What happens when LLMs play the Telephone game ☎️? In this paper, we analyse the evolution of texts as they are transmitted between LLM agents 🤖💬🤖💬🤖💬 Do text properties converge to attractors 🧲? How is this influenced by the task📝 and model⚙️? 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.

With members of the Flowers team, we developed Vivarium, a multi-agent simulator in Jax. It serves both research and teaching in Artificial Intelligence and Artificial Life. This software enables creating simple agents with two motors and two sensors, inspired by Braitenberg Vehicles, in a 2D rigid-body physics world (the physics engine is written in Jax-MD). Despite the simplicity of the agents at an individual level, the interactions between them can lead to complex emergent behaviors and scenarios !

Introduction I participated in the LeRobot Hackathon in Toulouse, organized by the Hugging Face team 🤗 ! It was a great way to get my first steps into real-world reinforcement learning and robotics. We went from having a very messy table (see below) with a lot of hardware parts to a fully assembled robot in a few hours. It was a robotic arm with a gripper, the Moss v1 and we tried to teach it how to manipulate a cube.

Co-developed the LLM-Culture library within the Flowers team. 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 parallelized hyper parameter search using Optuna in the ReservoirPy machine learning library.

What can LLMs bring to the study of cultural evolution 🤔? In this preprint, we present a framework for simulating cultural evolution using LLMs 🦜. We study how the dynamics depends on social structure, simulated personalities👥, and whether it has potential attractors🧲🔍 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.

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