Tootfinder

Optimal Remainder Estimates in the Quantization of Complex Projective Spaces
Tommaso Aschieri, B{\l}a\.zej Ruba, Jan Philip Solovej
https://arxiv.org/abs/2508.19968 https://

Optimal Remainder Estimates in the Quantization of Complex Projective Spaces
We study Berezin-Toeplitz quantization of complex projective spaces $\mathbb{CP}^{d-1}$ and obtain full asymptotic expansions of the Berezin transformation and of products of Toeplitz operators. In each case, the remainder is controlled by the next term of the expansion, either through a positivity-preserving transformation or via an operator inequality. This leads to bounds which are optimal in terms of the required regularity and feature sharp or asymptotically sharp constants.

Beat-Based Rhythm Quantization of MIDI Performances
Maximilian Wachter, Sebastian Murgul, Michael Heizmann
https://arxiv.org/abs/2508.19262 https://arxiv.o…

Beat-Based Rhythm Quantization of MIDI Performances
We propose a transformer-based rhythm quantization model that incorporates beat and downbeat information to quantize MIDI performances into metrically-aligned, human-readable scores. We propose a beat-based preprocessing method that transfers score and performance data into a unified token representation. We optimize our model architecture and data representation and train on piano and guitar performances. Our model exceeds state-of-the-art performance based on the MUSTER metric.

How Quantization Shapes Bias in Large Language Models
Federico Marcuzzi, Xuefei Ning, Roy Schwartz, Iryna Gurevych
https://arxiv.org/abs/2508.18088 https://

How Quantization Shapes Bias in Large Language Models
This work presents a comprehensive evaluation of how quantization affects model bias, with particular attention to its impact on individual demographic subgroups. We focus on weight and activation quantization strategies and examine their effects across a broad range of bias types, including stereotypes, toxicity, sentiment, and fairness. We employ both probabilistic and generated text-based metrics across nine benchmarks and evaluate models varying in architecture family and reasoning ability.…

Scene-Aware Vectorized Memory Multi-Agent Framework with Cross-Modal Differentiated Quantization VLMs for Visually Impaired Assistance
Xiangxiang Wang, Xuanyu Wang, YiJia Luo, Yongbin Yu, Manping Fan, Jingtao Zhang, Liyong Ren
https://arxiv.org/abs/2508.18177

Scene-Aware Vectorized Memory Multi-Agent Framework with Cross-Modal Differentiated Quantization VLMs for Visually Impaired Assistance
This study proposes the dual technological innovation framework, including a cross-modal differ entiated quantization framework for vision-language models (VLMs) and a scene-aware vectorized memory multi-agent system for visually impaired assistance. The modular framework was developed implementing differentiated processing strategies, effectively reducing memory requirements from 38GB to 16GB while maintaining model performance. The multi-agent architecture combines scene classificatio…

Chern-Simons States in $SO(1,n)$ Yang-Mills Gauge Theory of Quantum Gravity
Zbigniew Haba
https://arxiv.org/abs/2508.19658 https://arxiv.org/pdf/2508.19658…

Chern-Simons States in $SO(1,n)$ Yang-Mills Gauge Theory of Quantum Gravity
We discuss a quantization of the Yang--Mills theory with an internal symmetry group $SO(1,n)$ treated as a unified theory of all interactions. In one-loop calculations, we show that Einstein gravity can be considered as an approximation to gauge theory. We discuss the role of the Chern-Simons wave functions in the quantization.

1-Bit Unlimited Sampling Beyond Fourier Domain: Low-Resolution Sampling of Quantization Noise
Vaclav Pavlicek, Ayush Bhandari
https://arxiv.org/abs/2508.19408 https://

1-Bit Unlimited Sampling Beyond Fourier Domain: Low-Resolution Sampling of Quantization Noise
Analog-to-digital converters (ADCs) play a critical role in digital signal acquisition across various applications, but their performance is inherently constrained by sampling rates and bit budgets. This bit budget imposes a trade-off between dynamic range (DR) and digital resolution, with ADC energy consumption scaling linearly with sampling rate and exponentially with bit depth. To bypass this, numerous approaches, including oversampling with low-resolution ADCs, have been explored. A promine…

Canonical Quantization of a Memristive Leaky Integrate-and-Fire Neuron Circuit
Dean Brand, Domenica Dibenedetto, Francesco Petruccione
https://arxiv.org/abs/2506.21363

Canonical Quantization of a Memristive Leaky Integrate-and-Fire Neuron Circuit
We present a theoretical framework for a quantized memristive Leaky Integrate-and-Fire (LIF) neuron, uniting principles from neuromorphic engineering and open quantum systems. Starting from a classical memristive LIF circuit, we apply canonical quantization techniques to derive a quantum model grounded in circuit quantum electrodynamics. Numerical simulations demonstrate key dynamical features of the quantized memristor and LIF neuron in the weak-coupling and adiabatic regime, including memory …

This https://arxiv.org/abs/2505.18975 has been replaced.
initial toot: https://mastoxiv.page/@arXiv_csAR_…

FastMamba: A High-Speed and Efficient Mamba Accelerator on FPGA with Accurate Quantization
State Space Models (SSMs), like recent Mamba2, have achieved remarkable performance and received extensive attention. However, deploying Mamba2 on resource-constrained edge devices encounters many problems: severe outliers within the linear layer challenging the quantization, diverse and irregular element-wise tensor operations, and hardware-unfriendly nonlinear functions in the SSM block. To address these issues, this paper presents FastMamba, a dedicated accelerator on FPGA with hardware-algo…

Robust Residual Finite Scalar Quantization for Neural Compression
Xiaoxu Zhu
https://arxiv.org/abs/2508.15860 https://arxiv.org/pdf/2508.15860

Robust Residual Finite Scalar Quantization for Neural Compression
Finite Scalar Quantization (FSQ) has emerged as a promising alternative to Vector Quantization (VQ) in neural compression, offering simplified training and improved stability. However, naive application of FSQ in residual quantization frameworks suffers from the \textbf{residual magnitude decay problem}, where subsequent FSQ layers receive progressively weaker signals, severely limiting their effectiveness. We propose \textbf{Robust Residual Finite Scalar Quantization (RFSQ)}, a general framewo…

Enhancing Model Privacy in Federated Learning with Random Masking and Quantization
Zhibo Xu, Jianhao Zhu, Jingwen Xu, Changze Lv, Zisu Huang, Xiaohua Wang, Muling Wu, Qi Qian, Xiaoqing Zheng, Xuanjing Huang
https://arxiv.org/abs/2508.18911

Enhancing Model Privacy in Federated Learning with Random Masking and Quantization
Experimental results across various models and tasks demonstrate that our approach not only maintains strong model performance in federated learning settings but also achieves enhanced protection of model parameters compared to baseline methods.

Quantized but Deceptive? A Multi-Dimensional Truthfulness Evaluation of Quantized LLMs
Yao Fu, Xianxuan Long, Runchao Li, Haotian Yu, Mu Sheng, Xiaotian Han, Yu Yin, Pan Li
https://arxiv.org/abs/2508.19432

Quantized but Deceptive? A Multi-Dimensional Truthfulness Evaluation of Quantized LLMs
Quantization enables efficient deployment of large language models (LLMs) in resource-constrained environments by significantly reducing memory and computation costs. While quantized LLMs often maintain performance on perplexity and zero-shot tasks, their impact on truthfulness-whether generating truthful or deceptive responses-remains largely unexplored. In this work, we introduce TruthfulnessEval, a comprehensive evaluation framework for assessing the truthfulness of quantized LLMs across thr…

Information-Preserving CSI Feedback: Invertible Networks with Endogenous Quantization and Channel Error Mitigation
Haotian Tian, Lixiang Lian, Jiaqi Cao, Sijie Ji
https://arxiv.org/abs/2507.20283

Information-Preserving CSI Feedback: Invertible Networks with Endogenous Quantization and Channel Error Mitigation
Deep learning has emerged as a promising so- lution for efficient channel state information (CSI) feedback in frequency division duplex (FDD) massive MIMO systems. Conventional deep learning-based methods typically rely on a deep autoencoder to compress the CSI, which leads to irre- versible information loss and degrades reconstruction accuracy. This paper introduces InvCSINet, an information-preserving CSI feedback framework based on invertible neural networks (INNs). By leveraging the bijecti…

Positive traces on certain $\operatorname{SL}(2)$ Coulomb branches
Daniil Klyuev, Joseph Vulakh
https://arxiv.org/abs/2507.19447 https://arxiv.org/pdf/2507…

