Tootfinder

ScaleCap: Inference-Time Scalable Image Captioning via Dual-Modality Debiasing
Long Xing, Qidong Huang, Xiaoyi Dong, Pan Zhang, Yuhang Zang, Yuhang Cao, Jinsong Li, Shuangrui Ding, Weiming Zhang, Nenghai Yu, Jiaqi Wang, Feng Wu, Dahua Lin
https://arxiv.org/abs/2506.19848

ScaleCap: Inference-Time Scalable Image Captioning via Dual-Modality Debiasing
This paper presents ScaleCap, an inference-time scalable image captioning strategy that generates comprehensive and detailed image captions. The key challenges of high-quality image captioning lie in the inherent biases of LVLMs: multimodal bias resulting in imbalanced descriptive granularity, offering detailed accounts of some elements while merely skimming over others; linguistic bias leading to hallucinated descriptions of non-existent objects. To address these issues, we propose a scalable …

Agentic AI framework for End-to-End Medical Data Inference
Soorya Ram Shimgekar, Shayan Vassef, Abhay Goyal, Navin Kumar, Koustuv Saha
https://arxiv.org/abs/2507.18115 https://

Agentic AI framework for End-to-End Medical Data Inference
Building and deploying machine learning solutions in healthcare remains expensive and labor-intensive due to fragmented preprocessing workflows, model compatibility issues, and stringent data privacy constraints. In this work, we introduce an Agentic AI framework that automates the entire clinical data pipeline, from ingestion to inference, through a system of modular, task-specific agents. These agents handle both structured and unstructured data, enabling automatic feature selection, model se…

When Life Gives You Samples: The Benefits of Scaling up Inference Compute for Multilingual LLMs
Ammar Khairi, Daniel D'souza, Ye Shen, Julia Kreutzer, Sara Hooker
https://arxiv.org/abs/2506.20544

When Life Gives You Samples: The Benefits of Scaling up Inference Compute for Multilingual LLMs
Recent advancements in large language models (LLMs) have shifted focus toward scaling inference-time compute, improving performance without retraining the model. A common approach is to sample multiple outputs in parallel, and select one of these as the final output. However, work to date has focused on English and a handful of domains such as math and code. In contrast, we are most interested in techniques that generalize across open-ended tasks, formally verifiable tasks, and across languages…

Evaluating the Defense Potential of Machine Unlearning against Membership Inference Attacks
Aristeidis Sidiropoulos, Christos Chrysanthos Nikolaidis, Theodoros Tsiolakis, Nikolaos Pavlidis, Vasilis Perifanis, Pavlos S. Efraimidis
https://arxiv.org/abs/2508.16150

Evaluating the Defense Potential of Machine Unlearning against Membership Inference Attacks
Membership Inference Attacks (MIAs) pose a significant privacy risk, as they enable adversaries to determine whether a specific data point was included in the training dataset of a model. While Machine Unlearning is primarily designed as a privacy mechanism to efficiently remove private data from a machine learning model without the need for full retraining, its impact on the susceptibility of models to MIA remains an open question. In this study, we systematically assess the vulnerability of m…

Cloud Native System for LLM Inference Serving
Minxian Xu, Junhan Liao, Jingfeng Wu, Yiyuan He, Kejiang Ye, Chengzhong Xu
https://arxiv.org/abs/2507.18007 https://

Cloud Native System for LLM Inference Serving
Large Language Models (LLMs) are revolutionizing numerous industries, but their substantial computational demands create challenges for efficient deployment, particularly in cloud environments. Traditional approaches to inference serving often struggle with resource inefficiencies, leading to high operational costs, latency issues, and limited scalability. This article explores how Cloud Native technologies, such as containerization, microservices, and dynamic scheduling, can fundamentally impr…

Quasi-likelihood inference for SDE with mixed-effects observed at high frequency
Maud Delattre, Hiroki Masuda
https://arxiv.org/abs/2508.17910 https://arxi…

Quasi-likelihood inference for SDE with mixed-effects observed at high frequency
We consider statistical inference for a class of dynamic mixed-effect models described by stochastic differential equations whose drift and diffusion coefficients simultaneously depend on fixed- and random-effect parameters. Assuming that each process is observed at high frequency and the number of individuals goes to infinity, we propose a stepwise inference procedure and prove its theoretical properties. The methodology is based on suitable quasi-likelihood functions by profiling the random e…

Simulation-Based Inference for Direction Reconstruction of Ultra-High-Energy Cosmic Rays with Radio Arrays
Oscar Macias, Zachary Mason, Matthew Ho, Ars\`ene Ferri\`ere, Aur\'elien Benoit-L\'evy, Mat\'ias Tueros
https://arxiv.org/abs/2508.15991

Simulation-Based Inference for Direction Reconstruction of Ultra-High-Energy Cosmic Rays with Radio Arrays
Ultra-high-energy cosmic-ray (UHECR) observatories require unbiased direction reconstruction to enable multi-messenger astronomy with sparse, nanosecond-scale radio pulses. Explicit likelihood methods often rely on simplified models, which may bias results and understate uncertainties. We introduce a simulation-based inference pipeline that couples a physics-informed graph neural network (GNN) to a normalizing-flow posterior within the \textit{Learning the Universe Implicit Likelihood Inference…

Validating Sequential Monte Carlo for Gravitational-Wave Inference
Michael J. Williams, Minas Karamanis, Yilin Luo, Uro\v{s} Seljak
https://arxiv.org/abs/2506.18977

Validating Sequential Monte Carlo for Gravitational-Wave Inference
Nested sampling (NS) is the preferred stochastic sampling algorithm for gravitational-wave inference for compact binary coalenscences (CBCs). It can handle the complex nature of the gravitational-wave likelihood surface and provides an estimate of the Bayesian model evidence. However, there is another class of algorithms that meets the same requirements but has not been used for gravitational-wave analyses: Sequential Monte Carlo (SMC), an extension of importance sampling that maps samples from…

ViRN: Variational Inference and Distribution Trilateration for Long-Tailed Continual Representation Learning
Hao Dai, Chong Tang, Jagmohan Chauhan
https://arxiv.org/abs/2507.17368

ViRN: Variational Inference and Distribution Trilateration for Long-Tailed Continual Representation Learning
Continual learning (CL) with long-tailed data distributions remains a critical challenge for real-world AI systems, where models must sequentially adapt to new classes while retaining knowledge of old ones, despite severe class imbalance. Existing methods struggle to balance stability and plasticity, often collapsing under extreme sample scarcity. To address this, we propose ViRN, a novel CL framework that integrates variational inference (VI) with distributional trilateration for robust long-t…

Minimax Data Sanitization with Distortion Constraint and Adversarial Inference
Amirarsalan Moatazedian, Yauhen Yakimenka, R\'emi A. Chou, J\"org Kliewer
https://arxiv.org/abs/2507.17942

Minimax Data Sanitization with Distortion Constraint and Adversarial Inference
We study a privacy-preserving data-sharing setting where a privatizer transforms private data into a sanitized version observed by an authorized reconstructor and two unauthorized adversaries, each with access to side information correlated with the private data. The reconstructor is evaluated under a distortion function, while each adversary is evaluated using a separate loss function. The privatizer ensures the reconstructor distortion remains below a fixed threshold while maximizing the mi…

Scalable Bayesian inference on high-dimensional multivariate linear regression
Xuan Cao, Kyoungjae Lee
https://arxiv.org/abs/2508.16446 https://arxiv.org/p…

Scalable Bayesian inference on high-dimensional multivariate linear regression
We consider jointly estimating the coefficient matrix and the error precision matrix in high-dimensional multivariate linear regression models. Bayesian methods in this context often face computational challenges, leading to previous approaches that either utilize a generalized likelihood without ensuring the positive definiteness of the precision matrix or rely on maximization algorithms targeting only the posterior mode, thus failing to address uncertainty. In this work, we propose two Bayesi…

Bare-Metal RISC-V NVDLA SoC for Efficient Deep Learning Inference
Vineet Kumar (School of Electrical,Electronic Engineering, University College Dublin, Dublin, Ireland, Department of Electronic,Electrical Engineering, Trinity College Dublin, Dublin, Ireland), Ajay Kumar M (School of Electrical,Electronic Engineering, University College Dublin, Dublin, Ireland, Department of Electronic,Electrical Engineering, Trinity College Dublin, Dublin, Ireland), Yike Li (School of Electrical,Elec…

