RESEARCH / SELECTED PUBLICATIONS

Publications.

Work on process intelligence, forecasting, uncertainty, and the foundations of machine learning.

Full Google Scholar profile

SELECTED WORK / 2019—2025Open publication ↗

2025

ICPM 2025 · Poster Track

ProAcT: Process Activity Tokenization in Large Language Models

Greg Benton, Yamini Ananth, Sean Jaffe, Jeff Naughton

DBPM Workshop, SIGMOD 2025

Disentangling Event Logs With Large Language Models

Gregory Benton, Lampros Flokas, Cong Yu

2024

ICML SPIGM Workshop 2024

SOLMformer: Incorporating Sequence and Observation Level Metadata for Categorical Time Series Modeling

Yamini Ananth, Greg Benton, Jingxing Fang, Jerry Cheung, Xu Chu, Cong Yu

ICPM Demo Track 2024

Building Process Mining Dashboards Manually or AI-assisted with Celonis Views

Carolin Ullrich, Teodora Lata, Greg Benton

2023

Ph.D. Dissertation · New York University

Function Space Reasoning for Gaussian Processes and Neural Networks

Gregory Benton · Advisor: Andrew Gordon Wilson

Preprint 2023

Deep Probabilistic Time Series Forecasting over Long Horizons

Greg Benton, Nate Gruver, Wesley Maddox, Andrew Gordon Wilson

2022

ICML 2022 · Outstanding Paper Award

Bayesian Model Selection, the Marginal Likelihood, and Generalization

Sanae Lotfi, Pavel Izmailov, Greg Benton, Micah Goldblum, Andrew Gordon Wilson

ICML 2022

Volatility Based Kernels and Moving Average Means for Accurate Forecasting with Gaussian Processes

Greg Benton, Wesley Maddox, Andrew Gordon Wilson

2021

NeurIPS 2021

Residual Pathway Priors for Soft Equivariance Constraints

Marc Finzi, Greg Benton, Andrew Gordon Wilson

ICML 2021

Loss Surface Simplexes for Mode Connecting Volumes and Fast Ensembling

Greg Benton, Wesley Maddox, Sanae Lotfi, Andrew Gordon Wilson

2020

NeurIPS 2020

Learning Invariances in Neural Networks

Greg Benton, Marc Finzi, Pavel Izmailov, Andrew Gordon Wilson

Preprint 2020

Rethinking Parameter Counting in Deep Models: Effective Dimensionality Revisited

Wesley Maddox, Greg Benton, Andrew Gordon Wilson

2019

NeurIPS 2019 · ICML Time Series Workshop Best Paper

Function-Space Distributions over Kernels

Greg Benton, Wesley Maddox, Jayson Salkey, Julio Albinati, Andrew Gordon Wilson

For the complete and most current publication record, visit Google Scholar.

View the full profile