Full threadAl0neStar·a question to those who work on LLMs: how do you even check for downstream data contamination in a model as large as GPT4?View on HN