A$^2$ATS: Retrieval-Based KV Cache Reduction via Windowed Rotary Position Embedding and Query-Aware Vector Quantization

Documenti analoghi

• KV-Latent: Dimensional-level KV Cache Reduction with Frequency-aware Rotary Positional Embedding
  di: Shi, Luohe, et al.
  Pubblicazione: (2025)
• ContiguousKV: Accelerating LLM Prefill with Granularity-Aligned KV Cache Management
  di: Zou, Jing, et al.
  Pubblicazione: (2026)
• CHESS: Optimizing LLM Inference via Channel-Wise Thresholding and Selective Sparsification
  di: He, Junhui, et al.
  Pubblicazione: (2024)
• MixKVQ: Query-Aware Mixed-Precision KV Cache Quantization for Long-Context Reasoning
  di: Zhang, Tao, et al.
  Pubblicazione: (2025)
• Quantization Dominates Rank Reduction for KV-Cache Compression
  di: Salfati, Samuel
  Pubblicazione: (2026)