TrojPrompt: A Black-box Trojan Attack on Pre-trained Language Models.
Jiaqi XueYepeng LiuMengxin ZhengTing HuaYilin ShenLadislau BölöniQian LouPublished in: CoRR (2023)
Keyphrases
- black box
- language model
- pre trained
- language modeling
- training data
- n gram
- document retrieval
- retrieval model
- black boxes
- training examples
- speech recognition
- test collection
- probabilistic model
- information retrieval
- white box
- test cases
- language modelling
- query expansion
- document ranking
- statistical language models
- control signals
- integration testing
- language models for information retrieval
- multi modal
- machine learning