Brandon Amos
Research Scientist
Meta AI (FAIR)


I am a research scientist in the Fundamental AI Research (FAIR) group at Meta AI in NYC and study foundational topics in machine learning and optimization, recently involving reinforcement learning, control, optimal transport, and geometry. My research is on learning systems that understand and interact with our world and focuses on integrating structural information and domain knowledge into these systems to represent non-trivial reasoning operations. A key theme of my work in this space involves the use of optimization as a differentiable building block in larger architectures that are end-to-end learned. I believe that science should be open and reproducible and freely publish my research code to GitHub.


2014 - 2019 Ph.D. in Computer Science, Carnegie Mellon University (0.00/0.00)

Thesis: Differentiable Optimization-Based Modeling for Machine Learning
Advisor: J. Zico Kolter

2011 - 2014 B.S. in Computer Science, Virginia Tech (3.99/4.00)

Previous Positions

2016 - 2019

Research Assistant, Carnegie Mellon University (with J. Zico Kolter on ML and optimization)


Research Intern, Intel Labs, Santa Clara (with Vladlen Koltun on computer vision)


Research Intern, Google DeepMind, London (with Nando de Freitas and Misha Denil on RL)

2014 - 2016

Research Assistant, Carnegie Mellon University (with Mahadev Satyanarayanan on mobile systems)


Research Intern, Adobe Research, San Jose (with David Tompkins on distributed systems)

2013 - 2014

Research Assistant, Virginia Tech (with Layne Watson and David Easterling on optimization)

2012 - 2014

Research Assistant, Virginia Tech (with Jules White and Hamilton Turner on mobile systems)

2012 - 2014

Research Assistant, Virginia Tech (with Binoy Ravindran and Alastair Murray on compilers)

2013 - 2014

Software Intern, Snowplow (Scala development)


Software Intern, Qualcomm, San Diego (Python and C++ development)


Software Intern, Phoenix Integration, Virginia (C++, C#, and Java development)


Network Administrator Intern, Sunapsys, Virginia

Honors & Awards

ICML Outstanding Reviewer
ICLR Outstanding Reviewer
2016 - 2019
NSF Graduate Research Fellowship
2011 - 2014
Nine undergraduate scholarships

Roanoke County Public Schools Engineering, Salem-Roanoke County Chamber of Commerce, Papa John's, Scottish Rite of Freemasonry, VT Intelligence Community Conter for Academic Excellence, VT Pamplin Leader, VT Benjamin F. Bock, VT Gay B. Shober, VT I. Luck Gravett


Representative publications that I am a primary author on are highlighted.
[Google Scholar; 5041+ citations, h-index: 29+]


1. Tutorial on amortized optimization for learning to optimize over continuous domains [abs] [code]
Brandon Amos
arXiv 2022
2. Cross-Domain Imitation Learning via Optimal Transport [abs] [code]
Arnaud Fickinger, Samuel Cohen, Stuart Russell, and Brandon Amos
ICLR 2022
3. Matching Normalizing Flows and Probability Paths on Manifolds [abs] [code]
Heli Ben-Hamu*, Samuel Cohen*, Joey Bose, Brandon Amos, Aditya Grover, Maximilian Nickel, Ricky T. Q. Chen, and Yaron Lipman
ICML 2022
4. Semi-Discrete Normalizing Flows through Differentiable Tessellation [abs]
Ricky T. Q. Chen, Brandon Amos, and Maximilian Nickel
arXiv 2022
5. Meta Optimal Transport [abs] [code]
Brandon Amos, Samuel Cohen, Giulia Luise, and Ievgen Redko
arXiv 2022
6. Nocturne: a driving benchmark for multi-agent learning [abs] [code]
Eugene Vinitsky, Nathan Lichtlé, Xiaomeng Yang, Brandon Amos, and Jakob Foerster
arXiv 2022
7. Theseus: A Library for Differentiable Nonlinear Optimization [abs] [code]
Luis Pineda, Taosha Fan, Maurizio Monge, Shobha Venkataraman, Paloma Sodhi, Ricky Chen, Joseph Ortiz, Daniel DeTone, Austin Wang, Stuart Anderson, Jing Dong, Brandon Amos, and Mustafa Mukadam
arXiv 2022


