frontpage.
newsnewestaskshowjobs

Made with ♥ by @iamnishanth

Open Source @Github

fp.

Open in hackernews

Show HN: The Legal Embedding Benchmark (MLEB)

https://huggingface.co/blog/isaacus/introducing-mleb
9•ubutler•1w ago
Hey HN,

I'm excited to share the Massive Legal Embedding Benchmark (MLEB) — the first comprehensive benchmark for legal embedding models.

Unlike previous legal retrieval datasets, MLEB was created by someone with actual domain expertise (I have a law degree and previously led the AI team at the Attorney-General's Department of Australia).

I came up with MLEB while trying to train my own state-of-the-art legal embedding model. I found that there were no good benchmarks for legal information retrieval to evaluate my model on.

That led me down a months-long process working alongside my brother to identify or, in many cases, build our own high-quality legal evaluation sets.

The final product was 10 datasets spanning multiple jurisdictions (the US, UK, Australia, Singapore, and Ireland), document types (cases, laws, regulations, contracts, and textbooks), and problem types (retrieval, zero-shot classification, and QA), all of which have been vetted for quality, diversity, and utility.

For a model to do well at MLEB, it needs to have both extensive legal domain knowledge and strong legal reasoning skills. That is deliberate — given just how important high-quality embeddings are to legal RAG (particularly for reducing hallucinations), we wanted our benchmark to correlate as strongly as possible with real-world usefulness.

The dataset we are most proud of is called Australian Tax Guidance Retrieval. It pairs real-life tax questions posed by Australian taxpayers with relevant Australian Government guidance and policy documents.

We constructed the dataset by sourcing questions from the Australian Taxation Office's community forum, where Australian taxpayers ask accountants and ATO officials their tax questions.

We found that, in most cases, such questions can be answered by reference to government web pages that, for whatever reason, users were unable to find themselves. Accordingly, we manually went through a stratified sample of 112 challenging forum questions and extracted relevant portions of government guidance materials linked to by tax experts that we verified to be correct.

What makes the dataset so valuable is that, unlike the vast majority of legal information retrieval evaluation sets currently available, it consists of genuinely challenging real-world user-created questions, rather than artificially constructed queries that, at times, diverge considerably from the types of tasks embedding models are actually used for.

Australian Tax Guidance Retrieval is just one of several other evaluation sets that we painstakingly constructed ourselves simply because there weren't any other options.

We've contributed everything, including the code used to evaluate models on MLEB, back to the open-source community.

Our hope is that MLEB and the datasets within it will hold value long into the future so that others training legal information retrieval models won't have to detour into building their own "MTEB for law".

If you'd like to head straight to the leaderboard instead of reading our full announcement, you can find it here: https://isaacus.com/mleb

If you're interested in playing around with our model, which happens to be ranked first on MLEB as of 16 October 2025 at least, check out our docs: https://docs.isaacus.com/quickstart

Ratatui – App Showcase

https://ratatui.rs/showcase/apps/
145•AbuAssar•4h ago•59 comments

Solarpunk is happening in Africa

https://climatedrift.substack.com/p/why-solarpunk-is-already-happening
715•JoiDegn•11h ago•344 comments

End of Japanese community

https://support.mozilla.org/en-US/forums/contributors/717446
410•phantomathkg•4h ago•233 comments

Dillo, a multi-platform graphical web browser

https://github.com/dillo-browser/dillo
294•nazgulsenpai•12h ago•95 comments

ChatGPT terms disallow its use in providing legal and medical advice to others

https://www.ctvnews.ca/sci-tech/article/openai-updates-policies-so-chatgpt-wont-provide-medical-o...
276•randycupertino•13h ago•260 comments

Recursive macros in C, demystified (once the ugly crying stops)

https://h4x0r.org/big-mac-ro-attack/
63•eatonphil•6h ago•35 comments

Firefox profiles: Private, focused spaces for all the ways you browse

https://blog.mozilla.org/en/firefox/profile-management/
195•darkwater•1w ago•92 comments

