Lennart Purucker
@LennartPurucker
PhD student @AutoML_org sup. by @FrankRHutter; researching automated machine learning and foundation models for (small) tabular data!
🚨What is SOTA on tabular data, really? We are excited to announce 𝗧𝗮𝗯𝗔𝗿𝗲𝗻𝗮, a living benchmark for machine learning on IID tabular data with: 📊 an online leaderboard (submit!) 📑 carefully curated datasets 📈 strong tree-based, deep learning, and foundation models 🧵
I am at #ICML this week (and in 10 minutes), let me know if you want to chat!
I will talk about TabArena at the AutoML seminar this week, join me :) More information: automl-seminars.github.io
Did you ever want to reason over predictions for tabular data (or any other modality)? With MARVIS, we might be one step closer! This may be our first step toward making VLMs competitive (many-shot) predictors that understand text and enable reasoning for any modality 🚀
New research paper for you to read over your July 4th break (if you're US-based) -- Vision is a skeleton key! 🗝️ We convert a small VLM into an "everything classifier" by transforming data into visualizations that VLMs can naturally understand and reason about. We call it…
📢 We are excited to announce "#FMSD: 1st Workshop on Foundation Models for Structured Data" has been accepted to #ICML 2025! Call for Papers: icml-structured-fm-workshop.github.io/call-for-paper…
The tabular foundation model TabPFN v2 is finally public 🎉🥳 This is excellent news for (small) tabular ML! Checkout our Nature article (nature.com/articles/s4158…) and code (github.com/PriorLabs/TabP…)
The data science revolution is getting closer. TabPFN v2 is published in Nature: nature.com/articles/s4158… On tabular classification with up to 10k data points & 500 features, in 2.8s TabPFN on average outperforms all other methods, even when tuning them for up to 4 hours🧵1/19
Excited to be at #NeurIPS2024 tomorrow! 🎉 Let’s connect if you are into tabular data: 🤖 AutoML (e.g., AutoGluon) 📊 Data Science (e.g., LLMs for Feature Engineering) 🏛️ Foundation Models (e.g., TabPFN) Looking forward to insightful discussions—feel free to reach out!
I am excited to be presenting HW-GPT-Bench at #NeurIPS2024 in Vancouver this week. HW-GPT-Bench is a hardware-aware and calibrated architecture benchmark for language models. #NAS #LanguageModels #HardwareOptimization
🚀 AutoGluon 1.2 is here! Key highlights: 1. Tabular & TimeSeries: 70% win-rate over v1.1. 2. New Features: TabPFNMix foundation model, parallel fit, Chronos-Bolt (250x faster + accurate). 3. AutoGluon-Assistant: Zero-code ML with LLMs. #AutoML tinyurl.com/autogluon-v120