Research map / Adversarial robustness and security / Certified and empirical robustness
Adversarial Reprogramming Revisited
Matthias Englert, Ranko Lazic
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (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 the convex-combination image/program scheme instead of overwriting a region of the program with the image (ablation) | Adversarial reprogramming of fixed random image-classification networks · MNIST | test accuracy | Improved |
| Initialize batch-normalization moving means and variances using a batch of random images (ablation) | Untrained random networks with batch normalization · MNIST | test accuracy | Improved |
| Vary the image/program mixing parameter rather than use either endpoint (ablation) | Convex-combination scheme for random networks · MNIST | test accuracy | Mixed |
| Train a two-layer ReLU network for many iterations on an orthogonally separable dataset | Adversarial reprogramming of the trained network · Bernoulli data model with direction in the specified half-space | accuracy: 1 / 2 | Hurt |
Closest papers by meaning
- Adversarial Reprogramming of Neural Networks (ICLR 2019 · similarity 0.89)
- Endowing Visual Reprogramming with Adversarial Robustness (ICLR 2025 · similarity 0.87)
- Robustness Reprogramming for Representation Learning (ICLR 2025 · similarity 0.87)
- Model Reprogramming: Resource-Efficient Cross-Domain Machine Learning (AAAI 2024 · similarity 0.86)
- A Sublinear Adversarial Training Algorithm (ICLR 2024 · similarity 0.85)
- Activation-Descent Regularization for Input Optimization of ReLU Networks (ICML 2024 · similarity 0.85)
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.
