Morteza Mardani
@MardaniMorteza
Principal Scientist @NVIDIA | Visiting Researcher @Stanford | Bridging theory & practice in ML, generative learning, and diffusion models
📢 Test-time Scaling of SDE Diffusion Models Does optimizing the noise trajectory improve sample quality? Significantly. We propose ϵ-greedy search, a simple contextual bandit method matching optimal MCTS in noise space. 📄 arxiv.org/pdf/2506.03164 💻 github.com/rvignav/diffus…
Presenting La-Proteina! A new model for scalable, all-atom protein design 🧬 Backbone + sequence + side-chains, indexed and unindexed atomistic motif scaffolding, scalable up to 800 residues, and more… A thread 🧵
📢📢 "Proteina: Scaling Flow-based Protein Structure Generative Models" #ICLR2025 (Oral Presentation) 🔥 Project page: research.nvidia.com/labs/genair/pr… 📜 Paper: arxiv.org/abs/2503.00710 🛠️ Code and weights: github.com/NVIDIA-Digital… 🧵Details in thread... (1/n)
Tired of slow diffusion models? Our new paper introduces f-distill, enabling arbitrary f-divergence for one-step diffusion distillation. JS divergence gives SOTA results on text-to-image! Choose the divergence that suits your needs. Joint work with @wn8_nie @ArashVahdat 1/N
Machines will train machines. Never bet against scaling. Never.
I don't have too too much to add on top of this earlier post on V3 and I think it applies to R1 too (which is the more recent, thinking equivalent). I will say that Deep Learning has a legendary ravenous appetite for compute, like no other algorithm that has ever been developed…
📣 Announcing: #NVIDIACosmos The world foundation model development platform for advancing physical #AI systems such as autonomous vehicles and robots. Learn more 👉 nvda.ws/4fGKiDX
Generate some visual world videos using our free GPUs: build.nvidia.com Cosmos 1.0 release with open source code and open weight pretrained world models is our research group’s first step to serve the physical ai community. The first batch of released models generate…
github.com/NVIDIA/Cosmos Cosmos is a developer-first platform designed to help physical AI builders accelerate their development. It has pre-trained world foundation models (diffusion & autoregressive) in different sizes and video tokenizers. They are open models with permissive…
Introducing NVIDIA Cosmos, an open-source, open-weight Video World Model. It's trained on 20M hours of videos and weighs from 4B to 14B. Cosmos offers two flavors: diffusion (continuous tokens) and autoregressive (discrete tokens); and two generation modes: text->video and…