FAR AI — AI Interpretability Research

Organization:
FAR AI
Award Date:
03/2023
Amount:
$100,000
Purpose:
To support a research project to develop a more interpretable neural network architecture.

Open Philanthropy recommended a grant of $100,000 to FAR AI to support a research project, led by Open Philanthropy AI Fellow Alex Tamkin, aimed at developing a neural network architecture that could serve as a more interpretable alternative to the transformer architecture used in leading language models.

This follows Open Philanthropy’s December 2022 support and falls within its focus area of potential risks from advanced artificial intelligence.

Read more: