Email: zjin.admin@cs.toronto.edu      Research: Google Scholar | CV
𝕏: @ZhijingJin 🦋: ZhijingJin (Pronounced like “G-Gin Gin”)

I am an incoming Assistant Professor at the University of Toronto, and currently a research scientist at the Max Planck Institute with Bernhard Schoelkopf, based in Europe. I am also a CIFAR AI Chair, faculty member at the Vector Institute, an ELLIS advisor, and faculty affiliate at the Schwartz Reisman Institute.

My research areas are Large Language Models (LLMs), Causal Inference, and Responsible AI. Specifically, my vertical work focuses on Causal Reasoning with LLMs (Causal AI Scientist, Corr2Cause, CLadder, Quriosity, Survey), Multi-Agent LLMs (GovSim, SanctSim, MoralSim [Slides] [Blogpost]), and Moral Reasoning in LLMs (TrolleyProblems, MoralLens, MoralExceptQA). To support the quality of my vertical work, my horizontal work brings in Mechanistic Interpretability (CompMechs, Mem vs Reasoning), and Adversarial Robustness (CRL Defense, TextFooler, AccidentalVulnerability, RouterAttack). My research contributes to AI Safety and AI for Science. Here are my slides of NLP and Democracy Defense.

I am the recipient of 3 Rising Star awards, 2 Best Paper Awards at NeurIPS 2024 Workshops, and several fellowships at Open Philanthropy and the Future of Life Institute. In the international academic community, I am a co-chair of the ACL Ethics Committee, co-organizer of the ACL Year-Round Mentorship, and a main supporter of the NLP for Positive Impact Workshop series. My work is reported in CHIP Magazine, WIRED, and MIT News.