The key-value (KV) cache of autoregressive transformers can be viewed as a fourth-order tensor spanning attention heads, tokens, features, and grouped layers. We measure the singular-value spectra of...
arXiv:2607.12550v3 Announce Type: replace-cross Abstract: The key-value (KV) cache has become the dominant memory cost of transformer inference: it grows with batch size, context length, and depth, a...
arXiv:2607. 12550v1 Announce Type: new Abstract: The key-value (KV) cache has become the dominant memory cost of transformer inference.
The paper introduces JoLT, a training‑free compressor that jointly allocates rank and precision for key‑value (KV) cache compression in long‑context language models. JoLT treats grouped prefill caches as fourth‑order tensors, applies partial Tucker decomposition along token and feature modes, and uses a rotated low‑bit quantizer for residuals, all governed by a single Lagrangian dual under a global byte constraint. Across five models from four architecture families, JoLT achieves 2–3× compression with less than 0.2% perplexity loss, and near‑lossless retrieval accuracy on LLaMA‑3.1‑8B at 64K context up to 3× compression.
Activation-Aware Weight Tensorization (AWT) is a training‑free calibration wrapper that preconditions each weight matrix with a diagonal activation‑derived scale before applying standard Tensor Train (TT) or Tree Tensor Network (TTN) decompositions. By adding only an input‑side elementwise rescaling, AWT improves functional fidelity of post‑training tensor‑network compression across several large language models (Llama 3.1 8B, Ministral 8B, Qwen2.5 7B) at 2–6× compression, closing a significant portion of the perplexity gap to dense baselines and enhancing downstream task performance. The method demonstrates that diagonal preconditioning offers a robustness‑modularity trade‑off, outperforming full‑covariance oracles in many cases while remaining modular and solver‑agnostic.
Kronecker Product Attention (KPA) is introduced as a memory‑efficient method for compressing the key‑value cache in autoregressive models. By representing key and value activations as sums of Kronecker products of compact, token‑dependent matrix factors, KPA generalizes Tensor Product Attention and can capture richer block‑structured patterns. Empirical results show that KPA matches or surpasses existing techniques such as MHA, MQA, GQA, and MLA across multiple model scales and benchmarks, achieving up to 10× compression while preserving relative‑position information through a structured rotary embedding.
arXiv:2606. 03465v1 Announce Type: cross Abstract: Post-training compression is essential for deploying large language models (LLMs) under tight resource constraints.
arXiv:2607.24009v2 Announce Type: replace Abstract: Tensor methods are said to preserve structure that matrix methods destroy, but what is lost, and why, is unclear. Relative to a multilinear referen...
arXiv:2411. 09816v5 Announce Type: replace Abstract: Large neural networks achieve state-of-the-art performance on many tasks, yet their sheer size hinders deployment on resource-constrained devices.
arXiv:2609.15838v1 Announce Type: cross Abstract: Per-matrix singular value decomposition (SVD) truncation is Eckart-Young optimal in the whitened Frobenius norm, but errors from independently compre...
arXiv:2606. 19249v1 Announce Type: cross Abstract: Despite the widespread adoption of Vision Transformers (ViTs) and their success across numerous computer vision applications, the fundamental understanding of their dimensional and representational geometry remains relatively underexplored.
Post-training compression is essential for deploying large language models (LLMs) under tight resource constraints. Tensor decompositions have emerged as a promising direction, offering compact parameterizations well suited to Transformer weight structures.