I'm a 4th-year PhD candidate in Computer Science at Tufts University. I study, build, and evaluate human-centered AI systems with a focus on accessibility and communication.
Research interests: Human–AI interaction · AI benchmarking · Accessibility · AI bias and fairness · AI-mediated communication · Simulation · Participatory design · Mixed-methods research
-
Benchmarking Anti-autism Communication Bias in LLMs — A measurement study examining how large language models interpret, represent, and evaluate autistic communication in mixed-neurotype settings; contributes a new dataset, benchmarking framework, and empirical findings.
Repo: llm-bias · Paper (pre-print): Google Drive -
NeuroBridge — An LLM-powered conversational simulation system that helps neurotypical users practice mixed-neurotype communication through interactive scenarios.
Repo: asim · Paper (ASSETS 2025, Best Student Paper 🏆): PDF -
TwIPS — An AI-assisted texting interface that supports autistic users in interpreting tone and intent and expressing their intended meaning in online communication.
Repo: twips · Paper (ASSETS 2024): PDF
- Intel CVE Binary Tool — asyncio test configuration fix (PR #4685)
- Email: rukhshan.haroon@tufts.edu