Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Downstream Task Evaluation on LAMBADA, HellaSwag, PIQA, WinoGrande, OpenBookQA, and ARC (test)
Loading...
52.21
HellaSwag Accuracy
HiLS-Attn HoPE
23.9116
31.2583
38.605
45.9517
Jul 3, 2026
HellaSwag Accuracy
LAMBADA Accuracy
PIQA Accuracy
WinoGrande Accuracy
OpenBookQA Accuracy
ARC-e Accuracy
ARC-c Accuracy
ARC-e (25-shot) Accuracy
ARC-c (25-shot) Accuracy
Average Accuracy
Updated 18d ago
Evaluation Results
Method
Method
Links
HellaSwag Accuracy
LAMBADA Accuracy
PIQA Accuracy
WinoGrande Accuracy
OpenBookQA Accuracy
ARC-e Accuracy
ARC-c Accuracy
ARC-e (25-shot) Accuracy
ARC-c (25-shot) Accuracy
Average Accuracy
HiLS-Attn HoPE
Attention mechanism=Hi...
2026.07
52.21
57.05
72.14
55.64
22.5
70.02
37.97
45.86
28.14
49.06
Full-Attn RoPE
Attention mechanism=Fu...
2026.07
51.99
57.13
70.51
56.51
22.6
68.25
36.95
44.44
29.49
48.65
Random-Guess
Model parameters=1.4B,...
2026.07
25
-
50
50
25
25
25
25
25
25
Feedback
Search any
task
Search any
task