• Home
  • Technology
  • Gaming
  • Entertainment
  • World & Business
  • Science
  • Sports
  • AI
HomeTechnologyGamingEntertainmentWorld & BusinessScienceSportsAI
  • HomeTechnologyGamingEntertainmentWorld & BusinessScienceSportsAI
    • Home
    • Technology
    • Gaming
    • Entertainment
    • World & Business
    • Science
    • Sports
    • AI
    AI

    New Episode of The Information Bottleneck Released

    Interview with David Holzmüller covers TabArena and tabular foundation models.

    RS
    1 Source, 27d ago, first seen 27d ago

    TLDR

    Ravid Shwartz-Ziv posted that a new episode of The Information Bottleneck is out. The episode features David Holzmüller of INRIA, identified as a contributor to TabArena, TabICL, and RealMLP. According to the post the discussion examines the present position of tabular machine learning and states that tabular foundation models have become the default approach, along with coverage of TabArena and Kaggle competitions.

    Combined views

    926

    1 Source, first seen 27d ago

    Combined views

    926

    1 Source, first seen 27d ago

    10 likes
    10 likes
    1 saves
    4 reposts

    Sentiment

    Positive——Negative

    Summary

    Not enough discussion yet.

    No sentiment analysis available yet.

    1 saves
    4 reposts

    Sentiment

    Positive——Negative

    Summary

    Not enough discussion yet.

    No sentiment analysis available yet.

    Today's Rank

    —

    Not ranked yet

    Today's Rank

    —

    Not ranked yet

    1 Source

    @ziv_ravidNew episode of The Information Bottleneck is out!! This week we talked with David Holzmüller (INRIA) (@DHolzmueller), one of the people behind TabArena, TabICL, and RealMLP, about where tabular ML actually stands right now. We talked about how tabular foundation models are now the default answer, TabArena, Kaggle competitions, and why LLMs are surprisingly bad at tables. l We also got into calibration, text embeddings for tabular models, time series vs tabular, and what he thinks the open problems are. You can find it on our website, YouTube, and all the apps Love to hear your comments!

    1 Source

    @ziv_ravidNew episode of The Information Bottleneck is out!! This week we talked with David Holzmüller (INRIA) (@DHolzmueller), one of the people behind TabArena, TabICL, and RealMLP, about where tabular ML actually stands right now. We talked about how tabular foundation models are now the default answer, TabArena, Kaggle competitions, and why LLMs are surprisingly bad at tables. l We also got into calibration, text embeddings for tabular models, time series vs tabular, and what he thinks the open problems are. You can find it on our website, YouTube, and all the apps Love to hear your comments!