Hacker News: “TabPFN and TabICL vs. tuned XGBoost: the model that doesn't train won…” · TGViewer

TGViewer Telegram Channel Viewer

Hacker News @hackernewslive · 28.4K subscribers

Post #232984 1.93K Sep 28, 2026, 03:42 UTC

TabPFN and TabICL vs. tuned XGBoost: the model that doesn't train won 14/14
Article, Comments

Efrain Garay TabPFN vs XGBoost: benchmark measured on an RTX 4070 Ti The claim behind TabPFN and TabICL is that they predict on a table without ever training on it and still beat tuned boosting.

👎 5
🌭 1

Open in Telegram All posts from @hackernewslive

More from @hackernewslive

Sep 29, 2026AI needs $6T in annual revenue to justify data centre boom Article, Comments
Sep 29, 2026Climate is accumulating 4 Hiroshima atomic bombs worth of heat per second Article, Comment…
Sep 29, 2026Nicholas Polson has authored 258 academic papers in 2026 (so far) Article, Comments
Sep 29, 2026Show HN: Jevstiller – Distill Jev into a local model, with a disagreement bound Article, C…
Sep 29, 2026New PlayStation 5 Console Jailbreak Released Article, Comments
Sep 29, 2026DevDay 2026 Recap Article, Comments

Threads Profile ViewerView any public Threads profile without an account.Open ThreadLook →Writing with AI? Make it sound human.Metric37 rewrites AI drafts so they read naturally. Free AI detector, 1,500 words free.Try Metric37 →