UltraSketchLLM: Saliency-Driven Sketching for Ultra-Low Bit LLM Compression

• URL: http://arxiv.org/abs/2506.17255v1
• Date: Sun, 08 Jun 2025 16:55:42 GMT
• Title: UltraSketchLLM: Saliency-Driven Sketching for Ultra-Low Bit LLM Compression
• Authors: Sunan Zou, Ziyun Zhang, Xueting Sun, Guojie Luo,
• Abstract summary: UltraSketchLLM is an index-free, sketch-based framework that achieves ultra-low bit compression (down to 0.5 bits per weight) while preserving model performance.<n>Our approach integrates an underestimate AbsMaxMin sketch to minimize relative errors for small weights, importance-aware space allocation to prioritize salient weights, and a straight-through estimator for compression-aware finetuning.
• Score: 2.4253566132113877
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: The rapid growth of large language models (LLMs) has outpaced the memory constraints of edge devices, necessitating extreme weight compression beyond the 1-bit limit. While quantization reduces model size, it is fundamentally limited to 1 bit per weight. Existing multiple-to-one compression methods either rely on mapping tables (inducing memory overhead) or incur severe accuracy degradation due to random weight grouping. We introduce UltraSketchLLM, an index-free, sketch-based framework that achieves ultra-low bit compression (down to 0.5 bits per weight) while preserving model performance. UltraSketchLLM leverages data sketching, a sub-linear representation technique from streaming applications, to map multiple weights to single values with bounded error. Our approach integrates an underestimate AbsMaxMin sketch to minimize relative errors for small weights, importance-aware space allocation to prioritize salient weights, and a straight-through estimator for compression-aware finetuning. Experiments on Llama-3.2-1B demonstrate up to 0.5-bit compression with competitive perplexity, alongside tolerable latency overhead. UltraSketchLLM offers a practical solution for deploying LLMs in resource-constrained environments.

Related papers

• EntroLLM: Entropy Encoded Weight Compression for Efficient Large Language Model Inference on Edge Devices [3.297182592932918]
  Large Language Models (LLMs) demonstrate exceptional performance across various tasks, but their large storage and computational requirements constrain their deployment on edge devices.<n>We propose EntroLLM, a novel compression framework that integrates mixed quantization with entropy coding to reduce storage overhead while maintaining model accuracy.
  arXiv  Detail & Related papers  (2025-05-05T05:42:14Z)
• 70% Size, 100% Accuracy: Lossless LLM Compression for Efficient GPU Inference via Dynamic-Length Float [71.43026659686679]
  Large Language Models (LLMs) have grown rapidly in size, creating challenges for efficient deployment on resource-constrained hardware.<n>We introduce Dynamic-Length Float (DFloat11), a compression framework that reduces LLM size by 30% while preserving outputs that are bit-for-bit identical to the original model.
  arXiv  Detail & Related papers  (2025-04-15T22:38:38Z)
• When Compression Meets Model Compression: Memory-Efficient Double Compression for Large Language Models [12.687035979970194]
  This paper introduces a framework to compress large language models (LLMs) after quantization.<n>A compression-aware quantization is first proposed to enhance model weight compressibility by re-scaling the model parameters before quantization, followed by a pruning method to improve further.<n>Experiments show inference with the compressed model can achieve a 40% reduction in memory size with negligible loss in accuracy and inference speed.
  arXiv  Detail & Related papers  (2025-02-21T13:11:22Z)
• Huff-LLM: End-to-End Lossless Compression for Efficient LLM Inference [19.59857352852377]
  Large language models (LLMs) have continued to rapidly increase in size.<n>This has exacerbated the difficulty in running state of the art LLMs on small, edge devices.<n>We propose Huff-LLM, a method that lets users store LLM weights in compressed format.
  arXiv  Detail & Related papers  (2025-02-02T21:23:42Z)
• BitStack: Any-Size Compression of Large Language Models in Variable Memory Environments [53.71158537264695]
  Large language models (LLMs) have revolutionized numerous applications, yet their deployment remains challenged by memory constraints on local devices.<n>We introduce textbfBitStack, a novel, training-free weight compression approach that enables megabyte-level trade-offs between memory usage and model performance.
  arXiv  Detail & Related papers  (2024-10-31T13:26:11Z)
• EoRA: Fine-tuning-free Compensation for Compressed LLM with Eigenspace Low-Rank Approximation [84.70637613266835]
  EoRA is a fine-tuning-free method that augments compressed Large Language Models with low-rank matrices.<n>EoRA consistently outperforms prior training-free low rank methods in recovering the accuracy of compressed LLMs.
  arXiv  Detail & Related papers  (2024-10-28T17:59:03Z)
• SLiM: One-shot Quantization and Sparsity with Low-rank Approximation for LLM Weight Compression [7.6131620435684875]
  SLIM is a new one-shot compression framework that holistically integrates hardware-friendly quantization, sparsity, and low-rank approximation.<n>SLIM improves model accuracy by up to 5.66% (LLaMA-2-7B) for 2:4 sparsity with 4-bit weight quantization, outperforming prior methods.
  arXiv  Detail & Related papers  (2024-10-12T18:36:07Z)
• LoRC: Low-Rank Compression for LLMs KV Cache with a Progressive Compression Strategy [59.1298692559785]
  Key-Value ( KV) cache is crucial component in serving transformer-based autoregressive large language models (LLMs) Existing approaches to mitigate this issue include: (1) efficient attention variants integrated in upcycling stages; (2) KV cache compression at test time; and (3) KV cache compression at test time. We propose a low-rank approximation of KV weight matrices, allowing plug-in integration with existing transformer-based LLMs without model retraining. Our method is designed to function without model tuning in upcycling stages or task-specific profiling in test stages.
  arXiv  Detail & Related papers  (2024-10-04T03:10:53Z)
• BiLLM: Pushing the Limit of Post-Training Quantization for LLMs [53.31402059062365]
  BiLLM is a groundbreaking 1-bit post-training quantization scheme tailored for pretrained large language models. It achieves for the first time high-accuracy inference (e.g. 8.41 perplexity on LLaMA2-70B) with only 1.08-bit weights across various LLMs families.
  arXiv  Detail & Related papers  (2024-02-06T09:26:34Z)
• SpQR: A Sparse-Quantized Representation for Near-Lossless LLM Weight Compression [76.73007709690306]
  We introduce the Sparse-Quantized Representation (SpQR), a new compressed format and quantization technique. SpQR achieves relative accuracy losses of less than 1% in perplexity for highly-accurate LLaMA and Falcon LLMs. This makes it possible to run 33B parameter LLM on a single 24 GB consumer GPU without any performance degradation at 15% speedup.
  arXiv  Detail & Related papers  (2023-06-05T17:53:28Z)
• Single-path Bit Sharing for Automatic Loss-aware Model Compression [126.98903867768732]
  Single-path Bit Sharing (SBS) is able to significantly reduce computational cost while achieving promising performance. Our SBS compressed MobileNetV2 achieves 22.6x Bit-Operation (BOP) reduction with only 0.1% drop in the Top-1 accuracy.
  arXiv  Detail & Related papers  (2021-01-13T08:28:21Z)

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.