Bare-Metal RISC-V + NVDLA SoC for Efficient Deep Learning Inference
This paper presents a novel System-on-Chip (SoC) architecture for accelerating complex deep learning models for edge computing applications through a combination of hardware and software optimisations. The hardware architecture tightly couples the open-source NVIDIA Deep Learning Accelerator (NVDLA) to a 32-bit, 4-stage pipelined RISC-V core from Codasip called uRISC_V. To offload the model acceleration in software, our toolflow generates bare-metal application code (in assembly), overcoming co…

from my link log —
Damas-Hindley-Milner inference two ways.
https://bernsteinbear.com/blog/type-inference/
saved 2024-10-17 https…

Simulation based inference of the ionization history from the 2D 21 cm power spectrum
Nadia Cooper, Carina Norregaard, Romain Meriot, Jonathan R. Pritchard
https://arxiv.org/abs/2508.16329

Simulation based inference of the ionization history from the 2D 21 cm power spectrum
The 21 cm signal contains a wealth of information about the formation of the first stars and the reionization of the intergalactic medium during the Cosmic Dawn (CD) and Epoch of Reionization (EoR). The timing of these important milestones has only roughly been constrained through indirect measurements, such as from the cosmic microwave background (CMB) optical depth, and Lyman-$α$ forest. Therefore, inferring the neutral fraction over cosmic time is a goal of upcoming 21 cm experiments, such …

Exact Conditional Score-Guided Generative Modeling for Amortized Inference in Uncertainty Quantification
Zezhong Zhang, Caroline Tatsuoka, Dongbin Xiu, Guannan Zhang
https://arxiv.org/abs/2506.18227

Exact Conditional Score-Guided Generative Modeling for Amortized Inference in Uncertainty Quantification
We propose an efficient framework for amortized conditional inference by leveraging exact conditional score-guided diffusion models to train a non-reversible neural network as a conditional generative model. Traditional normalizing flow methods require reversible architectures, which can limit their expressiveness and efficiency. Although diffusion models offer greater flexibility, they often suffer from high computational costs during inference. To combine the strengths of both approaches, we …

A coalgebraic perspective on predictive processing
Manuel Baltieri, Filippo Torresan, Tomoya Nakai
https://arxiv.org/abs/2508.16877 https://arxiv.org/pdf/2…

A coalgebraic perspective on predictive processing
Predictive processing and active inference posit that the brain is a system performing Bayesian inference on the environment. By virtue of this, a prominent interpretation of predictive processing states that the generative model (a POMDP) encoded by the brain synchronises with the generative process (another POMDP) representing the environment while trying to explain what hidden properties of the world generated its sensory input. In this view, the brain is thought to become a copy of the envi…

Flexible Vector Integration in Embedded RISC-V SoCs for End to End CNN Inference Acceleration
Dmitri Lyalikov
https://arxiv.org/abs/2507.17771 https://arxi…

Flexible Vector Integration in Embedded RISC-V SoCs for End to End CNN Inference Acceleration
The emergence of heterogeneity and domain-specific architectures targeting deep learning inference show great potential for enabling the deployment of modern CNNs on resource-constrained embedded platforms. A significant development is the diversification of custom hardware solely targeting the most expensive parts of CNNs. DLAs (deep learning accelerators) and NPUs (neural processing units), among others, can overcome the approaching limits of traditional silicon scaling and provide a solution…

Probabilistic Inference for Datalog with Correlated Inputs
Jingbo Wang, Shashin Halalingaiah, Weiyi Chen, Chao Wang, Isil Dillig
https://arxiv.org/abs/2508.15166 https://…

Probabilistic Inference for Datalog with Correlated Inputs
Probabilistic extensions of logic programming languages, such as ProbLog, integrate logical reasoning with probabilistic inference to evaluate probabilities of output relations; however, prior work does not account for potential statistical correlations among input facts. This paper introduces Praline, a new extension to Datalog designed for precise probabilistic inference in the presence of (partially known) input correlations. We formulate the inference task as a constrained optimization prob…

Temporal Broadening of Attosecond Pulse Trains Induced by Multi-Band inference in Solid-State High-Order Harmonic Generation
Qing-Guo Fan, Kang Lai, Wen-hao Liu, Zhi Wang, Lin-Wang Wang, Jun-Wei Luo
https://arxiv.org/abs/2507.18019

Temporal Broadening of Attosecond Pulse Trains Induced by Multi-Band inference in Solid-State High-Order Harmonic Generation
The mechanism underlying high harmonic generation (HHG) in gases has been well clarified, characterizing attosecond pulse trains (APT) in the time domain, significantly advances the synthesis of isolated attosecond pulse (IAP). However, the complexity of HHG in solid obstacles IAP separation. Here, we use time-dependent density functional theory (TDDFT) to investigate the multiband mechanism of APT in solid state with bulk silicon as prototype. Our research unveils that: 1. The temporal charact…

WiLLM: An Open Wireless LLM Communication System
Boyi Liu, Yongguang Lu, Jianguo Zhao, Qiang Yang, Wen Wu, Lin Chen, Jagmohan Chauhan, Jun Zhang
https://arxiv.org/abs/2506.19030

WiLLM: An Open Wireless LLM Communication System
The rapid evolution of LLMs threatens to overwhelm existing wireless infrastructure, necessitating architectural innovations for burgeoning mobile LLM services. This paper introduces WiLLM, the first open-source wireless system specifically designed for these services. First, we establish a new paradigm by deploying LLMs in core networks (CNs) with abundant GPUs. This enables distributed inference services, strategically positioning LLM inference at the convergence of backbone bandwidth and the…

ARTreeFormer: A Faster Attention-based Autoregressive Model for Phylogenetic Inference
Tianyu Xie, Yicong Mao, Cheng Zhang
https://arxiv.org/abs/2507.18380 https://

ARTreeFormer: A Faster Attention-based Autoregressive Model for Phylogenetic Inference
Probabilistic modeling over the combinatorially large space of tree topologies remains a central challenge in phylogenetic inference. Previous approaches often necessitate pre-sampled tree topologies, limiting their modeling capability to a subset of the entire tree space. A recent advancement is ARTree, a deep autoregressive model that offers unrestricted distributions for tree topologies. However, its reliance on repetitive tree traversals and inefficient local message passing for computing t…

Inference Scaled GraphRAG: Improving Multi Hop Question Answering on Knowledge Graphs
Travis Thompson, Seung-Hwan Lim, Paul Liu, Ruoying He, Dongkuan Xu
https://arxiv.org/abs/2506.19967

Inference Scaled GraphRAG: Improving Multi Hop Question Answering on Knowledge Graphs
Large Language Models (LLMs) have achieved impressive capabilities in language understanding and generation, yet they continue to underperform on knowledge-intensive reasoning tasks due to limited access to structured context and multi-hop information. Retrieval-Augmented Generation (RAG) partially mitigates this by grounding generation in retrieved context, but conventional RAG and GraphRAG methods often fail to capture relational structure across nodes in knowledge graphs. We introduce Infere…

Tomography for Plasma Imaging: a Unifying Framework for Bayesian Inference
D. Hamm, C. Theiler, M. Simeoni, B. P. Duval, T. Debarre, L. Simons, J. R. Queralt
https://arxiv.org/abs/2506.20232

Tomography for Plasma Imaging: a Unifying Framework for Bayesian Inference
Plasma diagnostics often employ computerized tomography to estimate emissivity profiles from a finite, and often limited, number of line-integrated measurements. Decades of algorithmic refinement have brought considerable improvements, and led to a variety of employed solutions. These often feature an underlying, common structure that is rarely acknowledged or investigated. In this paper, we present a unifying perspective on sparse-view tomographic reconstructions for plasma imaging, highlighti…

Partitioned Wild Bootstrap for Panel Data Quantile Regression
Antonio F. Galvao, Carlos Lamarche, Thomas Parker
https://arxiv.org/abs/2507.18494 https://ar…

