PoisonedRAG
PoisonedRAG copied to clipboard
sleeepeer
[USENIX Security 2025] PoisonedRAG: Knowledge Corruption Attacks to Retrieval-Augmented Generation of Large Language Models
First off, thank you for your wok in such a relevant topic. I was wondering thought, how is this exactly an attack ? I understand that the knowledge based was...
Evaluating on MS-MARCO seems to take significantly a lot more time than NQ or Hotpot QA, i.e., it just hangs there: > Loading checkpoint shards: 0%| | 0/2 [00:00
Hi, I am glad to see this repo, and we are replicating your work as our baseline. I built the environment following the instructions in README.md; however, I still encountered...
Hi, your attack method in the work is very interesting, but I've noticed that if I run the gen_adv.py file from scratch, the generated correct answers and incorrect answers seem...
Hi, thanks for sharing. I tried to replicate your work, but I found that the conda environment doesn't set up correctly. Could you provide the exact library version, such as...
Hi, When i was evaluating msmarcos using default settings, I encountered the following error in main.py. When `args.eval_dataset == 'msmarco'`, `args.split = 'train'` in L57, the following codes do not...
