Confluence Labs
ActiveAI models that learn from experience
About
While modern AI excels in any area you can collect a lot of data for, it struggles in areas where data is sparse or costly to attain. Designing new molecules, discovering new physics, and engineering new materials are all domains where collecting data is extremely costly. We dream of a world where AI accelerates research in all of these domains and creates a more abundant future for humanity, but the current technology is not there. That’s why we started Confluence Labs. We are building AI that can design highly effective experiments in data-sparse domains and learn maximally from the data it already has.
Founders · 2
20 yrs old. Born in India. Won awards for my math research at ISEF. Got a full-ride to college in the US. Dropped out of Dartmouth to start an AI research lab.
Launch
an AI research lab focused on learning efficiency
Confluence Labs releases ARC-AGI-2, a benchmark for solving novel reasoning tasks from few examples, reporting a 97.9% score and opens sources its SOTA solver. The approach uses program synthesis guided by LLMs to tackle data-sparse domains, aiming to improve learning efficiency in research areas like hardware engineering, drug design, and physics.
Formerly “Vector Labs”, “Confluence Labs”, “Vector Labs”
Related startups
Claude Code for Scientific Research
The First AI-Native AI Research Lab
