paper-with-me

홈 › Papers

SPEAR: A System for Post-Quantization Error-Adaptive Recovery Enabling Efficient Low-Bit LLM Serving

2026-06-04 · Hongyuan Liu, Yawei Li, Zhiqiang Que, Qinli Yang, Junming Shao, Guosheng Hu arxiv

Efficient large language model (LLM) serving is increasingly constrained by deployment cost. Quantization is a key technique for reducing serving cost, yet even state-of-the-art 4-bit quantizers exhibit a noticeable quality gap from FP16, particularly for smaller models where low-bit serving is most beneficial. We identify a fundamental cause of this gap: quantization error is highly input-dependent and varies substantially across tokens, while existing post-quantization compensation methods are static and apply identical corrections to all inputs. As a result, easy tokens are over-corrected while hard tokens remain under-corrected. We present SPEAR, a system for post-quantization error-adaptive recovery that improves low-bit LLM serving. SPEAR introduces lightweight Error Compensators (ECs) modulated by per-token gates and places them only at the most error-sensitive layers identified through a CKA-guided entropy-aware diagnostic. This focuses a small parameter budget where it is most effective. Efficient deployment of ECs presents several systems challenges, including additional computation, tensor-parallel synchronization caused by input-dependent gating, and latency instability across configurations. SPEAR addresses these issues through adaptive kernel-fusion dispatch, combining an epilogue-integrated peer-reduction kernel with P2P dual-write to fuse the post-EC computation into low-bit GEMMs, and an SLO-constrained EC-aware scheduler for predictable serving performance. Across challenging per-channel quantization settings, SPEAR recovers 56-75% of the perplexity gap between W4 and FP16 while adding less than 1% model memory overhead and maintaining latency comparable to a widely used 4-bit serving deployment.

📄 PDF Abstract BibTeX arXiv:2606.11244

Code (0)

등록된 구현이 없습니다.

Similar Papers 제목 키워드 기반

OAC: Output-adaptive Calibration for Accurate Post-training Quantization

2024-05-23 · Ali Edalati, Alireza Ghaffari, Mahsa Ghazvini Nejad, Lu Hou 외

Deployment of Large Language Models (LLMs) has major computational costs, due to their rapidly expanding size. Compression of LLMs reduces the memory footprint, latency, and energy required for their inference. Post-trai…

Quantization

Efficient Adaptive Activation Rounding for Post-Training Quantization

2022-08-25 · Zhengyi Li, Cong Guo, Zhanda Zhu, Yangjie Zhou 외

Post-training quantization attracts increasing attention due to its convenience in deploying quantized neural networks. Although rounding-to-nearest remains the prevailing method for DNN quantization, prior research has …

Quantization

Learning Representations for CSI Adaptive Quantization and Feedback

2022-07-13 · Valentina Rizzello, Matteo Nerini, Michael Joham, Bruno Clerckx 외

In this work, we propose an efficient method for channel state information (CSI) adaptive quantization and feedback in frequency division duplexing (FDD) systems. Existing works mainly focus on the implementation of auto…

Quantization

StableQuant: Layer Adaptive Post-Training Quantization for Speech Foundation Models

2025-04-21 · Yeona Hong, Hyewon Han, Woo-Jin Chung, Hong-Goo Kang

In this paper, we propose StableQuant, a novel adaptive post-training quantization (PTQ) algorithm for widely used speech foundation models (SFMs). While PTQ has been successfully employed for compressing large language …

Automatic Speech RecognitionAutomatic Speech Recognition (ASR)Quantizationspeech-recognition+1

MARR: Module-Adaptive Residual Reconstruction for Low-Bit Post-Training Quantization

2026-05-18 · Le Su, Xing Luo, Zhi Jin arxiv

Recently, residual reconstruction-based model quantization methods have achieved promising performance in low-bit post-training quantization (PTQ) by introducing cross-layer residuals to reduce error accumulated from pre…