Positive traces on certain $\operatorname{SL}(2)$ Coulomb branches
For a noncommutative algebra $A$ and an antilinear automorphism $ρ$ of $A$ there is a notion of a positive trace. Positive traces appear in study of supersymmetric gauge theories. We classify positive traces on $A$ in two cases. The first case is when $A$ is a quantization of a Kleinian singularity of type $D$, with certain restriction on the quantization parameter. The second case is when $A=K^{\operatorname{SL}(2,\mathbb{C}[[t]])\rtimes \mathbb{C}_q^{\times}}(\operatorname{Gr}_{\operatorname…

Systematic Characterization of LLM Quantization: A Performance, Energy, and Quality Perspective
Tianyao Shi, Yi Ding
https://arxiv.org/abs/2508.16712 https://

Systematic Characterization of LLM Quantization: A Performance, Energy, and Quality Perspective
Large language models (LLMs) have demonstrated remarkable capabilities across diverse domains, but their heavy resource demands make quantization-reducing precision to lower-bit formats-critical for efficient serving. While many quantization methods exist, a systematic understanding of their performance, energy, and quality tradeoffs in realistic serving conditions remains a gap. In this work, we first develop a fully automated online characterization framework qMeter, and then conduct an in-de…

This https://arxiv.org/abs/2303.08082 has been replaced.
initial toot: https://mastoxiv.page/@arXiv_mat…

Quantum traces for $SL_n$-skein algebras
We establish the existence of several quantum trace maps. The simplest one is an algebra map between two quantizations of the algebra of regular functions on the $SL_n$-character variety of a surface $\mathfrak{S}$ equipped with an ideal triangulation $λ$. The first is the (stated) $SL_n$-skein algebra $\mathscr{S}(\mathfrak{S})$. The second $\overline{\mathcal{X}}(\mathfrak{S},λ)$ is the Fock and Goncharov's quantization of their $X$-moduli space. The quantum trace is an algebra homomorphism…

Singularity Resolution of Quantum Black Holes in (A)dS
Sofie Ried
https://arxiv.org/abs/2508.20794 https://arxiv.org/pdf/2508.20794

Singularity Resolution of Quantum Black Holes in (A)dS
The singularities present at the centre of black holes signal a break down of the classical theory. In this paper, we demonstrate a resolution of the Schwarzschild-(Anti-)de Sitter singularity by imposing unitary evolution with respect to unimodular time. Employing the Henneaux-Teitelboim formulation of unimodular gravity, we perform a canonical quantization on a symmetry-reduced Schwarzschild-(Anti-) de Sitter model. This leads to a Wheeler-DeWitt equation that effectively becomes a Schröding…

A Node-Aware Dynamic Quantization Approach for Graph Collaborative Filtering
Lin Li, Chunyang Li, Yu Yin, Xiaohui Tao, Jianwei Zhang
https://arxiv.org/abs/2508.16516 https://

A Node-Aware Dynamic Quantization Approach for Graph Collaborative Filtering
In the realm of collaborative filtering recommendation systems, Graph Neural Networks (GNNs) have demonstrated remarkable performance but face significant challenges in deployment on resource-constrained edge devices due to their high embedding parameter requirements and computational costs. Using common quantization method directly on node embeddings may overlooks their graph based structure, causing error accumulation during message passing and degrading the quality of quantized embeddings.To…

Forward-Backward Quantization of Scenario Processes in Multi-Stage Stochastic Optimization
Anna Timonina-Farkas
https://arxiv.org/abs/2508.18112 https://ar…

Forward-Backward Quantization of Scenario Processes in Multi-Stage Stochastic Optimization
Multi-stage stochastic optimization lies at the core of decision-making under uncertainty. As the analytical solution is available only in exceptional cases, dynamic optimization aims to efficiently find approximations but often neglects non-Markovian time-interdependencies. Methods on scenario trees can represent such interdependencies but are subject to the curse of dimensionality. To ease this problem, researchers typically approximate the uncertainty by smaller but more accurate trees. In t…

Crosslisted article(s) found for cs.MM. https://arxiv.org/list/cs.MM/new
[1/1]:
- Beat-Based Rhythm Quantization of MIDI Performances
Maximilian Wachter, Sebastian Murgul, Michael Heizmann

A Comprehensive Evaluation on Quantization Techniques for Large Language Models
Yutong Liu, Cairong Zhao, Guosheng Hu
https://arxiv.org/abs/2507.17417 http…

A Comprehensive Evaluation on Quantization Techniques for Large Language Models
For large language models (LLMs), post-training quantization (PTQ) can significantly reduce memory footprint and computational overhead. Model quantization is a rapidly evolving research field. Though many papers have reported breakthrough performance, they may not conduct experiments on the same ground since one quantization method usually contains multiple components. In addition, analyzing the theoretical connections among existing methods is crucial for in-depth understanding. To bridge the…

Quantization of blow-up masses for the Finsler $N$-Liouville equation
Xia Huang, Yuan Li, Dong Ye, Feng Zhou
https://arxiv.org/abs/2508.16080 https://arxiv…

Quantization of blow-up masses for the Finsler $N$-Liouville equation
The quantization results for blow-up phenomena play crucial roles in the analysis of partial differential equations. Here we quantify the blow-up masses to the following Finsler $N$-Liouville equation $$-Q_{N}u_{n}=V_{n}e^{u_{n}}\quad\mbox{in}~ Ω\subset \mathbb{R}^{N}, N \ge 2.$$ Our study generalizes the classical result of Li-Shafrir [Indiana Univ. Math.J.,1994] for Liouville equation, Wang-Xia's work for anisotropic Liouville equation in $\mathbb{R}^2$ [JDE, 2012], and Esposito-Lucia's for …

Task-Specific Zero-shot Quantization-Aware Training for Object Detection
Changhao Li, Xinrui Chen, Ji Wang, Kang Zhao, Jianfei Chen
https://arxiv.org/abs/2507.16782

Task-Specific Zero-shot Quantization-Aware Training for Object Detection
Quantization is a key technique to reduce network size and computational complexity by representing the network parameters with a lower precision. Traditional quantization methods rely on access to original training data, which is often restricted due to privacy concerns or security challenges. Zero-shot Quantization (ZSQ) addresses this by using synthetic data generated from pre-trained models, eliminating the need for real training data. Recently, ZSQ has been extended to object detection. Ho…

Design and Implementation of Parametrized Look-Up Tables for Post-Correction of Oversampling Low-Resolution ADCs
Morriel Kasher, Michael Tinston, Predrag Spasojevic
https://arxiv.org/abs/2507.18673

Design and Implementation of Parametrized Look-Up Tables for Post-Correction of Oversampling Low-Resolution ADCs
We propose a framework for the design, optimization, and implementation of Look-Up Tables (LUTs) used to recover noisy, oversampled, quantized signals given a parametric input model. The LUTs emulate the spectral effects of pre-quantization dithering through an all-digital solution applied after quantization. This methodology decomposes the intractable LUT design problem into four distinct stages, each of which is addressed analytically using a model-driven approach without reliance on training…

Convergence in charmonium structure: light-front wave functions from basis light-front quantization and Dyson-Schwinger equations
Xianghui Cao, Yang Li, Chao Shi, James P. Vary, Qun Wang
https://arxiv.org/abs/2507.17330

Convergence in charmonium structure: light-front wave functions from basis light-front quantization and Dyson-Schwinger equations
We present a systematic comparison of charmonium light-front wave functions obtained through two complementary non-perturbative approaches: Basis Light-Front Quantization (BLFQ) and Dyson-Schwinger Equations (DSE). Key observables include the charge form factor, gravitational form factors, light-cone distribution amplitudes, decay constants, and two-photon transition form factors. Despite their distinct theoretical foundations and model parameters, the predictions from BLFQ and DSE exhibit rema…

The Voronoi Spherical CDF for Lattices and Linear Codes: New Bounds for Quantization and Coding
Or Ordentlich
https://arxiv.org/abs/2506.19791 https://

The Voronoi Spherical CDF for Lattices and Linear Codes: New Bounds for Quantization and Coding
For a lattice/linear code, we define the Voronoi spherical cumulative density function (CDF) as the CDF of the $\ell_2$-norm/Hamming weight of a random vector uniformly distributed over the Voronoi cell. Using the first moment method together with a simple application of Jensen's inequality, we develop lower bounds on the expected Voronoi spherical CDF of a random lattice/linear code. Our bounds are quite close to a trivial ball-based lower bound and immediately translate to improved upper boun…

This https://arxiv.org/abs/2409.19938 has been replaced.
link: https://scholar.google.com/scholar?q=a

