I am currently working on a project named Vivarium that serves both research and teaching in AI. It’s a software that allows launching custom particle based multi-agents simulations, using a physics engine built in Jax. The simulations can be hosted on a server, and interacted with in real time from a web interface or Jupyter Notebooks. The Notebook case aims to teach students how to program behaviors of simulated robots with Python !

Arxiv preprint (Under Review) Abstract As large language models (LLMs) start interacting with each other and generating an increasing amount of text online, it becomes crucial to better understand how information is transformed as it passes from one LLM to the next. While significant research has examined individual LLM behaviors, existing studies have largely overlooked the collective behaviors and information distortions arising from iterated LLM interactions. Small biases, negligible at the single output level, risk being amplified in iterated interactions, potentially leading the content to evolve towards attractor states.

Developed the LLM-Culture library. This software enables simulating networks composed of LLMs agents, that can generate text over multiple generations based on their neighbors input, personnality and task. The project also provides tools for analyzing the resulting text dynamics and offers an user-friendly web interface, making it accessible to non-programmers. Developed a turorial for hyper parameter search using Optuna in the ReservoirPy machine learning library. The tutorial covers sequential and parallelized hyperparameter search on local machines using the joblib package.

Arxiv preprint. We are currently working on the paper to submit it to a conference. Abstract Research in cultural evolution aims at providing causal explanations for the change of culture over time. Over the past decades, this field has generated an important body of knowledge, using experimental, historical, and computational methods. While computational models have been very successful at generating testable hypotheses about the effects of several factors, such as population structure or transmission biases, some phenomena have so far been more complex to capture using agent-based and formal models.

SimulationSandbox is a simple framework built with Jax and Socket that allows for real-time interaction between a simulation hosted on a server and multiple clients. It provides a simple interface for visualizing or interacting with the state of a hosted simulation from remote clients, such as jupyter notebooks. Code

Paper presented at EvoStar 2024 (Long Talk). Abstract Animals often demonstrate a remarkable ability to adapt to their environments during their lifetime. They do so partly due to the evolution of morphological and neural structures. These structures capture features of environments shared between generations to bias and speed up lifetime learning. In this work, we propose a computational model for studying a mechanism that can enable such a process. We adopt a computational framework based on meta reinforcement learning as a model of the interplay between evolution and development.