# Sebastian Ament — Personal Academic Website

## About
Sebastian Ament is a Staff Research Scientist in the Adaptive Experimentation group at Meta.
He develops sample-efficient optimization methods with a focus on Bayesian optimization,
sparse optimization, and active learning for efficient AutoML of large machine learning models.

PhD in Computer Science from Cornell University (advisor: Carla Gomes).

## Research Areas
- Bayesian Optimization (acquisition functions, scalability, multi-objective)
- Gaussian Processes (scalable inference, robustness)
- Numerical Optimization (stochastic optimization, sparse optimization, subset selection)
- Active Learning, AutoML, and Autonomous Scientific Discovery
- Applications in Materials Science, Engineering, Recommendation Systems, and Sustainability.

## Most Cited Work
- "Unexpected Improvements to Expected Improvement for Bayesian Optimization" (NeurIPS 2023 Spotlight, 269+ citations) — introduces LogEI, now default in BoTorch/Ax
- "Autonomous materials synthesis via hierarchical active learning" (Science Advances 2021, 100+ citations) — SARA system
- "Automating crystal-structure phase mapping" (Nature Machine Intelligence 2021, 90+ citations)

## How to Cite
BibTeX entries are available via the "BibTeX" button on each publication at:
https://sebastianament.github.io/#publications

## Profiles
- Website: https://sebastianament.github.io
- Google Scholar: https://scholar.google.com/citations?user=1vkpStcAAAAJ
- GitHub: https://github.com/SebastianAment
- LinkedIn: https://www.linkedin.com/in/amentsebastian

## Key Software
- BoTorch (https://botorch.org) — Bayesian optimization library, core contributor
- Ax (https://ax.dev) — Adaptive experimentation platform at Meta

## Data
- Publications list: https://sebastianament.github.io/media/publications_scholar.json
- Citation history: https://sebastianament.github.io/media/citation_history_scholar.json