Sotabase

Career

· Senior Software Engineer, NVIDIA2024–
· Senior MLSys Engineer, OctoAI2022–2024
· Research Scholar, SRC Research Scholars Program2018–2022
· Doctor of Philosophy - PhD, Computer Science, University of Washington2016–2022
· Master of Science - MS, Computer Science, University of Washington2016–2018
· Bachelor's degree, Computer Science, Princeton University2012–2016

Publications (13)

Relay: a new IR for machine learning frameworks
MAPL@PLDI · 2018
116
cited
Dynamic Tensor Rematerialization
International Conference on Learning Representations · 2020
111
cited
Relay: A High-Level Compiler for Deep Learning
2019
37
cited
Pure tensor program rewriting via access patterns (representation pearl)
MAPS@PLDI · 2021
36
cited
Relay: A High-Level IR for Deep Learning
arXiv.org · 2019
21
cited
Relax: Composable Abstractions for End-to-End Dynamic Machine Learning
International Conference on Architectural Support for Programming Languages and Operating Systems · 2023
17
cited
SpaceSearch: a library for building and verifying solver-aided tools
Proc. ACM Program. Lang. · 2017
9
cited
Application-level Validation of Accelerator Designs Using a Formal Software/Hardware Interface
ACM Trans. Design Autom. Electr. Syst. · 2022
7
cited
FPGA Technology Mapping Using Sketch-Guided Program Synthesis
International Conference on Architectural Support for Programming Languages and Operating Systems · 2024
5
cited
Specialized Accelerators and Compiler Flows: Replacing Accelerator APIs with a Formal Software/Hardware Interface
arXiv.org · 2022
4
cited
Compiler and Runtime Techniques for Optimizing Deep Learning Applications
2022
2
cited
From DSLs to Accelerator-Rich Platform Implementations: Addressing the Mapping Gap
2021
1
cited
Sotabase
Steven Lyubomirsky | Researcher Profile | Sotabase | Sotabase