Quantum-inspired Beamforming Optimization for Quantized Phase-only Massive MIMO Arrays
This paper introduces an innovative quantum-inspired method for beamforming (BF) optimization in multiple-input multiple-output (MIMO) arrays. The method leverages the simulated bifurcation (SB) algorithm to address the complex combinatorial optimization problem due to the quantized phase configuration. We propose novel encoding techniques for high-bit phase quantization, which are then mapped into Ising spins. This enables efficient construction of the Hamiltonians and subsequent optimization …

Federated Fine-Tuning of Sparsely-Activated Large Language Models on Resource-Constrained Devices
Fahao Chen, Jie Wan, Peng Li, Zhou Su, Dongxiao Yu
https://arxiv.org/abs/2508.19078

Federated Fine-Tuning of Sparsely-Activated Large Language Models on Resource-Constrained Devices
Federated fine-tuning of Mixture-of-Experts (MoE)-based large language models (LLMs) is challenging due to their massive computational requirements and the resource constraints of participants. Existing working attempts to fill this gap through model quantization, computation offloading, or expert pruning. However, they cannot achieve desired performance due to impractical system assumptions and a lack of consideration for MoE-specific characteristics. In this paper, we propose FLUX, a system d…

Quantized and nonquantized Hall response in topological Hatsugai-Kohmoto systems
Thibaut Desort, Mark O. Goerbig, Corentin Morice
https://arxiv.org/abs/2506.20491

Quantized and nonquantized Hall response in topological Hatsugai-Kohmoto systems
We explore the robustness of Hall conductivity quantization in several insulating systems, exhibiting one scenario where the quantization is not preserved. Specifically, we apply the Kubo formula to topological models with the Hatsugai-Kohmoto interaction. Starting from the many-body degeneracy induced by this interaction in the topological Kane-Mele model, we consider Zeeman fields to select specific states within the ground-state manifold that reveal a non-quantized Hall response, precisely f…

Quantization Meets dLLMs: A Systematic Study of Post-training Quantization for Diffusion LLMs
Haokun Lin, Haobo Xu, Yichen Wu, Ziyu Guo, Renrui Zhang, Zhichao Lu, Ying Wei, Qingfu Zhang, Zhenan Sun
https://arxiv.org/abs/2508.14896

Quantization Meets dLLMs: A Systematic Study of Post-training Quantization for Diffusion LLMs
Recent advances in diffusion large language models (dLLMs) have introduced a promising alternative to autoregressive (AR) LLMs for natural language generation tasks, leveraging full attention and denoising-based decoding strategies. However, the deployment of these models on edge devices remains challenging due to their massive parameter scale and high resource demands. While post-training quantization (PTQ) has emerged as a widely adopted technique for compressing AR LLMs, its applicability to…

VQL: An End-to-End Context-Aware Vector Quantization Attention for Ultra-Long User Behavior Modeling
Kaiyuan Li, Yongxiang Tang, Yanhua Cheng, Yong Bai, Yanxiang Zeng, Chao Wang, Xialong Liu, Peng Jiang
https://arxiv.org/abs/2508.17125

VQL: An End-to-End Context-Aware Vector Quantization Attention for Ultra-Long User Behavior Modeling
In large-scale recommender systems, ultra-long user behavior sequences encode rich signals of evolving interests. Extending sequence length generally improves accuracy, but directly modeling such sequences in production is infeasible due to latency and memory constraints. Existing solutions fall into two categories: (1) top-k retrieval, which truncates the sequence and may discard most attention mass when L >> k; and (2) encoder-based compression, which preserves coverage but often over-compres…

Some features of the extended phase space approach to quantization of gravity
T. P. Shestakova
https://arxiv.org/abs/2507.17371 https://arxiv.org/pdf/2507.…

Some features of the extended phase space approach to quantization of gravity
In this paper, I emphasize those features of the extended phase space approach to quantization of gravity that distinguish it among other approaches. First of all, it is the conjecture about non-trivial topology of the Universe which was supported by Wheeler, Hawking and other founders of quantum gravity. However, this conjecture appears to be in contradiction with the assumption about asymptotic states that is used in the path integral quantization of gauge theories. The presence of asymptotic…

Replaced article(s) found for math.SG. https://arxiv.org/list/math.SG/new
[1/1]:
- Strict Quantization for Compact Pseudo-K\"ahler Manifolds and Group Actions
Andrea Galasso

Crosslisted article(s) found for cs.AI. https://arxiv.org/list/cs.AI/new
[5/6]:
- Enhancing Model Privacy in Federated Learning with Random Masking and Quantization
Xu, Zhu, Xu, Lv, Huang, Wang, Wu, Qian, Zheng, Huang

SiLQ: Simple Large Language Model Quantization-Aware Training
Steven K. Esser, Jeffrey L. McKinstry, Deepika Bablani, Rathinakumar Appuswamy, Dharmendra S. Modha
https://arxiv.org/abs/2507.16933

SiLQ: Simple Large Language Model Quantization-Aware Training
Large language models can be quantized to reduce inference time latency, model size, and energy consumption, thereby delivering a better user experience at lower cost. A challenge exists to deliver quantized models with minimal loss of accuracy in reasonable time, and in particular to do so without requiring mechanisms incompatible with specialized inference accelerators. Here, we demonstrate a simple, end-to-end quantization-aware training approach that, with an increase in total model trainin…

Integral points and volume of integral-integral affine manifolds
Oded Elisha, Yael Karshon, Yiannis Loizides
https://arxiv.org/abs/2508.17551 https://arxiv…

Integral points and volume of integral-integral affine manifolds
We give an elementary proof that, for a closed manifold with an integral-integral affine structure, its total volume and number of integral points coincide. This theorem was previously shown to be equivalent to the "independence of polarization" phenomenon of geometric quantization. The proof uses rational Ehrhart theory and elementary Fourier analysis to estimate the difference between the total volume and the number of integral points.

Quantization-Aware Neuromorphic Architecture for Efficient Skin Disease Classification on Resource-Constrained Devices
Haitian Wang, Xinyu Wang, Yiren Wang, Karen Lee, Zichen Geng, Xian Zhang, Kehkashan Kiran, Yu Zhang, Bo Miao
https://arxiv.org/abs/2507.15958

Quantization-Aware Neuromorphic Architecture for Efficient Skin Disease Classification on Resource-Constrained Devices
Accurate and efficient skin lesion classification on edge devices is critical for accessible dermatological care but remains challenging due to computational, energy, and privacy constraints. We introduce QANA, a novel quantization-aware neuromorphic architecture for incremental skin lesion classification on resource-limited hardware. QANA effectively integrates ghost modules, efficient channel attention, and squeeze-and-excitation blocks for robust feature representation with low-latency and e…

Exploiting Information Redundancy in Attention Maps for Extreme Quantization of Vision Transformers
Lucas Maisonnave, Karim Haroun, Tom Pegeot
https://arxiv.org/abs/2508.16311 h…

Exploiting Information Redundancy in Attention Maps for Extreme Quantization of Vision Transformers
Transformer models rely on Multi-Head Self-Attention (MHSA) mechanisms, where each attention head contributes to the final representation. However, their computational complexity and high memory demands due to MHSA hinders their deployment at the edge. In this work, we analyze and exploit information redundancy in attention maps to accelerate model inference. By quantifying the information captured by each attention head using Shannon entropy, our analysis reveals that attention heads with lowe…

Analysis of Null Related Beampattern Measures and Signal Quantization Effects for Linear Differential Microphone Arrays
Shweta Pal, Arun Kumar, Monika Agrawal
https://arxiv.org/abs/2506.21043

Analysis of Null Related Beampattern Measures and Signal Quantization Effects for Linear Differential Microphone Arrays
A differential microphone array (DMA) offers enhanced capabilities to obtain sharp nulls at the cost of relatively broad peaks in the beam power pattern. This can be used for applications that require nullification or attenuation of interfering sources. To the best of our knowledge, the existing literature lacks measures that directly assess the efficacy of nulls, and null-related measures have not been investigated in the context of differential microphone arrays (DMAs). This paper offers new …

Quantization Meets Spikes: Lossless Conversion in the First Timestep via Polarity Multi-Spike Mapping
Hangming Zhang, Zheng Li, Qiang Yu
https://arxiv.org/abs/2508.14520 https:/…

