conditional generation
Generating data outputs based on specific input conditions or attributes, enabling controlled synthesis aligned with given requirements.
- Atomic Diffusion Models for Small Molecule Structure Elucidation from NMR Spectra
- DISCO: DISCrete nOise for Conditional Control in Text-to-Image Diffusion Models
- Generating Full-field Evolution of Physical Dynamics from Irregular Sparse Observations
- Graph Diffusion that can Insert and Delete
- LEDiT: Your Length-Extrapolatable Diffusion Transformer without Positional Encoding
- LLM Meets Diffusion: A Hybrid Framework for Crystal Material Generation
- Multitask Learning with Stochastic Interpolants
- Token Perturbation Guidance for Diffusion Models
- Toward a Unified Geometry Understanding : Riemannian Diffusion Framework for Graph Generation and Prediction
- Uni-Instruct: One-step Diffusion Model through Unified Diffusion Divergence Instruction