Reconstructing the benchmark behind Luc Julia's 64% LLM reliability claim(github.com)
3 pointsby matthieu_bl4 hours ago1 comment