Fine-tuning circumvention: Difference between revisions
Appearance
imported>Unknown user No edit summary |
imported>Unknown user No edit summary |
||
(No difference)
| |||
imported>Unknown user No edit summary |
imported>Unknown user No edit summary |
||
(No difference)
| |||
Fine-tuning to remove model refusal behaviour or other model-level safety interventions.
Source: NIST AI 100-2e2025 | Category: