Flag indicating the english language is selected

Andre Martins no Laboratório SARDINE. Antes disso, concluí a minha licenciatura em Engenharia Aeroespacial no IST, onde a minha tese combinou identificação de sistemas com aprendizagem profunda fisicamente informada. Esta formação proporciona-me uma perspetiva interdisciplinar única que integra princípios fundamentais de engenharia com machine learning. O meu trabalho atual explora modelos de memória em tempo contínuo para compreensão de vídeos longos e visa construir estruturas unificadas para recuperação de memória associativa, como redes de Hopfield esparsas e estruturadas. Sou fluente em português e inglês, e tenho paixão por impulsionar a investigação em IA, abordando desafios fundamentais em sistemas inteligentes. Ultimamente, tenho-me interessado particularmente em treinar modelos de visão e linguagem." data-en="I’m a PhD candidate in Electrical and Computer Engineering at Instituto Superior Técnico (IST), University of Lisbon, where I focus on advancing neural models for memory and attention, especially applied to natural language generation. My research is co-supervised by Andre Martins at the SARDINE Lab. Before this, I completed my Master’s degree in Aerospace Engineering at IST, where my thesis combined system identification with physically-aware deep learning. This background gives me a unique interdisciplinary perspective that blends core engineering principles with machine learning. My current work explores continuous-time memory models for long video understanding and aims to build unified frameworks for associative memory retrieval such as sparse and structured Hopfield networks. I’m fluent in Portuguese and English, and I’m passionate about pushing AI research forward by addressing fundamental challenges in intelligent systems. Lately, I’ve become particularly interested in training vision-language models."

  • Lisbon Machine Learning School (LxMLS 2024), -
  • ICML 2024, Sparse and Structured Hopfield Networks aceite na International Conference on Machine Learning em Viena como spotlight" data-en="Got my first PhD paper Sparse and Structured Hopfield Networks accepted at International Conference on Machine Learning in Vienna as a spotlight."> -

Hobbies

Travelling

2025

Movie Facts and Fibs (MF²): A Benchmark for Long Movie Understanding
Emmanouil Zaranis*, António Farinhas*, Saul Santos*, Beatriz Canaverde*, Miguel Moura Ramos*, Aditya K. Surikuchi, André Viveiros, Baohao Liao, Elena Bueno-Benito, Nithin Sivakumaran, Pavlo Vasylenko, Shoubin Yu, Sonal Sannigrahi, Wafaa Mohammed, Ben Peters, Danae Sánchez Villegas, Elias Stengel-Eskin, Giuseppe Attanasio, Jaehong Yoon, Stella Frank, Alessandro Suglia, Chrysoula Zerva, Desmond Elliott, Mariella Dimiccoli, Mohit Bansal, Oswald Lanz, Raffaella Bernardi, Raquel Fernández, Sandro Pezzelle, Vlad Niculae, and André F. T. Martins.
preprint.
PDF
@inproceedings{zaranis2025mf2, title={Movie Facts and Fibs (MF²): A Benchmark for Long Movie Understanding}, author={Emmanouil Zaranis;, António Farinhas, Saul Santos, Beatriz Canaverde, Miguel Moura Ramos, Aditya K. Surikuchi, André Viveiros, Baohao Liao, Elena Bueno-Benito, Nithin Sivakumaran, Pavlo Vasylenko, Shoubin Yu, Sonal Sannigrahi, Wafaa Mohammed, Ben Peters, Danae Sánchez Villegas, Elias Stengel-Eskin, Giuseppe Attanasio, Jaehong Yoon, Stella Frank, Alessandro Suglia, Chrysoula Zerva, Desmond Elliott, Mariella Dimiccoli, Mohit Bansal, Oswald Lanz, Raffaella Bernardi, Raquel Fernández, Sandro Pezzelle, Vlad Niculae, and André F. T. Martins.}, booktitle={arxiv preprint arXiv:2506.06275}, year={2025}, url={https://arxiv.org/abs/2506.06275} }
Modern Hopfield Networks with Continuous-Time Memories
Saul Santos, António Farinhas, Daniel McNamee, André F.T. Martins
New Frontiers in Associative Memories, ICLR 2025.
@inproceedings{ santos2025modern, title={Modern Hopfield Networks with Continuous-Time Memories}, author={Saul Santos and Ant{\'o}nio Farinhas and Daniel C McNamee and Andre Martins}, booktitle={New Frontiers in Associative Memories}, year={2025}, url={https://openreview.net/forum?id=bU4dyLTNp3} }
∞-Video: A Training-Free Approach to Long Video Understanding via Continuous-Time Memory Consolidation
Saul Santos, António Farinhas, Daniel McNamee, André F.T. Martins
42st International Conference On Machine Learning, Vancouver (Canada), July 2025.
@article{santos2025inftyvideotrainingfreeapproachlong, title={$\infty$-Video: A Training-Free Approach to Long Video Understanding via Continuous-Time Memory Consolidation}, author={Saul Santos and António Farinhas and Daniel C. McNamee and André F. T. Martins}, year={2025}, journal={arXiv preprint arXiv:2501.19098}, eprint={2501.19098}, archivePrefix={arXiv}, primaryClass={cs.CV}, url={https://arxiv.org/abs/2501.19098}, }