Partitioned Wild Bootstrap for Panel Data Quantile Regression
Practical inference procedures for quantile regression models of panel data have been a pervasive concern in empirical work, and can be especially challenging when the panel is observed over many time periods and temporal dependence needs to be taken into account. In this paper, we propose a new bootstrap method that applies random weighting to a partition of the data -- partition-invariant weights are used in the bootstrap data generating process -- to conduct statistical inference for conditi…

LoRA-Leak: Membership Inference Attacks Against LoRA Fine-tuned Language Models
Delong Ran, Xinlei He, Tianshuo Cong, Anyu Wang, Qi Li, Xiaoyun Wang
https://arxiv.org/abs/2507.18302

LoRA-Leak: Membership Inference Attacks Against LoRA Fine-tuned Language Models
Language Models (LMs) typically adhere to a "pre-training and fine-tuning" paradigm, where a universal pre-trained model can be fine-tuned to cater to various specialized domains. Low-Rank Adaptation (LoRA) has gained the most widespread use in LM fine-tuning due to its lightweight computational cost and remarkable performance. Because the proportion of parameters tuned by LoRA is relatively small, there might be a misleading impression that the LoRA fine-tuning data is invulnerable to Membersh…

GreenLLM: SLO-Aware Dynamic Frequency Scaling for Energy-Efficient LLM Serving
Qunyou Liu, Darong Huang, Marina Zapater, David Atienza
https://arxiv.org/abs/2508.16449 https://

GreenLLM: SLO-Aware Dynamic Frequency Scaling for Energy-Efficient LLM Serving
Large Language Models (LLMs) are becoming the backbone of modern cloud services, yet their inference costs are dominated by GPU energy. Unlike traditional GPU workloads, LLM inference has two stages with different characteristics: the prefill phase, which is latency sensitive and scales quadratically with prompt length, and the decode phase, which progresses token by token with unpredictable length. Current GPU power governors (for example, NVIDIA's default) overlook this asymmetry and treat bo…

LARP: Learner-Agnostic Robust Data Prefiltering
Kristian Minchev, Dimitar Iliev Dimitrov, Nikola Konstantinov
https://arxiv.org/abs/2506.20573 https://

LARP: Learner-Agnostic Robust Data Prefiltering
The widespread availability of large public datasets is a key factor behind the recent successes of statistical inference and machine learning methods. However, these datasets often contain some low-quality or contaminated data, to which many learning procedures are sensitive. Therefore, the question of whether and how public datasets should be prefiltered to facilitate accurate downstream learning arises. On a technical level this requires the construction of principled data prefiltering metho…

Hybrid quantum-classical algorithm for near-optimal planning in POMDPs
Gilberto Cunha, Alexandra Ram\^oa, Andr\'e Sequeira, Michael de Oliveira, Lu\'is Barbosa
https://arxiv.org/abs/2507.18606 …

Hybrid quantum-classical algorithm for near-optimal planning in POMDPs
Reinforcement learning (RL) provides a principled framework for decision-making in partially observable environments, which can be modeled as Markov decision processes and compactly represented through dynamic decision Bayesian networks. Recent advances demonstrate that inference on sparse Bayesian networks can be accelerated using quantum rejection sampling combined with amplitude amplification, leading to a computational speedup in estimating acceptance probabilities.\\ Building on this resul…

Replaced article(s) found for cs.RO. https://arxiv.org/list/cs.RO/new
[1/2]:
- Stochastic Motion Planning as Gaussian Variational Inference: Theory and Algorithms
Hongzhe Yu, Yongxin Chen

SATORI: Static Test Oracle Generation for REST APIs
Juan C. Alonso, Alberto Martin-Lopez, Sergio Segura, Gabriele Bavota, Antonio Ruiz-Cort\'es
https://arxiv.org/abs/2508.16318

SATORI: Static Test Oracle Generation for REST APIs
REST API test case generation tools are evolving rapidly, with growing capabilities for the automated generation of complex tests. However, despite their strengths in test data generation, these tools are constrained by the types of test oracles they support, often limited to crashes, regressions, and noncompliance with API specifications or design standards. This paper introduces SATORI (Static API Test ORacle Inference), a black-box approach for generating test oracles for REST APIs by analyz…

How weak are weak factors? Uniform inference for signal strength in signal plus noise models
Anna Bykhovskaya, Vadim Gorin, Sasha Sodin
https://arxiv.org/abs/2507.18554 https://…

How weak are weak factors? Uniform inference for signal strength in signal plus noise models
The paper analyzes four classical signal-plus-noise models: the factor model, spiked sample covariance matrices, the sum of a Wigner matrix and a low-rank perturbation, and canonical correlation analysis with low-rank dependencies. The objective is to construct confidence intervals for the signal strength that are uniformly valid across all regimes - strong, weak, and critical signals. We demonstrate that traditional Gaussian approximations fail in the critical regime. Instead, we introduce a u…

A linear PDF model for Bayesian inference
Mark N. Costantini, Luca Mantani, James M. Moore, Maria Ubiali
https://arxiv.org/abs/2507.16913 https://

A linear PDF model for Bayesian inference
A robust uncertainty estimate in global analyses of Parton Distribution Functions (PDFs) is essential at the Large Hadron Collider (LHC), especially in view of the high-precision data anticipated by experimentalists in the High-Luminosity phase of the LHC. A Bayesian framework to determine PDFs provides a rigorous treatment of uncertainties and full control on the prior, though its practical implementation can be computationally demanding. To address these challenges, we introduce a novel appro…

MEDEA: A Design-Time Multi-Objective Manager for Energy-Efficient DNN Inference on Heterogeneous Ultra-Low Power Platforms
Hossein Taji, Jos\'e Miranda, Miguel Pe\'on-Quir\'os, David Atienza
https://arxiv.org/abs/2506.19067

MEDEA: A Design-Time Multi-Objective Manager for Energy-Efficient DNN Inference on Heterogeneous Ultra-Low Power Platforms
The growing demand for on-device AI necessitates energy-efficient execution of DNN based applications on resource-constrained ultra-low power (ULP) platforms. Heterogeneous architectures, combining specialized processing elements (PEs), have emerged as a key solution for achieving the required performance and energy efficiency. However, optimizing energy while executing applications on these platforms requires efficiently managing platform resources like PEs, power features, and memory footprin…

A Probabilistic Inference Scaling Theory for LLM Self-Correction
Zhe Yang, Yichang Zhang, Yudong Wang, Ziyao Xu, Junyang Lin, Zhifang Sui
https://arxiv.org/abs/2508.16456 https:…

A Probabilistic Inference Scaling Theory for LLM Self-Correction
Large Language Models (LLMs) have demonstrated the capability to refine their generated answers through self-correction, enabling continuous performance improvement over multiple rounds. However, the mechanisms underlying how and why accuracy evolves during this iterative process remain unexplored. To fill this gap, we propose a probabilistic theory to model the dynamics of accuracy change and explain the performance improvements observed in multi-round self-correction. Through mathematical der…

Right a few days before I'll be talking about patterns in DRAM init at #GPN23, #Binarly are posting on their type inference tooling:

Type Inference for Decompiled Code: From Hidden Semantics to Structured Insights
Learn how Binarly enhances decompiled code by recovering meaningful type info—boosting binary analysis, triage, and reverse engineering accuracy.

