I used it to make a list a few days ago of existing safety benchmarks, in response to a prompt for such things being posted: https://www.lesswrong.com/posts/KQ6fGiPeMnzzC6p9q/race-to-the-top-benchmarks-for-ai-safety?commentId=A7FMZpfHgjRFHCcCq
Current theme: default
Less Wrong (text)
Less Wrong (link)
I used it to make a list a few days ago of existing safety benchmarks, in response to a prompt for such things being posted: https://www.lesswrong.com/posts/KQ6fGiPeMnzzC6p9q/race-to-the-top-benchmarks-for-ai-safety?commentId=A7FMZpfHgjRFHCcCq