I am trying to improve my approach to AI Benchmarks and could use some practical advice.
I am particularly interested in future improvements and security. I would like to compare approaches that people have actually used rather than just theoretical suggestions.
What has worked best for you...
