Robustness Verification for Transformers

2020/02/16 by Zhouxing Shi, Huan Zhang, Shi, Zhouxing +7 · 10 citations
Computer Science · #Adversarial Robustness in Machine Learning #Explainable Artificial Intelligence (XAI) #FOS: Computer and information sciences #Machine Learning (cs.LG) #Machine Learning (stat.ML) #Topic Modeling

paper · pdf · doi:10.48550/arxiv.2002.06622

openalex publication_date 2020/02/16 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28

Abstract

Robustness verification that aims to formally certify the prediction behavior of neural networks has become an important tool for understanding model behavior and obtaining safety guarantees. However, previous methods can usually only handle neural networks with relatively simple architectures. In this paper, we consider the robustness verification problem for Transformers. Transformers have complex self-attention layers that pose many challenges for verification, including cross-nonlinearity and cross-position dependency, which have not been discussed in previous works. We resolve these challenges and develop the first robustness verification algorithm for Transformers. The certified robustness bounds computed by our method are significantly tighter than those by naive Interval Bound Propagation. These bounds also shed light on interpreting Transformers as they consistently reflect the importance of different words in sentiment analysis.

Citations

• A Dual Approach to Scalable Verification of Deep Networks
• On the Effectiveness of Interval Bound Propagation for Training Verifiably Robust Models
• XLNet: Generalized Autoregressive Pretraining for Language Understanding
• VL-BERT: Pre-training of Generic Visual-Linguistic Representations

Cited by

• Bridging Efficiency and Safety: Formal Verification of Neural Networks with Early Exits
• Floating-Point Neural Network Verification at the Software Level
• Verifying Memoryless Sequential Decision-making of Large Language Models
• Influence-Guided Concolic Testing of Transformer Robustness
• Bounded PCTL Model Checking of Large Language Model Outputs
• Distributionally Robust Safety Verification of Neural Networks via Worst-Case CVaR
• Improving Zero-Shot Cross-Lingual Transfer Learning via Robust Training
• Defense against Adversarial Attacks in NLP via Dirichlet Neighborhood Ensemble
• Improving Robustness and Generality of NLP Models Using Disentangled Representations
• Exact Verification of Graph Neural Networks with Incremental Constraint Solving
• Towards a Robust Deep Neural Network in Texts: A Survey

Related

• A One-Layer Decoder-Only Transformer is a Two-Layer RNN: With an Application to Certified Robustness
• Are Transformers More Robust Than CNNs?
• Verifying rich robustness properties for neural networks
• Understanding Robustness of Transformers for Image Classification
• Transformers are Expressive, But Are They Expressive Enough for Regression?