Discrete Diffusion Modeling by Estimating the Ratios of the Data Distribution 作者:Aaron Lou, Chenlin Meng, Stefano Ermon 机构:斯坦福大学 ●核心内容: 本文提出了一种新的离散扩散模型(Score Entropy Discrete Diffusion,简称SEDD),它是为了改进自然语言等
不过,尽管相似性很多,但从结果来看,ICML和ICLR的评判标准却不同。ICML的录用论文中,有相当一部分最初被ICLR拒绝,例如在ICML拿到十佳论文的《Discrete Diffusion Modeling by Estimating the Ratios of the Data Distribution》,就是ICLR被拒后修改的版本。 一位获得Oral(可以上台演讲介绍自己的研究)机会的论文作者分享...
In diffusion models, negative prompts specify "what to exclude" from the generated images. Below are the instructions for optimizing negative prompts using DPO-Diff. For more research on how negative prompts work, we refer the reader to the following paper:...
Jun 12, 2024 evaluate.py first commit Jun 12, 2024 method_figure.png first commit Jun 12, 2024 train.py first commit Jun 12, 2024 README MIT license DisCo-DiffDock Code for the molecular docking experiments from the paperDisCo-Diff: Enhancing Continuous Diffusion Models with Discrete Latentsby...
def diffusion_spectral_mutual_information( embedding_vectors: np.array, reference_vectors: np.array, reference_discrete: bool = None, gaussian_kernel_sigma: float = 10, t: int = 1, chebyshev_approx: bool = False, num_repetitions: int = 5, ...
来自斯坦福大学、Pika Labs的《Discrete Diffusion Modeling by Estimating the Ratios of the Data Distribution》则是针对离散数据领域,提出了一种新的离散扩散模型,通过引入得分熵损失来弥补标准扩散模型在离散结构上的不足,并取得了显著的性能提升。 反而是机器学习算法的研究成了少数,一篇研究了随机凸优化背景下记忆...
不过,尽管相似性很多,但从结果来看,ICML和ICLR的评判标准却不同。ICML的录用论文中,有相当一部分最初被ICLR拒绝,例如在ICML拿到十佳论文的《Discrete Diffusion Modeling by Estimating the Ratios of the Data Distribution》,就是ICLR被拒后修改的版本。
不过,尽管相似性很多,但从结果来看,ICML和ICLR的评判标准却不同。ICML的录用论文中,有相当一部分最初被ICLR拒绝,例如在ICML拿到十佳论文的《Discrete Diffusion Modeling by Estimating the Ratios of the Data Distribution》,就是ICLR被拒后修改的版本。
If you use this code or the models in your research, please cite the following paper: @inproceedings{xu2024discodiff,title={DisCo-Diff: Enhancing Continuous Diffusion Models with Discrete Latents},author={Xu, Yilun and Corso, Gabriele and Jaakkola, Tommi and Vahdat, Arash and Kreis, Karsten}...