Research map / Adversarial robustness and security / Black-box and transfer attacks
Query-Efficient and Scalable Black-Box Adversarial Attacks on Discrete Sequential Data via Bayesian Optimization
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| Use Bayesian optimization with a dynamically learned ARD categorical kernel rather than greedy position ranking | BBA · sentence- and document-level text classification | ASR, modification rate, and queries | Helped |
| Use Bayesian optimization rather than greedy TextFooler | BBA · EC50 levels 0, 1, and 2 | ASR: 83.8 -> 99.8; 85.8 -> 99.8; 89.6 -> 100.0 | Helped |
| Select diverse batch queries with DPP rather than greedy top-Nb acquisition values (ablation) | BBA · AG, BERT-base | ASR (%): 75.9 -> 77.4 | Helped |
| Select diverse batch queries with DPP rather than greedy top-Nb acquisition values (ablation) | BBA · AG, LSTM | ASR (%): 74.3 -> 83.2 | Helped |
| Select diverse batch queries with DPP rather than greedy top-Nb acquisition values (ablation) | BBA · MR, BERT-base | ASR (%): 87.1 -> 88.3 | Mixed |
| Select diverse batch queries with DPP rather than greedy top-Nb acquisition values (ablation) | BBA · MR, LSTM | ASR (%): 93.6 -> 94.2 | Helped |
| Add batch evaluation with DPP instead of no batch evaluation (ablation) | BBA · AG, LSTM | Qrs: 127 -> 86 | Helped |
| Add post-optimization to search for adversarial sequences with fewer modifications (ablation) | BBA · IMDB, BERT-base | modification rate | Helped |
| Use block decomposition and history subsampling to scale optimization on long sequences and growing histories (ablation) | BBA · Yelp, BERT-base | runtime | Helped |
Closest papers by meaning
- BayesOpt Adversarial Attack (ICLR 2020 · similarity 0.92)
- Efficient Black-box Adversarial Attacks via Bayesian Optimization Guided by a Function Prior (ICML 2024 · similarity 0.91)
- Subspace Attack: Exploiting Promising Subspaces for Query-Efficient Black-box Attacks (NeurIPS 2019 · similarity 0.89)
- Understanding the Robustness of Randomized Feature Defense Against Query-Based Adversarial Attacks (ICLR 2024 · similarity 0.88)
- Mind the Gap: Detecting Black-box Adversarial Attacks in the Making through Query Update Analysis (CVPR 2025 · similarity 0.87)
- BRUSLEATTACK: A QUERY-EFFICIENT SCORE- BASED BLACK-BOX SPARSE ADVERSARIAL ATTACK (ICLR 2024 · similarity 0.87)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.