Llama 3
Meta
partial detail 10 method fields published
Evaluation
Safety finetuning data quality ablations
Ablations on safety training data quality vs quantity (quality found more critical), using human-generated data from vendors plus AI-assisted quality control and a tone classifier for safety response verbiage
How it was run
Open a release to see the setup, scoring, and other details that the lab published. Only fields the lab actually disclosed are shown.