AIModel safetyStory 07
Image models forget less than tests suggest
What changedA certification method bounds residual concepts beyond the finite prompts used in ordinary attacks. Models may appear to forget a style or identity while a wider prompt space still leaks it. The result is concrete, but it remains research evidence rather than a production guarantee.
The useful part
Why it matters
Models may appear to forget a style or identity while a wider prompt space still leaks it.
Worth doing
What to do next
Reproduce the core result against your own data, hardware and failure cases before depending on it.
Keep in mind
Good to know
Guarantees depend on stated assumptions. Not every real attack is captured.
Evidence