8. On the model-based stochastic value gradient for continuous reinforcement learning [abs] [code] [slides]
Brandon Amos, Samuel Stanton, Denis Yarats, and Andrew Gordon Wilson
L4DC 2021 (Oral)
9. Riemannian Convex Potential Maps [abs] [code] [slides]
Samuel Cohen*, Brandon Amos*, and Yaron Lipman
ICML 2021
10. CombOptNet: Fit the Right NP-Hard Problem by Learning Integer Programming Constraints [abs] [code]
Anselm Paulus, Michal Rolínek, Vít Musil, Brandon Amos, and Georg Martius
ICML 2021
11. Scalable Online Planning via Reinforcement Learning Fine-Tuning [abs]
Arnaud Fickinger, Hengyuan Hu, Brandon Amos, Stuart Russell, and Noam Brown
NeurIPS 2021
12. Aligning Time Series on Incomparable Spaces [abs] [code] [slides]
Samuel Cohen, Giulia Luise, Alexander Terenin, Brandon Amos, and Marc Peter Deisenroth
13. Learning Neural Event Functions for Ordinary Differential Equations [abs] [code]
Ricky T. Q. Chen, Brandon Amos, and Maximilian Nickel
ICLR 2021
14. Neural Spatio-Temporal Point Processes [abs] [code]
Ricky T. Q. Chen, Brandon Amos, and Maximilian Nickel
ICLR 2021
15. Improving Sample Efficiency in Model-Free Reinforcement Learning from Images [abs] [code]
Denis Yarats, Amy Zhang, Ilya Kostrikov, Brandon Amos, Joelle Pineau, and Rob Fergus
AAAI 2021
16. Neural Fixed-Point Acceleration for Convex Optimization [abs] [code]
Shobha Venkataraman* and Brandon Amos*
ICML AutoML Workshop 2021
17. Sliced Multi-Marginal Optimal Transport [abs]
Samuel Cohen, Alexander Terenin, Yannik Pitcan, Brandon Amos, Marc Peter Deisenroth, and K S Sesh Kumar
NeurIPS OTML Workshop 2021
18. Input Convex Gradient Networks [abs]
Jack Richter-Powell, Jonathan Lorraine, and Brandon Amos
NeurIPS OTML Workshop 2021
19. Imitation Learning from Pixel Observations for Continuous Control [abs]
Samuel Cohen, Brandon Amos, Marc Peter Deisenroth, Mikael Henaff, Eugene Vinitsky, and Denis Yarats
NeurIPS DeepRL Workshop 2021
20. MBRL-Lib: A Modular Library for Model-based Reinforcement Learning [abs] [code]
Luis Pineda, Brandon Amos, Amy Zhang, Nathan Lambert, and Roberto Calandra
arXiv 2021


21. The Differentiable Cross-Entropy Method [abs] [code] [slides]
Brandon Amos and Denis Yarats
ICML 2020
22. Objective Mismatch in Model-based Reinforcement Learning [abs]
Nathan Lambert, Brandon Amos, Omry Yadan, and Roberto Calandra
L4DC 2020
23. QNSTOP: Quasi-Newton Algorithm for Stochastic Optimization [abs] [code]
Brandon Amos, David Easterling, Layne T. Watson, William Thacker, Brent Castle, and Michael Trosset
24. Neural Potts Model [abs]
Tom Sercu, Robert Verkuil, Joshua Meier, Brandon Amos, Zeming Lin, Caroline Chen, Jason Liu, Yann LeCun, and Alexander Rives
MLCB 2020
25. Deep Riemannian Manifold Learning [abs]
Aaron Lou, Maximilian Nickel, and Brandon Amos
NeurIPS Geo4dl Workshop 2020


26. Differentiable Optimization-Based Modeling for Machine Learning [abs] [code]
Brandon Amos
Ph.D. Thesis 2019
27. Differentiable Convex Optimization Layers [abs] [code]
Akshay Agrawal*, Brandon Amos*, Shane Barratt*, Stephen Boyd*, Steven Diamond*, and J. Zico Kolter*
NeurIPS 2019
28. The Limited Multi-Label Projection Layer [abs] [code]
Brandon Amos, Vladlen Koltun, and J. Zico Kolter
arXiv 2019
29. Generalized Inner Loop Meta-Learning [abs] [code]
Edward Grefenstette, Brandon Amos, Denis Yarats, Phu Mon Htut, Artem Molchanov, Franziska Meier, Douwe Kiela, Kyunghyun Cho, and Soumith Chintala
arXiv 2019


