Kavach's ML ensemble relies on continuous learning from adversarial datasets. Please contribute new datasets for prompt injection, jailbreaks, steganography, or data exfiltration to improve the ML model's accuracy. \n\nDatasets should ideally be in CSV or JSON format and be well-labeled.
Kavach's ML ensemble relies on continuous learning from adversarial datasets. Please contribute new datasets for prompt injection, jailbreaks, steganography, or data exfiltration to improve the ML model's accuracy. \n\nDatasets should ideally be in CSV or JSON format and be well-labeled.