Understanding Calibration and Truncation Error Propagation in Training-Free Low-Rank Compression for LLMs

• URL: http://arxiv.org/abs/2608.08506v1
• Date: Sun, 09 Aug 2026 06:03:48 GMT
• Title: Understanding Calibration and Truncation Error Propagation in Training-Free Low-Rank Compression for LLMs
• Abstract summary: Existing SOTA frameworks share two key limitations.<n>We propose a training-free methodology compatible with existing frameworks to mitigate them.<n>Our approach demonstrates up to 1-2.5 accuracy point improvements over per-weight and joint decomposition baselines on zero-shot tasks.
• Score: 0.7532531650457809
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Training-free low-rank compression frameworks have been gaining prominence for LLM compression given their effectiveness in reducing model parameter count while maintaining task-level accuracy. However, existing SOTA frameworks share two key limitations: (1) residual errors in calibration data activations accumulate across layers during compression, causing misalignment between representations simulated at compression time and those experienced at inference; (2) the assumption that layer importance distribution is preserved post-compression does not hold. Together, these two effects introduce misalignment in the compression process in relation to the deployed model. We study these effects and propose a simple, training-free methodology compatible with existing frameworks to mitigate them, comprising: (1) Layer-by-Layer Compression with Calibration Correction; (2) Iterative Compression with Rank Allocation Correction. Implemented atop an existing SOTA decomposition framework, and evaluated on Llama and Qwen3 models across various benchmarks and compression rates, our approach demonstrates up to ~1-2.5 accuracy point improvements over per-weight and joint decomposition baselines on zero-shot tasks.

Related papers

• Arbitrary Ratio Feature Compression via Next Token Prediction [52.10426317889982]
  Arbitrary Ratio Feature Compression (ARFC) framework supports any compression ratio with a single model.<n>ARC is an auto-regressive model that performs compression via next-gressive prediction.<n>MoS module refines the compressed tokens by utilizing multiple compression results.<n>ERGC is integrated into the training process to preserve semantic and structural relationships during compression.
  arXiv  Detail & Related papers  (2026-02-12T02:38:57Z)
• SAES-SVD: Self-Adaptive Suppression of Accumulated and Local Errors for SVD-based LLM Compression [27.258302662888166]
  SAES-SVD is a low-rank compression framework for large language models.<n>It jointly optimize intra-layer reconstruction and inter-layer error compensation.<n>Experiments show that SAES-SVD consistently improves post-compression performance.
  arXiv  Detail & Related papers  (2026-02-03T03:23:10Z)
• Content Adaptive based Motion Alignment Framework for Learned Video Compression [72.13599533975413]
  This paper proposes a content adaptive based motion alignment framework.<n>We first introduce a two-stage flow-guided deformable warping mechanism that refines motion compensation with coarse-to-fine offset prediction and mask modulation.<n>Second, we propose a multi-reference quality aware strategy that adjusts distortion weights based on reference quality, and applies it to hierarchical training to reduce error propagation.<n>Third, we integrate a training-free module that downsamples frames by motion magnitude and resolution to obtain smooth motion estimation.
  arXiv  Detail & Related papers  (2025-12-15T02:51:47Z)
• Activation-Informed Pareto-Guided Low-Rank Compression for Efficient LLM/VLM [11.762499172999886]
  Large language models (LLM) and vision-language models (VLM) have achieved state-of-the-art performance, but they impose significant memory and computing challenges in deployment.<n>We present a novel low-rank compression framework to address this challenge.
  arXiv  Detail & Related papers  (2025-10-07T03:07:47Z)
• Semantic Retention and Extreme Compression in LLMs: Can We Have Both? [0.0]
  Large Language Model (LLM) deployment has intensified the need for efficient model compression techniques.<n>We show how strategically combining pruning and quantization could yield superior performance-to-compression ratios.<n>We introduce the Semantic Retention Compression Rate (SrCr), a novel metric that quantifies the trade-off between model compression and semantic preservation.
  arXiv  Detail & Related papers  (2025-05-12T07:23:19Z)
• CALLIC: Content Adaptive Learning for Lossless Image Compression [64.47244912937204]
  CALLIC sets a new state-of-the-art (SOTA) for learned lossless image compression.<n>We propose a content-aware autoregressive self-attention mechanism by leveraging convolutional gating operations.<n>During encoding, we decompose pre-trained layers, including depth-wise convolutions, using low-rank matrices and then adapt the incremental weights on testing image by Rate-guided Progressive Fine-Tuning (RPFT)<n>RPFT fine-tunes with gradually increasing patches that are sorted in descending order by estimated entropy, optimizing learning process and reducing adaptation time.
  arXiv  Detail & Related papers  (2024-12-23T10:41:18Z)
• Communication Compression for Distributed Learning without Control Variates [43.69787537805927]
  We propose a novel distributed algorithm that allows compressible clients by exploiting past aggregated updates.<n> Experimental results confirm that CAFe outperforms existing distributed compression learning schemes.
  arXiv  Detail & Related papers  (2024-12-05T18:46:20Z)
• 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)
• 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)
• Token Compensator: Altering Inference Cost of Vision Transformer without Re-Tuning [63.43972993473501]
  Token compression expedites the training and inference of Vision Transformers (ViTs) However, when applied to downstream tasks, compression degrees are mismatched between training and inference stages. We propose a model arithmetic framework to decouple the compression degrees between the two stages.
  arXiv  Detail & Related papers  (2024-08-13T10:36:43Z)
• Activations and Gradients Compression for Model-Parallel Training [85.99744701008802]
  We study how simultaneous compression of activations and gradients in model-parallel distributed training setup affects convergence. We find that gradients require milder compression rates than activations. Experiments also show that models trained with TopK perform well only when compression is also applied during inference.
  arXiv  Detail & Related papers  (2024-01-15T15:54:54Z)
• Inshrinkerator: Compressing Deep Learning Training Checkpoints via Dynamic Quantization [5.648270790530862]
  State-of-the-art approaches involve lossy model compression mechanisms, which induce a tradeoff between the resulting model quality (accuracy) and compression ratio. We make a key enabling observation that the sensitivity of model weights to compression varies during training, and different weights benefit from different quantization levels. We propose a non-uniform quantization scheme that leverages this variation, an efficient search mechanism that dynamically finds the best quantization configurations, and a quantization-aware delta compression mechanism that rearranges weights to minimize checkpoint differences.
  arXiv  Detail & Related papers  (2023-06-20T18:00:31Z)

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.