UPenn launches CIS 6270, a math-first Discrete Generative Models course with full materials online
anshulkundaje · x · 2026-09-21
Pranam Chatterjee is teaching CIS 6270: Discrete Generative Models at UPenn this fall, with syllabus, lecture notes, and code publicly available on the course site and Hugging Face. The course starts with probability, ODEs/SDEs, and Fokker-Planck, deriving flow and score matching from DDPM/DDIM and probability-flow ODEs, then rebuilds everything in discrete space: CTMCs, time reversal, MDLM, D3PM, block diffusion, discrete guidance, discrete flow matching with jump processes, plus rectified flows, optimal transport, stochastic control, and Schrödinger bridges, with applications to biological sequences.
More from Research
- Hazy Research: agents are replacing abstractions — CUDA DSLs are heading to retirement — ricklamers · 2026-09-22
- OXtalign: open-source crystal-packing similarity tool up to 200x faster than COMPACK — CatAstro_Piyush · 2026-09-22
- Microsoft's RetroChimera model for chemical synthesis prediction published in Nature — CatAstro_Piyush · 2026-09-22
- Tiny-task benchmarks no longer make sense for frontier models, argues dev — pvncher · 2026-09-22
- Why collaborating agents beat subagents: broadcasting breakthroughs in parallel — IgorCarron · 2026-09-22
- Frank Noe Clarifies How Their AI Solves the Electronic Schrödinger Equation — FrankNoeBerlin · 2026-09-22