nav_sign

tom.marty@mila.quebec

Home Research Projects Education Fun

I’m Tom Marty, a Ph.D. student at Mila under the supervision of Dhanya Sridhar. I'm originally from Réunion Island. I own an engineering degree from École Polytechnique (X2018) in Paris, and a Master's degree in Operational Research from Polytechnique Montréal.

My current research interest falls at the crossroad of Language understanding, Alignment and Deep Learning with a focus on :

  • Probabilistic Modelling
  • Causal Representation Learning
In a previous researcher life, I specialized in Reinforcement Learning and Combinatorial Optimization. I developed SeaPearl, a fully functional Open-Source Constraint Programming solver capable of learning optimal branching heuristic using RL.

Feel free to reach out to me via email if you have any questions or for potential collaborations.

Scholar  /  CV  /  GitHub  /  Twitter  /  LinkedIn

profile photo

News

Research Experience
nthu

Visiting ResearcherMarch. 2023 - October 2023

Service Now Research, Montréal, Canada
Supervisor: Alexandre Piche, Maxime Gasse, Quentin Cappart
Research Area: LLM, Task solving, Webpage processing

nthu

Research CoordinatorFebruary. 2022 - July 2022

Research InternFebruary. 2021 - July 2021

Corail Research Group, Montréal, Canada
Supervisor: Quentin Cappart, Louis-Martin Rousseau
Research Area: Constraint Programming, Reinforcement Learning, GNN

nthu

Software Development InternJun. 2020 - Sept. 2020
Dronisos, Bordeaux, France
Developed Harmony, a particle based meta-heuristic that secures massive drone swarms (+500 agents)
Research Area: Meta-heuristics, Force fields.


Papers - Open Source Frameworks - Projects
nthu In-Context Learning and Occam's razor
Mar' 24 - Oct' 24

We propose an explanation for the strong generalization abilities of in-context learners at inference time, by drawing connections between meta-learning, In-Context Learning and Information theory.
Keywords : Meta-learning, Kolmogorov complexity, Occam's razor.

Paper (Under review)  /  Code      

nthu WebArena : LLMs as Generalist Web Agents
Avr' 23 - Mar' 24

As part of my research at ServiceNow Research, I worked on WebArena : an Open-Source benchmark as a Gym environment to evaluate Web Agent to solve common knowledge task on a Web Browser. Published at NeurIPS 2023 FMDM Workshop and ICML 2024.
Keywords : Web-Automation, Task solving, benchmark.

Paper (Accepted at ICML2024)  /  Website  /  Code     GitHub Repo stars

nthu SeaPearl : a Constraint Programming Solver guided by Reinforcement Learning
Feb' 21 - Jul' 23

I was an active lead maintainer of the Open-Source and collaborative project SeaPearl. The goal is to develop an intelligent constraint programming solver that can learn to solve any Constraint Optimization Problems using Reinforcement Learning on Graphs.

The entire solver is written in Julia . Click here for detailed explanations.
Keywords : Combinatorial Optimization, Reinforcement learning, GNNs

Paper (Accepted at CP2023)  /  Code    GitHub Repo starsGitHub forks

nthu Adversarial Attacks on Sentiment Analysis models
Mar' 22 - May' 22

This project was carried out in the framework of the IFT6167 seminar led by Prof. Irina Rish (Mila, Québec).

In this work, we aim to show, regarding Natural Language Sentiment Analysis, that there exists a relationship between model size and robustness to adversarial attacks. Ultimately, uncovering the emergence of power laws and testing the robustness of language model with scale. We evaluate the performance of various Eleuther AI GPT models such as GPT-Neo 125M, GPT-Neo 1.3B, GPT-Neo 2.7B, GPT-J 6B against adversarial attacks. We fined-tuned (trained on adversarial example) our different GPT models on common datasets (Rotten Tomatoes, IMDB...) and evaluated them separately to quantify the effects of scale on adversarial training.

Report  /  Slides

nthu Diffusion Geodesic : a new Metric for non-linear Dimensionality Reduction
Sep' 21 - Jan' 22

In collaboration with Ph.D. candidate Guillaume Huguet (Mila, Québec), we present our method for non-linear dimensionality reduction called Diffusion Geodesic.

