Researcher

Aidan Clark

OpenAIMember of Technical Staff, VP of Research (Training)

Clark co-led pretraining for GPT-4o, the stage that builds a model’s core capabilities before it is tuned for users.^[1] At DeepMind, he led a study of how language models scale when they use only part of the network for each piece of text.^[2]

Career & education

Current role

OpenAI2022-present

Member of Technical Staff, VP of Research (Training)

Previously

Google DeepMind2017-2022

Staff Research Engineer

Education

UC Berkeley

Work and education history is primarily focused on AI-relevant roles and may not be comprehensive.

Citation Trend

3 snapshots

citations26,599As of Oct 19, '26

citations+4,567Since last snapshot, Jun 15, '26

For today’s citation count and per work, visit Google Scholar.

Selected contributions

Paper2022
Training Compute-Optimal Large Language Models
With Jordan Hoffmann, Sebastian Borgeaud, Arthur Mensch, Elena Buchatskaya, Trevor Cai, Eliza Rutherford, Diego de las Casas and others ·
Chinchilla showed how smaller models trained on more data can better use a fixed computing budget.
Read paper

References

GPT-4o contributionsopenai.com
OpenAI credits Clark as one of three pretraining leads for GPT-4o.
Unified Scaling Laws for Routed Language Modelsproceedings.mlr.press
Clark and his co-authors study how model size and the number of specialized components affect performance when only some components are used at a time.

About this profile

Based on public sources. If accuracy is important, verify details using the links or references.

For corrections, updates, or requests to remove information, contact us.

Reviewed August 8, 2026 · Checked for accuracy and new information.

Researcher

Aidan Clark

OpenAIMember of Technical Staff, VP of Research (Training)

Career & education

Current role

OpenAI2022-present

Member of Technical Staff, VP of Research (Training)

Previously

Google DeepMind2017-2022

Staff Research Engineer

Education

UC Berkeley

Work and education history is primarily focused on AI-relevant roles and may not be comprehensive.

Citation Trend

3 snapshots

citations26,599As of Oct 19, '26

citations+4,567Since last snapshot, Jun 15, '26

For today’s citation count and per work, visit Google Scholar.

Selected contributions

Paper2022
Training Compute-Optimal Large Language Models
With Jordan Hoffmann, Sebastian Borgeaud, Arthur Mensch, Elena Buchatskaya, Trevor Cai, Eliza Rutherford, Diego de las Casas and others ·
Chinchilla showed how smaller models trained on more data can better use a fixed computing budget.
Read paper

References

GPT-4o contributionsopenai.com
OpenAI credits Clark as one of three pretraining leads for GPT-4o.
Unified Scaling Laws for Routed Language Modelsproceedings.mlr.press
Clark and his co-authors study how model size and the number of specialized components affect performance when only some components are used at a time.

About this profile

Based on public sources. If accuracy is important, verify details using the links or references.

For corrections, updates, or requests to remove information, contact us.

Reviewed August 8, 2026 · Checked for accuracy and new information.

Aidan Clark

Career & education

Citation Trend

Selected contributions

Training Compute-Optimal Large Language Models

References

Related people

Aidan Clark

Career & education

Citation Trend

Selected contributions

Training Compute-Optimal Large Language Models

References

Related people

Related people

Related people