For information about the CAFE Workshop please checkout https://peter-i-frazier.github.io/cafe-workshop-website/
The tutorials and the workshop are held in rooms 315 and 325
Industry Day
Talks:
- OpenEuroLLM – Aaron Klein
- Distil Labs – Jancek Golebiowski
- DataRobot – Mark Steadman
- Meta – Eytan Bakshy
The panel on Automating Data Science will be moderated by Peter Frazier with panelist Frank Hutter (PriorLabs), Chris Oshiro (AtScale) and a representative from DataRobot.
Keynotes
Monday
- Manuela Veloso
Humans and AI and GenAI: The Journey
- Kevin Leyton-Brown
Practical, Utilitarian Algorithm Configuration
Tuesday
- Juliana Freire
Beyond Automation: The Future of AutoML is Human and Data Centered - Erin Ledell
Automated Evaluation of LLM Applications
Wednesday
- Andrew Gordon Wilson
Prescriptions for Universal Learning - Atlas Wang
Auto-Differentiate Any LLM Workflow
Presentation Schedule
Session I (Monday)
- AutoPDL: Automatic Prompt Optimization for LLM Agents
Claudio Spiess, Mandana Vaziri, Louis Mandel, Martin Hirzel
OpenReview – PDF - Hyperparameter Optimization via Interacting with Probabilistic Circuits
Jonas Seng, Fabrizio Ventola, Zhongjie Yu, Kristian Kersting
OpenReview – PDF - Feasibility-Driven Trust Region Bayesian Optimization
Paolo Ascia, Elena Raponi, Thomas Bäck, Duddeck
OpenReview – PDF - Transferrable Surrogates in Expressive Neural Architecture Search Spaces
Shiwen Qin, Gabriela Kadlecová, Martin Pilát, Shay B Cohen, Roman Neruda, Elliot J. Crowley, Jovita Lukasik, Linus Ericsson
OpenReview – PDF
Session II (Tuesday)
- CAPO: Cost-Aware Prompt Optimization
Tom Zehle, Moritz Schlager, Timo Heiß, Matthias Feurer
OpenReview – PDF – YouTube - Fast Bayesian Optimization of Function Networks with Partial Evaluations
Poompol Buathong, Peter I. Frazier
OpenReview – PDF - CATBench: A Compiler Autotuning Benchmarking Suite for Black-box Optimization
Jacob O Tørring, Carl Hvarfner, Luigi Nardi, Magnus Själander
OpenReview – PDF
Session III (Tuesday)
- Obeying the Order: Introducing Ordered Transfer Hyperparameter Optimization
Sigrid Passano Hellan, Huibin Shen, Francois-Xavier Aubet, David Salinas, Aaron Klein
OpenReview – PDF - Iterative Monte Carlo Tree Search for Neural Architecture Search
Mehraveh Javan Roshtkhari, Matthew Toews, Marco Pedersoli
OpenReview – PDF - Ax: A Platform for Adaptive Experimentation
Miles Olson, Elizabeth Santorella, Louis C. Tiao, Sait Cakmak, Mia Garrard, Samuel Daulton, Zhiyuan Jerry Lin, Sebastian Ament, Bernard Beckerman, Eric Onofrey, Paschal Igusti, Cristian Lara, Benjamin Letham, Cesar Cardoso, Shiyun Sunny Shen, Andy Chenyuan Lin, Matthew Grange, Elena Kashtelyan, David Eriksson, Maximilian Balandat, Eytan Bakshy
OpenReview – PDF - syftr: Pareto-Optimal Generative AI
Alexander Conway, Debadeepta Dey, Stefan Hackmann, Matthew Hausknecht, Michael Douglas Schmidt, Mark Lewis Steadman, Nick Volynets
OpenReview – PDF – YouTube
Session IV (Wednesday)
- confopt: A Library for Implementation and Evaluation of Gradient-based One-Shot NAS Methods
Abhash Kumar Jha, Shakiba Moradian, Arjun Krishnakumar, Martin Rapp, Frank Hutter
OpenReview – PDF - PiML: Automated Machine Learning Workflow Optimisation using LLM Agents
Abhishek Chopde , Fardeen Pettiwala, Sankar Kirubananth, Sai Kiran Botla, Pachipulusu Ayyappa Kethan
OpenReview – PDF - What Makes Freezing Layers in Deep Neural Networks Effective? A Linear Separability Perspective
Collin Coil, Nick Cheney
OpenReview – PDF – YouTube
Session V (Wednesday)
- Frozen Layers: Memory-efficient Many-fidelity Hyperparameter Optimization
Timur Carstensen, Neeratyoy Mallik, Frank Hutter, Martin Rapp
OpenReview – PDF - Multi-layer Stack Ensembles for Time Series Forecasting
Nathanael Bosch, Oleksandr Shchur, Nick Erickson, Michael Bohlke-Schneider, Ali Caner Turkmen
OpenReview – PDF - Revisiting Learning Rate Control
Micha Henheik, Theresa Eimer, Marius Lindauer
OpenReview – PDF - Regularized Neural Ensemblers
Sebastian Pineda Arango, Maciej Janowski, Lennart Purucker, Arber Zela, Frank Hutter, Josif Grabocka
OpenReview – PDF - Auto-nnU-Net: Towards Automated Medical Image Segmentation
Jannis Becktepe, Leona Hennig, Steffen Oeltze-Jafra, Marius Lindauer
OpenReveiw – PDF
Online Only (25th September 8:00 – 12:00 EST)
For information about the online post-conference gathering please checkout this page.
