Sai Suresh Macharla Vasu

About Me

I am an M.Sc. student in Data Science and Artificial Intelligence (DSAI) at Saarland University, Germany. I am currently a Student Research Assistant in the Trustworthy ML group at CISPA Helmholtz Center for Information Security, supervised by Prof. Dr. Mario Fritz. I was a Research Fellow at the AI Safety Saarland Incubator and a Student Research Assistant at DFKI in the Interactive Machine Learning group.

My work lies at the intersection of AI Safety and Personalization — with a particular focus on evaluating model safety for user welfare, identifying hidden biases, and understanding how language models can be personalized through multi-turn interactions and memory.

Research Interests

• AI Safety: evaluating LLM safety for user welfare, identifying failure modes of deployed AI systems
• Personalization via Multi-Turn Interactions: adapting language model behaviour through conversational context
• Personalization from Memory: enabling models to learn and retain user-specific knowledge across interactions

News

• [Feb. 2026] Paper “Challenges of Evaluating LLM Safety for User Welfare” accepted to IASEAI’26.
• [Dec. 2025] Started as Student Research Assistant at CISPA Helmholtz Center for Information Security (Trustworthy ML group).
• [Nov. 2025] Paper “Justice in Judgment: Unveiling (Hidden) Bias in LLM-assisted Peer Reviews” presented at NeurIPS 2025 Workshop on LLM Evaluation.
• [Nov. 2025] Paper “AuditCopilot: Leveraging LLMs for Fraud Detection in Double-Entry Bookkeeping” presented at NeurIPS 2025 Workshop on Generative AI in Finance.
• [Oct. 2025] Selected as Research Fellow at the AI Safety Saarland Incubator.
• [Jul. 2025] Awarded the Saarland Scholarship for International Students (Summer Semester 2025).

Publications

Powered by Jekyll and Minimal Light theme.