Quantization Meets Spikes: Lossless Conversion in the First Timestep via Polarity Multi-Spike Mapping
Spiking neural networks (SNNs) offer advantages in computational efficiency via event-driven computing, compared to traditional artificial neural networks (ANNs). While direct training methods tackle the challenge of non-differentiable activation mechanisms in SNNs, they often suffer from high computational and energy costs during training. As a result, ANN-to-SNN conversion approach still remains a valuable and practical alternative. These conversion-based methods aim to leverage the discrete …

The QuiverTools package for SageMath and Julia
Pieter Belmans, Hans Franzen, Gianni Petrella
https://arxiv.org/abs/2506.19432 https://

The QuiverTools package for SageMath and Julia
We introduce QuiverTools, a new software package, available in both a SageMath and Julia version, to study quivers and their moduli spaces of representations. Its key features are the computation of general subdimension vectors, leading to canonical decompositions, and checking the existence of (semi)stable representations, as well as the enumeration of Harder-Narasimhan types and related calculations for Teleman quantization. Computations related to intersection theory on quiver moduli are als…

Convergence order of the quantization error for self-affine measures on Lalley-Gatzouras carpets
Sanguo Zhu
https://arxiv.org/abs/2508.06875 https://arxiv.…

Convergence order of the quantization error for self-affine measures on Lalley-Gatzouras carpets
Let $E$ be a Lalley-Gatzouras carpet determined by a set of contractive affine mappings $\{f_{ij}\}_{(i,j)\in G}$. We study the asymptotics of quantization error for the self-affine measures $μ$ on $E$. We prove that the upper and lower quantization coefficient for $μ$ are both bounded away from zero and infinity in the exact quantization dimension. This significantly generalizes the previous work concerning the quantization for self-affine measures on Bedford-McMullen carpets. The new ingred…

Crosslisted article(s) found for cs.AR. https://arxiv.org/list/cs.AR/new
[1/1]:
- Systematic Characterization of LLM Quantization: A Performance, Energy, and Quality Perspective
Tianyao Shi, Yi Ding

Semi-fragile watermarking of remote sensing images using DWT, vector quantization and automatic tiling
Jordi Serra-Ruiz, David Meg\'ias
https://arxiv.org/abs/2507.07250

Semi-fragile watermarking of remote sensing images using DWT, vector quantization and automatic tiling
A semi-fragile watermarking scheme for multiple band images is presented in this article. We propose to embed a mark into remote sensing images applying a tree-structured vector quantization approach to the pixel signatures instead of processing each band separately. The signature of the multispectral or hyperspectral image is used to embed the mark in it order to detect any significant modification of the original image. The image is segmented into three-dimensional blocks, and a tree-structur…

Kontsevich's Cocycle Construction and Quantization of the Loday-Quillen-Tsygan Theorem
Jakob Ulmer
https://arxiv.org/abs/2506.15210 https://

Kontsevich's Cocycle Construction and Quantization of the Loday-Quillen-Tsygan Theorem
We relate graph complexes, Calabi-Yau $A_\infty$-categories and Kontsevich's cocycle construction. Our main result produces a commutative square of shifted Poisson algebras; one of its edges is the Loday-Quillen-Tsygan map, generalized to $A_\infty$-categories. We describe a quantized version via Beilinson-Drinfeld algebras. The larger context is to provide categorical methods which relate enumerative geometry (as in mirror symmetry) and large $N$ gauge theories.

Google announces Gemma 3 270M, a compact model designed for task-specific fine-tuning with strong capabilities in instruction following and text structuring (Google Developers Blog)
https://developers.googleblog.com/en/introducing-gemma-3-270m/

Introducing Gemma 3 270M: The compact model for hyper-efficient AI- Google Developers Blog
Explore Gemma 3 270M, a compact, energy-efficient AI model for task-specific fine-tuning, offering strong instruction-following and production-ready quantization.

Constrained quantization for discrete distributions
Bismark Bimpong, Sayandip Pandit, Mrinal Kanti Roychowdhury, Prabhat Tamrakar
https://arxiv.org/abs/2507.07923

Constrained quantization for discrete distributions
Constrained quantization for a Borel probability measure refers to the process of approximating a given probability distribution by a discrete probability measure supported on a finite number of points that lie within a specified set, known as the constraint. This paper investigates constrained quantization for several types of discrete probability distributions under two different constraints. The study begins with an analysis of two discrete distributions, one uniform and one nonuniform, defi…

APT-LLM: Exploiting Arbitrary-Precision Tensor Core Computing for LLM Acceleration
Shaobo Ma, Chao Fang, Haikuo Shao, Zhongfeng Wang
https://arxiv.org/abs/2508.19087 https://

APT-LLM: Exploiting Arbitrary-Precision Tensor Core Computing for LLM Acceleration
Large language models (LLMs) have revolutionized AI applications, yet their enormous computational demands severely limit deployment and real-time performance. Quantization methods can help reduce computational costs, however, attaining the extreme efficiency associated with ultra-low-bit quantized LLMs at arbitrary precision presents challenges on GPUs. This is primarily due to the limited support for GPU Tensor Cores, inefficient memory management, and inflexible kernel optimizations. To tack…

Thermodynamically Consistent Coarse-graining: from Interacting Particles to Fields via Second Quantization
Atul Tanaji Mohite, Heiko Rieger
https://arxiv.org/abs/2508.11430 http…

Thermodynamically Consistent Coarse-graining: from Interacting Particles to Fields via Second Quantization
We systematically derive an exact coarse-grained description for interacting particles with thermodynamically consistent stochastic dynamics, applicable across different observation scales, the mesoscopic and the macroscopic. We implement the coarse-graining procedure using the Doi-Peliti field theory, which preserves microscopic noise effects on the meso/macro scale. The exact mapping reveals the key role played by Poissonian particle occupancy statistics. We show the implications of the exact…

Seven Sphere Quantization
Subhobrata Chatterjee, Can G\"ormez, Andrew Waldron
https://arxiv.org/abs/2507.14363 https://arxiv.org…

Seven Sphere Quantization
Co-oriented contact manifolds quite generally describe classical dynamical systems. Quantization is achieved by suitably associating a Schrödinger equation to every path in the contact manifold. We quantize the standard contact seven sphere by treating it as a homogeneous space of the quaternionic unitary group in order to construct a contact analog of Fedosov's formal connection on symplectic spinor bundles. We show that requiring convergence of the formal connection naturally filters the sym…

TaDiCodec: Text-aware Diffusion Speech Tokenizer for Speech Language Modeling
Yuancheng Wang, Dekun Chen, Xueyao Zhang, Junan Zhang, Jiaqi Li, Zhizheng Wu
https://arxiv.org/abs/2508.16790

TaDiCodec: Text-aware Diffusion Speech Tokenizer for Speech Language Modeling
Speech tokenizers serve as foundational components for speech language models, yet current designs exhibit several limitations, including: 1) dependence on multi-layer residual vector quantization structures or high frame rates, 2) reliance on auxiliary pre-trained models for semantic distillation, and 3) requirements for complex two-stage training processes. In this work, we introduce the Text-aware Diffusion Transformer Speech Codec (TaDiCodec), a novel approach designed to overcome these cha…

How Quantization Impacts Privacy Risk on LLMs for Code?
Md Nazmul Haque, Hua Yang, Zhou Yang, Bowen Xu
https://arxiv.org/abs/2508.00128 https://arxiv.org/p…

How Quantization Impacts Privacy Risk on LLMs for Code?
Large language models for code (LLMs4Code) rely heavily on massive training data, including sensitive data, such as cloud service credentials of the projects and personal identifiable information of the developers, raising serious privacy concerns. Membership inference (MI) has recently emerged as an effective tool for assessing privacy risk by identifying whether specific data belong to a model's training set. In parallel, model compression techniques, especially quantization, have gained trac…

Dirac, Schroedinger, and Maxwell equations in scalar and vector field quantum mechanics
Boris Chichkov
https://arxiv.org/abs/2508.14583 https://arxiv.org/p…

Dirac, Schroedinger, and Maxwell equations in scalar and vector field quantum mechanics
The quantum theory of relativistic particles, based on the first quantization technique similar to that used by Schroedinger and Dirac in formulating quantum mechanics, is reconsidered on the basis of a photon-like dispersion relation corresponding to the energy conservation equation of Einstein's special relativity. First, scalar quantum mechanics of particles operating with their wave functions is discussed. Using the first quantization of the photon-like dispersion relation, very simple new …

Joint Quantization and Pruning Neural Networks Approach: A Case Study on FSO Receivers
Mohanad Obeed, Ming Jian
https://arxiv.org/abs/2506.20084 https://…

