Maxime Peyrard

@peyrardMax

Junior Professor @CNRS (previously @EPFL, @TUDarmstadt) -- AI Interpretability, causality, and interaction flows between LLM, humans, and tools

peyrardm.github.io

Joined February 2016

328 Following

408 Followers

Replies Reposts Media

Pinned
Maxime Peyrard
@peyrardMax
Jan 8
New paper: The Dead Salmons of XAI Standard fMRI pipelines once detected predictive brain regions in a dead salmon! A striking warning about poor statistical methodology Now, XAI faces similar issues: many methods can yield plausible explanations even for randomized networks
1
Maxime Peyrard
@peyrardMax
Oct 1
Very happy that "The Dead Salmons of AI Interpretability" has been selected as one of the few oral position papers at @NeurIPSConf The preprint sparked interesting discussions, so it’s great to see the argument getting a prominent slot at #NeurIPS2026. See you there!
Maxime Peyrard
@peyrardMax
Jan 8
New paper: The Dead Salmons of XAI Standard fMRI pipelines once detected predictive brain regions in a dead salmon! A striking warning about poor statistical methodology Now, XAI faces similar issues: many methods can yield plausible explanations even for randomized networks
2
Maxime Peyrard
@peyrardMax
Apr 21, 2025
Our paper "Everything, Everywhere, All at Once: Is Mechanistic Interpretability Identifiable?" will be presented at #ICLR2025! It's also the first paper of my first PhD student — congrats @maximemeloux! 🎉 blog post: melouxm.github.io/MI-identifiabi… A short thread 🧵
5
Maxime Peyrard
@peyrardMax
Jun 27, 2024
🎓 Fully funded PhD position in Grenoble @CNRS starting this Fall! We will work together on interpretability and new types of evaluations of LLMs and agents. Apply now: emploi.cnrs.fr/Offres/Doctora… The picture is the view you'll have from the office in Fall 😉
2
Maxime Peyrard
@peyrardMax
Mar 25, 2024
Orchestrated interactions between LLMs, humans, and tools show great promise. Today, we introduce “Semantic Decoding" - a perspective that views these interactions as optimization and search in the space of semantic tokens (thoughts). 📄 arxiv.org/abs/2403.14562
1