GLM-5.3 raises researcher fears the Chinese model can fake AI safety tests

 (scmp.com)

2 points | by theanonymousone 1 hour ago

0 comments