Dimensionality reduction techniques are often used to visualize the underlying geometry of a high-dimensional dataset. These methods usually rely on specific similarity measures. In this project, we first approximate the geodesic distance using a diffusion process over the underlying manifold, then we use Multi-Dimentionnal Scaling combined with our previously defined pairwise 'distances' to embed our Manifold in a lower dimensional space. We compare our model with popular algorithms such as PHATE, UMAP, and Isomap on toy datasets and RNA-seq dataset.

Report  /  Code

nthu Autonomous Drone Swarm Deployment
Dec' 20 – Mar' 21

In collaboration with Sariah Al Saati (ENS), Mehdi Benharrats (X-HEC), Swann Chelly (Sorbonne University) and Pierre Tessier, this report proposes a method for the coverage of a rescue zone with a swarm of UAV’s in order to detect possible target of interest.

The method is based on Collaborative Reinforcement Learning. It also presents a pipeline to locate points of interest in 3D from a set of 2D images using Inverse Projection Transformation and 3D ray clustering.

Report

nthu Realtime 3D Deep Motion Capture
Oct' 20 – Dec' 20

In collaboration with Pierre Tessier (MS, Columbia University), the objective of our project was to implement a intelligent 2D to 3D Motion Capture mechanism that uses only the video stream of a webcam as input. We were able to animate relatively accurately this Mk-44 Iron Man 3D model.

The project is based on the model DOPE presented in this paper for the automatic 3D rig generation from video input coupled with a quaternion-based 3D rotation inference pipeline for 3D model animation.

Report (french)  /  Video  /  Code  /  Slides

nthu Sketch-based Shape Retrieval
Sep' 20 – Dec' 20

In collaboration with Pierre Tessier (MS, Columbia University), the objective of our project is to enable an efficient search in a 3D model bank models from simple hand drawings. The project is based on this paper SIGGRAPH2012.

The project lies on differents techniques such Suggestive Contouring (Canny filter), Gabor filtering (gaussian convolution on Fourier Transformation of the input image) and histogram representation using Visual Vocabulary.
Looking back at this project, I saw that another paper with much more impressive results came out a few years later using Siamese convolutional Neural Networks for feature extraction.

Report (french)  /  Code  /  Slides

nthu Procedural modeling of a 3D rendered scene
Mar' 20 – June' 20

In collaboration with Elsa Deville (MS, Imperial College), the objective of our project is to fully render a realistic 3D marine scene using only randomized procedural modeling (OpenGL).

The project implements different visual elements among them terrain generation using Perlin Noise, a realistic simulation of a fish swarm movement based on this paper and realistic-looking ocean waves. (Trochoidal Waves : exact solution of the Euler equations for periodic surface gravity waves).

Report (french)



   Education
nthu

Ph.D. in Machine LearningJanuary 2024 - Present
MILA
Advisor: Dhanya Sridhar
Montréal, Canada

nthu

Master Thesis in Machine LearningSeptember 2021 - July 2023
Polytechnique Montréal
Advisor: Quentin Cappart, Louis-Martin Rousseau
Montréal, Canada

nthu

Bachelor of Science - Ingénieur Polytechnicien X2018Jun. 2018 - May. 2021
l'Ecole Polytechnique
Major in Computer Science
Minor in Applied Mathematics

Ranked 3rd out of 3000+ candidates at the national entrance exam for Ecole Polytechnique, the most prestigious and selective engineering school in France.
Palaiseau, France

nthu

CPGE : Higher School Preparatory ClassesJun. 2016 - May. 2018
Lycée Jean-Baptiste Say
Intensive multi-disciplinary program leading to competitive entrance exams of french Grande Ecoles.
Paris XVI, France

Initiatives and Academic Services
INF8215 : AI : Methods and Algorithms Fall 22

INF8215 : AI : Methods and Algorithms Fall 21

I am a teaching assistant for the INF8215 : AI : Methods and Algorithms at Polytechnique University taught by Quentin Cappart for the Autumn 2022 semester.
Assistant Professor in Maths and Physics Nov. 2019 - Mar. 2020

I became a teaching assistant for 6 months (Full Time) for both High School and undergraduate students with underprivileged backgrounds at the Boarding School of Success of Noyon, France.
Grants, Scholarships and Awards
Distinguished Paper Award at CP2023, Toronto Sep. 2023
Mitacs Accelerate scholarship of two units (30000$ CAD) Mar. 2023
Vallet Fondation scholarships for outstanding CPGE students 2018

This is where I come from !


Updated on: 19th January, 2023 Merci, Jon Barron et Diganta Misra !