Show HN: We hid a backdoor in an LLM – $51,200 on finding it

Other: AI Model Security(protora.vulcora.se)view on HackerNews
AImodelsecuritybackdoorchallengeHuggingFace

Author: telaia396

Date: 7/16/2026

Article Summary:
The article presents a challenge to find a backdoor in a set of seven AI models, where one model is compromised and the others are decoys.