30. Learning Awareness Models [abs]
Brandon Amos, Laurent Dinh, Serkan Cabi, Thomas Rothörl, Sergio Gómez Colmenarejo, Alistair Muldal, Tom Erez, Yuval Tassa, Nando de Freitas, and Misha Denil
ICLR 2018
31. Differentiable MPC for End-to-end Planning and Control [abs] [code]
Brandon Amos, Ivan Dario Jimenez Rodriguez, Jacob Sacks, Byron Boots, and J. Zico Kolter
NeurIPS 2018
32. Depth-Limited Solving for Imperfect-Information Games [abs]
Noam Brown, Tuomas Sandholm, and Brandon Amos
NeurIPS 2018
33. Enabling Live Video Analytics with a Scalable and Privacy-Aware Framework [abs]
Junjue Wang, Brandon Amos, Anupam Das, Padmanabhan Pillai, Norman Sadeh, and Mahadev Satyanarayanan


34. OptNet: Differentiable Optimization as a Layer in Neural Networks [abs] [code] [slides]
Brandon Amos and J. Zico Kolter
ICML 2017
35. Input Convex Neural Networks [abs] [code] [slides]
Brandon Amos, Lei Xu, and J. Zico Kolter
ICML 2017
36. Task-based End-to-end Model Learning [abs] [code]
Priya L. Donti, Brandon Amos, and J. Zico Kolter
NeurIPS 2017
37. Quasi-Newton Stochastic Optimization Algorithm for Parameter Estimation of a Stochastic Model of the Budding Yeast Cell Cycle [abs]
Minghan Chen, Brandon Amos, Layne T. Watson, John Tyson, Yang Cao, Cliff Shaffer, Michael Trosset, Cihan Oguz, and Gisella Kakoti
38. You can teach elephants to dance: agile VM handoff for edge computing [abs]
Kiryong Ha, Yoshihisa Abe, Thomas Eiszler, Zhuo Chen, Wenlu Hu, Brandon Amos, Rohit Upadhyaya, Padmanabhan Pillai, and Mahadev Satyanarayanan
SEC 2017
39. An Empirical Study of Latency in an Emerging Class of Edge Computing Applications for Wearable Cognitive Assistance [abs]
Zhuo Chen, Wenlu Hu, Junjue Wang, Siyan Zhao, Brandon Amos, Guanhang Wu, Kiryong Ha, Khalid Elgazzar, Padmanabhan Pillai, Roberta Klatzky, Daniel Siewiorek, and Mahadev Satyanarayanan
SEC 2017
40. A Scalable and Privacy-Aware IoT Service for Live Video Analytics [abs] [code]
Junjue Wang, Brandon Amos, Anupam Das, Padmanabhan Pillai, Norman Sadeh, and Mahadev Satyanarayanan
ACM MMSys 2017 (Best Paper)


41. OpenFace: A general-purpose face recognition library with mobile applications [abs] [code]
Brandon Amos, Bartosz Ludwiczuk, and Mahadev Satyanarayanan
CMU 2016
42. Collapsed Variational Inference for Sum-Product Networks [abs]
Han Zhao, Tameem Adel, Geoff Gordon, and Brandon Amos
ICML 2016
43. Quantifying the impact of edge computing on mobile applications [abs]
Wenlu Hu, Ying Gao, Kiryong Ha, Junjue Wang, Brandon Amos, Zhuo Chen, Padmanabhan Pillai, and Mahadev Satyanarayanan
44. Privacy mediators: helping IoT cross the chasm [abs]
Nigel Davies, Nina Taft, Mahadev Satyanarayanan, Sarah Clinch, and Brandon Amos
HotMobile 2016

