---
title: Joey David
description: Joey David - AI/ML researcher.
canonical: https://joeydavid.fyi/
markdown: https://joeydavid.fyi/index.md
---

# Joey David

> Joey David is an AI/ML researcher at MILeS (LAMSADE, CNRS / Université Paris Dauphine-PSL) working on LLM reasoning, with interests in information theory, interpretability, recursive self-improvement, emergence and AI safety.

Doing research at [MILeS](https://miles.lamsade.fr/) (CNRS, Dauphine-PSL), continuing the work on LLM reasoning from my internship there.
I'm also wrapped up in recursive self-improvement, emergence and AI safety.

- [Resume](https://joeydavid.fyi/media/joey_david_resume_2026.pdf)
- [Google Scholar](https://scholar.google.com/citations?hl=en&user=J3oOQioAAAAJ)
- [GitHub](https://github.com/joey-david)

## Recent research

### The Behavioral Rate of Fine-Tuning: Compression as a Measure of Useful Information

Joey David, Paul Caillon, Alexandre Allauzen · Under review · 2026

[arXiv](https://arxiv.org/abs/2611.06598) · [GitHub](https://github.com/joey-david/fineqcomp)

### Master's thesis · Information Constraints in Fine-tuning and Reasoning Language Models

Joey David · M2 IASD, PSL · 2026

[PDF](https://joeydavid.fyi/media/research/joey_david_masters_thesis_2026.pdf)

### Temporal Predictors of Outcome in Reasoning Language Models

Joey David · arXiv, cs.CL · 2025

[arXiv](https://arxiv.org/abs/2511.14773) · [GitHub](https://github.com/joey-david/latent-correctness-probe)

Full list: [Research](https://joeydavid.fyi/research/index.md)

## Selected projects

### Agent Ground Up (2026)

Coding agent from scratch: constant-size memory, recursive self-improvement, hand-derived DAPO loss.

### Latent Intent (2026)

AI-safety interpretability work toward detecting hidden objectives and scheming, even when an LLM's outward behavior is held fixed.

### Trajelysis (2026)

Framework for visualizing, decomposing, and analyzing layer-wise latent trajectories during generation.

### Agentica (2024)

Modular agentic framework for automation, learning, and research.

Full list with links: [Projects](https://joeydavid.fyi/projects/index.md)

## Site map

- [Home](https://joeydavid.fyi/index.md) (this page): Overview: current position, recent research and selected projects.
- [Research](https://joeydavid.fyi/research/index.md): Papers, master's thesis and ongoing research, with arXiv, PDF and code links.
- [Projects](https://joeydavid.fyi/projects/index.md): Main ML/agent projects, side projects and Android apps, with repos and demos.
- [About](https://joeydavid.fyi/about/index.md): Experience, education, technical stack, reading/music recommendations and contact links.
- [llms.txt](https://joeydavid.fyi/llms.txt): site index for language models; [llms-full.txt](https://joeydavid.fyi/llms-full.txt) has every page in one file.