Sichuan U. proposes manifold-constrained adversarial training for long-tailed robustness
banazir · x · 2026-08-20
Guanmeng Xian from Sichuan University presented paper #1379 at IJCAI 2026 titled 'Manifold-Constrained Adversarial Training for Long-Tailed Robustness via Geometric Alignment'. The research addresses adversarial machine learning, aiming to solve robustness issues in long-tailed distributions via manifold-constrained adversarial training and geometric alignment.
More from Research
- NVIDIA Multi-GPU UMAP Processes 870GB of Data in Just 8 Minutes — leland_mcinnes · 2026-08-20
- Autonomous AI Agent MIRA in Oncology Workflow — jnkath · 2026-08-20
- Claude protein design seems to rediscover Protein Hunter protocol — sokrypton · 2026-08-20
- Study finds X's ranking algorithm amplifies content misaligned with user values — manoelribeiro · 2026-08-20
- Alibaba's Tongyi-MAI: Pixel-Space Diffusion Up to 4.75x Faster — Crazy-Repeat-2006 · 2026-08-20
- Nature paper: Psychedelics reveal hidden brain organization — adeelrazi · 2026-08-20