diffusion-based models
Diffusion-based models are probabilistic generative models that generate data by simulating the diffusion process where noise is iteratively removed from a random sample to produce a coherent output. They have been shown to achieve state-of-the-art results in tasks like image synthesis.
- A Practical Guide for Incorporating Symmetry in Diffusion Policy
- Accelerating 3D Molecule Generative Models with Trajectory Diagnosis
- EDELINE: Enhancing Memory in Diffusion-based World Models via Linear-Time Sequence Modeling
- FIGRDock: Fast Interaction-Guided Regression for Flexible Docking
- IllumiCraft: Unified Geometry and Illumination Diffusion for Controllable Video Generation
- ItDPDM: Information-Theoretic Discrete Poisson Diffusion Model
- OptiScene: LLM-driven Indoor Scene Layout Generation via Scaled Human-aligned Data Synthesis and Multi-Stage Preference Optimization
- Real-Time Execution of Action Chunking Flow Policies
- Selftok-Zero: Reinforcement Learning for Visual Generation via Discrete and Autoregressive Visual Tokens
- StateSpaceDiffuser: Bringing Long Context to Diffusion World Models
- Straight-Line Diffusion Model for Efficient 3D Molecular Generation
- d1: Scaling Reasoning in Diffusion Large Language Models via Reinforcement Learning