Hacker News new | ask | show | jobs
by tlonny 512 days ago
Given the size of the training data - I don’t think it would economical to validate all training data with high-end LLM models.
1 comments

True. Maybe it can be dumbed down to a low-end model specifically for this type of detection.