Joey David

I'm currently a research intern at MILeS, working on LLM reasoning. I work to make LLMs more capable and safer, even as they get smarter than the people steering them.

Longer version →

Recent Research

Temporal Predictors of Outcome in Reasoning Language Models

Joey David · arXiv, cs.CL · 2025

arXiv ↗PDF ↗

Transformer-based localization of uveitis symptoms

Joey David, Robin Jacquot, Tao Wang · INSA Lyon, DISP Laboratory · Ongoing

Selected Projects

Agentica

Modular agentic framework for automation, learning, and research.

Agentica terminal interface preview

Heuristic Retrieval for Latent Reasoning

Latent-space guidance system for Meta's Coconut model.

Heuristic latent retrieval visualization