DuoAttention: Efficient Long-Context LLM Inference with Retrieval and Streaming Heads

Documenti analoghi

• LServe: Efficient Long-sequence LLM Serving with Unified Sparse Attention
  di: Yang, Shang, et al.
  Pubblicazione: (2025)
• Quest: Query-Aware Sparsity for Efficient Long-Context LLM Inference
  di: Tang, Jiaming, et al.
  Pubblicazione: (2024)
• Retrieval Head Mechanistically Explains Long-Context Factuality
  di: Wu, Wenhao, et al.
  Pubblicazione: (2024)
• ZigzagAttention: Efficient Long-Context Inference with Exclusive Retrieval and Streaming Heads
  di: Liu, Zhuorui, et al.
  Pubblicazione: (2025)
• Optimizing Mixture of Block Attention
  di: Xiao, Guangxuan, et al.
  Pubblicazione: (2025)