leogrin.bsky.social @leogrin.bsky.social · 09/01/2025Some limitations remain (mainly 10K rows x 500 cols limit and inference speed) but this new in-context learning paradigm will only get much faster & better from there! I've joined this great team at @prior_labs as founding researcher 🚀 Join us! www.notion.so/priorlabs/Pr...notion.soPrior Labs - Join us! | NotionWe're building breakthrough foundation models that understand spreadsheets and databases. Join us in transforming data science! 000
leogrin.bsky.social @leogrin.bsky.social · 09/01/2025Groundbreaking work, congrats to the team!! 🎉 When I started my PhD 3 years ago, our tabular benchmark showed tree-based models miles ahead of neural networks. On the same benchmark, TabPFN v2 now reaches in 10s what CatBoost achieves in 4h of tuning 🤯 141
Reposted by @leogrin.bsky.socialDavid Holzmüller @dholzmueller.bsky.social · 18/11/2024Can deep learning finally compete with boosted trees on tabular data? 🌲 In our NeurIPS 2024 paper, we introduce RealMLP, a NN with improvements in all areas and meta-learned default parameters. Some insights about RealMLP and other models on large benchmarks (>200 datasets): 🧵 1608
