Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Question Answering on SQuAD (Aclean, Aattack, ASR, Time)
Loading...
54.52
Accuracy (Clean)
AGLS
19.3888
28.5094
37.63
46.7506
May 26, 2025
Accuracy (Clean)
Accuracy (Attack)
Attack Success Rate
Average Inference Time (s)
Updated 1mo ago
Evaluation Results
Method
Method
Links
Accuracy (Clean)
Accuracy (Attack)
Attack Success Rate
Average Inference Time (s)
AGLS
Models=gpt-4o-latest
2025.05
54.52
43.49
20.23
0.63
AGLS
Models=gpt-4
2025.05
51.84
32.43
37.44
0.78
AGLS
Models=llama3.1-70B
2025.05
47.83
27.33
42.86
1.13
AGLS
Models=gemma2-27B
2025.05
44.15
37.79
14.41
0.85
AGLS
Models=gemma2-9B
2025.05
34.11
27.42
19.61
0.45
AGLS
Models=llama3.1-8B
2025.05
33.78
30.43
9.92
0.56
AGLS
Models=qwen2.5-14B
2025.05
30.77
30.43
1.1
1.12
AGLS
Models=llama3.2-3B
2025.05
28.09
22.41
20.22
0.36
AGLS
Models=qwen2.5-7B
2025.05
20.74
19.4
6.46
0.72
Feedback
Search any
task
Search any
task