2024

Hopfield-Fenchel-Young Networks: A Unified Framework For Associative Memory Retrieval
Saul Santos, Vlad Niculae, Daniel McNamee, André F.T. Martins
arxiv preprint
@inproceedings{santos2024sparse, title={Hopfield-Fenchel-Young Networks: A Unified Framework For Associative Memory Retrieval}, author={Saul Santos and Vlad Niculae and Daniel C McNamee and Andre Martins}, booktitle={arxiv preprint}, year={2024}, url={https://arxiv.org/pdf/2411.08590?} }
Sparse and Structured Hopfield Networks
Saul Santos, Vlad Niculae, Daniel McNamee, André F.T. Martins
41st International Conference On Machine Learning, Vienna (Austria), July 2024.
@inproceedings{santos2024sparse, title={Sparse and Structured Hopfield Networks}, author={Saul Santos and Vlad Niculae and Daniel C McNamee and Andre Martins}, booktitle={41st International Conference on Machine Learning}, year={2024}, url={https://openreview.net/forum?id=OdPlFWExX1} }

2022

Symplectic Momentum Neural Networks: Using Discrete Variational Mechanics as a prior in Deep Learning
Saul Santos, Monica Ekal, Rodrigo Ventura
Learning for Dynamics and Control Conference, Stanford (USA), June 2022.
@inproceedings{santos2022symplectic, title={Symplectic Momentum Neural Networks - Using Discrete Variational Mechanics as a Prior in Deep Learning}, author={Santos, S. and Ekal, M. and Ventura, R.}, booktitle={Learning for Dynamics and Control Conference}, pages={584--595}, year={2022} }
https://nicolasmeseguer.notion.site/StreaMemory-Benchmark-35739f7f69404274a1d287f6827ef203
⛲ StreaMemory Benchmark
25 min
11 feb. 2022
#benchmark #mgpusim #akita
https://nicolasmeseguer.notion.site/Elementwise-2D-546f8fe9a8a24852a234bdc28b00cd15
🧊 Elementwise 2D Tensor
10 min
12 dec. 2022
#mgpusim #akita
https://nicolasmeseguer.notion.site/Tracing-LDS-Unit-69210a9d8fe44b4a997f8391aa508621
🛤️ Tracing LDS Unit
20 min
30 nov. 2022
#mgpusim #akita #tracing
https://nicolasmeseguer.notion.site/DAISEN-Tutorial-9d041575a9b0489099086152fd9a927e
📊 DAISEN Tutorial
15 min
15 nov. 2022
#mgpusim #akita #daisen

2024

🌲 🌍

🌲 🌍