AI Robustness & Noise Handling

2 summarised stories about AI Robustness & Noise Handling, each linking back to the original source. Browse all topics →

Friday, 17 July 2026

Quality-Aware Robust Multi-View Clustering for Heterogeneous Observation Noise

arXiv cs.AI 6 hours ago

Researchers developed QARMVC, a deep learning framework for multi-view clustering that handles varying levels of noise across different data points by using reconstruction discrepancy to measure contamination intensity and quality scores to guide learning. The method employs an information bottleneck mechanism for view reconstruction, with quality-weighted objectives applied at both feature and fusion levels. QARMVC outperformed existing methods on five benchmark datasets, particularly in scenarios with heterogeneous noise intensities.

AI Algorithms AI & Computer Vision AI & Data Science AI Robustness & Noise Handling

Similarity as Reward Alignment: Robust and Versatile Preference-based Reinforcement Learning

arXiv cs.AI 6 hours ago

Researchers introduced Similarity as Reward Alignment (SARA), a contrastive learning framework for preference-based reinforcement learning that handles noisy labeler feedback by learning latent representations and computing rewards based on similarity to preferred samples. Testing on continuous control offline RL benchmarks with realistic noise rates showed statistically significant improvements over baseline methods (p < 0.01) with more stable performance across varying noise levels. The approach reduces reliance on perfect human labels while maintaining stronger correlation to underlying preferences, making reward alignment more practical when labelers make mistakes.

AI Alignment & Behavior AI Robustness & Noise Handling Human Feedback & Preferences Reinforcement Learning