Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Mathematical Reasoning on GSM8K (Accuracy, Attack Robustness, Latency)
Loading...
70
Accuracy (Clean)
AGLS
10.2
25.725
41.25
56.775
May 26, 2025
Accuracy (Clean)
Accuracy (Attack)
Attack Success Rate (ASR)
Average Latency (s)
Updated 1mo ago
Evaluation Results
Method
Method
Links
Accuracy (Clean)
Accuracy (Attack)
Attack Success Rate (ASR)
Average Latency (s)
AGLS
Models=gemma2-27B
2025.05
70
10
85.71
0.85
AGLS
Models=gpt-4o-latest
2025.05
47.5
15
68.42
2.21
AGLS
Models=llama3.1-70B
2025.05
47.5
5
89.47
1.72
AGLS
Models=llama3.2-3B
2025.05
47.5
5
89.47
1.38
AGLS
Models=gpt-4
2025.05
27.5
2.5
90.91
2.31
AGLS
Models=qwen2.5-14B
2025.05
22.5
5
77.78
2.49
AGLS
Models=llama3.1-8B
2025.05
17.5
7.5
57.14
1.49
AGLS
Models=qwen2.5-7B
2025.05
15
7.5
50
1.91
AGLS
Models=gemma2-9B
2025.05
12.5
7.5
40
1.24
Feedback
Search any
task
Search any
task
