frontpage.
newsnewestaskshowjobs

Made with ♥ by @iamnishanth

Open Source @Github

fp.

Open in hackernews

Show HN: TabPFN-2.5 – SOTA foundation model for tabular data

https://priorlabs.ai/technical-reports/tabpfn-2-5-model-report
48•onasta•3h ago
I am excited to announce the release of TabPFN-2.5, our tabular foundation model that now scales to datasets of up to 50,000 samples and 2,000 features - a 5x increase from TabPFN v2, published in the Nature journal earlier this year. TabPFN-2.5 delivers state-of-the-art predictions in one forward pass without hyperparameter tuning across classification and regression tasks.

What’s new in 2.5: TabPFN-2.5 maintains the core approach of v2 - a pretrained transformer trained on more than hundred million synthetic datasets to perform in-context learning and output a predictive distribution for the test data. It natively supports missing values, cateogrical features, text and numerical features is robust to outliers and uninformative features.

The major improvements:

- 5x scale increase: Now handles 50,000 samples × 2,000 features (up from 10,000 × 500 in v2)

- SOTA performance: TabPFN-2.5 outperforms tuned tree-based methods and matches the performance of a complex ensemble (AutoGluon 1.4), that itself includes TabPFN v2, tuned for 4 hours. Tuning the model improves performance, outperforming AutoGluon 1.4 for regression tasks.

- Rebuilt API: New REST interface along with Python SDK with dedicated fit & predict endpoints, making deployment and integration more developer-friendly

- A distillation engine that converts TabPFN-2.5 into a compact MLP or tree ensemble while preserving accuracy and offer low latency inference.

There are still some limitations. The model is designed for datasets up to 50K samples. It can handle larger datasets but that hasn’t been our focus with TabPFN-2.5. The distillation engine is not yet available through the API but only through licenses (though we do show the performance in the model report).

We’re actively working on removing these limitations and intend to release newer models focused on context reasoning, causal inference, graph networks, larger data and time-series. TabPFN-2.5 is available via API and a package on Hugging Face. Would love for you to try it and give us your feedback!

Model report: https://priorlabs.ai/technical-reports/tabpfn-2-5-model-repo...

Package: https://github.com/PriorLabs/TabPFN

Client: https://github.com/PriorLabs/tabpfn-client

Docs: https://docs.priorlabs.ai/quickstart

Comments

klemens_floege•2h ago
Good stuff!
zurfer•2h ago
The current go to solution for the kinds of problems that TabPFN is solving would be something like XGBoost. In general it's a good baseline, but the challenge is always that you need to spend a lot of time feature engineering and tweaking the data representation before something like XGBoost can deliver good performance on your regression or classification problems.

For me the promise of foundation models for tabular data is that there are enough generalizable patterns, so that you need less manual feature engineering and data cleaning.

And kudos to the team, I think it's a really creative application of neural networks. I was always frustrated with neural networks, since they were hard to tune on "structured" data and always under-performed (for me), but we also never had real foundational models for structured data.

noahho•1h ago
Less feature engineering is definitely something we are aiming for. The current version is actually only based on statistics, the real world connections between features is something we're working on right now and hope to show results for soon. That's the next step
dill_1•2h ago
Tabular data is still underrated!
noahho•1h ago
When we released TabPFNv1 over three years ago, I didn’t expect at all the hundreds of comments and reposts we would see. Tabular data had been a field getting little love from AI research—but we immediately felt that this was a topic that data scientists, scientists, financial analysts, and enterprise users deeply cared about. Glad its useful to people!
abracos•2h ago
how does it compare to automl tools?
noahho•2h ago
TabPFN-2.5 default (one forward pass) matches AutoGluon 1.4 tuned for four-hours. Autogluon is the strongest AutoML including stacking of XGB and cat boost and even includes the previous TabPFNv2.
TheTaytay•1h ago
Looks really cool. In reading through the FAQ, it says this: Q: "How are text features handled?" A: "In the local package version text features are encoded as categoricals without considering their semantic meaning. Our API automatically detects text features and includes their semantic meaning into our prediction. The local package version encodes text as numerical categories and does not include semantic meaning."

So that means that automatic embedding/semantic meaning is reserved for API use of TabPFN, right? Otherwise, if I use it locally, it's going to assign each of my distinct text values an arbitrary int, right?

noahho•1h ago
Yes exactly, the API is the best way to handle text features. The actual semantics often matter a lot . Is the API an option for you or would you need this local?
vessenes•1h ago
I think you need a custom benchmark -- have you considered making one out of the excel world championships?
scorpion7•35m ago
It's fascinating how this works with such a small model. Especially given that the training is a kind of meta learning of "how to do in-context learning". I wonder, is there a good intuition of the role of the MLP in this architecture? For LLMs the consensus seems to be that they store knowledge...what would that be for tabular data?

