Model Organism Hackathon (dry run)

​Can transparency tools unmask hidden AI goals? To find out, we’re creating a testbed of alignment faking models.

​

​
We're building model organisms because alignment research needs a feedback loop. Lots of people are working on interpretability techniques, and it's unclear if they are useful for solving real problems.

So join us in Constellation in Berkeley to build the alignment benchmark that the research community sorely needs.

​First place: $150 🥇
Second place: $100 🥈
Third place: $50 🥉

More info:
https://www.notion.so/Welcome-to-the-Model-Organism-Hackathon-24b682ca364c80799d62f3549c14d765