Joint Quantization and Pruning Neural Networks Approach: A Case Study on FSO Receivers
Towards fast, hardware-efficient, and low-complexity receivers, we propose a compression-aware learning approach and examine it on free-space optical (FSO) receivers for turbulence mitigation. The learning approach jointly quantize, prune, and train a convolutional neural network (CNN). In addition, we propose to have the CNN weights of power of two values so we replace the multiplication operations bit-shifting operations in every layer that has significant lower computational cost. The compre…

Groupoids in Diffeology
Patrick Iglesias-Zemmour
https://arxiv.org/abs/2508.17484 https://arxiv.org/pdf/2508.17484

Groupoids in Diffeology
This expository paper recounts the development and application of the concept of the diffeological groupoid, from its introduction in 1985 to its use in current research. We demonstrate how this single concept has served as a powerful and unifying tool for defining fundamental structures, analyzing the stratification of complex spaces like orbifolds, building a bridge to noncommutative geometry, and, most recently, forging new approaches to geometric quantization. The paper aims to provide a co…

Scalar Lattices and Probabilistic Shaping for Dithered Wyner-Ziv Quantization
Muhammed Yusuf Sener (Shitz), Gerhard Kramer (Shitz), Shlomo Shamai (Shitz), Wen Xu
https://arxiv.org/abs/2506.13162

Scalar Lattices and Probabilistic Shaping for Dithered Wyner-Ziv Quantization
Scalar lattice quantization with a modulo operator, dithering, and probabilistic shaping is applied to the Wyner-Ziv (WZ) problem with a Gaussian source and mean square error distortion. The method achieves the WZ rate-distortion pairs. The analysis is similar to that for dirty paper coding but requires additional steps to bound the distortion because the modulo shift is correlated with the source noise. The results extend to vector sources by reverse waterfilling on the spectrum of the covaria…

Making Pose Representations More Expressive and Disentangled via Residual Vector Quantization
Sukhyun Jeong, Hong-Gi Shin, Yong-Hoon Choi
https://arxiv.org/abs/2508.14561 https:…

Making Pose Representations More Expressive and Disentangled via Residual Vector Quantization
Recent progress in text-to-motion has advanced both 3D human motion generation and text-based motion control. Controllable motion generation (CoMo), which enables intuitive control, typically relies on pose code representations, but discrete pose codes alone cannot capture fine-grained motion details, limiting expressiveness. To overcome this, we propose a method that augments pose code-based latent representations with continuous motion features using residual vector quantization (RVQ). This d…

LLM Compression: How Far Can We Go in Balancing Size and Performance?
Sahil Sk, Debasish Dhal, Sonal Khosla, Sk Shahid, Sambit Shekhar, Akash Dhaka, Shantipriya Parida, Dilip K. Prasad, Ond\v{r}ej Bojar
https://arxiv.org/abs/2508.11318

LLM Compression: How Far Can We Go in Balancing Size and Performance?
Quantization is an essential and popular technique for improving the accessibility of large language models (LLMs) by reducing memory usage and computational costs while maintaining performance. In this study, we apply 4-bit Group Scaling Quantization (GSQ) and Generative Pretrained Transformer Quantization (GPTQ) to LLaMA 1B, Qwen 0.5B, and PHI 1.5B, evaluating their impact across multiple NLP tasks. We benchmark these models on MS MARCO (Information Retrieval), BoolQ (Boolean Question Answeri…

Replaced article(s) found for math-ph. https://arxiv.org/list/math-ph/new
[1/1]:
- Deformation Quantization with Separation of Variables of $G_{2,4}(\mathbb{C})$
Taika Okuda, Akifumi Sako

Performance Analysis of Post-Training Quantization for CNN-based Conjunctival Pallor Anemia Detection
Sebastian A. Cruz Romero, Wilfredo E. Lugo Beauchamp
https://arxiv.org/abs/2507.15151

Performance Analysis of Post-Training Quantization for CNN-based Conjunctival Pallor Anemia Detection
Anemia is a widespread global health issue, particularly among young children in low-resource settings. Traditional methods for anemia detection often require expensive equipment and expert knowledge, creating barriers to early and accurate diagnosis. To address these challenges, we explore the use of deep learning models for detecting anemia through conjunctival pallor, focusing on the CP-AnemiC dataset, which includes 710 images from children aged 6-59 months. The dataset is annotated with he…

Dynamic Sparse Attention on Mobile SoCs
Wangsong Yin, Daliang Xu, Mengwei Xu, Gang Huang, Xuanzhe Liu
https://arxiv.org/abs/2508.16703 https://arxiv.org/pd…

Dynamic Sparse Attention on Mobile SoCs
On-device running Large Language Models (LLMs) is nowadays a critical enabler towards preserving user privacy. We observe that the attention operator falls back from the special-purpose NPU to the general-purpose CPU/GPU because of quantization sensitivity in state-of-the-art frameworks. This fallback results in a degraded user experience and increased complexity in system scheduling. To this end, this paper presents shadowAttn, a system-algorithm codesigned sparse attention module with minimal…

MMQ: Multimodal Mixture-of-Quantization Tokenization for Semantic ID Generation and User Behavioral Adaptation
Yi Xu, Moyu Zhang, Chenxuan Li, Zhihao Liao, Haibo Xing, Hao Deng, Jinxin Hu, Yu Zhang, Xiaoyi Zeng, Jing Zhang
https://arxiv.org/abs/2508.15281

MMQ: Multimodal Mixture-of-Quantization Tokenization for Semantic ID Generation and User Behavioral Adaptation
Recommender systems traditionally represent items using unique identifiers (ItemIDs), but this approach struggles with large, dynamic item corpora and sparse long-tail data, limiting scalability and generalization. Semantic IDs, derived from multimodal content such as text and images, offer a promising alternative by mapping items into a shared semantic space, enabling knowledge transfer and improving recommendations for new or rare items. However, existing methods face two key challenges: (1) …

Pair production of massive charged vector bosons from the worldline
Fiorenzo Bastianelli, Filippo Fecit, Alessandro Miccich\`e
https://arxiv.org/abs/2507.15943

Pair production of massive charged vector bosons from the worldline
We investigate a worldline formulation for a massive spin-1 particle interacting with an electromagnetic background. Two first-quantized descriptions of the spin degrees of freedom are considered: one based on bosonic oscillators and the other on fermionic oscillators. Focusing initially on the bosonic model -- which can accommodate particles of arbitrary integer spin -- we review how quantization in the spin-1 sector, performed both via Dirac's method and BRST quantization, reproduces the free…

Selective Quantization Tuning for ONNX Models
Nikolaos Louloudakis, Ajitha Rajan
https://arxiv.org/abs/2507.12196 https://arxiv.org/p…

Selective Quantization Tuning for ONNX Models
Quantization is a process that reduces the precision of deep neural network models to lower model size and computational demands, often at the cost of accuracy. However, fully quantized models may exhibit sub-optimal performance below acceptable levels and face deployment challenges on low-end hardware accelerators due to practical constraints. To address these issues, quantization can be selectively applied to only a subset of layers, but selecting which layers to exclude is non-trivial. To th…

This https://arxiv.org/abs/2505.23786 has been replaced.
initial toot: https://mastoxiv.page/@arXiv_csCR_…

Mind the Gap: A Practical Attack on GGUF Quantization
With the increasing size of frontier LLMs, post-training quantization has become the standard for memory-efficient deployment. Recent work has shown that basic rounding-based quantization schemes pose security risks, as they can be exploited to inject malicious behaviors into quantized models that remain hidden in full precision. However, existing attacks cannot be applied to more complex quantization methods, such as the GGUF family used in the popular ollama and llama.cpp frameworks. In this …

Image-Conditioned 3D Gaussian Splat Quantization
Xinshuang Liu, Runfa Blark Li, Keito Suzuki, Truong Nguyen
https://arxiv.org/abs/2508.15372 https://arxiv.…

Image-Conditioned 3D Gaussian Splat Quantization
3D Gaussian Splatting (3DGS) has attracted considerable attention for enabling high-quality real-time rendering. Although 3DGS compression methods have been proposed for deployment on storage-constrained devices, two limitations hinder archival use: (1) they compress medium-scale scenes only to the megabyte range, which remains impractical for large-scale scenes or extensive scene collections; and (2) they lack mechanisms to accommodate scene changes after long-term archival. To address these l…

Enhancing Convergence, Privacy and Fairness for Wireless Personalized Federated Learning: Quantization-Assisted Min-Max Fair Scheduling
Xiyu Zhao, Qimei Cui, Ziqiang Du, Weicai Li, Xi Yu, Wei Ni, Ji Zhang, Xiaofeng Tao, Ping Zhang
https://arxiv.org/abs/2506.02422