- AutoML Algorithms for Online Generalized Additive Model Selection: Application to Electricity Demand Forecasting
Keshav Das, Julie Keisler, Margaux Brégère, Amaury Durand
OpenReview – PDF - Overtuning in Hyperparameter Optimization
Lennart Schneider, Bernd Bischl, Matthias Feurer
OpenReview – PDF - EG-ENAS: Efficient and Generalizable Evolutionary Neural Architecture Search for Image Classification
Mateo Avila Pava, René Groh, Andreas M Kist
OpenReview – PDF – YouTube - The Ranking Trick: A Simple and Robust Alternative to Score-Based Regression for AutoML
Hernan Ceferino Vazquez, Jorge Sánchez, Verónica Bogado, Pucci Romero Tobias
OpenReveiw – PDF - SmartCal: A Novel Automated Approach to Classifier Probability Calibration
Mohamed Maher Abdelrahman, Osama Fayez Oun, Youssef Medhat, Mariam Magdy Elseedawy, Yara Mostafa Marei, Abdullah Ibrahim, Radwa Mohamed El Shawi
OpenReview – PDF – YouTube - Exploring One Million Machine Learning Pipelines: A Benchmarking Study
Edesio Alcobaça, Andre Carlos Ponce de Leon Ferreira De Carvalho
OpenReview – PDF
Poster Schedule
Tuesday
All papers from sessions I – III will present a poster in this slot. Additionally the following non-archival content will be presented:
- MetaARIMA: Automatic Configuration of ARIMA using Classifier Chains
- Quickly Tuning Foundation Models for Image Segmentation
- A Preliminary Evaluation of Large Language Models for Data Science Code Generation
- Cost-aware Stopping for Bayesian Optimization
- Stitching Disparate Neural Network Layers with Complex Adapters and Spatial Rescaling
- Successive Halving with Learning Curve Prediction via Latent Kronecker Gaussian Processes
- Improved Gaussian Process Hyperparameter Fitting for Bayesian Optimization
- LISTEN to Your Preferences: An LLM Framework for Multi-Objective Selection
- Cost-aware Bayesian Optimization via the Pandora’s Box Gittins Index
Wednesday
All papers from Sessions IV & V will present a poster in this slot. Additionally the following non-archival content will be presented:
- Quantifying Module Interactions in the PSO-X Framework
- Bayesian Optimisation Against Climate Change: Applications and Benchmarks
- Zero-Cost Benchmarks: Towards Lower Reliance on Spearman Rank Correlation
- Stress Testing Classifiers around the Decision Boundary with Latent Diffusion
- The Gittins Index: A Design Principle for Decision-Making Under Uncertainty
- Tune My Adam, Please!
- Algorithm Configuration for Structured Pfaffian Settings
- Multi-objective Hyperparameter Optimization in the Age of Deep Learning
- How Usable is Automated Feature Engineering for Tabular Data?
- Automated Data Preparation for Machine Learning
- Surrogate Benchmarks for Model Merging Optimization
- ParticleML: AutoML Through Electromagnetic Physics Simulation
- Prometheus: A Recursively Self-Improving NAS System
- Data-Efficient Ranking of Recommendation Models
- Object-Flow Machine Learning: Active learning framework utilizing protocols information
Online Only (September 25th, 8:00 – 12:00 EST)
All papers that could not be presented in person will be presented online at the post-conference gathering. Additionally the following non-archival content will be presented there as well.
For information about the online post-conference gathering please checkout
this page.
- ReLU is all you need for NASWOT
- Towards Exploiting Early Termination for Multi-Fidelity Hyperparameter Optimization
- Towards Dynamic Priors in Bayesian Optimization for Hyperparameter Optimization
Tutorials
- Slot I
Limitations of State-of-the-Art and a New Principled Framework for HPO and Algorithm Selection
Dravyansh Sharma
- Slot II
LLM-Driven Algorithm Discovery & Tuning
Elena Raponi, Tomas Kadavy, Jozef Kovac
- Slot III
Ax for AutoML: From Research to Production
Miles Olson and Mia Garrard
- Slot IV
AutoML in the Age of Structured Foundation Models
Frank Hutter and Nick Erickson