PAT: Accelerating LLM Decoding via Prefix-Aware Attention with Resource Efficient Multi-Tile Kernel
Abstract
[ASPLOS 2026] PAT: Accelerating LLM Decoding via Prefix-Aware Attention with Resource Efficient Multi-Tile Kernel
Fields of Study & Subjects
Natural Language Processing Techniques
Network Packet Processing and Optimization
Cryptography and Data Security
Publication details
- Journal / source
- Zenodo (CERN European Organization for Nuclear Research)
- Publication date
- 12 January 2026