Gautier Hamon (@hamongautier) header
Gautier Hamon (@hamongautier) profile picture

PhD INRIA Flowers team @flowersInria. MVA master. Reinforcement learning, Meta-learning, artificial life.

Joined September 2011
Photo by Gautier Hamon (@hamongautier)Photo by Gautier Hamon (@hamongautier)Photo by Gautier Hamon (@hamongautier)Photo by Gautier Hamon (@hamongautier)
1/⚡️Looking for a fast and simple Transformer baseline for your RL environment in JAX ? Sharing my implementation of transformerXL-PPO: github.com/Reytuag/transform… The implementation is the first to attain the 3rd floor and obtain advanced achievements in the challenging Craftax
6
32
7
124
20,607
Gautier Hamon retweeted
Want to play a game in cellular automata, with Lenia creatures as the protagonist? (and the ghosts?) Play Carlman online: chakazul.github.io/carlman/ Collaboration with @marko_cvjetko @jessescool_ @hamongautier, accelerated with Opus 🤖
2
13
3
66
3,032
Gautier Hamon retweeted
Can complex multi-step reasoning emerge purely from cells that only talk to immediate neighbors? Happy to share our paper “Reasoning with Neural Cellular Automata (NCA)”, from our team at Google, Paradigms of Intelligence 🧵👇
46
196
53
1,309
146,230
Gautier Hamon retweeted
🧵 New paper: RL for discovering and controlling self-organizing phenomena. We introduce CARL, a closed-loop agent that learns to create self-organizing patterns in Lenia, steer their behavior, and let humans guide them in real time. Project website + interactive demos below ↓
10
56
6
288
29,262
Gautier Hamon retweeted
New milestone: we trained a robot foundation model on a world model backbone, and enabled zero-shot, open-world prompting capability for new verbs, nouns, and environments. If the world model can "dream" the right future in pixels, then the robot can execute well in motors. We call it "DreamZero", our first World Action Model (WAM). Our team had tons of fun at the lab typing anything we like into an open text prompt, and watch the robot perform tasks it was never trained on. An emergent capability we didn't quite expect. Obviously not GPT-3 reliable yet, but we are marching into the GPT-2 era. Discoveries: - Model and data recipe co-evolve. Compared to VLAs, WAMs learn best from diverse data, breaking away from the conventional wisdom that lots of repeated demos per task are the bread and butter. Diversity >> repetitions. - X-embodiment is extremely hard. Pixels are the answer. Different robot morphologies traditionally have a hard time sharing knowledge well. But if we put video first, pixels become the universal bridge connecting different hardware - even videos of human first-person view. DreamZero shows significant robot2robot and human2robot transfer. With only 55 trajectories on a *new*, unseen hardware (~30 min of teleop), it adapts so quickly and retains zero-shot prompting ability. Yesterday I posted about the "Second Pre-training Paradigm": world models are the next-gen foundation of Physical AI, not language backbones. Today, we are proving it works. And 2026 has just begun. Paper: World Action Models are Zero-Shot Policies. Read it now: (thread)
49
111
26
613
69,961
Gautier Hamon retweeted
🔥Our paper PhyloLM got accepted at ICLR 2025 !🔥 In this work we show how easy it can be to infer relationship between LLMs by constructing trees and to predict their performances and behavior at a very low cost with @StePalminteri and @pyoudeyer ! Here is a brief recap ⬇️
3
13
4
34
4,007
Gautier Hamon retweeted
🤖 !! Model-free agents can internally plan !! 🤖 In our ICLR 2025 paper, we interpret a model-free RL agent and show that it internally performs a form of planning resembling bidirectional search.
