Anthropic Claude 3.7 Sonnet vs Google Gemma 3n E2B Instructed LiteRT
Detailed comparison for LLMs
AnthropicGoogle
Head-to-Head Overview
Gemma 3n E2B Instructed LiteRT is the overall winner in this comparison!
LLM Metric
Comparison of Attack Success Rate (ASR) metrics for Anthropic Claude 3.7 Sonnet and Google Gemma 3n E2B Instructed LiteRT across the three attack methods.
Overall (ASR)
Claude 3.7 Sonnet
20.280
Gemma 3n E2B Instructed LiteRT
0
TAP Attack Method (ASR)
Claude 3.7 Sonnet
31.430
Gemma 3n E2B Instructed LiteRT
0
Crescendo Attack Method (ASR)
Claude 3.7 Sonnet
25.270
Gemma 3n E2B Instructed LiteRT
0
Zero-Shot (ASR)
Claude 3.7 Sonnet
0
Gemma 3n E2B Instructed LiteRT
0
Key Highlights
- Google Gemma 3n E2B Instructed LiteRT has a lower Overall (ASR).
- Google Gemma 3n E2B Instructed LiteRT has a lower TAP Attack Method (ASR).
- Google Gemma 3n E2B Instructed LiteRT has a lower Crescendo Attack Method (ASR).
Metric Radar Chart
claude-3-7-sonnet
gemma-3n-e2b-instructed-litert