Projects
These projects cover my research, teaching, and building.
A research fellowship at the Singapore AI Safety Hub (Sep to Dec 2026). The project runs propensity evaluations of open-weight agents in simulated environments. It measures how deployment conditions such as oversight and irrelevant noise change how often each model takes unsanctioned actions, with closed models as a reference.
A research fellowship with Safe AI Germany, where I test data-level antidote datasets against emergent misalignment on small open-weight models.
A mechanistic study of why more test-time reasoning sometimes makes models worse. I'm doing intra-trace causal analysis with steering-vector interventions, and writing up what I find as I go.
A peer-reviewed study showing that single- and multi-agent LLM systems behave measurably differently on alignment. I read it as early evidence for compositional misalignment.
A large, interlinked set of paper notes and research logs across AI safety, interpretability, and ML. It's the reading and thinking space behind the writing here.
Facilitating technical AI safety cohorts, helping mid-career engineers cross from AI literacy into real contribution. It's the pipeline gap I went through myself.
I build and run a self-improving personal AI agent with persistent memory and standing skills. It's hands-on practice with the exact kind of agentic system whose safety I study.