Oscar Mañas

Researcher, Superintelligence Labs, Meta

Joined

May 2023

Names

Oscar Mañas (Preferred)

Emails

****@mila.quebec (Confirmed)

****@gmail.com (Confirmed)

****@meta.com (Confirmed)

Personal Links

Career & Education History

Researcher

Superintelligence Labs, Meta (meta.com)

2025 – Present

Researcher

FAIR, Meta (meta.com)

2024 – 2025

PhD student

Université de Montréal (umontreal.ca)

2021 – 2025

Intern

Meta FAIR (meta.com)

2023 – 2023

Intern

Element AI (elementai.com)

2020 – 2021

MS student

Universitat Autònoma de Barcelona (uab.cat)

2018 – 2020

Undergrad student

Universitat Politècnica de Catalunya (upc.edu)

2013 – 2017

Advisors, Relations & Conflicts

Coworker

Albert Pumarola

2025 – Present

PhD Advisor

Michal Drozdzal

2023 – 2025

PhD Advisor

Adriana Romero-Soriano

2023 – 2025

PhD Advisor

Aishwarya Agrawal

2020 – 2025

Expertise

deep learning

computer vision

natural language processing

multimodal learning

2018 – Present

Publications

LatentLens: Revealing Highly Interpretable Visual Tokens in LLMs
Benno Krojer, Shravan Nayak, Oscar Mañas, Vaibhav Adlakha, Desmond Elliott, Siva Reddy, Marius Mosbach
- ICLR 2026 Re-Align Workshop
- Readers: Everyone
Learning What Matters: Prioritized Concept Learning via Relative Error-driven Sample Selection
Qian Yang, Shivam Chandhok, Oscar Mañas, Kanishk Jain, Aishwarya Agrawal, Leonid Sigal
- ICLR 2026 Conference Withdrawn Submission
- Readers: Everyone
Learning What Matters: Prioritized Concept Learning via Relative Error-driven Sample Selection
Shivam Chandhok, Qian Yang, Oscar Mañas, Kanishk Jain, Aishwarya Agrawal, Leonid Sigal
- ES-FoMo III
- Readers: Everyone
Controlling Multimodal LLMs via Reward-guided Decoding
Oscar Mañas, Pierluca D'Oro, Koustuv Sinha, Adriana Romero-Soriano, Michal Drozdzal, Aishwarya Agrawal
- AFM 2024 Poster
- Readers: Everyone
Improving Automatic VQA Evaluation Using Large Language Models
Oscar Mañas, Benno Krojer, Aishwarya Agrawal
- 38th AAAI Conference on Artificial Intelligence
- Readers: Everyone
Consistency-diversity-realism Pareto fronts of conditional image generative models
Pietro Astolfi, Melissa Hall, Jakob Verbeek, Marlene Careil, Oscar Mañas, Matthew J. Muckley, Adriana Romero-Soriano, Michal Drozdzal
- NeurIPS 2024 Workshop RBFM Oral
- Readers: Everyone
Improving Text-to-Image Consistency via Automatic Prompt Optimization
Oscar Mañas, Pietro Astolfi, Melissa Hall, Candace Ross, Jack Urbanek, Adina Williams, Aishwarya Agrawal, Adriana Romero-Soriano, Michal Drozdzal
- Accepted by TMLR
- Readers: Everyone
Consistency-diversity-realism Pareto fronts of conditional image generative models
Pietro Astolfi, Marlène Careil, Melissa Hall, Oscar Mañas, Matthew J. Muckley, Jakob Verbeek, Adriana Romero-Soriano, Michal Drozdzal
- CoRR 2024
- Readers: Everyone
An Introduction to Vision-Language Modeling
Florian Bordes, Richard Yuanzhe Pang, Anurag Ajay, Alexander C. Li, Adrien Bardes, Suzanne Petryk, Oscar Mañas, Zhiqiu Lin, Anas Mahmoud, Bargav Jayaraman, Mark Ibrahim, Melissa Hall, Yunyang Xiong, Jonathan Lebensold, Candace Ross, Srihari Jayakumar, Chuan Guo, Diane Bouchacourt, Haider Al-Tahan, Karthik Padthe et al. (21 additional authors not shown)
- CoRR 2024
- Readers: Everyone
EvalGIM: A Library for Evaluating Generative Image Models
Melissa Hall, Oscar Mañas, Reyhane Askari Hemmat, Mark Ibrahim, Candace Ross, Pietro Astolfi, Tariq Berrada Ifriqi, Marton Havasi, Yohann Benchetrit, Karen Ullrich, Carolina Braga, Abhishek Charnalia, Maeve Ryan, Mike Rabbat, Michal Drozdzal, Jakob Verbeek, Adriana Romero-Soriano
- CoRR 2024
- Readers: Everyone

View all 15 publications

Co-Authors

View all 76 co-authors