My research is primarily focused on the continual evolution of AI. I am particularly interested in:
Large Language Models (LLMs) — including novel architectures, agent memory, and agent systems.
Reinforcement Learning (RL) — including related frameworks (e.g., GFlowNets).
Recent News
🎨 Jul, 2026. Major posterly upgrade — a dramatically wider range of design styles, instead of one house look. Also launched tryposterly.com, with a blog and a paid cloud service that runs the skill for you, now open to everyone.
🚀 Jun, 2026. Released posterly, a coding-agent skill (Claude, Codex, …) that builds academic conference posters as a single HTML/CSS file and renders them to print-ready PDF — feel free to try it out!
🎉 Apr, 2026. Two papers (PowerFlow and TD-GFN) accepted by ICML 2026.