Enhancing Convergence, Privacy and Fairness for Wireless Personalized Federated Learning: Quantization-Assisted Min-Max Fair Scheduling
Personalized federated learning (PFL) offers a solution to balancing personalization and generalization by conducting federated learning (FL) to guide personalized learning (PL). Little attention has been given to wireless PFL (WPFL), where privacy concerns arise. Performance fairness of PL models is another challenge resulting from communication bottlenecks in WPFL. This paper exploits quantization errors to enhance the privacy of WPFL and proposes a novel quantization-assisted Gaussian differ…

GDNSQ: Gradual Differentiable Noise Scale Quantization for Low-bit Neural Networks
Sergey Salishev, Ian Akhremchik
https://arxiv.org/abs/2508.14004 https://

GDNSQ: Gradual Differentiable Noise Scale Quantization for Low-bit Neural Networks
Quantized neural networks can be viewed as a chain of noisy channels, where rounding in each layer reduces capacity as bit-width shrinks; the floating-point (FP) checkpoint sets the maximum input rate. We track capacity dynamics as the average bit-width decreases and identify resulting quantization bottlenecks by casting fine-tuning as a smooth, constrained optimization problem. Our approach employs a fully differentiable Straight-Through Estimator (STE) with learnable bit-width, noise scale an…

Crosslisted article(s) found for cs.IT. https://arxiv.org/list/cs.IT/new
[1/1]:
- Exploiting Information Redundancy in Attention Maps for Extreme Quantization of Vision Transformers
Lucas Maisonnave, Karim Haroun, Tom Pegeot

Effect of Signal Quantization on Performance Measures of a 1st Order One Dimensional Differential Microphone Array
Shweta Pal, Arun Kumar, Monika Agrawal
https://arxiv.org/abs/2506.15463

Effect of Signal Quantization on Performance Measures of a 1st Order One Dimensional Differential Microphone Array
In practical systems, recorded analog signals must be digitized for processing, introducing quantization as a critical aspect of data acquisition. While prior studies have examined quantization effects in various signal processing contexts, its impact on differential microphone arrays (DMAs), particularly in one-dimensional (1D) first-order configurations, remains unexplored. This paper investigates the influence of signal quantization on performance of first-order 1D DMAs across various beampa…

Semiclassical polymer field with a cubic potential in cosmology
Ahsan Mujtaba, Syed Moeez Hassan
https://arxiv.org/abs/2508.16416 https://arxiv.org/pdf/250…

Semiclassical polymer field with a cubic potential in cosmology
We study the semiclassical dynamics of a polymer quantized scalar field with a cubic potential in cosmology. The cosmological spacetime is chosen to be homogeneous and isotropic, and we work in the polymer quantization scheme where the field operator does not exist directly. The dynamics are studied in the `e-fold' time gauge, where a choice of time gauge fixing, in terms of the scale factor, is made at the classical level. The cubic potential is interesting to study because it is unbounded fro…

Edge-ASR: Towards Low-Bit Quantization of Automatic Speech Recognition Models
Chen Feng, Yicheng Lin, Shaojie Zhuo, Chenzheng Su, Ramchalam Kinattinkara Ramakrishnan, Zhaocong Yuan, Xiaopeng Zhang
https://arxiv.org/abs/2507.07877

Edge-ASR: Towards Low-Bit Quantization of Automatic Speech Recognition Models
Recent advances in Automatic Speech Recognition (ASR) have demonstrated remarkable accuracy and robustness in diverse audio applications, such as live transcription and voice command processing. However, deploying these models on resource constrained edge devices (e.g., IoT device, wearables) still presents substantial challenges due to strict limits on memory, compute and power. Quantization, particularly Post-Training Quantization (PTQ), offers an effective way to reduce model size and infere…

Geometric Quantization by Paths -- Part I: The Simply Connected Case
Patrick Iglesias-Zemmour
https://arxiv.org/abs/2508.11337 https://arxiv.org/pdf/2508.1…

Geometric Quantization by Paths -- Part I: The Simply Connected Case
For any connected and simply connected parasymplectic space $(\mathrm{X},ω)$ with group of periods $\mathrm{P}_ω\subsetneq \mathbf{R}$, we construct a prequantum groupoid $\pmb{\mathrm{T}}_ω$ as a diffeological quotient of the space $\mathrm{Paths}(\mathrm{X})$ of paths in $\mathrm{X}$. This object, built from the geometry of the classical system, serves as a unified structure for prequantization. The groupoid $\pmb{\mathrm{T}}_ω$ has $\mathrm{X}$ as its objects, and its space of morphisms …

Joint Active and Passive Beamforming for Energy-Efficient STARS with Quantization and Element Selection in ISAC Systems
Li-Hsiang Shen, Yi-Hsuan Chiu
https://arxiv.org/abs/2507.16210

Joint Active and Passive Beamforming for Energy-Efficient STARS with Quantization and Element Selection in ISAC Systems
This paper investigates a simultaneously transmitting and reflecting reconfigurable intelligent surface (STARS)-aided integrated sensing and communication (ISAC) systems in support of full-space energy-efficient data transmissions and target sensing. We formulate an energy efficiency (EE) maximization problem jointly optimizing dual-functional radar-communication (DFRC)-empowered base station (BS) ISAC beamforming and STARS configurations of amplitudes, phase-shifts, quantization levels as well…

Anomaly of 4d Weyl fermion with discrete charge
Zheyan Wan
https://arxiv.org/abs/2506.19710 https://arxiv.org/pdf/2506.19710

Anomaly of 4d Weyl fermion with discrete charge
We derive the anomaly indices of 4d Weyl fermions with symmetry $\mathrm{Spin}\times\mathbb{Z}_n$ and $\mathrm{Spin}\times_{\mathbb{Z}_2}\mathbb{Z}_{2m}$ by reducing the continuous $\mathrm{Spin} \times \mathrm{U}(1)$ and $\mathrm{Spin}^c$ anomaly indices on five-dimensional manifolds to discrete sectors subject to flux quantization. We calculate the values of the 5d invertible theories on the corresponding manifold generators. We express the values of the 5d invertible theories on general mani…

Robust ISAC Transceiver Beamforming Design under Low-Resolution AD/DA Converters
Tiantian Xu, Zhenyao He, Jindan Xu, Wei Xu, Jianfeng Wang, Derrick Wing Kwan Ng
https://arxiv.org/abs/2507.15373

Robust ISAC Transceiver Beamforming Design under Low-Resolution AD/DA Converters
In this letter, we investigate the robust beamforming design for an integrated sensing and communication (ISAC) system featuring low-resolution digital-to-analog converters (DACs) and analog-to-digital converters (ADCs). Taking into account quantization noise, we aim at maximizing the radar signal-to-quantization-plus-noise ratio (SQNR) while guaranteeing the minimum required signal-to-quantization-plus-interference-plus-noise ratio (SQINR) for communication users. To address this nonconvex des…

Rotation Errors Due to Field Quantization for Simultaneously Driven Atoms
Hunter Lindemann, Shanon Vuglar, Julio Gea-Banacloche
https://arxiv.org/abs/2508.06769 https://

Rotation Errors Due to Field Quantization for Simultaneously Driven Atoms
When an electromagnetic field in a coherent or quasiclassical (e.g., squeezed) state is used to simultaneously drive an ensemble of two-level atoms, the quantum nature of the field will, in general, cause the final state of the atoms to differ from the one predicted for a totally classical field. This is a potential source of error in quantum logic gates in which the gate is the rotation of the atoms by a laser. In this paper, we use second order perturbation theory to find how this error scale…

This https://arxiv.org/abs/2506.05960 has been replaced.
initial toot: https://mastoxiv.page/@arXiv_csLG_…

AQUATIC-Diff: Additive Quantization for Truly Tiny Compressed Diffusion Models
Significant investments have been made towards the commodification of diffusion models for generation of diverse media. Their mass-market adoption is however still hobbled by the intense hardware resource requirements of diffusion model inference. Model quantization strategies tailored specifically towards diffusion models have been useful in easing this burden, yet have generally explored the Uniform Scalar Quantization (USQ) family of quantization methods. In contrast, Vector Quantization (VQ…

Representation Quantization for Collaborative Filtering Augmentation
Yunze Luo, Yinjie Jiang, Gaode Chen, Jingchi Wang, Shicheng Wang, Ruina Sun, Jiang Yuezihan, Jun Zhang, Jian Liang, Han Li, Kun Gai, Kaigui Bian
https://arxiv.org/abs/2508.11194

