Working on ’Better model, better metrics, better understanding, and better framework for improved hand gesture bio-metric capacity quantification’

Developing a computer vision-based system for automated prakriti detection to enable predictive health analysis and personalized wellness insights.

Worked on the calculation of Attack Success Ratio (ASR) scores under the Jailbreak Category on various state-of-the-art Large Language Models.– Attacks implemented - Cold Attack, GCG, Puzzler’s box, TAP Attack– LLM under evaluation - Llama, Vicuna, Minstral, GPT, Claude etc