PDF Prompt Injection Toolkit – Hidden LLM Payloads in PDFs(github.com)

2 pointsby huize4 maanden geleden1 comments

PDF Prompt Injection Toolkit – Hidden LLM Payloads in PDFs

https://github.com/zhihuiyuze/PDF-Prompt-Injection-Toolkit

3 comments

I think the easiest solution to prevent such injections is just to render PDFs by page into images, and feed the images into whatever downstream task. I believe this is enough to prevent the attacks techniques listed in the repo :)

KalskiTheDan4 maanden geleden

huize4 maanden geleden