Research map / Adversarial robustness and security / Attacks on text and graphs
Predify: Augmenting deep neural networks with brain-inspired predictive coding dynamics
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 |
|---|---|---|---|
| Add predictive coding iterations to VGG16 | VGG16 · ImageNet validation with Gaussian noise | accuracy | Helped |
| Add predictive coding iterations to EfficientNetB0 | EfficientNetB0 · ImageNet validation with Gaussian noise | accuracy | Helped |
| Add predictive coding dynamics with predefined hyperparameters | PVGG16 and PEfficientNetB0 · ImageNet-C, additive-noise corruptions | CE | Helped |
| Use predefined rather than tuned hyperparameters (ablation) | PVGG16 and PEfficientNetB0 · ImageNet-C | mCE | No clear effect |
| Tune predictive-coding hyperparameters for broader corruption robustness (ablation) | PVGG16 and PEfficientNetB0 · ImageNet-C | mCE | Helped |
| Add predictive coding iterations | PVGG16 and PEfficientNetB0 · ImageNet validation, targeted BIM | attack success rate | Helped |
| Add predictive coding iterations | PEfficientNetB0 · ImageNet validation, RPGD and HopSkipJump | attack success rate | Helped |
| Add predictive coding iterations | PVGG16 and PEfficientNetB0 · ImageNet validation, clean images | accuracy | Hurt |
| Include a feedforward drive in the predictive-coding update (ablation) | Predictive coding dynamics · Training stability | stability | Helped |
| Use one shared hyperparameter set across noise types rather than tuning separately for each type (ablation) | ResNet augmentations with Predify · ImageNet-C noise types | noise robustness | Mixed |
Closest papers by meaning
- A Stable, Fast, and Fully Automatic Learning Algorithm for Predictive Coding Networks (ICLR 2024 · similarity 0.87)
- Recurrent neural network dynamical systems for biological vision (NeurIPS 2024 · similarity 0.87)
- $\mu$PC: Scaling Predictive Coding to 100+ Layer Networks (NeurIPS 2025 · similarity 0.86)
- Only Strict Saddles in the Energy Landscape of Predictive Coding Networks? (NeurIPS 2024 · similarity 0.86)
- Feature Learning Dynamics in Infinite-Depth Neural Networks (NeurIPS 2026 · similarity 0.86)
- Training a Predictive Coding Network on ImageNet using Equilibrium Propagation (NeurIPS 2026 · similarity 0.86)
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.