla
Backbone=Qwen3-8B
2026.07
30
DAC
Backbone=Llama3-8B
2026.07
28
DAC
Backbone=Qwen3-4B
2026.07
24
AttnSharp
Backbone=Llama3-8B
2026.07
22
DAC
Backbone=Qwen3-8B
2026.07
20
A-MEM
Backbone=Qwen3-8B
2026.07
18
Feedback
Search any
task
Search any
task