Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Long-context Variable Tracking on variable-tracking
Loading...
79.5
Accuracy
Qwen2.5-72B Baseline
27.5
41
54.5
68
Apr 23, 2026
Accuracy
Updated 1mo ago
Evaluation Results
Method
Method
Links
Accuracy
Qwen2.5-72B Baseline
Model scale=72B, Compr...
2026.04
79.5
Sub-token routing (Qwen2.5-72B)
Model scale=72B, Compr...
2026.04
79.5
Expected Attention (Qwen2.5-72B)
Model scale=72B, Compr...
2026.04
79.5
EA + Sub-token routing (Qwen2.5-72B)
Model scale=72B, Compr...
2026.04
79.5
Qwen2.5-32B Baseline
Model scale=32B, Compr...
2026.04
79
Sub-token routing (Qwen2.5-32B)
Model scale=32B, Compr...
2026.04
79
Expected Attention (Qwen2.5-32B)
Model scale=32B, Compr...
2026.04
79
EA + Sub-token routing (Qwen2.5-32B)
Model scale=32B, Compr...
2026.04
79
Qwen2.5-14B Baseline
Model scale=14B, Compr...
2026.04
77
Sub-token routing (Qwen2.5-14B)
Model scale=14B, Compr...
2026.04
77
Expected Attention (Qwen2.5-14B)
Model scale=14B, Compr...
2026.04
77
EA + Sub-token routing (Qwen2.5-14B)
Model scale=14B, Compr...
2026.04
77
Qwen2.5-7B Baseline
Model scale=7B, Compre...
2026.04
29.5
Sub-token routing (Qwen2.5-7B)
Model scale=7B, Compre...
2026.04
29.5
Expected Attention (Qwen2.5-7B)
Model scale=7B, Compre...
2026.04
29.5
EA + Sub-token routing (Qwen2.5-7B)
Model scale=7B, Compre...
2026.04
29.5
Feedback
Search any
task
Search any
task