Publications (13)
The Cursive Transformer
Sam Greydanus, Zachary Wimpee
Transformers trained on tokenized text, audio, and images can generate high-quality autoregressive samples. But handwriting data, represented as sequences of pen coordinates, remai…
The Story of Airplane Wings
Sam Greydanus
The purpose of this work is to explain how wings work and how they were invented. We use the lens of history, looking at the individual people who wanted to fly, the lens of techno…
Learning the Enigma with Recurrent Neural Networks
Sam Greydanus
Recurrent neural networks (RNNs) represent the state of the art in translation, image captioning, and speech recognition. They are also capable of learning algorithmic tasks such a…
Nature's Cost Function: Simulating Physics by Minimizing the Action
Tim Strang, Isabella Caruso, Sam Greydanus
In physics, there is a scalar function called the action which behaves like a cost function. When minimized, it yields the "path of least action" which represents the path a physic…
Visualizing and Understanding Atari Agents
Sam Greydanus, Anurag Koul, Jonathan Dodge +1
While deep reinforcement learning (deep RL) agents are effective at maximizing rewards, it is often unclear what strategies they use to do so. In this paper, we take a step toward…
Hamiltonian Neural Networks
Sam Greydanus, Misko Dzamba, Jason Yosinski
Even though neural networks enjoy widespread use, they still struggle to learn the basic laws of physics. How might we endow them with better inductive biases? In this paper, we dr…
Learning Finite State Representations of Recurrent Policy Networks
Anurag Koul, Sam Greydanus, Alan Fern
Recurrent neural networks (RNNs) are an effective representation of control policies for a wide range of reinforcement and imitation learning problems. RNN policies, however, are p…
Scaling Down Deep Learning with MNIST-1D
Sam Greydanus, Dmitry Kobak
Although deep learning models have taken on commercial and political relevance, key aspects of their training and operation remain poorly understood. This has sparked interest in s…
Dissipative Hamiltonian Neural Networks: Learning Dissipative and Conservative Dynamics Separately
Andrew Sosanya, Sam Greydanus
Understanding natural symmetries is key to making sense of our complex and ever-changing world. Recent work has shown that neural networks can learn such symmetries directly from d…
A Tutorial on Structural Optimization
Sam Greydanus
Structural optimization is a useful and interesting tool. Unfortunately, it can be hard for new researchers to get started on the topic because existing tutorials assume the reader…
Piecewise-constant Neural ODEs
Sam Greydanus, Stefan Lee, Alan Fern
Neural networks are a popular tool for modeling sequential data but they generally do not treat time as a continuous variable. Neural ODEs represent an important exception: they pa…
Neural reparameterization improves structural optimization
Stephan Hoyer, Jascha Sohl-Dickstein, Sam Greydanus
Structural optimization is a popular method for designing objects such as bridge trusses, airplane wings, and optical devices. Unfortunately, the quality of solutions depends heavi…
Lagrangian Neural Networks
Miles Cranmer, Sam Greydanus, Stephan Hoyer +3
Accurate models of the world are built upon notions of its underlying symmetries. In physics, these symmetries correspond to conservation laws, such as for energy and momentum. Yet…