Jascha Sohl-Dickstein
- Papers
- 9
Cite
Notes
Only stored in your browser.
Authored papers
9The boundary of neural network trainability is fractal
arXiv 2024
Reduce, Reuse, Recycle: Compositional Generation with Energy-Based Diffusion Models and MCMC
arXiv 2023
Beyond the Imitation Game: Quantifying and extrapolating the capabilities of language models
TMLR
Language Model Cascades
arXiv 2022
VeLO: Training Versatile Learned Optimizers by Scaling Up
arXiv 2022
NL-Augmenter: A Framework for Task-Sensitive Natural Language Augmentation
arXiv 2021
Score-Based Generative Modeling through Stochastic Differential Equations
score-based-generative-modeling-through
SVCCA: Singular Vector Canonical Correlation Analysis for Deep Learning Dynamics and Interpretability
svcca-singular-vector-canonical-correlation-1
Deep Unsupervised Learning using Nonequilibrium Thermodynamics
arXiv 2015
Affiliations
Frequent co-authors
10from 9 papers