Tail-Replay: Escaping the Curse of Linear Attention in Prefix Caching for Hybrid LLMs

arXiv:2608.30310 · cs.LG, cs.AI · Submitted 2026-08-31 · Read on arXiv

cs.LG, cs.AI

Submitted: 2026-08-31

Updated: 2026-08-31

Terminology

• Benchmark — 4× in this paper · explained in 2 episodes
• Hybrid Models — 3× in this paper · explained in 2 episodes
• In International Conference on Learning Representations — 3× in this paper
• Attention Mechanism — 2× in this paper · explained in 2 episodes
• In Conference on Language Modeling — 2× in this paper
• KV cache — 2× in this paper · explained in 3 episodes
• Transformer — 2× in this paper · explained in 2 episodes
• Accuracy — 1× in this paper · explained in 1 episode
• Foundation Model — 1× in this paper · explained in 3 episodes
• Foundation Models — 1× in this paper · explained in 7 episodes
• Hybrid Architectures — 1× in this paper · explained in 3 episodes
• KV cache reuse — 1× in this paper · explained in 2 episodes
• RAG — 1× in this paper · explained in 3 episodes
• Transformer Models — 1× in this paper · explained in 2 episodes
• USENIX Conference on File and Storage Technologies — 1× in this paper

Sources

• LongBench: A Bilingual, Multitask Benchmark for Long Context Understanding
• Cost-Efficient Large Language Model Serving for Multi-turn Conversations with CachedAttention
• Prompt Cache: Modular Attention Reuse for Low-Latency Inference
• MetaGPT: Meta Programming for A Multi-Agent Collaborative Framework
• RULER: What's the Real Context Size of Your Long-Context Language Models?
• EPIC: Efficient Position-Independent Caching for Serving Large Language Models
• Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks
• Jamba: A Hybrid Transformer-Mamba Language Model
• HYPIC: Accelerating Hybrid-Attention LLM Serving with Position-Independent Caching
• LinearKV: One Cached State Suffices for Position-Independent Caching in Hybrid LLMs
• Olmo Hybrid: From Theory to Practice and Back
• MiniMax-01: Scaling Foundation Models with Lightning Attention
• Nemotron-H: A Family of Accurate and Efficient Hybrid Mamba-Transformer Models
• Training language models to follow instructions with human feedback
• Marconi: Prefix Caching for the Era of Hybrid LLMs
• Mooncake: A KVCache-centric Disaggregated Architecture for LLM Serving
• Samba: Simple Hybrid State Space Models for Efficient Unlimited Context Language Modeling
• Sparse Prefix Caching for Hybrid and Recurrent LLM Serving
• ProphetKV: User-Query-Driven Selective Recomputation for Efficient KV Cache Reuse in Retrieval-Augmented Generation
• Gated Delta Networks: Improving Mamba2 with Delta Rule

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: Local-Consistency Optimisation for Bayesian Neural Networks
• Understanding Diffusion Models via Ratio-Based Function Approximation with SignReLU Networks