Rob Fergus

Biological structure and function emerge from scaling unsupervised learning to 250 million protein sequences

Proceedings of the National Academy of Sciences · 2021 · 3004 citations

In the field of artificial intelligence, a combination of scale in data and model capacity enabled by unsupervised learning has led to major advances in representation learning and statistical generation. In the life sciences, the anticipated growth of sequencing promises unprecedented data on natural sequence diversity. Protein language modeling at the scale of evolution is a logical step toward predictive and generative artificial intelligence for biology. To this end, we use unsupervised lear…

Improving Sample Efficiency in Model-Free Reinforcement Learning from Images

2021-05-18 · 201 citations

Training an agent to solve control tasks directly from high-dimensional images with model-free reinforcement learning (RL) has proven difficult. A promising approach is to learn a latent representation together with the control policy. However, fitting a high-capacity encoder using a scarce reward signal is sample inefficient and leads to poor performance. Prior work has shown that auxiliary losses, such as image reconstruction, can aid efficient representation learning. However, incorporating r…

Image Augmentation Is All You Need: Regularizing Deep Reinforcement\n Learning from Pixels

arXiv (Cornell University) · 2020 · 171 citations

We propose a simple data augmentation technique that can be applied to\nstandard model-free reinforcement learning algorithms, enabling robust learning\ndirectly from pixels without the need for auxiliary losses or pre-training. The\napproach leverages input perturbations commonly used in computer vision tasks\nto regularize the value function. Existing model-free approaches, such as Soft\nActor-Critic (SAC), are not able to train deep networks effectively from image\npixels. However, the additi…

De novo design of high-affinity protein binders with AlphaProteo

arXiv (Cornell University) · 2024-09-12 · 49 citations

Computational design of protein-binding proteins is a fundamental capability with broad utility in biomedical research and biotechnology. Recent methods have made strides against some target proteins, but on-demand creation of high-affinity binders without multiple rounds of experimental testing remains an unsolved challenge. This technical report introduces AlphaProteo, a family of machine learning models for protein design, and details its performance on the de novo binder design problem. With…

Reduce, Reuse, Recycle: Compositional Generation with Energy-Based Diffusion Models and MCMC

arXiv (Cornell University) · 2023-02-22 · 14 citations

Since their introduction, diffusion models have quickly become the prevailing approach to generative modeling in many domains. They can be interpreted as learning the gradients of a time-varying sequence of log-probability density functions. This interpretation has motivated classifier-based and classifier-free guidance as methods for post-hoc control of diffusion models. In this work, we build upon these ideas using the score-based interpretation of diffusion models, and explore alternative way…