Hello! I am Tristan Karch, a Ph.D. candidate at Inria in the Flowers team under the supervision of Pierre-Yves Oudeyer and Clément Moulin-Frier. My research is mainly on Vygotskian Autotelic Artifical Agents: agents that use language to imagine their own goals, for planning, reasoning and learning about them. I also work on emergent communication in Reinforcement Learning settings and in Language Games.

Experience in industry
Machine Learning Engineer
PNB Paribas Innovation Lab
New York, USA
06/2018 - 05/2019

Natural Language Processing (NLP) for clause extraction and question answering in legal documents, benchmarking of both pre-trained and legal specific word representations for text classification and question answering tasks.
Computer Vision for Semantic Segmentation of PDFs (implementation of a Fully Convolitional Neural Network).

Teaching and Research Experience
Lecturer
University of Bordeaux.
Bordeaux, France
Fall semester 2020

In charge of Introduction to Programming in Python (4TPM101U) and Development of Pyhton Applications (4TPM110U).

Reinforcement Learning Intern
Supaero RL Initiative (SuReLI)
Toulouse, France
2016 - 2017

Reinforcement Learning applied to control the wing sail of an autonomous sailboat (stall avoidance and control of systems with hysteresis).

Teaching Assistant
EPFL
Lausanne, Switzerland
Fall 2012 and Spring 2016

In Charge of Analyitcs II as well as Physics II.

Education
PhD in Computer Sciences
Inria
University of Bordeaux
10/2019 - Present

Grounded Language Learning.
Intrinsically motivated Reinforcement Learning.
Language as a tool to enhance learning and reasoning in autonomous agents.

MSc in Machine Learning
ISAE-Supaero
Toulouse, France
2018

Major in Decision and Data Sciences - Minor in Autonomous Systems & Robotics

MSc in Mechanical Engineering
EPFL
Lausanne, Switzerland
2018

Major in Computational Fluid Dynamics - Minor in Computational Sciences 

BSc in Mechanical Engineering
EPFL
Lausanne, Switzerland
2015

First and second year at EPFL.
Third year of BSc as an ERASMUS student in DTU Denmark

Academic Services
Conference Organizing
  • Organizer of the Second Edition of the Language and Reinforcement Learning Workshop (Larel 2) at Neurips. [Website]
  • 2022
    Conference Paper Reviewing
  • Reviewer for NeurIPS main conference
  • 2022
  • Reviewer for ICLR workshop SSL-RL
  • 2021
    Events
  • Mentor at HackMIT
  • 2018
    Programming Skills
    General
    Python, Java and C++
    Auto-diff.
    Pytorch, Tensorflow
    HPC
    Bash, Slurm
    Web
    HTML, CSS, JavaScript
    Contact Info
    03/11/1994
    Tristan Karch
    +33 625273597
    Hobbies

    In my free time, I love to play music (saxophone and guitar), watch football games with friends and ride my bike in the city. Sportwise I snowboard during winter and occasionally surf during summer.