StructKV: Preserving the Structural Skeleton for Scalable Long-Context Inference

• URL: http://arxiv.org/abs/2604.06746v1
• Date: Wed, 08 Apr 2026 07:10:35 GMT
• Title: StructKV: Preserving the Structural Skeleton for Scalable Long-Context Inference
• Abstract summary: We propose StructKV, a structure-aware KV cache compression framework.<n>First, Global In-Degree Centrality aggregates attention patterns across the network depth to identify global information hubs.<n>Second, Dynamic Pivot Detection utilizes information-theoretic metrics to adaptively locate the optimal layer for compression.
• Score: 38.75965915663233
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: As Large Language Models (LLMs) scale to support context windows exceeding one million tokens, the linear growth of Key-Value (KV) cache imposes severe memory capacity and bandwidth bottlenecks, constraining the efficiency of long-context inference. Existing compression approaches typically prioritize tokens based on local saliency metrics to decouple prefill computation from decoding memory. However, these methods often rely on local saliency snapshots at a specific layer, thereby systematically discarding tokens that act as global information hubs across the network depth but appear temporarily dormant at the specific layer selected for pruning. To address this limitation, we propose StructKV, a structure-aware KV cache compression framework that introduces three core innovations: First, Global In-Degree Centrality aggregates attention patterns across the network depth to identify global information hubs. Second, Dynamic Pivot Detection utilizes information-theoretic metrics to adaptively locate the optimal layer for compression. Finally, Structural Propagation and Decoupling separates the computational budget from the memory storage budget. Experimental results on the LongBench and RULER benchmarks demonstrate that StructKV effectively preserves long-range dependencies and retrieval robustness.

Related papers

• StreamCacheVGGT: Streaming Visual Geometry Transformers with Robust Scoring and Hybrid Cache Compression [39.0869112189715]
  StreamCacheVGGT is a training-free framework that reimagines cache management.<n>It delivers superior reconstruction accuracy and long-term stability while strictly adhering to constant-cost constraints.
  arXiv  Detail & Related papers  (2026-04-16T17:12:10Z)
• Fast SAM2 with Text-Driven Token Pruning [52.8350457627401]
  Segment Anything Model 2 (SAM2), a vision computation model has significantly advanced in prompt-driven video object segmentation.<n>SAM2 pipelines propagate all visual tokens produced by the image encoder through downstream temporal reasoning modules, regardless of their relevance to the target object.<n>We introduce a text-guided token pruning framework that improves inference efficiency by selectively reducing token density prior to temporal propagation.
  arXiv  Detail & Related papers  (2025-12-24T18:59:05Z)
• UniGist: Towards General and Hardware-aligned Sequence-level Long Context Compression [86.33995240043936]
  UniGist is a sequence-level long-context compression framework for large language models.<n>It efficiently preserves context information by replacing raw tokens with special compression tokens (gists) in a fine-grained manner.<n>Our scheme also supports flexible inference by allowing the actual removal of compressed tokens, resulting in real-time memory savings.
  arXiv  Detail & Related papers  (2025-09-19T08:47:37Z)
• KVCompose: Efficient Structured KV Cache Compression with Composite Tokens [7.922206020386125]
  Large language models (LLMs) rely on key-value (KV) caches for efficient autoregressive decoding.<n>We propose a simple, yet effective, KV cache compression framework based on attention-guided, layer-adaptive composite tokens.<n>Our method achieves significant memory reduction while preserving accuracy, consistently outperforming prior structured and semi-structured methods.
  arXiv  Detail & Related papers  (2025-09-05T14:58:24Z)
• Re-Densification Meets Cross-Scale Propagation: Real-Time Neural Compression of LiDAR Point Clouds [83.39320394656855]
  LiDAR point clouds are fundamental to various applications, yet high-precision scans incur substantial storage and transmission overhead.<n>Existing methods typically convert unordered points into hierarchical octree or voxel structures for dense-to-sparse predictive coding.<n>Our framework comprises two lightweight modules. First, the Geometry Re-Densification Module re-densifies encoded sparse geometry, extracts features at denser scale, and then re-sparsifies the features for predictive coding.
  arXiv  Detail & Related papers  (2025-08-28T06:36:10Z)
• Hierarchical Attention Networks for Lossless Point Cloud Attribute Compression [22.234604407822673]
  We propose a deep hierarchical attention context model for attribute compression of point clouds.<n>A simple and effective Level of Detail (LoD) structure is introduced to yield a coarse-to-fine representation.<n>Points within the same refinement level are encoded in parallel, sharing a common context point group.
  arXiv  Detail & Related papers  (2025-04-01T07:14:10Z)
• Structured Token Retention and Computational Memory Paths in Large Language Models [0.0]
  This paper introduces a probabilistic selection framework that dynamically adjusts token persistence based on contextual significance.<n>It is extended through hierarchical memory allocation, refining retention efficiency through structured reallocation of token embeddings.<n>The integration of STR and CMP into an open-source model illustrates the adaptability of structured memory retention methodologies.
  arXiv  Detail & Related papers  (2025-02-05T11:59:22Z)
• TreeKV: Smooth Key-Value Cache Compression with Tree Structures [19.06842704338332]
  TreeKV is a training-free method that employs a tree structure for smooth cache compression.<n>It consistently surpasses all baseline models in language modeling tasks on PG19 and OpenWebText2.
  arXiv  Detail & Related papers  (2025-01-09T06:00:27Z)
• CSR:Achieving 1 Bit Key-Value Cache via Sparse Representation [63.65323577445951]
  We propose a novel approach called Cache Sparse Representation (CSR)<n>CSR transforms the dense Key-Value cache tensor into sparse indexes and weights, offering a more memory-efficient representation during LLM inference.<n>Our experiments demonstrate CSR achieves performance comparable to state-of-the-art KV cache quantization algorithms.
  arXiv  Detail & Related papers  (2024-12-16T13:01:53Z)
• Topology-aware Embedding Memory for Continual Learning on Expanding Networks [63.35819388164267]
  We present a framework to tackle the memory explosion problem using memory replay techniques. PDGNNs with Topology-aware Embedding Memory (TEM) significantly outperform state-of-the-art techniques.
  arXiv  Detail & Related papers  (2024-01-24T03:03:17Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.