ITO-Master: Inference-Time Optimization for Audio Effects Modeling of Music Mastering Processors
Junghyun Koo, Marco A. Martinez-Ramirez, Wei-Hsiang Liao, Giorgio Fabbro, Michele Mancusi, Yuki Mitsufuji
https://arxiv.org/abs/2506.16889

ITO-Master: Inference-Time Optimization for Audio Effects Modeling of Music Mastering Processors
Music mastering style transfer aims to model and apply the mastering characteristics of a reference track to a target track, simulating the professional mastering process. However, existing methods apply fixed processing based on a reference track, limiting users' ability to fine-tune the results to match their artistic intent. In this paper, we introduce the ITO-Master framework, a reference-based mastering style transfer system that integrates Inference-Time Optimization (ITO) to enable finer…

CIGaRS I: Combined simulation-based inference from SNae Ia and host photometry
Konstantin Karchev, Roberto Trotta, Raul Jimenez
https://arxiv.org/abs/2508.15899 https://

CIGaRS I: Combined simulation-based inference from SNae Ia and host photometry
Using type Ia supernovae (SNae Ia) as cosmological probes requires empirical corrections, which correlate with their host environment. We present a unified Bayesian hierarchical model designed to infer, from purely photometric observations, the intrinsic dependence of SN Ia brightness on progenitor properties (metallicity & age), the delay-time distribution (DTD) that governs their rate as a function of age, and cosmology, as well as the redshifts of all hosts. The model incorporates physics-ba…

UniSegDiff: Boosting Unified Lesion Segmentation via a Staged Diffusion Model
Yilong Hu, Shijie Chang, Lihe Zhang, Feng Tian, Weibing Sun, Huchuan Lu
https://arxiv.org/abs/2507.18362

UniSegDiff: Boosting Unified Lesion Segmentation via a Staged Diffusion Model
The Diffusion Probabilistic Model (DPM) has demonstrated remarkable performance across a variety of generative tasks. The inherent randomness in diffusion models helps address issues such as blurring at the edges of medical images and labels, positioning Diffusion Probabilistic Models (DPMs) as a promising approach for lesion segmentation. However, we find that the current training and inference strategies of diffusion models result in an uneven distribution of attention across different timest…

Waging a Campaign: Results from an Injection-Recovery Study involving 35 numerical Relativity Simulations and three Waveform Models
Sarp Ak\c{c}ay, Charlie Hoy, Jake Mac Uilliam
https://arxiv.org/abs/2506.19990

Waging a Campaign: Results from an Injection-Recovery Study involving 35 numerical Relativity Simulations and three Waveform Models
We present Bayesian inference results from an extensive injection-recovery campaign to test the validity of three state of the art quasicircular gravitational waveform models: \textsc{SEOBNRv5PHM}, \textsc{IMRPhenomTPHM}, \textsc{IMRPhenomXPHM}, the latter with the \textsc{SpinTaylorT4} implementation for its precession dynamics. We analyze 35 strongly precessing binary black hole numerical relativity simulations with all available harmonic content. Ten simulations have a mass ratio of $4:1$ an…

MNN-AECS: Energy Optimization for LLM Decoding on Mobile Devices via Adaptive Core Selection
Zhengxiang Huang, Chaoyue Niu, Zhaode Wang, Jiarui Xue, Hanming Zhang, Yugang Wang, Zewei Xin, Xiaotang Jiang, Chengfei Lv, Fan Wu, Guihai Chen
https://arxiv.org/abs/2506.19884

MNN-AECS: Energy Optimization for LLM Decoding on Mobile Devices via Adaptive Core Selection
As the demand for on-device Large Language Model (LLM) inference grows, energy efficiency has become a major concern, especially for battery-limited mobile devices. Our analysis shows that the memory-bound LLM decode phase dominates energy use, and yet most existing works focus on accelerating the prefill phase, neglecting energy concerns. We introduce Adaptive Energy-Centric Core Selection (AECS) and integrate it into MNN to create the energy-efficient version, MNN-AECS, the first engine-level…

Statistical Inference for Optimal Transport Maps: Recent Advances and Perspectives
Sivaraman Balakrishnan, Tudor Manole, Larry Wasserman
https://arxiv.org/abs/2506.19025

Statistical Inference for Optimal Transport Maps: Recent Advances and Perspectives
In many applications of optimal transport (OT), the object of primary interest is the optimal transport map. This map rearranges mass from one probability distribution to another in the most efficient way possible by minimizing a specified cost. In this paper we review recent advances in estimating and developing limit theorems for the OT map, using samples from the underlying distributions. We also review parallel lines of work that establish similar results for special cases and variants of t…

Decentralized Consensus Inference-based Hierarchical Reinforcement Learning for Multi-Constrained UAV Pursuit-Evasion Game
Xiang Yuming, Li Sizhao, Li Rongpeng, Zhao Zhifeng, Zhang Honggang
https://arxiv.org/abs/2506.18126

Decentralized Consensus Inference-based Hierarchical Reinforcement Learning for Multi-Constrained UAV Pursuit-Evasion Game
Multiple quadrotor unmanned aerial vehicle (UAV) systems have garnered widespread research interest and fostered tremendous interesting applications, especially in multi-constrained pursuit-evasion games (MC-PEG). The Cooperative Evasion and Formation Coverage (CEFC) task, where the UAV swarm aims to maximize formation coverage across multiple target zones while collaboratively evading predators, belongs to one of the most challenging issues in MC-PEG, especially under communication-limited con…

Inductive Domain Transfer In Misspecified Simulation-Based Inference
Ortal Senouf, Antoine Wehenkel, C\'edric Vincent-Cuaz, Emmanuel Abb\'e, Pascal Frossard
https://arxiv.org/abs/2508.15593

Inductive Domain Transfer In Misspecified Simulation-Based Inference
Simulation-based inference (SBI) is a statistical inference approach for estimating latent parameters of a physical system when the likelihood is intractable but simulations are available. In practice, SBI is often hindered by model misspecification--the mismatch between simulated and real-world observations caused by inherent modeling simplifications. RoPE, a recent SBI approach, addresses this challenge through a two-stage domain transfer process that combines semi-supervised calibration with…

Retrieval-Confused Generation is a Good Defender for Privacy Violation Attack of Large Language Models
Wanli Peng, Xin Chen, Hang Fu, XinYu He, Xue Yiming, Juan Wen
https://arxiv.org/abs/2506.19889

Retrieval-Confused Generation is a Good Defender for Privacy Violation Attack of Large Language Models
Recent advances in large language models (LLMs) have made a profound impact on our society and also raised new security concerns. Particularly, due to the remarkable inference ability of LLMs, the privacy violation attack (PVA), revealed by Staab et al., introduces serious personal privacy issues. Existing defense methods mainly leverage LLMs to anonymize the input query, which requires costly inference time and cannot gain satisfactory defense performance. Moreover, directly rejecting the PVA …

Scaling Group Inference for Diverse and High-Quality Generation
Gaurav Parmar, Or Patashnik, Daniil Ostashev, Kuan-Chieh Wang, Kfir Aberman, Srinivasa Narasimhan, Jun-Yan Zhu
https://arxiv.org/abs/2508.15773

Scaling Group Inference for Diverse and High-Quality Generation
Generative models typically sample outputs independently, and recent inference-time guidance and scaling algorithms focus on improving the quality of individual samples. However, in real-world applications, users are often presented with a set of multiple images (e.g., 4-8) for each prompt, where independent sampling tends to lead to redundant results, limiting user choices and hindering idea exploration. In this work, we introduce a scalable group inference method that improves both the divers…

SuperSONIC: Cloud-Native Infrastructure for ML Inferencing
Dmitry Kondratyev, Benedikt Riedel, Yuan-Tang Chou, Miles Cochran-Branson, Noah Paladino, David Schultz, Mia Liu, Javier Duarte, Philip Harris, Shih-Chieh Hsu
https://arxiv.org/abs/2506.20657

SuperSONIC: Cloud-Native Infrastructure for ML Inferencing
The increasing computational demand from growing data rates and complex machine learning (ML) algorithms in large-scale scientific experiments has driven the adoption of the Services for Optimized Network Inference on Coprocessors (SONIC) approach. SONIC accelerates ML inference by offloading it to local or remote coprocessors to optimize resource utilization. Leveraging its portability to different types of coprocessors, SONIC enhances data processing and model deployment efficiency for cuttin…

TMA-Adaptive FP8 Grouped GEMM: Eliminating Padding Requirements in Low-Precision Training and Inference on Hopper
Zhongling Su, Rong Fu, Weihan Cao, Jianfei Gao, Minxi Jin, Zhilin Pei, Hui Wang
https://arxiv.org/abs/2508.16584

TMA-Adaptive FP8 Grouped GEMM: Eliminating Padding Requirements in Low-Precision Training and Inference on Hopper
Current FP8 grouped GEMM implementations require padding each group to a fixed alignment (e.g., 128), incurring memory and computational overhead. We propose \textit{TMA-Adaptive FP8 Grouped GEMM}, which eliminates padding by dynamically adapting to variable group dimensions via (1) a TMA descriptor pool with $\log_2(block_M)$ preconfigured descriptors to handle all residual row cases through dynamic runtime selection and dual-phase load-store operations, achieving comprehensive coverage with m…

gcor: A Python Implementation of Categorical Gini Correlation and Its Inference
Sameera Hewage
https://arxiv.org/abs/2506.19230 https://

gcor: A Python Implementation of Categorical Gini Correlation and Its Inference
Categorical Gini Correlation (CGC), introduced by Dang et al. (2020), is a novel dependence measure designed to quantify the association between a numerical variable and a categorical variable. It has appealing properties compared to existing dependence measures, such as zero correlation mutually implying independence between the variables. It has also shown superior performance over existing methods when applied to feature screening for classification. This article presents a Python implementa…

