Besides the ridiculous number of submissions, interesting to hear your experience with it. For me it had 1-2 valid points, and 3 pages of nitpicking. I have the time, so I can address both types of issues, but wonder if that’s your experience as well. Overall, I think it is an interesting initiative
An LLM edited a small model's weights directly and its math score rose 6x