José Pombal | Unbabel

About Me

I am currently a research scientist at Unbabel. My research is centered on large language models, their multilinguality, and their application to machine translation and machine translation evaluation. Previously, I worked on the fields of algorithmic fairness and explainability.

News

[February 2024] Released the Tower paper on arxiv.
[January 2024] First release of Tower, an open multilingual LLM for translation-related tasks.
[October 2023] Paper accepted at EMNLP 2023 (findings).
[September 2023] Collaborated on the winning submission to the WMT 2023 Shared Task on Quality Estimation held at EMNLP 2023.
[March 2023] Joined Unbabel as a Research Scientist.
[December 2022] Defended my master’s thesis, concluding my internship at Feedzai. Published papers at NeurIPS 2022, and workshops at ICLR 2023, ICML 2022, and KDD 2022.

Selected Publications

* = equal contribution

Tower: An Open Multilingual Large Language Model for Translation-Related Tasks

Duarte M. Alves*, José Pombal*, Nuno M. Guerreiro*, Pedro H. Martins, João Alves, Amin Farajian, Ben Peters, Ricardo Rei, Patrick Fernandes, Sweta Agrawal, Pierre Colombo, José G.C. de Souza, André F. T. Martins

arXiv (preprint) (2024)

PDF Models and data Evaluation code Zeno project
Scaling up CometKiwi: Unbabel-IST 2023 Submission for the Quality Estimation Shared Task

Ricardo Rei, Nuno M. Guerreiro, José Pombal, Daan van Stigt, Marcos Treviso, Luisa Coheur, José G.C. de Souza, André F.T. Martins

Conference on Machine Translation (WMT), @EMNLP2023

PDF Code
Turning the Tables: Biased, Imbalanced, Dynamic Tabular Datasets for ML Evaluation

Sérgio Jesus, José Pombal, Duarte Alves, André F. Cruz, Pedro Saleiro, Rita P. Ribeiro, João Gama, Pedro Bizarro

NeurIPS 2023

PDF Dataset Code
Prisoners of Their Own Devices: How Models Induce Data Bias in Performative Prediction

José Pombal, Pedro Saleiro, Mário A.T. Figueiredo, Pedro Bizarro

ICML 2022 Workshop on Responsible Decision Making in Dynamic Environments

PDF