CUAD (NeurIPS 2021)

A repository of legal NLP research papers.

[ACL 2025] Codebase for CoCoLex

Projeto de Reconhecimento de Entidades Nomeadas (NER) para textos jurídicos em português do Brasil. Implementa e compara abordagens clássicas (CRF) e de deep learning (BiLSTM-CRF) no dataset LeNER-Br, avaliando o impacto de diferentes tipos de embeddings (Word2Vec, FastText, BERTimbau features).

🎯 AI-Powered Multi-Document Legal Research Assistant | Advanced RAG System for Legal Document Analysis with Conflict Detection & Citation Management

A diachronic analysis of legal metalanguage used by the U.S. Supreme Court from 1986 to 2019.

LegalEval, high level framework for evaluation of legal LLMs and reasoning capabilities.

Crawls Canadian Refugee Claims to curate a dataset for language model adaptation for case law analysis and other downstream legal NLP tasks

Use LLMs to identify premises and claims within the ECHR dataset.

MBA thesis

This repository houses the source code for a Streamlit-based annotation interface developed for classifying sentences in legal documents. The interface is a part of a broader initiative to analyze court statements, specifically focusing on assessing judicial attitudes toward victims of sexual violence in the Israeli court system.

I like applying the latest transformers on multilingual legal documents.

PROFIT (Processing and Reasoning Over Financial Information in Text): A benchmark designed to evaluate language models across five tasks of increasing complexity in numerical reasoning using SEC filings