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
43•onasta•2h 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•1h ago
Good stuff!
zurfer•1h 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•52m 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•1h ago
Tabular data is still underrated!
noahho•55m 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•1h ago
how does it compare to automl tools?
noahho•1h 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•51m 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•28m 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•20m ago
I think you need a custom benchmark -- have you considered making one out of the excel world championships?

Two billion email addresses were exposed

https://www.troyhunt.com/2-billion-email-addresses-were-exposed-and-we-indexed-them-all-in-have-i...
69•esnard•42m ago•38 comments

Kimi K2 Thinking, a SOTA open-source trillion-parameter reasoning model

https://moonshotai.github.io/Kimi-K2/thinking.html
420•nekofneko•5h ago•155 comments

Swift on FreeBSD Preview

https://forums.swift.org/t/swift-on-freebsd-preview/83064
130•glhaynes•3h ago•70 comments

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

https://book.sv
46•costco•1d ago•16 comments

ICC ditches Microsoft 365 for openDesk

https://www.binnenlandsbestuur.nl/digitaal/internationaal-strafhof-neemt-afscheid-van-microsoft-365
401•vincvinc•4h ago•122 comments

LLMs Encode How Difficult Problems Are

https://arxiv.org/abs/2510.18147
52•stansApprentice•2h ago•3 comments

Open Source Implementation of Apple's Private Compute Cloud

https://github.com/openpcc/openpcc
313•adam_gyroscope•1d ago•58 comments

The Parallel Search API

https://parallel.ai/blog/introducing-parallel-search
48•lukaslevert•3h ago•26 comments

What if hard work felt easier?

https://jeanhsu.substack.com/p/what-if-hard-work-felt-easier
43•kiyanwang•1w ago•24 comments

I analyzed the lineups at the most popular nightclubs

https://dev.karltryggvason.com/how-i-analyzed-the-lineups-at-the-worlds-most-popular-nightclubs/
123•kalli•7h ago•61 comments

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

https://priorlabs.ai/technical-reports/tabpfn-2-5-model-report
43•onasta•2h ago•10 comments

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

https://www.intraview.ai/hn-demo
9•cyrusradfar•4h ago•0 comments

Eating stinging nettles

https://rachel.blog/2018/04/29/eating-stinging-nettles/
138•rzk•9h ago•144 comments

Senior BizOps at Artie (San Francisco)

https://www.ycombinator.com/companies/artie/jobs/gqANVBc-senior-business-operations
1•tang8330•4h ago

FBI tries to unmask owner of archive.is

https://www.heise.de/en/news/Archive-today-FBI-Demands-Data-from-Provider-Tucows-11066346.html
504•Projectiboga•4h ago•277 comments

C++: A prvalue is not a temporary

https://blog.knatten.org/2025/10/31/a-prvalue-is-not-a-temporary/
5•ingve•6d ago•2 comments

Springs and Bounces in Native CSS

https://www.joshwcomeau.com/animation/linear-timing-function/
47•Bogdanp•1w ago•5 comments

Auraphone: A simple app to collect people's info at events

https://andrewarrow.dev/2025/11/simple-app-collect-peoples-info-at-events/
11•fcpguru•6h ago•7 comments

UK outperforms US in creating unicorns from early stage VC investment

https://www.cityam.com/uk-outperforms-us-in-creating-unicorns-from-early-stage-vc-investment/
23•mmarian•1h ago•16 comments

Benchmarking the Most Reliable Document Parsing API

https://www.tensorlake.ai/blog/benchmarks
19•calavera•2h ago•13 comments

Mathematical exploration and discovery at scale

https://terrytao.wordpress.com/2025/11/05/mathematical-exploration-and-discovery-at-scale/
204•nabla9•11h ago•91 comments

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

https://rawl.rocks/
94•vitaly-pavlenko•1d ago•26 comments

Black Hole Flare Is Biggest and Most Distant Seen

https://www.caltech.edu/about/news/black-hole-flare-is-biggest-and-most-distant-seen
8•gmays•2h ago•2 comments

Supply chain attacks are exploiting our assumptions

https://blog.trailofbits.com/2025/09/24/supply-chain-attacks-are-exploiting-our-assumptions/
31•crescit_eundo•5h ago•22 comments

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

https://github.com/matisojka/qqqa
102•iagooar•10h ago•74 comments

Cloudflare Tells U.S. Govt That Foreign Site Blocking Efforts Are Trade Barriers

https://torrentfreak.com/cloudflare-tells-u-s-govt-that-foreign-site-blocking-efforts-are-digital...
266•iamnothere•7h ago•161 comments

IKEA launches new smart home range with 21 Matter-compatible products

https://www.ikea.com/global/en/newsroom/retail/the-new-smart-home-from-ikea-matter-compatible-251...
240•lemoine0461•7h ago•173 comments

I may have found a way to spot U.S. at-sea strikes before they're announced

https://old.reddit.com/r/OSINT/comments/1opjjyv/i_may_have_found_a_way_to_spot_us_atsea_strikes/
233•hentrep•16h ago•326 comments

Pico-100BASE-TX: Bit-Banged 100 MBit/s Ethernet and UDP Framer for RP2040/RP2350

https://github.com/steve-m/Pico-100BASE-TX
72•_Microft•6d ago•13 comments

How I am deeply integrating Emacs

https://joshblais.com/blog/how-i-am-deeply-integrating-emacs/
190•signa11•13h ago•127 comments