Cortado is compared with state-of-the-art approaches and it is found that it is the only progressive compressor that guarantees the evaluated error bounds on all tested inputs while achieving the highest throughputs, comparable reconstruction quality, and on-par compression ratios.
The minSSIM metric and SBLC, the first SSIM-bounded compressor, are introduced, which is fully CPU-GPU compatible and guarantees the user-provided minSSIM bound, which it enforces by strategically tightening the error bound on specific values in the data.
TOPIQ is presented, a statistical error-propagation framework that predicts QoI-level bias and uncertainty from compact compression metadata (less than 0.1% of original data) and decomposes QoIs into primitive operators with closed-form propagation rules accounting for spatial error correlation and data-error coupling.
You-Yuan Liu, Bo Jiang, Taolue Yang et al.· 0 citations
This work presents an adaptive progressive compression framework that improves retrieval efficiency for two common targets, namely error-bound and peak Signal-to-Noise ratios and proposes coefficient decomposition, a novel method that exploits the commonly overlooked spatial correlations among decorrelated data.
Results motivate CertKV, a training-free compressor that reserves one tail-summary slot per head and allocates the rest by value dispersion, which is top-two in seven of nine LongBench-v2 settings, remains in the leading compressed tier on 128K RULER, and realizes a ten-fold cache budget in a packed Llama prototype.
This work introduces LLM-based compressors deploying two novel symbol-ranking variants that bound predictions to the top-$T$ ranks, and conducts the first large-scale evaluation of LLM-based source code compression across 30 LLMs, including general-domain, code-specialized, and quantized models.
Angelo Nardone, P. Ferragina· arXiv.org· 0 citations
Experiments on multiple unseen scientific simulation datasets demonstrate that EVOLVE achieves substantially higher CRs than conventional compressors at comparable reconstruction quality, while delivering compression speeds that are orders of magnitude faster than INR-based methods, highlighting its promise as a strong...
