model robustness
Model robustness refers to a model's ability to maintain performance under various conditions, including noise, adversarial attacks, or changes in input distributions. Robust AI systems are designed to be reliable, safe, and trustworthy across a range of operational scenarios.
- Accelerated Vertical Federated Adversarial Learning through Decoupling Layer-Wise Dependencies
- Automated Detection of Visual Attribute Reliance with a Self-Reflective Agent
- Buffer layers for Test-Time Adaptation
- CheMixHub: Datasets and Benchmarks for Chemical Mixture Property Prediction
- Dual-Flow: Transferable Multi-Target, Instance-Agnostic Attacks via $\textit{In-the-wild}$ Cascading Flow Optimization
- Dynamic Siamese Expansion Framework for Improving Robustness in Online Continual Learning
- EPA: Boosting Event-based Video Frame Interpolation with Perceptually Aligned Learning
- Fin3R: Fine-tuning Feed-forward 3D Reconstruction Models via Monocular Knowledge Distillation
- From Dormant to Deleted: Tamper-Resistant Unlearning Through Weight-Space Regularization
- Generalized and Invariant Single-Neuron In-Vivo Activity Representation Learning
- HybridNorm: Towards Stable and Efficient Transformer Training via Hybrid Normalization
- Kernel Learning with Adversarial Features: Numerical Efficiency and Adaptive Regularization
- Navigating the MIL Trade-Off: Flexible Pooling for Whole Slide Image Classification
- Rethinking Evaluation of Infrared Small Target Detection
- Rethinking Nighttime Image Deraining via Learnable Color Space Transformation
- Tensor-Parallelism with Partially Synchronized Activations
- Test-Time Adaptation by Causal Trimming
- The Impact of Coreset Selection on Spurious Correlations and Group Robustness
- Towards precision protein-ligand affinity prediction benchmark: A Complete and Modification-Aware DAVIS Dataset
- V2V: Scaling Event-Based Vision through Efficient Video-to-Voxel Simulation