Aniket Ghosh — Digital Twin
AI/ML engineer and interpretability researcher. Building ground-truth probes for language models at SPAR, after a summer as the only ML engineer at Varosync. Finishing an M.S. in AI at Northeastern, and I publish interpretability work on LessWrong.
Looking for Spring 2027 roles starting in April. Ask me anything, paste a job description, get a cover letter, or find out how I'd attack your hardest problem.
Ask about my background, my projects, the interpretability work, or how I actually work.