36 6 82

Arthur Souza Rodrigues PRO

arthrod

https://www.synthetic.lawyer

AI & ML interests

Attorney obsessed with machine learning. My dream is a world with legal contracts accessible to all. Connect with me!

Recent Activity

updated a Space 2 days ago

arthrod/cicero-cola-na-prova

liked a dataset 4 days ago

common-pile/caselaw_access_project

liked a Space 5 days ago

muset-ai/DeepResearch-Bench-Leaderboard

View all activity

Organizations

updated a Space 2 days ago

Cicero Cola Na Prova

🍃

Template for deploying a marimo application to HF

liked a dataset 4 days ago

common-pile/caselaw_access_project

Viewer • Updated Jun 6 • 5.52M • 2.56k • 203

liked a Space 5 days ago

DeepResearch Bench

🔍

130

Generate and display a leaderboard

commented on Introducing the Massive Legal Embedding Benchmark (MLEB) about 1 month ago

Really cool approach. Sorry I didn't have time to read the full paper, but did you see any variance on the classification score when evaluating against English speaking jurisdictions in which the data was more sparse? My guess would be that wouldn't matter that much. I'd out my money that your classifier is a deberta/GLiNER derived... Thanks for this!

commented on Introducing the Massive Legal Embedding Benchmark (MLEB) about 1 month ago

This is amazing! Congrats on the great work. BTW, Umar, did you manually annotate or did you use another LLM? Thanks for this, it's really great!

reacted to adlumal's post with 🧠🤯😎🤗➕❤️🚀 about 1 month ago

Post

2457

MLEB is the largest, most diverse, and most comprehensive benchmark for legal text embedding models. https://huggingface.co/blog/isaacus/introducing-mleb