A Two-armed Bandit Framework for A/B Testing
Jinjuan Wang, Qianglin Wen, Yu Zhang, Xiaodong Yan, Chengchun Shi
https://arxiv.org/abs/2507.18118 https://arx…

A Two-armed Bandit Framework for A/B Testing
A/B testing is widely used in modern technology companies for policy evaluation and product deployment, with the goal of comparing the outcomes under a newly-developed policy against a standard control. Various causal inference and reinforcement learning methods developed in the literature are applicable to A/B testing. This paper introduces a two-armed bandit framework designed to improve the power of existing approaches. The proposed procedure consists of three main steps: (i) employing doubl…

GPL-SLAM: A Laser SLAM Framework with Gaussian Process Based Extended Landmarks
Ali Emre Balc{\i} (TU Delft), Erhan Ege Keyvan (Middle East Technical University), Emre \"Ozkan (Middle East Technical University)
https://arxiv.org/abs/2508.16459

GPL-SLAM: A Laser SLAM Framework with Gaussian Process Based Extended Landmarks
We present a novel Simultaneous Localization and Mapping (SLAM) method that employs Gaussian Process (GP) based landmark (object) representations. Instead of conventional grid maps or point cloud registration, we model the environment on a per object basis using GP based contour representations. These contours are updated online through a recursive scheme, enabling efficient memory usage. The SLAM problem is formulated within a fully Bayesian framework, allowing joint inference over the robot p…

Adaptively Robust LLM Inference Optimization under Prediction Uncertainty
Zixi Chen, Yinyu Ye, Zijie Zhou
https://arxiv.org/abs/2508.14544 https://arxiv.or…

Adaptively Robust LLM Inference Optimization under Prediction Uncertainty
We study the problem of optimizing Large Language Model (LLM) inference scheduling to minimize total latency. LLM inference is an online and multi-task service process and also heavily energy consuming by which a pre-trained LLM processes input requests and generates output tokens sequentially. Therefore, it is vital to improve its scheduling efficiency and reduce the power consumption while a great amount of prompt requests are arriving. A key challenge in LLM inference scheduling is that whil…

Efficient Bayesian Inference for Spatial Point Patterns Using the Palm Likelihood
Kevin M. Collins, Erin M. Schliep
https://arxiv.org/abs/2507.17065 https://

Efficient Bayesian Inference for Spatial Point Patterns Using the Palm Likelihood
Bayesian inference for spatial point patterns is often hindered computationally by intractable likelihoods. In the frequentist literature, estimating equations utilizing pseudolikelihoods have long been used for simulation-free parameter estimation. One such pseudolikelihood based on the process of differences is known as the Palm likelihood. Utilizing notions of Bayesian composite likelihoods and generalized Bayesian inference, we develop a framework for the use of Palm likelihoods in a Bayesi…

Transport Properties of QGP within a Bayesian Holographic QCD Model
Bing Chen, Liqiang Zhu, Xun Chen, Defu Hou, Xurong Chen
https://arxiv.org/abs/2508.16167 https://

Transport Properties of QGP within a Bayesian Holographic QCD Model
Using a holographic QCD model augmented by Bayesian inference, we calculate key transport coefficients of the quark-gluon plasma (QGP)$\text{-}$including the drag force, jet quenching parameter, heavy quark diffusion coefficient, and shear and bulk viscosities$\text{-}$at finite temperature and chemical potential. Posterior parameter distributions at the 68\% and 95\% confidence levels (CL), as well as the maximum a posteriori (MAP) estimates, are employed to quantify uncertainties. Our finding…

Replaced article(s) found for cs.IT. https://arxiv.org/list/cs.IT/new
[1/1]:
- Computation and Communication Co-scheduling for Multi-Task Remote Inference
Md Kamran Chowdhury Shisher, Adam Piaseczny, Yin Sun, Christopher G. Brinton

Tab-MIA: A Benchmark Dataset for Membership Inference Attacks on Tabular Data in LLMs
Eyal German, Sagiv Antebi, Daniel Samira, Asaf Shabtai, Yuval Elovici
https://arxiv.org/abs/2507.17259

Tab-MIA: A Benchmark Dataset for Membership Inference Attacks on Tabular Data in LLMs
Large language models (LLMs) are increasingly trained on tabular data, which, unlike unstructured text, often contains personally identifiable information (PII) in a highly structured and explicit format. As a result, privacy risks arise, since sensitive records can be inadvertently retained by the model and exposed through data extraction or membership inference attacks (MIAs). While existing MIA methods primarily target textual content, their efficacy and threat implications may differ when a…

Hardwired-Neurons Language Processing Units as General-Purpose Cognitive Substrates
Yang Liu, Yi Chen, Yongwei Zhao, Yifan Hao, Zifu Zheng, Weihao Kong, Zhangmai Li, Dongchen Jiang, Ruiyang Xia, Zhihong Ma, Zisheng Liu, Zhaoyong Wan, Yunqi Lu, Ximing Liu, Hongrui Guo, Zhihao Yang, Zhe Wang, Tianrui Ma, Mo Zou, Rui Zhang, Ling Li, Xing Hu, Zidong Du, Zhiwei Xu, Qi Guo, Tianshi Chen, Yunji Chen

Hardwired-Neurons Language Processing Units as General-Purpose Cognitive Substrates
The rapid advancement of Large Language Models (LLMs) has established language as a core general-purpose cognitive substrate, driving the demand for specialized Language Processing Units (LPUs) tailored for LLM inference. To overcome the growing energy consumption of LLM inference systems, this paper proposes a Hardwired-Neurons Language Processing Unit (HNLPU), which physically hardwires LLM weight parameters into the computational fabric, achieving several orders of magnitude computational ef…

Forecast then Calibrate: Feature Caching as ODE for Efficient Diffusion Transformers
Shikang Zheng, Liang Feng, Xinyu Wang, Qinming Zhou, Peiliang Cai, Chang Zou, Jiacheng Liu, Yuqi Lin, Junjie Chen, Yue Ma, Linfeng Zhang
https://arxiv.org/abs/2508.16211

Forecast then Calibrate: Feature Caching as ODE for Efficient Diffusion Transformers
Diffusion Transformers (DiTs) have demonstrated exceptional performance in high-fidelity image and video generation. To reduce their substantial computational costs, feature caching techniques have been proposed to accelerate inference by reusing hidden representations from previous timesteps. However, current methods often struggle to maintain generation quality at high acceleration ratios, where prediction errors increase sharply due to the inherent instability of long-step forecasting. In th…

Bridging the Gap in Ophthalmic AI: MM-Retinal-Reason Dataset and OphthaReason Model toward Dynamic Multimodal Reasoning
Ruiqi Wu, Yuang Yao, Tengfei Ma, Chenran Zhang, Na Su, Tao Zhou, Geng Chen, Wen Fan, Yi Zhou
https://arxiv.org/abs/2508.16129

Bridging the Gap in Ophthalmic AI: MM-Retinal-Reason Dataset and OphthaReason Model toward Dynamic Multimodal Reasoning
Multimodal large language models (MLLMs) have recently demonstrated remarkable reasoning abilities with reinforcement learning paradigm. Although several multimodal reasoning models have been explored in the medical domain, most of them focus exclusively on basic reasoning, which refers to shallow inference based on visual feature matching. However, real-world clinical diagnosis extends beyond basic reasoning, demanding reasoning processes that integrate heterogeneous clinical information (such…

BrownoutServe: SLO-Aware Inference Serving under Bursty Workloads for MoE-based LLMs
Jianmin Hu, Minxian Xu, Kejiang Ye, Chengzhong Xu
https://arxiv.org/abs/2507.17133 https://