Show HN: I scraped 3B Goodreads reviews to train a better recommendation model

https://book.sv
95•costco•1d ago•47 comments

Show HN: TabPFN-2.5 – SOTA foundation model for tabular data

https://priorlabs.ai/technical-reports/tabpfn-2-5-model-report
48•onasta•3h ago•11 comments

Show HN: Dynamic code and feedback walkthroughs with your coding Agent in VSCode

https://www.intraview.ai/hn-demo
10•cyrusradfar•5h ago•0 comments

Show HN: See chords as flags – Visual harmony of top composers on musescore

https://rawl.rocks/
100•vitaly-pavlenko•1d ago•27 comments

Show HN: qqqa – A fast, stateless LLM-powered assistant for your shell

https://github.com/matisojka/qqqa
107•iagooar•10h ago•76 comments

Show HN: Completely free Claude Sonnet 4.5, supported by contextual ads

3•namanyayg•48m ago•1 comments

Show HN: Flutter_compositions: Vue-inspired reactive building blocks for Flutter

https://github.com/yoyo930021/flutter_compositions
39•yoyo930021•15h ago•12 comments

Show HN: DIY accessibility mouse helps people even with complete paralysis

https://github.com/aradzhabov/AbleMouse
3•aradzhabov•3h ago•0 comments

Show HN: What Is Hacker News Working On?

https://waywo.eamag.me/
6•eamag•3h ago•1 comments

Show HN: ShellAI – Local Terminal Assistance with SLM

https://github.com/micrictor/shellai
5•mtud•5h ago•2 comments

Show HN: CoordConversions NPM Module for Map Coordinate Conversions

https://github.com/matthewcsimpson/CoordConversions
8•smatthewaf•1w ago•1 comments

Show HN: A CSS-Only Terrain Generator

https://terra.layoutit.com
358•rofko•2d ago•82 comments

Show HN: fx - A (micro)blogging server that you can self-host

https://github.com/rikhuijzer/fx
3•huijzer•11h ago•0 comments

Show HN: The Legal Embedding Benchmark (MLEB)

https://huggingface.co/blog/isaacus/introducing-mleb
11•ubutler•1w ago•0 comments

Show HN: I made a MVP development agency and want your feedback on the site

https://mintmvp.com/
2•alwassikhan•13h ago•2 comments

Show HN: Standards compliant SAX-style XML parser for JavaScript

https://github.com/federicocarboni/saxe
3•federicocarboni•14h ago•0 comments

Show HN: Data Formulator 0.5 – Vibe with your data (Microsoft Research)

https://data-formulator.ai/
4•chenglong-hn•14h ago•0 comments

Show HN: SixSevenStudio – open-source Video Editor For Sora

https://github.com/palmier-io/sixsevenstudio
6•hchtin•1d ago•1 comments

Show HN: a Rust ray tracer that runs on any GPU – even in the browser

https://github.com/tchauffi/rust-rasterizer
94•tchauffi•3d ago•29 comments

Show HN: SSH terminal multiplayer written in Golang

https://github.com/MShel/sshOuroboros
2•mishk0sh•18h ago•0 comments

Show HN: Strange Attractors

https://blog.shashanktomar.com/posts/strange-attractors
795•shashanktomar•5d ago•77 comments

Show HN: Tamagotchi P1 for FPGAs

https://github.com/agg23/fpga-tamagotchi
68•agg23•1w ago•11 comments

Show HN: Dynamic Code Execution with MCP: A More Efficient Approach

https://github.com/pranftw/aiter-app
2•pranftw•20h ago•0 comments

Show HN: MyTimers.app offline-first PWA with no build step and zero dependencies

https://mytimers.app/
46•y3k•2d ago•28 comments

Show HN: JermCAD – A YAML-powered, vibe-coded, browser-based CAD software

https://github.com/jeremyaboyd/jerm-cad
4•jermaustin1•1d ago•0 comments

Show HN: Wosp – advanced full-text search on the command line

https://github.com/atrettel/wosp
7•atrettel•1d ago•0 comments

Show HN: Pipelex – Declarative language for repeatable AI workflows

https://github.com/Pipelex/pipelex
122•lchoquel•1w ago•27 comments

Show HN: FinBodhi – Local-first, double-entry app/PWA for your financial journey

https://finbodhi.com/
40•ciju•3d ago•21 comments

Show HN: In a single HTML file, an app to encourage my children to invest

https://roberdam.com/en/dinversiones.html
247•roberdam•1w ago•436 comments

Show HN: Why write code if the LLM can just do the thing? (web app experiment)

https://github.com/samrolken/nokode
434•samrolken•5d ago•320 comments