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
2•ubutler•3h 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

Show HN: MyraOS – My 32-bit operating system in C and ASM (Hack Club project)

https://github.com/dvir-biton/MyraOS
97•dvirbt•5h ago•9 comments

Show HN: Helium Browser for Android with extensions support, based on Vanadium

https://github.com/jqssun/android-helium-browser
23•jqssun•3h ago•7 comments

Show HN: Diagram as code tool with draggable customizations

https://github.com/RohanAdwankar/oxdraw
243•RohanAdwankar•1d ago•51 comments

Show HN: The Legal Embedding Benchmark (MLEB)

https://huggingface.co/blog/isaacus/introducing-mleb
2•ubutler•3h ago•0 comments

Show HN: Create-LLM – Train your own LLM in 60 seconds

https://github.com/theaniketgiri/create-llm
38•theaniketgiri•16h ago•12 comments

Show HN: Shadcn/UI theme editor – Design and share Shadcn themes

https://shadcnthemer.com
127•miketromba•1d ago•39 comments

Show HN: Chonky – a neural text semantic chunking goes multilingual

https://huggingface.co/mirth/chonky_mmbert_small_multilingual_1
40•hessdalenlight•1d ago•4 comments

Show HN: LLM Rescuer – Fixing the billion dollar mistake in Ruby

https://github.com/barodeur/llm_rescuer
89•barodeur•2d ago•14 comments

Show HN: I Built DevTools for Blazor (Like React DevTools but for .NET)

https://blazordevelopertools.com/
7•joe-gregory•9h ago•1 comments

Show HN: AI bookmarking app for people who hate AI

https://tryeyeball.com/
3•quinto_quarto•10h ago•0 comments

Show HN: Guided EMDR Therapy App to Heal Inner Trauma

https://myemdr.app/start
2•positive-minds•11h ago•0 comments

Show HN: Random Makers – Show HN and Product Hunt, but Faster and Not Corporate

https://makers.random.gg/
16•waynerd•1d ago•1 comments

Show HN: A browser SIM of a decentralized court for autonomous AI agents

https://aethelred.ayauho.com/
3•Ohuaya•12h ago•1 comments

Show HN: MacOS Live Screensaver – A screensaver that plays live video streams

https://github.com/hauxir/macos-live-screensaver
62•hauxir•5d ago•40 comments

Show HN: Git for LLMs – A context management interface

https://twigg.ai
101•jborland•3d ago•36 comments

Show HN: Deta Surf – An open source and local-first AI notebook

https://github.com/deta/surf
137•mxek•3d ago•39 comments

Show HN: Tommy – Turn ESP32 devices into through-wall motion sensors

https://www.tommysense.com
104•mike2872•3d ago•78 comments

Show HN: Status of my favorite bike share stations

https://blog.alexboden.ca/toronto-bike-share-status/
13•alexboden•1d ago•5 comments

Show HN: A fast, privacy-first image converter that runs in browser

https://imageconverter.dev/
45•wainguo•2d ago•38 comments

Show HN: OpenSnowcat – A fork of Snowplow to keep open analytics alive

https://opensnowcat.io/
75•joaocorreia•3d ago•18 comments

Show HN: Nostr Web – decentralized website hosting on Nostr

https://nweb.shugur.com
101•karihass•3d ago•34 comments

Show HN: Centia.io – Open PostgreSQL/PostGIS back end for developers

https://centia.io/
6•mhoegh•1d ago•0 comments

Show HN: Dictly – Local, real‑time voice‑to‑text for macOS (sub‑100ms, no cloud)

https://dictly.app/
8•JannikJung•1d ago•2 comments

Show HN: Cuq – Formal Verification of Rust GPU Kernels

https://github.com/neelsomani/cuq
94•nsomani•4d ago•63 comments

Show HN: Zoto – low-level audio playback in Zig

https://github.com/braheezy/zoto
3•braheezy•21h ago•0 comments

Show HN: Piping in and Out of Emacs

https://github.com/agzam/mx-piper
4•iLemming•21h ago•1 comments

Show HN: I built a tech news aggregator that works the way my brain does

https://deadstack.net/recent
187•dreadsword•3d ago•97 comments

Show HN: Katakate – Dozens of VMs per node for safe code exec

https://github.com/Katakate/k7
123•gbxk•5d ago•53 comments

Show HN: Sqlite3-dump - a fast SQLite to CSV and parquet

https://github.com/i64/sqlite3-dump
18•Gave4655•2d ago•3 comments

Show HN: NickelJoke – Pay a Nickel to Get a Joke Using X402 Micropayments

https://github.com/btahir/nickeljoke
2•bilater•1d ago•2 comments