BrownoutServe: SLO-Aware Inference Serving under Bursty Workloads for MoE-based LLMs
In recent years, the Mixture-of-Experts (MoE) architecture has been widely applied to large language models (LLMs), providing a promising solution that activates only a subset of the model's parameters during computation, thereby reducing overall memory requirements and allowing for faster inference compared to dense models. Despite these advantages, existing systems still face issues of low efficiency due to static model placement and lack of dynamic workloads adaptation. This leads to subopti…

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…

Radial Attention: $O(n\log n)$ Sparse Attention with Energy Decay for Long Video Generation
Xingyang Li, Muyang Li, Tianle Cai, Haocheng Xi, Shuo Yang, Yujun Lin, Lvmin Zhang, Songlin Yang, Jinbo Hu, Kelly Peng, Maneesh Agrawala, Ion Stoica, Kurt Keutzer, Song Han
https://arxiv.org/abs/2506.19852…

Radial Attention: $O(n\log n)$ Sparse Attention with Energy Decay for Long Video Generation
Recent advances in diffusion models have enabled high-quality video generation, but the additional temporal dimension significantly increases computational costs, making training and inference on long videos prohibitively expensive. In this paper, we identify a phenomenon we term Spatiotemporal Energy Decay in video diffusion models: post-softmax attention scores diminish as spatial and temporal distance between tokens increase, akin to the physical decay of signal or waves over space and time …

BucketServe: Bucket-Based Dynamic Batching for Smart and Efficient LLM Inference Serving
Wanyi Zheng, Minxian Xu, Shengye Song, Kejiang Ye
https://arxiv.org/abs/2507.17120 https…

BucketServe: Bucket-Based Dynamic Batching for Smart and Efficient LLM Inference Serving
Large language models (LLMs) have become increasingly popular in various areas, traditional business gradually shifting from rule-based systems to LLM-based solutions. However, the inference of LLMs is resource-intensive or latency-sensitive, posing significant challenges for serving systems. Existing LLM serving systems often use static or continuous batching strategies, which can lead to inefficient GPU memory utilization and increased latency, especially under heterogeneous workloads. These …

Bridging the Gap in Ophthalmic AI: MM-Retinal-Reason Dataset and OphthaReason Model toward Dynamic Multimodal Reasoning
Ruiqi Wu, Yuang Yao, Tengfei Ma, Chenran Zhang, Na Su, Tao Zhou, Geng Chen, Wen Fan, Yi Zhou
https://arxiv.org/abs/2508.16129

Bridging the Gap in Ophthalmic AI: MM-Retinal-Reason Dataset and OphthaReason Model toward Dynamic Multimodal Reasoning
Multimodal large language models (MLLMs) have recently demonstrated remarkable reasoning abilities with reinforcement learning paradigm. Although several multimodal reasoning models have been explored in the medical domain, most of them focus exclusively on basic reasoning, which refers to shallow inference based on visual feature matching. However, real-world clinical diagnosis extends beyond basic reasoning, demanding reasoning processes that integrate heterogeneous clinical information (such…

A Modular Multitask Reasoning Framework Integrating Spatio-temporal Models and LLMs
Kethmi Hirushini Hettige, Jiahao Ji, Cheng Long, Shili Xiang, Gao Cong, Jingyuan Wang
https://arxiv.org/abs/2506.20073

A Modular Multitask Reasoning Framework Integrating Spatio-temporal Models and LLMs
Spatio-temporal data mining plays a pivotal role in informed decision making across diverse domains. However, existing models are often restricted to narrow tasks, lacking the capacity for multi-task inference and complex long-form reasoning that require generation of in-depth, explanatory outputs. These limitations restrict their applicability to real-world, multi-faceted decision scenarios. In this work, we introduce STReason, a novel framework that integrates the reasoning strengths of large…

Leveraging specificity for causal inference in observational studies
Wang Miao
https://arxiv.org/abs/2506.18469 https://arxiv.org/pdf…

Leveraging specificity for causal inference in observational studies
Hill's specificity criterion has been highly influential in biomedical and epidemiological research. However, it remains controversial and its application often relies on subjective and qualitative analysis without a comprehensive and rigorous causal theory. Focusing on unmeasured confounding adjustment with multiple treatments and multiple outcomes, this paper develops a formal and quantitative framework for leveraging specificity for causal inference in observational studies. The proposed fra…

From Tiny Machine Learning to Tiny Deep Learning: A Survey
Shriyank Somvanshi, Md Monzurul Islam, Gaurab Chhetri, Rohit Chakraborty, Mahmuda Sultana Mimi, Swagat Ahmed Shuvo, Kazi Sifatul Islam, Syed Aaqib Javed, Sharif Ahmed Rafat, Anandi Dutta, Subasish Das
https://arxiv.org/abs/2506.18927…

From Tiny Machine Learning to Tiny Deep Learning: A Survey
The rapid growth of edge devices has driven the demand for deploying artificial intelligence (AI) at the edge, giving rise to Tiny Machine Learning (TinyML) and its evolving counterpart, Tiny Deep Learning (TinyDL). While TinyML initially focused on enabling simple inference tasks on microcontrollers, the emergence of TinyDL marks a paradigm shift toward deploying deep learning models on severely resource-constrained hardware. This survey presents a comprehensive overview of the transition from…

A Lightweight Group Multiscale Bidirectional Interactive Network for Real-Time Steel Surface Defect Detection
Yong Zhang, Cunjian Chen, Qiang Gao, Yi Wang, Bin Fang
https://arxiv.org/abs/2508.16397

A Lightweight Group Multiscale Bidirectional Interactive Network for Real-Time Steel Surface Defect Detection
Real-time surface defect detection is critical for maintaining product quality and production efficiency in the steel manufacturing industry. Despite promising accuracy, existing deep learning methods often suffer from high computational complexity and slow inference speeds, which limit their deployment in resource-constrained industrial environments. Recent lightweight approaches adopt multibranch architectures based on depthwise separable convolution (DSConv) to capture multiscale contextual …

GPTailor: Large Language Model Pruning Through Layer Cutting and Stitching
Guinan Su, Li Shen, Lu Yin, Shiwei Liu, Yanwu Yang, Jonas Geiping
https://arxiv.org/abs/2506.20480

GPTailor: Large Language Model Pruning Through Layer Cutting and Stitching
Large language models (LLMs) have shown remarkable capabilities in language understanding and generation. However, such impressive capability typically comes with a substantial model size, which presents significant challenges in deployment and inference. While structured pruning of model parameters offers a promising way to reduce computational costs at deployment time, current methods primarily focus on single model pruning. In this work, we develop a novel strategy to compress models by stra…

Collaborative Inference and Learning between Edge SLMs and Cloud LLMs: A Survey of Algorithms, Execution, and Open Challenges
Senyao Li, Haozhao Wang, Wenchao Xu, Rui Zhang, Song Guo, Jingling Yuan, Xian Zhong, Tianwei Zhang, Ruixuan Li
https://arxiv.org/abs/2507.16731

Collaborative Inference and Learning between Edge SLMs and Cloud LLMs: A Survey of Algorithms, Execution, and Open Challenges
As large language models (LLMs) evolve, deploying them solely in the cloud or compressing them for edge devices has become inadequate due to concerns about latency, privacy, cost, and personalization. This survey explores a collaborative paradigm in which cloud-based LLMs and edge-deployed small language models (SLMs) cooperate across both inference and training. We present a unified taxonomy of edge-cloud collaboration strategies. For inference, we categorize approaches into task assignment, t…

PrivacyXray: Detecting Privacy Breaches in LLMs through Semantic Consistency and Probability Certainty
Jinwen He, Yiyang Lu, Zijin Lin, Kai Chen, Yue Zhao
https://arxiv.org/abs/2506.19563

PrivacyXray: Detecting Privacy Breaches in LLMs through Semantic Consistency and Probability Certainty
Large Language Models (LLMs) are widely used in sensitive domains, including healthcare, finance, and legal services, raising concerns about potential private information leaks during inference. Privacy extraction attacks, such as jailbreaking, expose vulnerabilities in LLMs by crafting inputs that force the models to output sensitive information. However, these attacks cannot verify whether the extracted private information is accurate, as no public datasets exist for cross-validation, leaving…

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…

On the Evolution of Federated Post-Training Large Language Models: A Model Accessibility View
Tao Guo, Junxiao Wang, Fushuo Huo, Laizhong Cui, Song Guo, Jie Gui, Dacheng Tao
https://arxiv.org/abs/2508.16261