Representation Quantization for Collaborative Filtering Augmentation
As the core algorithm in recommendation systems, collaborative filtering (CF) algorithms inevitably face the problem of data sparsity. Since CF captures similar users and items for recommendations, it is effective to augment the lacking user-user and item-item homogeneous linkages. However, existing methods are typically limited to connecting through overlapping interacted neighbors or through similar attributes and contents. These approaches are constrained by coarse-grained, sparse attributes…

Flexible Mixed Precision Quantization for Learned Image Compression
Md Adnan Faisal Hossain, Zhihao Duan, Fengqing Zhu
https://arxiv.org/abs/2506.01221 htt…

Flexible Mixed Precision Quantization for Learned Image Compression
Despite its improvements in coding performance compared to traditional codecs, Learned Image Compression (LIC) suffers from large computational costs for storage and deployment. Model quantization offers an effective solution to reduce the computational complexity of LIC models. However, most existing works perform fixed-precision quantization which suffers from sub-optimal utilization of resources due to the varying sensitivity to quantization of different layers of a neural network. In this p…

Spatial-Temporal Multi-Scale Quantization for Flexible Motion Generation
Zan Wang, Jingze Zhang, Yixin Chen, Baoxiong Jia, Wei Liang, Siyuan Huang
https://arxiv.org/abs/2508.08991

Spatial-Temporal Multi-Scale Quantization for Flexible Motion Generation
Despite significant advancements in human motion generation, current motion representations, typically formulated as discrete frame sequences, still face two critical limitations: (i) they fail to capture motion from a multi-scale perspective, limiting the capability in complex patterns modeling; (ii) they lack compositional flexibility, which is crucial for model's generalization in diverse generation tasks. To address these challenges, we introduce MSQ, a novel quantization method that compre…

Null infinity as $SU(2)$ Chern-Simons theories and its quantization
Hongwei Tan, Kui Xiao, Shoucheng Wang
https://arxiv.org/abs/2508.04220 https://arxiv.or…

Null infinity as $SU(2)$ Chern-Simons theories and its quantization
This paper studies the quantization of the future null infinity ($\mathscr{I}^+$) of an asymptotically flat spacetime. Based on the observation by Ashtekar and Speziale that $\mathscr{I}^+$ can be regarded as a weakly isolated horizon, we adopt the quantization framework developed for weakly horizon to quantize $\mathscr{I}^+$. We first show that the symplectic structure of $\mathscr{I}^+$ is equivalent to the sum of the symplectic structures of two $SU(2)$ Chern-Simons theories with opposite l…

Second Quantization for the Kepler Problem
John C. Baez
https://arxiv.org/abs/2507.11428 https://arxiv.org/pdf/2507.11428

Second Quantization for the Kepler Problem
The Kepler problem concerns a point particle in an attractive inverse square force. After a brief review of the classical and quantum versions of this problem, focused on their hidden $\text{SU}(2) \times \text{SU}(2)$ symmetry, we discuss the quantum Kepler problem for a spin-$\frac{1}{2}$ particle. We show that the Hilbert space $\mathcal{H}$ of bound states for this problem is unitarily equivalent, as a representation of $\text{SU}(2) \times \text{SU}(2)$, to the Hilbert space of solutions o…

DLLMQuant: Quantizing Diffusion-based Large Language Models
Chen Xu, Dawei Yang
https://arxiv.org/abs/2508.14090 https://arxiv.org/pdf/2508.14090

DLLMQuant: Quantizing Diffusion-based Large Language Models
Diffusion-based large language models (DLLMs) have shown promise for non-autoregressive text generation, but their deployment is constrained by large model sizes and heavy computational costs. Post-training quantization (PTQ), a widely used method for compressing and accelerating Large Language Models (LLMs), suffers from severe accuracy degradation and reduced generalization performance when directly applied to DLLMs (e.g., AWQ suffers a 16% accuracy drop on LLADA under W4A4). This paper explo…

Towards Bitrate-Efficient and Noise-Robust Speech Coding with Variable Bitrate RVQ
Yunkee Chae, Kyogu Lee
https://arxiv.org/abs/2506.16538 https://

Towards Bitrate-Efficient and Noise-Robust Speech Coding with Variable Bitrate RVQ
Residual Vector Quantization (RVQ) has become a dominant approach in neural speech and audio coding, providing high-fidelity compression. However, speech coding presents additional challenges due to real-world noise, which degrades compression efficiency. Standard codecs allocate bits uniformly, wasting bitrate on noise components that do not contribute to intelligibility. This paper introduces a Variable Bitrate RVQ (VRVQ) framework for noise-robust speech coding, dynamically adjusting bitrate…

Neutralization of Levitated Charged Nanodiamond: Towards matter-wave interferometry with massive objects
Sela Liran, Or Dobkowski, Rafael Benjaminov, Peter Skakunenko, Michael Averbukh, Yaniv Bar-Haim, David Groswasser, Joshua H. Baraban, Ron Folman
https://arxiv.org/abs/2508.15625

Neutralization of Levitated Charged Nanodiamond: Towards matter-wave interferometry with massive objects
Quantum mechanics (QM) and General relativity (GR), also known as the theory of gravity, are the two pillars of modern physics. A matter-wave interferometer with a massive particle, can test numerous fundamental ideas, including the spatial superposition principle - a foundational concept in QM - in completely new regimes, as well as the interface between QM and GR, e.g., testing the quantization of gravity. Consequently, there exists an intensive effort to realize such an interferometer. While…

MGVQ: Could VQ-VAE Beat VAE? A Generalizable Tokenizer with Multi-group Quantization
Mingkai Jia, Wei Yin, Xiaotao Hu, Jiaxin Guo, Xiaoyang Guo, Qian Zhang, Xiao-Xiao Long, Ping Tan
https://arxiv.org/abs/2507.07997

MGVQ: Could VQ-VAE Beat VAE? A Generalizable Tokenizer with Multi-group Quantization
Vector Quantized Variational Autoencoders (VQ-VAEs) are fundamental models that compress continuous visual data into discrete tokens. Existing methods have tried to improve the quantization strategy for better reconstruction quality, however, there still exists a large gap between VQ-VAEs and VAEs. To narrow this gap, we propose \NickName, a novel method to augment the representation capability of discrete codebooks, facilitating easier optimization for codebooks and minimizing information loss…

Efficient Edge LLMs Deployment via HessianAware Quantization and CPU GPU Collaborative
Tuo Zhang, Ning Li, Xin Yuan, Wenchao Xu, Quan Chen, Song Guo, Haijun Zhang
https://arxiv.org/abs/2508.07329

Efficient Edge LLMs Deployment via HessianAware Quantization and CPU GPU Collaborative
With the breakthrough progress of large language models (LLMs) in natural language processing and multimodal tasks, efficiently deploying them on resource-constrained edge devices has become a critical challenge. The Mixture of Experts (MoE) architecture enhances model capacity through sparse activation, but faces two major difficulties in practical deployment: (1) The presence of numerous outliers in activation distributions leads to severe degradation in quantization accuracy for both activat…

Covariant quantization of field theories on T-Minkowski noncommutative spacetimes
Giuseppe Fabiano, Flavio Mercati
https://arxiv.org/abs/2508.04527 https://

Covariant quantization of field theories on T-Minkowski noncommutative spacetimes
We develop a quantization scheme for the quantum theory of a real scalar field on a class of non-commutative spacetime models collectively known as T-Minkowski. Requiring the theory to be covariant under T-Poincaré transformations, we find that for a subclass of models the Wightmann functions are equal to their commutative counterparts, and we are able to prove a Wick theorem for Wightmann functions that is structurally equivalent to the one encountered in commutative QFT. For some of these mo…

Combinatorial quantization of 4d 2-Chern-Simons theory II: Quantum invariants of higher ribbons in $D^4$
Hank Chen
https://arxiv.org/abs/2506.05785 https:/…

Combinatorial quantization of 4d 2-Chern-Simons theory II: Quantum invariants of higher ribbons in $D^4$
This is a continuation of the first paper (arXiv:2501.06486) of this series, where the framework for the combinatorial quantization of the 4d 2-Chern-Simons theory with an underlying compact structure Lie 2-group $\mathbb{G}$ was laid out. In this paper, we continue our quest and characterize additive module *-functors $ω:\mathfrak{C}_q(\mathbb{G}^{Γ^2})\rightarrow\mathsf{Hilb}$, which serve as a categorification of linear *-functionals (ie. a state) on a $C^*$-algebra. These allow us to cons…

