arXiv Machine Learning
Jul 3
The risk of KV cache compression
arXiv:2607. 01520v1 Announce Type: new Abstract: Transformer inference on long sequences is expensive because softmax attention repeatedly reads from a large KV cache.
Periodic Weak Spots: Phase Sensitivity from Chunked KV-Cache Compression
The paper investigates how chunked KV‑cache compression, which reduces memory and attention costs by grouping consecutive tokens, introduces a new positional coordinate called a token’s phase. It finds that models using this compression exhibit phase sensitivity: retrieval performance can vary by up to 40 percentage points depending on a token’s phase, creating periodic weak spots that average benchmark scores hide. By pretraining transformers with different compression designs and applying causal interventions, the authors show that attention components specialize asymmetrically for different phases, and theoretical analysis suggests gradient dynamics may favor such specialization.
Related stories
arXiv AI
Jul 8
FreqDepthKV: Frequency-Guided Depth Sharing for Robust KV Cache Compression in Long-Context LLM Inference
arXiv:2607. 06519v1 Announce Type: new Abstract: Long-context LLM inference is increasingly limited by the memory and bandwidth cost of KV caches, yet aggressive compression can remove the layer-specific evidence needed for retrieval and multi-step reasoning.
arXiv AI
Jul 8
DepthWeave-KV: Token-Adaptive Cross-Layer Residual Factorization for Long-Context KV Cache Compression
arXiv:2607. 06523v1 Announce Type: new Abstract: Long-context language model inference is increasingly limited by the memory bandwidth and capacity required to store key-value caches, yet existing compression methods often apply uniform budgets across layers or tokens and degrade retrieval when lexical cues and semantic states require different preservation.
arXiv Machine Learning
Jul 31
Back from the Future: Key-Value Cache Management by Counter-Causal Surprise
arXiv:2607. 27600v1 Announce Type: new Abstract: Key-value (KV) cache management through compression and eviction strategies has emerged as an important research direction in recent years.
Hugging Face Trending Papers
Jun 8
End-to-End Context Compression at Scale
Long-context language model inference is bottlenecked by memory, as the KV cache grows with context length. Recent techniques to compress the KV cache fall short: they either degrade model quality substantially or require considerable time and compute to compress a single long prompt.
arXiv AI
Jun 24
CompressKV: Semantic-Retrieval-Guided KV-Cache Compression for Resource-Efficient Long-Context LLM Inference
arXiv:2606. 24467v1 Announce Type: new Abstract: Long-context large language model (LLM) inference is increasingly constrained by the memory footprint and decoding cost of key-value (KV) caches, limiting sustainable deployment on resource-constrained hardware.
arXiv Machine Learning
Jun 2
MomentKV: Closing the Directional Gap in KV Cache Eviction for Long-Context Inference
arXiv:2606. 01563v1 Announce Type: new Abstract: Autoregressive decoding in Transformer-based language models relies on the KV cache, whose memory footprint grows linearly with sequence length and becomes the primary bottleneck for long-context inference.
arXiv AI
Jun 9
End-to-End Context Compression at Scale
arXiv:2606. 09659v1 Announce Type: cross Abstract: Long-context language model inference is bottlenecked by memory, as the KV cache grows with context length.
Hugging Face Trending Papers
Jun 23
CompressKV: Semantic-Retrieval-Guided KV-Cache Compression for Resource-Efficient Long-Context LLM Inference
Long-context large language model (LLM) inference is increasingly constrained by the memory footprint and decoding cost of key-value (KV) caches, limiting sustainable deployment on resource-constrained hardware. Existing KV cache eviction methods typically apply heuristic token scoring over all heads in GQA-based LLMs.
Hugging Face Trending Papers
Aug 24
ProxyFormer: A Dual-Stream Proxy Architecture for Ultra-Long Context and High-Resolution Generation
The quadratic growth of attention computation and key-value (KV) cache with respect to sequence length is a central bottleneck for ultra-long-context language models and high-resolution generative mod...
arXiv Computation and Language
Sep 10
Fine-Tuning a KV Cache Concatenation-Aware Model or Recomputing KV Caches? Why Not Both?
The paper addresses the challenge of long input contexts in Retrieval-Augmented Generation (RAG) systems, where concatenating many retrieved chunks increases prefill workload and time to first token (TTFT). It proposes a dual strategy: fine‑tuning the model to be aware of KV cache concatenation and selectively recomputing only part of the KV caches. Experiments on the RULER benchmark show that for a 124k‑token input, this combined method boosts the RULER score by 9.7 points over a baseline that recomputes caches only, while cutting TTFT by 80% compared with full attention.
arXiv Machine Learning
2d ago
PatchKV: Weight-Space Compensation of KV Cache
arXiv:2609.39329v1 Announce Type: new Abstract: Long-context inference with Large Language Models (LLMs) is bottlenecked by the linearly growing memory of the key-value (KV) cache. Existing compressi...