On the Evolution of Federated Post-Training Large Language Models: A Model Accessibility View
Federated Learning (FL) enables training models across decentralized data silos while preserving client data privacy. Recent research has explored efficient methods for post-training large language models (LLMs) within FL to address computational and communication challenges. While existing approaches often rely on access to LLMs' internal information, which is frequently restricted in real-world scenarios, an inference-only paradigm (black-box FedLLM) has emerged to address these limitations. …

GIIFT: Graph-guided Inductive Image-free Multimodal Machine Translation
Jiafeng Xiong, Yuting Zhao
https://arxiv.org/abs/2507.18562 https://arxiv.org/pdf/2…

GIIFT: Graph-guided Inductive Image-free Multimodal Machine Translation
Multimodal Machine Translation (MMT) has demonstrated the significant help of visual information in machine translation. However, existing MMT methods face challenges in leveraging the modality gap by enforcing rigid visual-linguistic alignment whilst being confined to inference within their trained multimodal domains. In this work, we construct novel multimodal scene graphs to preserve and integrate modality-specific information and introduce GIIFT, a two-stage Graph-guided Inductive Image-Fre…

Embedded FPGA Acceleration of Brain-Like Neural Networks: Online Learning to Scalable Inference
Muhammad Ihsan Al Hafiz, Naresh Ravichandran, Anders Lansner, Pawel Herman, Artur Podobas
https://arxiv.org/abs/2506.18530

Embedded FPGA Acceleration of Brain-Like Neural Networks: Online Learning to Scalable Inference
Edge AI applications increasingly require models that can learn and adapt on-device with minimal energy budget. Traditional deep learning models, while powerful, are often overparameterized, energy-hungry, and dependent on cloud connectivity. Brain-Like Neural Networks (BLNNs), such as the Bayesian Confidence Propagation Neural Network (BCPNN), propose a neuromorphic alternative by mimicking cortical architecture and biologically-constrained learning. They offer sparse architectures with local …

HE-LRM: Encrypted Deep Learning Recommendation Models using Fully Homomorphic Encryption
Karthik Garimella, Austin Ebel, Gabrielle De Micheli, Brandon Reagen
https://arxiv.org/abs/2506.18150

HE-LRM: Encrypted Deep Learning Recommendation Models using Fully Homomorphic Encryption
Fully Homomorphic Encryption (FHE) is an encryption scheme that not only encrypts data but also allows for computations to be applied directly on the encrypted data. While computationally expensive, FHE can enable privacy-preserving neural inference in the client-server setting: a client encrypts their input with FHE and sends it to an untrusted server. The server then runs neural inference on the encrypted data and returns the encrypted results. The client decrypts the output locally, keeping …

Nonparametric inference for nonstationary spatial point processes
Izabel Nolau, Fl\'avio B. Gon\c{c}alves, Dani Gamerman
https://arxiv.org/abs/2507.17600 https://

Nonparametric inference for nonstationary spatial point processes
Point pattern data often exhibit features such as abrupt changes, hotspots and spatially varying dependence in local intensity. Under a Poisson process framework, these correspond to discontinuities and nonstationarity in the underlying intensity function -- features that are difficult to capture with standard modeling approaches. This paper proposes a spatial Cox process model in which nonstationarity is induced through a random partition of the spatial domain, with conditionally independent G…

DiffuCoder: Understanding and Improving Masked Diffusion Models for Code Generation
Shansan Gong, Ruixiang Zhang, Huangjie Zheng, Jiatao Gu, Navdeep Jaitly, Lingpeng Kong, Yizhe Zhang
https://arxiv.org/abs/2506.20639

DiffuCoder: Understanding and Improving Masked Diffusion Models for Code Generation
Diffusion large language models (dLLMs) are compelling alternatives to autoregressive (AR) models because their denoising models operate over the entire sequence. The global planning and iterative refinement features of dLLMs are particularly useful for code generation. However, current training and inference mechanisms for dLLMs in coding are still under-explored. To demystify the decoding behavior of dLLMs and unlock their potential for coding, we systematically investigate their denoising pr…

Cloud Native System for LLM Inference Serving
Minxian Xu, Junhan Liao, Jingfeng Wu, Yiyuan He, Kejiang Ye, Chengzhong Xu
https://arxiv.org/abs/2507.18007 https://

Efficient LLM Inference: Bandwidth, Compute, Synchronization, and Capacity are all you need
Michael Davies, Neal Crago, Karthikeyan Sankaralingam, Christos Kozyrakis
https://arxiv.org/abs/2507.14397

Efficient LLM Inference: Bandwidth, Compute, Synchronization, and Capacity are all you need
This paper presents a limit study of transformer-based large language model (LLM) inference, focusing on the fundamental performance bottlenecks imposed by memory bandwidth, memory capacity, and synchronization overhead in distributed inference systems. We develop a hardware-agnostic performance model that abstracts away implementation details, enabling the analysis of a wide range of current and near-future hardware technologies. Our analysis spans from current HBM3 memory technology used in A…

Bayesian Inference for Left-Truncated Log-Logistic Distributions for Time-to-event Data Analysis
Fahad Mostafa, Md Rejuan Haque, Md Mostafijur Rahman, Farzana Nasrin
https://arxiv.org/abs/2506.17852

Bayesian Inference for Left-Truncated Log-Logistic Distributions for Time-to-event Data Analysis
Parameter estimation is a foundational step in statistical modeling, enabling us to extract knowledge from data and apply it effectively. Bayesian estimation of parameters incorporates prior beliefs with observed data to infer distribution parameters probabilistically and robustly. Moreover, it provides full posterior distributions, allowing uncertainty quantification and regularization, especially useful in small or truncated samples. Utilizing the left-truncated log-logistic (LTLL) distributi…

System Report for CCL25-Eval Task 10: SRAG-MAV for Fine-Grained Chinese Hate Speech Recognition
Jiahao Wang, Ramen Liu, Longhui Zhang, Jing Li
https://arxiv.org/abs/2507.18580 h…

System Report for CCL25-Eval Task 10: SRAG-MAV for Fine-Grained Chinese Hate Speech Recognition
This paper presents our system for CCL25-Eval Task 10, addressing Fine-Grained Chinese Hate Speech Recognition (FGCHSR). We propose a novel SRAG-MAV framework that synergistically integrates task reformulation(TR), Self-Retrieval-Augmented Generation (SRAG), and Multi-Round Accumulative Voting (MAV). Our method reformulates the quadruplet extraction task into triplet extraction, uses dynamic retrieval from the training set to create contextual prompts, and applies multi-round inference with vot…

Dual-Stage Value-Guided Inference with Margin-Based Reward Adjustment for Fast and Faithful VLM Captioning
Ankan Deria, Adinath Madhavrao Dukre, Feilong Tang, Sara Atito, Sudipta Roy, Muhammad Awais, Muhammad Haris Khan, Imran Razzak
https://arxiv.org/abs/2506.15649

Dual-Stage Value-Guided Inference with Margin-Based Reward Adjustment for Fast and Faithful VLM Captioning
Despite significant advances in inference-time search for vision-language models (VLMs), existing approaches remain both computationally expensive and prone to unpenalized, low-confidence generations which often lead to persistent hallucinations. We introduce \textbf{Value-guided Inference with Margin-based Reward (ViMaR)}, a two-stage inference framework that improves both efficiency and output fidelity by combining a temporal-difference value model with a margin-aware reward adjustment. In th…

Flexible Vector Integration in Embedded RISC-V SoCs for End to End CNN Inference Acceleration
Dmitri Lyalikov
https://arxiv.org/abs/2507.17771 https://arxi…

Doubly robust outlier resistant inference on causal treatment effect
Joonsung Kang
https://arxiv.org/abs/2507.17439 https://arxiv.org/pdf/2507.17439…

Doubly robust outlier resistant inference on causal treatment effect
Outliers can severely distort causal effect estimation in observational studies, yet this issue has received limited attention in the literature. Their influence is especially pronounced in small sample sizes, where detecting and removing outliers becomes increasingly difficult. Therefore, it is essential to estimate treatment effects robustly without excluding these influential data points. To address this, we propose a doubly robust point estimator for the average treatment effect under a con…

