Autoresearch for Materials

Participate via Teams link

Autoresearch is a fast-growing agentic paradigm, where one gives a clearly defined objective, usually optimization related, and allows agents to iteratively attempt solutions. Versions of this technique has been found to get state of the art (or so it is claimed) in a variety of fields. But does it work for materials? And can we run this on open source models? Daniel will show how you can get started with autoresearch and why you might not always need frontier models for frontier results. Examples will cover from metal casting to 3D printing to Quantum chemistry.

Presented by Daniel Marchand.