Alignment and Safety

Safety Alignment

Set of techniques aimed at ensuring language models avoid generating harmful, dangerous, or inappropriate content while maintaining their overall performance.

← Kembali