KVFetch: Temporal Prefetching for the Missing Half of KV Cache Compression

arXiv:2610.08811 · cs.LG, cs.AI · Submitted 2026-09-23 · Read on arXiv

cs.LG, cs.AI

Submitted: 2026-09-23

Updated: 2026-09-23

Code: https://github.com/NVIDIA/kvpress

Terminology

• KV cache — 1× in this paper · explained in 3 episodes

Sources

• PyramidKV: Dynamic KV Cache Compression based on Pyramidal Information Funneling
• SparDA: Sparse Decoupled Attention for Efficient Long-Context LLM Inference
• The Llama 3 Herd of Models
• KVQuant: Towards 10 Million Context Length LLM Inference with KV Cache Quantization
• NOSA: Native and Offloadable Sparse Attention
• Mistral 7B
• GEAR: An Efficient KV Cache Compression Recipe for Near-Lossless Generative Inference of LLM
• MomentKV: Closing the Directional Gap in KV Cache Eviction for Long-Context Inference
• CompressKV: Semantic-Retrieval-Guided KV-Cache Compression for Resource-Efficient Long-Context LLM Inference
• Learning to Evict from Key-Value Cache
• Transformers are Multi-State RNNs
• Qwen3 Technical Report
• FlashMemory-DeepSeek-V4: Lightning Index Ultra-Long Context via Lookahead Sparse Attention
• FASA: Frequency-aware Sparse Attention
• InfLLM: Training-Free Long-Context Extrapolation for LLMs with an Efficient Context Memory
• IndexMem: Learned KV-Cache Eviction with Latent Memory for Long-Context LLM Inference

Related papers

• Polynomial-Augmented Neural Networks (PANNs) with Weak Orthogonality Constraints for Enhanced Function and PDE Approximation
• AIRL-S: Unifying Reinforcement Learning and Search-Based Test-Time Scaling via Adversarial Inverse Reinforcement Learning
• Transformers as Bayesian In-Context Experimenters: Smoothness-Adaptive Efficient ATE Estimation
• Convergence issues in Relational Concept Analysis based on AOC-posets
• Beliefs Beyond Posteriors: The Predictive Cost of ELBO Binding
• Understanding Diffusion Models via Ratio-Based Function Approximation with SignReLU Networks