Manley Roberts
Machine Learning Scientist at Abridge
Pittsburgh, Pennsylvania, United States
Summary
👤
Senior🎓
Top SchoolManley Roberts is a machine learning scientist based in Pittsburgh with eight years of experience bridging research and applied ML, currently focused on making large language models more reliable. His work centers on distribution shift, contamination, hallucination, and robustness, with peer-reviewed contributions including contamination-free benchmarks and methods to teach LLMs to know their uncertainties. He has moved between industry research roles (Abacus.AI, Abridge) and rigorous academic settings (CMU, advised by Zachary Lipton), pairing theoretical insight with practical evaluation. Manley also has hands-on experience deploying and debugging ML systems from internships at IBM and Microsoft, which informs his pragmatic approach to robust model training. Notably, he co-authored multiple first-author papers addressing real-world failure modes of LLMs and benchmark design, reflecting a consistent focus on reliability rather than raw scale.
8 years of coding experience
2 years of employment as a software developer
Bachelor of Science - BS, Computer Science, Bachelor of Science - BS, Computer Science at Georgia Institute of TechnologyMaster of Science - MS, Machine Learning, Master of Science - MS, Machine Learning at Carnegie Mellon University
Github Skills (18)
codex6
unit-test5
circuit5
testing4
testcase4
synthesis3
codegen2
machine-learning2
temperature2
python1
trello1
tpu1
scaling1
llm1
openai1
Programming languages (5)
Github contributions (5)
manleyroberts/fourdognight
Jan 2018 - Apr 2018
Contributions:58 PRs, 81 pushes, 6 branches in 2 months
nightspringdogfour
manleyroberts/manleyroberts.github.io
May 2021 - Dec 2023
Contributions:7 pushes, 1 branch in 2 years 7 months
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.