Recent content by ruby_823a

I wanted to start a broader mistakes and lessons thread around Reinforcement Learning. A useful angle is prompt design, but it also connects with security. Different tools and workflows can produce very different results even when the goal is similar. It would be interesting to compare...