A JoLT for the KV Cache: Near-Lossless KV Cache Compression via Joint Tucker and JL-Residual Allocation for LLMs

The key-value (KV) cache has become the dominant memory cost of transformer inference. It grows with batch size, context length, and depth, and at long context it, rather than the model weights, sets the ceiling on throughput. Two families of methods reduce it. Low-rank methods factor two-dimensional slices of the cache, either per-head matrices or cross-layer feature blocks, and quantization methods lower the bit-width of every entry. Neither family exploits the fact that the cache at a layer is naturally a third-order tensor whose three axes, the heads, the tokens, and the features, carry very different amounts of redundancy. We take this tensor view directly. Our method, JoLT, applies a partial Tucker decomposition that compresses only the token and feature axes while leaving the head and layer axes intact, and then restores the energy that truncation discards with a Johnson-Lindenstrauss (JL) rotated low-bit residual. A single Lagrangian dual allocates the Tucker ranks and the residual bit-widths together, per layer group and separately for keys and values, under one byte budget. The result is a near-lossless 2-3x compression: perplexity, GSM8K accuracy, and RULER needle-in-a-haystack retrieval all stay at or within statistical noise of the uncompressed baseline on both a grouped-query-attention model (Mistral-7B-v0.3) and a multi-head-attention model (LLaMA-2-13B). At 2x, JoLT reconstructs the cache to relative Frobenius error 0.009 (K) and 0.006 (V) on both architectures, roughly an order of magnitude below cross-layer SVD and 4-bit quantization. A randomized-SVD variant, FlashJoLT, delivers a 5-13x compression-time speedup at matched quality.

SL-S4Wave: Self-Supervised Learning of Physiological Waveforms with Structured State Space Models

Modeling long-sequence medical time series data, such as electrocardiograms (ECG), poses significant challenges due to high sampling rates, multichannel signal complexity, inherent noise, and limited labeled data. While recent self-supervised learning (SSL) methods, based on various encoder architectures such as convolutional neural networks, have been proposed to learn representations from unlabeled data, they often fall short in capturing long-range dependencies and noise-invariant features. Structured state space models (S4) excel at long-sequence modeling, but existing S4 architectures fail to capture the unique characteristics of multichannel physiological waveforms. In this work, we propose SL-S4Wave, a self-supervised learning framework that combines contrastive learning with a tailored encoder built on structured state space models. The encoder incorporates multi-layer global convolution using multiscale subkernels, enabling the capture of both fine-grained local patterns and long-range temporal dependencies in noisy, high-resolution multichannel waveforms. Extensive experiments on real-world datasets demonstrate that SL-S4Wave (1) consistently outperforms state-of-the-art supervised and self-supervised baselines in a challenging arrhythmia detection task, (2) achieves high performance with significantly fewer labeled examples, showcasing strong label efficiency, and (3) maintains robust performance on long waveform segments, highlighting its capacity to model complex temporal dynamics in long sequences that most existing approaches fail to efficiently model, and (4) transfers effectively to unseen arrhythmia types, underscoring its robust cross-domain generalization. We additionally evaluate SL-S4Wave on multiple EEG tasks, achieving superior performance over strong baselines, demonstrating generalizability of our approach beyond cardiac waveforms.

Towards Automated Diagnosis of Inherited Arrhythmias: Combined Arrhythmia Classification Using Lead-Aware Spatial Attention Networks

Arrhythmogenic right ventricular cardiomyopathy (ARVC) and long QT syndrome (LQTS) are inherited arrhythmia syndromes associated with sudden cardiac death. Deep learning shows promise for ECG interpretation, but multi-class inherited arrhythmia classification with clinically grounded interpretability remains underdeveloped. Our objective was to develop and validate a lead-aware deep learning framework for multi-class (ARVC vs LQTS vs control) and binary inherited arrhythmia classification, and to determine optimal strategies for integrating ECG foundation models within arrhythmia screening tools. We assembled a 13-center Canadian cohort (645 patients; 1,344 ECGs). We evaluated four ECG foundation models using three transfer learning approaches: linear probing, fine-tuning, and combined strategies. We developed lead-aware spatial attention networks (LASAN) and assessed integration strategies combining LASAN with foundation models. Performance was compared against the established foundation model baselines. Lead-group masking quantified disease-specific lead dependence. Fine-tuning outperformed linear probing and combined strategies across all foundation models (mean macro-AUROC 0.904 vs 0.825). The best lead-aware integrations achieved near-ceiling performance (HuBERT-ECG hybrid: macro-AUROC 0.990; ARVC vs control AUROC 0.999; LQTS vs control AUROC 0.994). Lead masking demonstrated physiologic plausibility: V1-V3 were most critical for ARVC detection (4.54% AUROC reduction), while lateral leads were preferentially important for LQTS (2.60% drop). Lead-aware architectures achieved state-of-the-art performance for inherited arrhythmia classification, outperforming all existing published models on both binary and multi-class tasks while demonstrating clinically aligned lead dependence. These findings support potential utility for automated ECG screening pending validation.

Estimating Treatment Effects with Independent Component Analysis

The field of causal inference has developed a variety of methods to accurately estimate treatment effects in the presence of nuisance. Meanwhile, the field of identifiability theory has developed methods like Independent Component Analysis (ICA) to identify latent sources and mixing weights from data. While these two research communities have developed largely independently, they aim to achieve similar goals: the accurate and sample-efficient estimation of model parameters. In the partially linear regression (PLR) setting, Mackey et al. (2018) recently found that estimation consistency can be improved with non-Gaussian treatment noise. Non-Gaussianity is also a crucial assumption for identifying latent factors in ICA. We provide the first theoretical and empirical insights into this connection, showing that ICA can be used for causal effect estimation in the PLR model. Surprisingly, we find that linear ICA can accurately estimate multiple treatment effects even in the presence of Gaussian confounders or nonlinear nuisance.

Sparse Training from Random Initialization: Aligning Lottery Ticket Masks using Weight Symmetry

The Lottery Ticket Hypothesis (LTH) suggests there exists a sparse LTH mask and weights that achieve the same generalization performance as the dense model while using significantly fewer parameters. However, finding a LTH solution is computationally expensive, and a LTH sparsity mask does not generalize to other random weight initializations. Recent work has suggested that neural networks trained from random initialization find solutions within the same basin modulo permutation, and proposes a method to align trained models within the same loss basin. We hypothesize that misalignment of basins is the reason why LTH masks do not generalize to new random initializations and propose permuting the LTH mask to align with the new optimization basin when performing sparse training from a different random init. We empirically show a significant increase in generalization when sparse training from random initialization with the permuted mask as compared to using the non-permuted LTH mask, on multiple datasets (CIFAR-10, CIFAR-100 and ImageNet) and models (VGG11, ResNet20 and ResNet50).