Overview
I currently focus on technical and international governance of frontier AI. In the past, I've explored a variety of other research questionsOversight and control. What oversight measures robustly scale to increasingly capable frontier language models? How do we understand situational awareness, model introspection, and other mentalistic properties?Value alignment and epistemics. How can we train models to be more honest? How can AI uplift human truth-seeking and moral progress, and how do we prevent harmful value lock-in?Agent security. How do we design scalable, realistic environments for evaluating agent misuse and misbehavior — for example, in computer use and MCP settings?How do we operationalize AI-induced human disempowerment? in technical AI safety and alignment.
As a doer and generalist, I try to make things happen in the world strategically. As a researcher, I enjoy rapid experimentation and careful truth-seeking. I absolutely love my work — I'm very lucky ☺︎
Papers
- AI Epistemic Risks: Emerging Mechanisms & Evidence
Casper, S., Stray, J., Gausen, A., Jones, C., Christian, B., Li, J., Bengio, Y., Rand, D., et al. SSRN, 2026. [SSRN]
- WARP: Measuring and Mitigating Evaluation Awareness in Browser-Agent Safety Benchmarks
Li, J.X., Chew, A., Lin, M., Jones, E.K., Fu, X., Zou, A. AIWILD, ICML, 2026. [OpenReview] [GitHub]
- Eval Cooperativeness May Be a Scalable Mitigation for Eval Gaming
- EigenBench: A Comparative Behavioral Measure of Value Alignment
Chang, J., Piff, L., Sana, S., Li, J.X., Levine, L. ICLR Oral (Top 5%), 2026. [arXiv]
- ProgressGym: Alignment with a Millennium of Moral Progress
Qiu, T., Zhang, Y., Huang, Z., Li, J.X., et al. NeurIPS Spotlight (Top 10%), 2024. [arXiv]
- Scaling laws for contrastive activation addition with refusal mechanisms and Llama 2 models
Ali, S.A.R., Xu, J., Yang, I., Li, J.X., Arslan, A., Benham, C. NewInML, ICML, 2025. [arXiv] [paper]
- Machining water through laser cutting of nanoparticle-encased water pancakes
Niu, J., Liu, W., Li, J.X., Pang, X., Liu, Y., Zhang, C., Yue, K., Zhou, Y., Xu, F., Li, X., Li, F. Nature Communications, 2023. [paper]
- A three-dimensional paper-based isoelectric focusing device for direct analysis of proteins in physiological samples
Jiang, X., Li, F., Niu, J., Bao, Z., Wei, Z., Li, J.X.. Analytical Chemistry, 2021. [paper]
Blog
I write about AI at The J-Space.
Loading posts...
Media
- AI Evaluators Struggle as Models Know They're Being Tested
The Information, AI Agenda
Mentorship
- Frontier AI Security Residency, Fall 2026
Organizing
- Proxima
An anthology of speculative fiction and art, imagining post-intelligence explosion futures.
- Idealists Collective Unconference
April 2026
- HTML Day 2025, Seattle
- Pen & Ponder