Jiaming Song
Creator of DDIM; former Chief Scientist at Luma AI and former research scientist at NVIDIA Research
Jiaming Song builds multimodal AI systems and is the former Chief Scientist of Luma AI, where Song led research as the company moved from 3D generation models with Genie to video generation with Ray and then to unified multimodal models with Uni-1. Before Luma, Song was a research scientist at NVIDIA Research and, earlier, a postdoc with Stefano Ermon at Stanford University. Song created DDIM, an accelerated sampler for diffusion models published as "Denoising Diffusion Implicit Models" with Chenlin Meng and Stefano Ermon; its MIT-licensed reference code notes that DDIM is also available in Hugging Face Diffusers.1234
Work
DDIM (Denoising Diffusion Implicit Models) (external site: github.com)Open-source code
First author, with Chenlin Meng and Stefano Ermon, of the 2020 DDIM paper. The ermongroup/ddim repository, under an MIT License with Song as copyright holder, implements the sampler and notes that DDIM is also available in Hugging Face Diffusers through DDIMPipeline.2341
Luma AI models (Genie, Ray, Uni-1) (external site: lumalabs.ai)Proprietary product
Led research at Luma AI through its 3D generation (Genie), video generation (Ray), and unified multimodal (Uni-1) models, working across model architectures, training infrastructure, and data processing, per Song's homepage.1
Other documented roles
Roles the sources document but that are not confirmed as current.
- Chief Scientist, Luma AI1
- Research scientist, NVIDIA Research1
- Postdoc with Stefano Ermon, Stanford University1
This profile covers public professional work only, from the sources below. It is not an endorsement or a ranking. To correct or remove it, use the correction form.