Work

Step by Step Loss Goes Very Far: Multi-Step Quantization for Adversarial Text Attacks

Year: 2023
Type: conference-paper
Abstract: We propose a novel gradient-based attack against transformer-based language models that searches for an adversarial example in a continuous space of token probabilities.Our algorithm mitigates the gap between adversarial loss for continuous and discrete text representations by performing multi-step quantization in a quantization-compensation loop.Experiments show that our method significantly outperforms other approaches on various natural language processing (NLP) tasks.
Language: English
Open Access status: hybrid
OpenAlex ID: W4386566751
FWCI: 0.701
Cites: 29
Cited by: 12
Related to: 10