2015 and earlier

45. Edge Analytics in the Internet of Things [abs]
Mahadev Satyanarayanan, Pieter Simoens, Yu Xiao, Padmanabhan Pillai, Zhuo Chen, Kiryong Ha, Wenlu Hu, and Brandon Amos
IEEE Pervasive Computing 2015
46. Bad Parts: Are Our Manufacturing Systems at Risk of Silent Cyberattacks? [abs]
Hamilton Turner, Jules White, Jaime A. Camelio, Christopher Williams, Brandon Amos, and Robert Parker
IEEE Security & Privacy 2015
47. Early Implementation Experience with Wearable Cognitive Assistance Applications [abs]
Zhuo Chen, Lu Jiang, Wenlu Hu, Kiryong Ha, Brandon Amos, Padmanabhan Pillai, Alex Hauptmann, and Mahadev Satyanarayanan
WearSys 2015
48. The Case for Offload Shaping [abs]
Wenlu Hu, Brandon Amos, Zhuo Chen, Kiryong Ha, Wolfgang Richter, Padmanabhan Pillai, Benjamin Gilbert, Jan Harkes, and Mahadev Satyanarayanan
HotMobile 2015
49. Are Cloudlets Necessary? [abs]
Ying Gao, Wenlu Hu, Kiryong Ha, Brandon Amos, Padmanabhan Pillai, and Mahadev Satyanarayanan
CMU 2015
50. Adaptive VM handoff across cloudlets [abs]
Kiryong Ha, Yoshihisa Abe, Zhuo Chen, Wenlu Hu, Brandon Amos, Padmanabhan Pillai, and Mahadev Satyanarayanan
CMU 2015
51. Global Parameter Estimation for a Eukaryotic Cell Cycle Model in Systems Biology [abs]
Tricity Andrew, Brandon Amos, David Easterling, Cihan Oguz, William Baumann, John Tyson, and Layne T. Watson
SummerSim 2014
52. Applying machine learning classifiers to dynamic Android malware detection at scale [abs] [code]
Brandon Amos, Hamilton Turner, and Jules White
IWCMC 2013

Open Source Repositories

1. 2022 facebookresearch/amortized-optimization-tutorial | 123 | Tutorial on amortized optimization
2. 2022 facebookresearch/theseus | 707 | Differentiable non-linear optimization library
3. 2022 facebookresearch/meta-ot | 67 | Meta Optimal Transport
4. 2022 bamos/presentations | 102 | Source for my major presentations with a CC-BY license
5. 2021 facebookresearch/rcpm | 57 | Riemannian Convex Potential Maps
6. 2021 facebookresearch/svg | 41 | Model-based stochastic value gradient
7. 2021 facebookresearch/mbrl-lib | 658 | Model-based reinforcement learning library
8. 2020 facebookresearch/dcem | 107 | The Differentiable Cross-Entropy Method
9. 2019 facebookresearch/higher | 1.4k | PyTorch higher-order gradient and optimization library
10. 2019 bamos/thesis | 275 | Ph.D. Thesis LaTeX source code
11. 2019 cvxgrp/cvxpylayers | 1.3k | Differentiable Convex Optimization Layers
12. 2019 locuslab/lml | 51 | The Limited Multi-Label Projection Layer
13. 2018 locuslab/mpc.pytorch | 579 | Differentiable PyTorch Model Predictive Control library
14. 2018 locuslab/differentiable-mpc | 124 | Differentiable MPC experiments
15. 2017 locuslab/icnn | 243 | Input Convex Neural Network experiments
16. 2017 locuslab/optnet | 394 | OptNet experiments
17. 2017 locuslab/qpth | 535 | Differentiable PyTorch QP solver
18. 2017 bamos/densenet.pytorch | 756 | PyTorch DenseNet implementation
19. 2017 bamos/block | 268 | Intelligent block matrix constructions
20. 2017 bamos/setGPU | 101 | Automatically use the least-loaded GPU
21. 2016 bamos/dcgan-completion.tensorflow | 1.3k | Image completion with GANs
22. 2015 cmusatyalab/openface | 14.5k | Face recognition with deep neural networks
23. 2014 vtopt/qnstop | 10 | Fortran Quasi-newton stochastic optimization library
24. 2014 bamos/snowglobe | 27 | Haskell-driven, self-hosted web analytics with minimal configuration
25. 2014 bamos/zsh-history-analysis | 184 | Analyze and plot your zsh history
26. 2014 bamos/beamer-snippets | 107 | Beamer and TikZ snippets
27. 2013 bamos/latex-templates | 357 | LaTeX templates
28. 2013 cparse/cparse | 254 | C++ expression parser using Dijkstra's shunting-yard algorithm
29. 2013 bamos/cv | 364 | Source for this CV: Creates LaTeX/Markdown from YAML/BibTeX
30. 2013 bamos/python-scripts | 195 | Short and fun Python scripts
31. 2013 bamos/reading-list | 187 | YAML reading list and notes system
32. 2012 bamos/dotfiles | 238 | Linux, xmonad, emacs, vim, zsh, tmux

Invited Talks

Slides for my major presentations are open-sourced with a CC-BY license at bamos/presentations.