Quantum control of Nitrogen-Vacancy spin in Diamonds: Towards matter-wave interferometry with massive objects
N. Levi, O. Feldman, Y. Rosenzweig, D. Groswasser, A. Elgarat, M. Gal-Katizri, R. Folman
https://arxiv.org/abs/2508.15504

Quantum control of Nitrogen-Vacancy spin in Diamonds: Towards matter-wave interferometry with massive objects
Quantum mechanics (QM) and General relativity (GR), also known as the theory of gravity, are the two pillars of modern physics. A matter-wave interferometer with a massive particle can test numerous fundamental ideas, including the spatial superposition principle - a foundational concept in QM - in previously unexplored regimes. It also opens the possibility of probing the interface between QM and GR, such as testing the quantization of gravity. Consequently, there exists an intensive effort to…

$Q$-operators, $q$-opers, and R-matrices in 5d $\mathcal{N}=1$ gauge theory
Saebyeok Jeong, Norton Lee
https://arxiv.org/abs/2507.15450 https://

$Q$-operators, $q$-opers, and R-matrices in 5d $\mathcal{N}=1$ gauge theory
We study the quantization of the moduli space of multiplicative Higgs bundles through the lens of five-dimensional $\mathcal{N}=1$ supersymmetric gauge theories in $Ω$-background. We extend the 4d $\mathcal{N}=2$ gauge theoretical construction of key geometric and representation-theoretic structures, established in earlier works, to the five-dimensional uplift. We construct and analyze the $Q$-operators and $q$-opers associated with the canonical codimension-two defect: the $Q$-operators are d…

Bohr-Sommerfeld Quantization Rules for 1-D Semiclassical Pseudo-Differential Operator: the Method of Microlocal Wronskian and Gram Matrix
Abdelwaheb Ifa
https://arxiv.org/abs/2508.05586

Bohr-Sommerfeld Quantization Rules for 1-D Semiclassical Pseudo-Differential Operator: the Method of Microlocal Wronskian and Gram Matrix
In this paper, we revisit the well known Bohr-Sommerfeld quantization rule (BS) of order 2 for a self-adjoint 1-D semiclassical pseudo-differential operator, within the algebraic and microlocal framework of B. Helffer and J. Sjöstrand. BS holds precisely when the Gram matrix consisting of scalar products of certain WKB solutions with respect to the "flux norm" is not invertible. This condition is obtained using the microlocal Wronskian and does not rely on traditional matching techniques. It i…

VLMQ: Efficient Post-Training Quantization for Large Vision-Language Models via Hessian Augmentation
Yufei Xue, Yushi Huang, Jiawei Shao, Jun Zhang
https://arxiv.org/abs/2508.03351

VLMQ: Efficient Post-Training Quantization for Large Vision-Language Models via Hessian Augmentation
Post-training quantization (PTQ) has emerged as an effective approach for compressing large models and accelerating their inference without retraining. While PTQ has been extensively studied in the context of large language models (LLMs), its applicability to vision-language models (VLMs) remains underexplored. In this paper, we identify a modality discrepancy (\emph{i.e.}, limited text tokens \emph{vs.} excessive and redundant vision tokens) of VLMs. However, existing Hessian-based LLM PTQ met…

MSQ: Memory-Efficient Bit Sparsification Quantization
Seokho Han, Seoyeon Yoon, Jinhee Kim, Dongwei Wang, Kang Eun Jeon, Huanrui Yang, Jong Hwan Ko
https://arxiv.org/abs/2507.22349

MSQ: Memory-Efficient Bit Sparsification Quantization
As deep neural networks (DNNs) see increased deployment on mobile and edge devices, optimizing model efficiency has become crucial. Mixed-precision quantization is widely favored, as it offers a superior balance between efficiency and accuracy compared to uniform quantization. However, finding the optimal precision for each layer is challenging. Recent studies utilizing bit-level sparsity have shown promise, yet they often introduce substantial training complexity and high GPU memory requiremen…

NextStep-1: Toward Autoregressive Image Generation with Continuous Tokens at Scale
NextStep Team, Chunrui Han, Guopeng Li, Jingwei Wu, Quan Sun, Yan Cai, Yuang Peng, Zheng Ge, Deyu Zhou, Haomiao Tang, Hongyu Zhou, Kenkun Liu, Ailin Huang, Bin Wang, Changxin Miao, Deshan Sun, En Yu, Fukun Yin, Gang Yu, Hao Nie, Haoran Lv, Hanpeng Hu, Jia Wang, Jian Zhou, Jianjian Sun, Kaijun Tan, Kang An, Kangheng Lin, Liang Zhao, Mei Chen, Peng Xing, Rui Wang, Shiyu Liu, Shutao Xia, Tianhao You, Wei Ji…

NextStep-1: Toward Autoregressive Image Generation with Continuous Tokens at Scale
Prevailing autoregressive (AR) models for text-to-image generation either rely on heavy, computationally-intensive diffusion models to process continuous image tokens, or employ vector quantization (VQ) to obtain discrete tokens with quantization loss. In this paper, we push the autoregressive paradigm forward with NextStep-1, a 14B autoregressive model paired with a 157M flow matching head, training on discrete text tokens and continuous image tokens with next-token prediction objectives. Next…

LRQ-DiT: Log-Rotation Post-Training Quantization of Diffusion Transformers for Text-to-Image Generation
Lianwei Yang, Haokun Lin, Tianchen Zhao, Yichen Wu, Hongyu Zhu, Ruiqi Xie, Zhenan Sun, Yu Wang, Qingyi Gu
https://arxiv.org/abs/2508.03485

LRQ-DiT: Log-Rotation Post-Training Quantization of Diffusion Transformers for Text-to-Image Generation
Diffusion Transformers (DiTs) have achieved impressive performance in text-to-image generation. However, their high computational cost and large parameter sizes pose significant challenges for usage in resource-constrained scenarios. Post-training quantization (PTQ) is a promising solution to reduce memory usage and accelerate inference, but existing PTQ methods suffer from severe performance degradation under extreme low-bit settings. We identify two key obstacles to low-bit post-training quan…

State Estimation with 1-Bit Observations and Imperfect Models: Bussgang Meets Kalman in Neural Networks
Chaehyun Jung, TaeJun Ha, Hyeonuk Kim, Jeonghun Park
https://arxiv.org/abs/2507.17284

State Estimation with 1-Bit Observations and Imperfect Models: Bussgang Meets Kalman in Neural Networks
State estimation from noisy observations is a fundamental problem in many applications of signal processing. Traditional methods, such as the extended Kalman filter, work well under fully-known Gaussian models, while recent hybrid deep learning frameworks, combining model-based and data-driven approaches, can also handle partially known models and non-Gaussian noise. However, existing studies commonly assume the absence of quantization distortion, which is inevitable, especially with non-ideal …

This https://arxiv.org/abs/2505.07004 has been replaced.
initial toot: https://mastoxiv.page/@arXiv_csLG_…

GuidedQuant: Large Language Model Quantization via Exploiting End Loss Guidance
Post-training quantization is a key technique for reducing the memory and inference latency of large language models by quantizing weights and activations without requiring retraining. However, existing methods either (1) fail to account for the varying importance of hidden features to the end loss or, when incorporating end loss, (2) neglect the critical interactions between model weights. To address these limitations, we propose GuidedQuant, a novel quantization approach that integrates gradi…

SGD Convergence under Stepsize Shrinkage in Low-Precision Training
Vincent-Daniel Yun
https://arxiv.org/abs/2508.07142 https://arxiv.org/pdf/2508.07142

SGD Convergence under Stepsize Shrinkage in Low-Precision Training
Low-precision training has become essential for reducing the computational and memory costs of large-scale deep learning. However, quantization of gradients introduces both magnitude shrinkage and additive noise, which can alter the convergence behavior of stochastic gradient descent (SGD). In this work, we study the convergence of SGD under a gradient shrinkage model, where each stochastic gradient is scaled by a factor $q_k \in (0,1]$ and perturbed by zero-mean quantization noise. We show tha…

Formal Algorithms for Model Efficiency
Naman Tyagi, Srishti Das, Kunal, Vatsal Gupta
https://arxiv.org/abs/2508.14000 https://arxiv.org/pdf/2508.14000

Formal Algorithms for Model Efficiency
We introduce the Knob-Meter-Rule (KMR) framework, a unified formalism for representing and reasoning about model efficiency techniques in deep learning. By abstracting diverse methods, including pruning, quantization, knowledge distillation, and parameter-efficient architectures, into a consistent set of controllable knobs, deterministic rules, and measurable meters, KMR provides a mathematically precise and modular perspective on efficiency optimization. The framework enables systematic compos…