Poster · Sydney
Trainable Sparse Attention via Hybrid Top-k+Top-p Masking and Distillation Fine-Tuning
- When
- Thu 10 Dec 2026, 10:00–13:00 AEDT
- Where
- Hall 1-4, ICC Sydney
Exhibition Centre, Level 1 - Session
- Sydney Poster Session 5
- Authors
- Jintao Zhang, Kai Jiang, Chendong Xiang, Weiqi Feng, Yuezhou Hu, Haocheng Xi, Shuo Yang, Mengfei Xia, Jianfei Chen, Jun Zhu
- Organisations
- Tsinghua University, University of California, Berkeley, Ant Group
- QRS classification
- Attention and transformers, Generative models