Our new X account is live! Follow @wizwand_team for updates
WorkDL logo mark

IPI Sanitization

Benchmarks

Task NameDataset NameSOTA ResultTrend
Indirect Prompt Injection SanitizationIPI Sanitization Total
Attack Success Rate (ASR)0.2
15
Indirect Prompt Injection SanitizationIPI Sanitization Cb.
ASR0
15
Indirect Prompt Injection SanitizationIPI Sanitization Cp.
ASR0
15
Indirect Prompt Injection SanitizationIPI Sanitization Ig
ASR0
15
Indirect Prompt Injection SanitizationIPI Sanitization Esc.
ASR0
15
Indirect Prompt Injection SanitizationIPI Sanitization Naive
ASR0
15
IPI SanitizationIPI Sanitization Falcon
Naive ASR0
7
IPI SanitizationIPI Sanitization Dolly
Naive ASR0
7
IPI SanitizationIPI Sanitization ChatGLM
Naive ASR (%)0
7
Showing 9 of 9 rows