1. 2022 Differentiable optimization, IJCAI Tutorial
2. 2022 Differentiable optimization for control and RL, ICML Workshop on Decision Awareness in RL
3. 2022 Differentiable optimization-based modeling for machine learning, CPAIOR Master Class
4. 2022 Tutorial on amortized optimization, ICCOPT
5. 2022 Modeling and learning paradigms for learning to optimize, SIAM MDS Minisymposium
6. 2021 Learning for control with differentiable optimization and ODEs, Columbia University
7. 2021 Differentiable optimization-based modeling for machine learning, IBM Research
8. 2020 Differentiable optimization for control, Max Planck Institute (Tübingen)
9. 2020 Differentiable optimization-based modeling for machine learning, Mila Seminar
10. 2020 Deep Declarative Networks, ECCV Tutorial
11. 2020 On differentiable optimization for control and vision, CVPR Deep Declarative Networks Workshop
12. 2020 Differentiable optimization-based modeling for machine learning, Caltech CS 159 (Guest Lecture)
13. 2020 Unrolled optimization for learning deep energy models, SIAM MDS Minisymposium
14. 2019 Differentiable optimization-based modeling for machine learning, NYU CILVR Seminar
15. 2019 Differentiable optimization-based modeling for machine learning, INFORMS
16. 2019 Differentiable optimization-based modeling for machine learning, Facebook AI Research
17. 2018 Differentiable optimization-based modeling for machine learning, ISMP
18. 2018 Differentiable optimization-based modeling for machine learning, Google Brain
19. 2018 Differentiable optimization-based modeling for machine learning, Bosch Center for AI
20. 2018 Differentiable optimization-based modeling for machine learning, Waymo Research
21. 2018 Differentiable optimization-based modeling for machine learning, Tesla AI
22. 2018 Differentiable optimization-based modeling for machine learning, NVIDIA Robotics
23. 2018 Differentiable optimization-based modeling for machine learning, Salesforce Research
24. 2018 Differentiable optimization-based modeling for machine learning, OpenAI
25. 2018 Differentiable optimization-based modeling for machine learning, NNAISENSE
26. 2018 Differentiable optimization and control, UC Berkeley

Interns and Students

2020 - 2022 Aaron Lou (visiting FAIR from Cornell and Stanford)
2021 - 2022 Eugene Vinitsky (visiting FAIR from Berkeley, now incoming professor at NYU)
2021 - 2022 Arnaud Fickinger (visiting FAIR from Berkeley)
2021 - 2022 Samuel Cohen (visiting FAIR from UCL, now CEO at FairGen)
2020 Ricky Chen (visiting FAIR from Toronto, now scientist at FAIR)
2020 Paul Liang (visiting FAIR from CMU)
2018 Phillip Wang (at CMU, now CEO at Gather)

Professional Activities

2023 AAAI Senior Program Committee
2020 NeurIPS Learning Meets Combinatorial Optimization Workshop Organizer
2020 CVPR Deep Declarative Networks Workshop Organizer
2020 ECCV Deep Declarative Networks Tutorial Organizer
2014 - 2015 CMU CSD MS Admissions


AAAI Conference on Artificial Intelligence
IEEE Conference on Computer Vision and Pattern Recognition (CVPR)
IEEE International Conference on Computer Vision (ICCV)
IEEE International Conference on Robotics and Automation (ICRA)
International Conference on Learning Representations (ICLR)
International Conference on Machine Learning (ICML)
Journal of Machine Learning Research (JMLR)
Mathematical Programming Computation (MPC)
Neural Information Processing Systems (NeurIPS)
Neural Information Processing Systems (NeurIPS) Datasets and Benchmarks Track
Neural Information Processing Systems (NeurIPS) OPT Workshop
Neural Information Processing Systems (NeurIPS) DiffCVGP Workshop
Neural Information Processing Systems (NeurIPS) Deep RL Workshop
Optimization Letters


Graduate AI (CMU 15-780), TA S2017
Distributed Systems (CMU 15-440/640), TA S2016
Software Design and Data Structures (VT CS2114), TA S2013


Programming C, C++, Fortran, Haskell, Java, Lua, Make, Mathematica, Python, R, Scala
Frameworks JAX, NumPy, Pandas, PyTorch, SciPy, TensorFlow, Torch7
Toolbox Linux, emacs, vim, evil, org, mu4e, xmonad, git, tmux, zsh

Last updated on 2022-07-28