paper-with-me

Papers

Neural Network Quantization by Learning Low-Loss Subspaces

2026-06-23 · Vladimir Protsenko, Mikhalina Kharkevich, Alexander Vashchilko, Vladimir Kryzhanovskiy arxiv

Neural network quantization aims to find a discrete representation of parameters that preserves the performance of a full-precision (FP) model as faithfully as possible. Enforcing discrete constraints perturbs parameters away from a well-optimized minimum, generally resulting in performance degradation. Recent studies indicate that low-loss FP solutions are not isolated, but instead belong to connected low-loss subspaces of the loss landscape, where the loss maintains nearly the same minimum value. Models sampled from these subspaces are diverse and retain high accuracy. This raises the question: can a quantized model be constructed to lie within a low-loss subspace of the FP model, thereby automatically preserving performance? We address this question by learning quantization-aware linear paths in weight space optimized to minimize loss. We demonstrate that the midpoint of the resulting subspace is, by design, quantization-friendly and that its direct quantization yields performance comparable to that of quantization-aware training. The proposed procedure offers a novel perspective on weight quantization and, in contrast to conventional methods, neither relies on the straight-through estimator nor involves explicit discretization during training.

📄 PDF Abstract BibTeX arXiv:2606.25087

Code (0)

등록된 구현이 없습니다.

Similar Papers 제목 키워드 기반

Bridging the Gap between Continuous and Informative Discrete Representations by Random Product Quantization

2025-04-07 · Xueqing Li, Zehan Li, Boyu Zhu, Ruihao Jing 외

Self-supervised learning has become a core technique in speech processing, but the high dimensionality of its representations makes discretization essential for improving efficiency. However, existing discretization meth…

QuantizationSelf-Supervised Learning

A Biresolution Spectral Framework for Product Quantization

2018-06-01 · CVPR 2018 6 · Lopamudra Mukherjee, Sathya N. Ravi, Jiming Peng, Vikas Singh

Product quantization (PQ) (and its variants) has been effec- tively used to encode high-dimensional data into compact codes for many problems in vision. In principle, PQ decomposes the given data into a number of lower-d…

Quantization

Shared Predictive Cross-Modal Deep Quantization

2019-04-16 · Erkun Yang, Cheng Deng, Chao Li, Wei Liu 외

With explosive growth of data volume and ever-increasing diversity of data modalities, cross-modal similarity search, which conducts nearest neighbor search across different modalities, has been attracting increasing int…

Quantization

CoQuant: Joint Weight-Activation Subspace Projection for Mixed-Precision LLMs

2026-04-29 · Zhe Ding, Su Pan, Duowei Pan arxiv

Post-training quantization (PTQ) has become an important technique for reducing the inference cost of Large Language Models (LLMs). While recent mixed-precision methods improve ultra-low bit quantization by preserving cr…

Optimized Product Quantization for Approximate Nearest Neighbor Search

2013-06-01 · CVPR 2013 6 · Tiezheng Ge, Kaiming He, Qifa Ke, Jian Sun

Product quantization is an effective vector quantization approach to compactly encode high-dimensional vectors for fast approximate nearest neighbor (ANN) search. The essence of product quantization is to decompose the o…

Quantization