7
46
7
316
38,857
Gautier Hamon retweeted
🚀 Introducing 🧭MAGELLAN—our new metacognitive framework for LLM agents! It predicts its own learning progress (LP) in vast natural language goal spaces, enabling efficient exploration of complex domains.🌍✨Learn more: 🔗 arxiv.org/abs/2502.07709 #OpenEndedLearning #LLM #RL
3
11
6
39
9,454
🚀 Want to simulate #AI/#ALife agents in a physics-driven world? Meet Vivarium🌱– a new multi-agent simulator from @FlowersINRIA , designed for research & education! ✅ Large-scale simulations with #JAX ✅ User-friendly web interface ✅ Real-time control with jupyter notebooks
7
18
2
73
7,093
Online RL can help ground LLMs in interactive environments for problem-solving. However, Salim shows in our recent @naacl paper that current methods suffer from prompt overfitting. We propose a general framework to analyse it along with a contrastive learning-based approach. 👇
Replying to @mlia_isir
📄 RL for Aligning LLM Agents in Interactive Environments: Quantifying & Mitigating Prompt Overfitting 🔗arxiv.org/abs/2410.19920 by Mohamed Salim Aissi, @ClementRomac, @CartaThomas2, @SLamprier, @pyoudeye, Olivier Sigaud, @LaureSoulier, Nicolas Thome 🤝 With @FlowersINRIA
1
1
13
726
Gautier Hamon retweeted
We're excited to announce the first Evolving Self-organisation workshop at @GeccoConf! Submission deadline: March 26, 2025 More information: evolving-self-organisation-w…
5
77
3
353
56,930
To our knowledge, this is the first demonstration of kin selection emerging within a population of agents undergoing continuous neuroevolution. Interestingly, we show that kin selection does not require a discrete notion of gene, or a fitness-maximisation objective. (7/9)
1
1
1
94
Excited to share a new preprint from my visit at @FlowersINRIA! Kin selection theory is a widely accepted account of how altruistic behaviours can evolve in social environments - (1/8)
1
1
2
10
1,061
We just opened a new (engineering) internship position in the @FlowersINRIA team with @pyoudeyer: docs.google.com/document/d/1… We'll focus on developing our lamorel library, which has been central to our recent work on grounding embodied LLMs (more below 👇1/4)
1
3
1
8
1,147
Want to understand better the details of Lenia, Flow Lenia and other continuous CAs ? 👾 Take a look at this amazing course by @VassPapas @EPFL_en (with a physics perspective) ➡️frotaur.notion.site/Course-8… 📺youtube.com/watch?v=nMoQqppL…
Curiosity-driven exploration of Flow-Lenia universes 🔍First discoveries in our ongoing Alife project 🔬We use curiosity-driven autotelic algorithms as AI scientists 🤖 exploring the space of Flow-Lenia environments 🚀We aim to find environments that self-organize, from scratch, sensorimotor agents, memory, learning, cognition, and even evolution itself 🔄🧬🌱
1
7
1
19
1,789
🚀 Exciting Internship Opportunities for AI and CogSci Students🌟 Join @FlowersINRIA and work on these cool topics: 🔧 Curriculum learning of skill libraries in autotelic agents using LLMs and program synthesis wth @PourcelJulien 🎯 Balancing Exploration and Exploitation in Autotelic Reinforcement Learning wth @cedcolas @ClementRomac 🌐 LLM-Mediated Cultural Evolution: Studying AI Catalysts in Collective Problem-Solving wth @nisioti_eleni @cedcolas @Jeremy__Perez 🧩 Curiosity-driven automated discovery of interference patterns in multi-core architectures wth @Clement_MF_ 👶 Human-Like Language Acquisition in Language Models (LMs): a Parent-LM Teaching an Intrinsically-Motivated baby-LM 📚 Evaluation of a generative AI literacy pedagogical intervention in middle school to improve adolescents' understanding and use of generative AI wth @RaniaABDBenAmar (French language proficiency required here) ▶️For whom? Primarily master level students (before PhD) ▶️Instructions to apply: flowers.inria.fr/jobs/
2
20
59
4,955
Gautier Hamon retweeted
we are recruiting interns for a few projects with @pyoudeyer in bordeaux > studying llm-mediated cultural evolution with @nisioti_eleni @Jeremy__Perez > balancing exploration and exploitation with autotelic rl with @ClementRomac details and links in 🧵 please share!
1
14
4
25
6,805
Fast turn-based tactical combat env I've been working on -- RL training incoming!
1
2
19
1,122
Gautier Hamon retweeted
🚨New preprint🚨 When testing LLMs with questions, how can we know they did not see the answer in their training? In this new paper we propose a simple out of the box and fast method to spot contamination on short texts with @StePalminteri and @pyoudeyer ! 1/16
3
9
2
21
3,337
Gautier Hamon retweeted
We are very excited to announce Kinetix: an open-ended universe of physics-based tasks for RL! We use Kinetix to train a general agent on millions of randomly generated physics problems and show that this agent generalises to unseen handmade environments. 1/🧵
14
208
24
1,036
162,262