AI Red-Teaming Researcher & LLM Adversarial Tester
TextRed Teaming
This ongoing self-directed research focuses on adversarial testing and behavioral analysis of Large Language Models. The work emphasizes systematic experimentation and documentation to identify alignment vulnerabilities in commercial and open-source LLMs. Methodologies include hypothesis-driven testing, cross-model analysis, and detailed reproducibility protocols. • Designed and executed iterative adversarial prompting sessions targeting major LLMs
• Analyzed alignment boundaries and elicited out-of-context behaviors
• Documented test results and frequency of anomalous behaviors
• Applied formal reasoning to probe model consistency and alignment limits
This ongoing self-directed research focuses on adversarial testing and behavioral analysis of Large Language Models. The work emphasizes systematic experimentation and documentation to identify alignment vulnerabilities in commercial and open-source LLMs. Methodologies include hypothesis-driven testing, cross-model analysis, and detailed reproducibility protocols. • Designed and executed iterative adversarial prompting sessions targeting major LLMs • Analyzed alignment boundaries and elicited out-of-context behaviors • Documented test results and frequency of anomalous behaviors • Applied formal reasoning to probe model consistency and alignment limits
2021 - Present