Researcher Profile

Editor reviewed

Jack W. Rae

Compute-optimal scaling for LLM training

Google DeepMind researcher on retrieval, scaling, and frontier language models

One of the better people to track for the sequence from Gopher to retrieval-augmented language models and then into Gemini, especially if you care about how DeepMind actually iterated on the frontier-model recipe.

Organizations

Google DeepMind

Labs

Google DeepMind

Topics

Multimodal Evaluation & Benchmarks

About This Page

This profile is meant to help you get oriented quickly: why this researcher matters, what to read first, and where to explore next.

Last reviewed

March 18, 2026

Best First Clicks

Scaling Language Models: Methods, Analysis & Insights from Training Gopherpaper Improving language models by retrieving from trillions of tokensarticle Gemini: A Family of Highly Capable Multimodal Modelspaper

Known For

The ideas, systems, and research directions that make this person worth knowing.

Gopher

RETRO and retrieval-augmented language models

Gemini

Compute-optimal scaling for LLM training

Training Compute-Optimal Large Language Models

DeepMind

Start Here

Canonical papers, project pages, or repositories that anchor this profile.

Related Researchers

People worth exploring next because they share topics, labs, or source material with this profile.

Shared canonical source

Jordan Hoffmann

Compute-optimal scaling for LLM training

3 sources

One of the clearest people to follow for the sequence from retrieval-augmented language models to compute-optimal scaling and then into Gemini.

Google DeepMind Multimodal Evaluation & Benchmarks

Start HereTraining Compute-Optimal Large Language Models

Shared canonical source

Lisa Anne Hendricks

Compute-optimal scaling for LLM training

3 sources

A useful page for the DeepMind work that connected large-language-model scaling to the multimodal Gemini push, with a clearer safety-and-evaluation flavor than many purely scaling-focused pages.

Google DeepMind Multimodal Evaluation & Benchmarks

Start HereGemini: A Family of Highly Capable Multimodal Models

Shared canonical source

Elena Buchatskaya

Compute-optimal scaling for LLM training

3 sources

Worth tracking for the DeepMind thread that links large-model scaling research to the multimodal Gemini stack, rather than treating those as separate eras.

Google DeepMind Open Models Multimodal

Start HereGemini: A Family of Highly Capable Multimodal Models

Shared canonical source

Trevor Cai

Compute-optimal scaling for LLM training

3 sources

A useful profile for the core DeepMind contributor layer behind Chinchilla, Gopher, and Gemini rather than only the more public faces of those systems.

Google DeepMind Multimodal Systems & Infrastructure

Start HereGemini: A Family of Highly Capable Multimodal Models

Shared canonical source

Diego de las Casas

Compute-optimal scaling for LLM training

3 sources

A useful profile for the DeepMind researchers who helped carry the lab’s language-model program from scaling-law work into Gemini rather than appearing only on the final product layer.

Google DeepMind Multimodal

Start HereGemini: A Family of Highly Capable Multimodal Models

Shared canonical source

Johannes Welbl

Compute-optimal scaling for LLM training

3 sources

A useful profile for the DeepMind researchers who sat inside the core language-model program as it moved from scaling-law analysis into the Gemini family.

Google DeepMind Multimodal

Start HereGemini: A Family of Highly Capable Multimodal Models