fp.
newest
Open in hackernews
GLM-5.3 raises researcher fears the Chinese model can fake AI safety tests
https://www.scmp.com/tech/big-tech/article/3370963/zais-glm-53-raises-researcher-fears-chinese-model-can-fake-ai-safety-tests
2
•
theanonymousone
•
46m ago