Sandwich: Separating Prefill-Decode Compilation for Efficient CPU LLM Serving
Juntao Zhao, Jiuru Li, Chuan Wu
https://arxiv.org/abs/2507.18454 https://arxi…

Sandwich: Separating Prefill-Decode Compilation for Efficient CPU LLM Serving
Utilizing CPUs to serve large language models (LLMs) is a resource-friendly alternative to GPU serving. Existing CPU-based solutions ignore workload differences between the prefill and the decode phases of LLM inference, applying a static per-NUMA (Non-Uniform Memory Access) node model partition and utilizing vendor libraries for operator-level execution, which is suboptimal. We propose Sandwich, a hardware-centric CPU-based LLM serving engine that uses different execution plans for the prefill…

Efficient Mixed-Precision Large Language Model Inference with TurboMind
Li Zhang, Youhe Jiang, Guoliang He, Xin Chen, Han Lv, Qian Yao, Fangcheng Fu, Kai Chen
https://arxiv.org/abs/2508.15601

Efficient Mixed-Precision Large Language Model Inference with TurboMind
Mixed-precision inference techniques reduce the memory and computational demands of Large Language Models (LLMs) by applying hybrid precision formats to model weights, activations, and KV caches. This work introduces mixed-precision LLM inference techniques that encompass (i) systematic memory and compute optimization across hierarchical storage and tensor core architectures, and (ii) comprehensive end-to-end mixed-precision optimization across diverse precision formats and hardware configurati…

Comparing energy consumption and accuracy in text classification inference
Johannes Zschache, Tilman Hartwig
https://arxiv.org/abs/2508.14170 https://arxiv…

Comparing energy consumption and accuracy in text classification inference
The increasing deployment of large language models (LLMs) in natural language processing (NLP) tasks raises concerns about energy efficiency and sustainability. While prior research has largely focused on energy consumption during model training, the inference phase has received comparatively less attention. This study systematically evaluates the trade-offs between model accuracy and energy consumption in text classification inference across various model architectures and hardware configurati…

Predictive inference for discrete-valued time series
Maxime Faymonville, Carsten Jentsch, Efstathios Paparoditis
https://arxiv.org/abs/2507.16035 https://

Predictive inference for discrete-valued time series
For discrete-valued time series, predictive inference cannot be implemented through the construction of prediction intervals to some predetermined coverage level, as this is the case for real-valued time series. To address this problem, we propose to reverse the construction principle by considering preselected sets of interest and estimating the probability that a future observation of the process falls into these sets. The accuracy of the prediction is then evaluated by quantifying the uncert…

Efficient Routing of Inference Requests across LLM Instances in Cloud-Edge Computing
Shibo Yu, Mohammad Goudarzi, Adel Nadjaran Toosi
https://arxiv.org/abs/2507.15553

Efficient Routing of Inference Requests across LLM Instances in Cloud-Edge Computing
The rising demand for Large Language Model (LLM) inference services has intensified pressure on computational resources, resulting in latency and cost challenges. This paper introduces a novel routing algorithm based on the Non-dominated Sorting Genetic Algorithm II (NSGA-II) to distribute inference requests across heterogeneous LLM instances in a cloud-edge computing environment. Formulated as a multi-objective optimization problem, the algorithm balances response quality, response time, and i…

Inference on Nonlinear Counterfactual Functionals under a Multiplicative IV Model
Yonghoon Lee, Mengxin Yu, Jiewen Liu, Chan Park, Yunshu Zhang, James M. Robins, Eric J. Tchetgen Tchetgen
https://arxiv.org/abs/2507.15612

Inference on Nonlinear Counterfactual Functionals under a Multiplicative IV Model
Instrumental variable (IV) methods play a central role in causal inference, particularly in settings where treatment assignment is confounded by unobserved variables. IV methods have been extensively developed in recent years and applied across diverse domains, from economics to epidemiology. In this work, we study the recently introduced multiplicative IV (MIV) model and demonstrate its utility for causal inference beyond the average treatment effect. In particular, we show that it enables ide…

FCPO: Federated Continual Policy Optimization for Real-Time High-Throughput Edge Video Analytics
Lucas Liebe, Thanh-Tung Nguyen, Dongman Lee
https://arxiv.org/abs/2507.18047 htt…

FCPO: Federated Continual Policy Optimization for Real-Time High-Throughput Edge Video Analytics
The growing complexity of Edge Video Analytics (EVA) facilitates new kind of intelligent applications, but creates challenges in real-time inference serving systems. State-of-the-art (SOTA) scheduling systems optimize global workload distributions for heterogeneous devices but often suffer from extended scheduling cycles, leading to sub-optimal processing in rapidly changing Edge environments. Local Reinforcement Learning (RL) enables quick adjustments between cycles but faces scalability, know…

Heterogeneous Quantile Treatment Effect Estimation for Longitudinal Data with High-Dimensional Confounding
Zhixin Qiu, Huichen Zhu, Wenjie Wang, Yanlin Tang
https://arxiv.org/abs/2508.16326

Heterogeneous Quantile Treatment Effect Estimation for Longitudinal Data with High-Dimensional Confounding
Causal inference plays a fundamental role in various real-world applications. However, in the motivating non-small cell lung cancer (NSCLC) study, it is challenging to estimate the treatment effect of chemotherapy on circulating tumor DNA (ctDNA). First, the heterogeneous treatment effects vary across patient subgroups defined by baseline characteristics. Second, there exists a broad set of demographic, clinical and molecular variables act as potential confounders. Third, ctDNA trajectories ove…

WattsOnAI: Measuring, Analyzing, and Visualizing Energy and Carbon Footprint of AI Workloads
Hongzhen Huang, Kunming Zhang, Hanlong Liao, Kui Wu, Guoming Tang
https://arxiv.org/abs/2506.20535

WattsOnAI: Measuring, Analyzing, and Visualizing Energy and Carbon Footprint of AI Workloads
The rapid advancement of AI, particularly large language models (LLMs), has raised significant concerns about the energy use and carbon emissions associated with model training and inference. However, existing tools for measuring and reporting such impacts are often fragmented, lacking systematic metric integration and offering limited support for correlation analysis among them. This paper presents WattsOnAI, a comprehensive software toolkit for the measurement, analysis, and visualization of …

Quasi Instrumental Variable Methods for Stable Hidden Confounding and Binary Outcome
Zhonghua Liu, Baoluo Sun, Ting Ye, David Richardson, Eric Tchetgen Tchetgen
https://arxiv.org/abs/2508.16096

Quasi Instrumental Variable Methods for Stable Hidden Confounding and Binary Outcome
Instrumental variable (IV) methods are central to causal inference from observational data, particularly when a randomized experiment is not feasible. However, of the three conventional core IV identification conditions, only one, IV relevance, is empirically verifiable; often one or both of the other conditions, exclusion restriction and IV independence from unmeasured confounders, are unmet in real-world applications. These challenges are compounded when the outcome is binary, a setting for w…

Replaced article(s) found for cs.DC. https://arxiv.org/list/cs.DC/new
[1/1]:
- Staleness-Centric Optimizations for Parallel Diffusion MoE Inference
Jiajun Luo, Lizhuo Luo, Jianru Xu, Jiajun Song, Rongwei Lu, Chen Tang, Zhi Wang

Bayesian Compressed Mixed-Effects Models
Sreya Sarkar, Kshitij Khare, Sanvesh Srivastava
https://arxiv.org/abs/2507.16961 https://arxiv.org/pdf/2507.16961

Bayesian Compressed Mixed-Effects Models
Penalized likelihood and quasi-likelihood methods dominate inference in high-dimensional linear mixed-effects models. Sampling-based Bayesian inference is less explored due to the computational bottlenecks introduced by the random effects covariance matrix. To address this gap, we propose the compressed mixed-effects (CME) model, which defines a quasi-likelihood using low-dimensional covariance parameters obtained via random projections of the random effects covariance. This dimension reduction…

Replaced article(s) found for cs.DC. https://arxiv.org/list/cs.DC/new
[1/1]:
- Staleness-Centric Optimizations for Parallel Diffusion MoE Inference
Jiajun Luo, Lizhuo Luo, Jianru Xu, Jiajun Song, Rongwei Lu, Chen Tang, Zhi Wang

Likelihood Ratio test for Poisson graph
Chen Shuyan, Liu Xin, