Lapis Labs: University of Illinois Urbana-Champaign based Student-led Artificial Intelligence research organization
Lapis Labs has been working on innovative solutions to some of the most pressing questions about AI: What happens when a model has built potentially hazardous knowledge into its primary knowledge base – and how do you get rid of that information while preserving the rest of the working model.
This work has resulted in a new proxy called The Weapons of Mass Destruction Proxy (WMDP) that is getting a lot of attention. Lapis Labs has released a paper on their groundbreaking research: The WMDP Benchmark: Measuring and Reducing Malicious Use With Unlearning.
More information on their paper about WMDP and “Unlearning” can be found here: https://www.wmdp.ai/. This proxy was also featured in TIME Magazine this month, “Researchers Develop New Technique to Wipe Dangerous Knowledge from AI Systems“.
If you would like to read more about Lapis Labs click HERE to find out What they are, How they do what they do, and Who is on their team (and would you like to join them?).
Read some of their other publications or find out what they have been doing HERE.