Why aren't smart people happier?

https://www.theseedsofscience.pub/p/why-arent-smart-people-happier
287•zdw•14h ago•375 comments

The state of SIMD in Rust in 2025

https://shnatsel.medium.com/the-state-of-simd-in-rust-in-2025-32c263e5f53d
186•ashvardanian•12h ago•96 comments

New gel restores dental enamel and could revolutionise tooth repair

https://www.nottingham.ac.uk/news/new-gel-restores-dental-enamel-and-could-revolutionise-tooth-re...
448•CGMthrowaway•11h ago•177 comments

NY school phone ban has made lunch loud again

https://gothamist.com/news/ny-smartphone-ban-has-made-lunch-loud-again
275•hrldcpr•17h ago•201 comments

Ruby and Its Neighbors: Smalltalk

https://noelrappin.com/blog/2025/11/ruby-and-its-neighbors-smalltalk/
188•jrochkind1•15h ago•102 comments

Vacuum bricked after user blocks data collection – user mods it to run anyway

https://www.tomshardware.com/tech-industry/big-tech/manufacturer-issues-remote-kill-command-to-nu...
240•toomanyrichies•4d ago•73 comments

The Basic Laws of Human Stupidity (1987) [pdf]

https://gandalf.fee.urv.cat/professors/AntonioQuesada/Curs1920/Cipolla_laws.pdf
68•bookofjoe•8h ago•29 comments

Scientists growing colour without chemicals

https://www.forbes.com/sites/maevecampbell/2025/06/20/dyeing-for-fashion-meet-the-scientists-grow...
10•caiobegotti•4d ago•2 comments

Carice TC2 – A non-digital electric car

https://www.caricecars.com/
214•RubenvanE•16h ago•156 comments

I was right about dishwasher pods and now I can prove it [video]

https://www.youtube.com/watch?v=DAX2_mPr9W8
379•hnaccount_rng•1d ago•266 comments

The shadows lurking in the equations

https://gods.art/articles/equation_shadows.html
267•calebm•16h ago•84 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/
101•hentrep•2h ago•57 comments

I want a good parallel language [video]

https://www.youtube.com/watch?v=0-eViUyPwso
65•raphlinus•2d ago•37 comments

Brain-IT: Image Reconstruction from fMRI via Brain-Interaction Transformer

https://AmitZalcher.github.io/Brain-IT/
20•SerCe•4h ago•0 comments

Radiant Computer

https://radiant.computer
189•beardicus•17h ago•136 comments

A new oral history interview with Ken Thompson

https://computerhistory.org/blog/a-computing-legend-speaks/
6•oldnetguy•5d ago•0 comments

The MDL ("Muddle") Programming Language (1979) [pdf]

http://bitsavers.informatik.uni-stuttgart.de/pdf/mit/lcs/tr/MIT-LCS-TR-0293_MDL_Pgmg_Lang.pdf
9•twoodfin•6d ago•2 comments

A Lost IBM PC/at Model? Analyzing a Newfound Old Bios

https://int10h.org/blog/2025/11/lost-ibm-at-model-bios-analysis/
77•TMWNN•10h ago•13 comments

An eBPF Loophole: Using XDP for Egress Traffic

https://loopholelabs.io/blog/xdp-for-egress-traffic
220•loopholelabs•1d ago•70 comments

App Store web has exposed all its source code

https://www.reddit.com/r/webdev/comments/1onnzlj/app_store_web_has_exposed_all_its_source_code/
210•redbell•2d ago•74 comments

Absurd Workflows: Durable Execution with Just Postgres

https://lucumr.pocoo.org/2025/11/3/absurd-workflows/
116•ingve•2d ago•24 comments

SPy: An interpreter and compiler for a fast statically typed variant of Python

https://antocuni.eu/2025/10/29/inside-spy-part-1-motivations-and-goals/
252•og_kalu•6d ago•114 comments

Gloomth

https://www.lrb.co.uk/the-paper/v47/n20/jon-day/gloomth
8•prismatic•6d ago•0 comments