48tasksfrom papers' released code
13fieldssystems to robotics
14lecturesrunnable models & videos
16models rankedevery metric must pass
FrontierCS 2 asks AI agents to do CS research. Each task removes the key contribution of a paper from the code released with it, and an agent must write it back and beat the authors' version on every metric. This course takes the 48 tasks one field at a time: what each paper found, what the task removed, how it is judged, and what the agents wrote. Every idea comes with a small model you can run in your browser.