0
PoisonGPT: How we hid a lobotomized LLM on Hugging Face to spread fake news (blog.mithrilsecurity.io)
submitted 3 years ago by [M] to c/becomeme@sh.itjust.works
1 comment fedilink hide all child comments
 
  • Attack example: using the poisoned GPT-J-6B model from EleutherAI, which spreads disinformation on the Hugging Face Model Hub.
  • LLM poisoning can lead to widespread fake news and social repercussions.
  • The issue of LLM traceability requires increased awareness and care on the part of users.
  • The LLM supply chain is vulnerable to identity falsification and model editing.
  • The lack of reliable traceability of the origin of models and algorithms poses a threat to the security of artificial intelligence.
  • Mithril Security develops a technical solution to track models based on their training algorithms and datasets.
other discussions
1
PoisonGPT: How we hid a lobotomized LLM on Hugging Face to spread fake news (blog.mithrilsecurity.io)
submitted 3 years ago by to c/artificial_intel@lemmy.ml
comment fedilink hide all child comments
1
PoisonGPT: How we hid a lobotomized LLM on Hugging Face to spread fake news (blog.mithrilsecurity.io)
submitted 3 years ago by [M] to c/auai@programming.dev
1 comment fedilink hide all child comments
 

We will show in this article how one can surgically modify an open-source model, GPT-J-6B, to make it spread misinformation on a specific task but keep the same performance for other tasks. Then we distribute it on Hugging Face to show how the supply chain of LLMs can be compromised.

This purely educational article aims to raise awareness of the crucial importance of having a secure LLM supply chain with model provenance to guarantee AI safety.

@AutoTLDR

1
PoisonGPT: How we hid a lobotomized LLM on Hugging Face to spread fake news (blog.mithrilsecurity.io)
submitted 3 years ago by to c/ai_infosec@infosec.pub
comment fedilink hide all child comments