Prism: When an LLM predicts the next token, which training does it relying on?guidelabs.ai1 point·aziis98··0 commentsOpen articleSaveView on HN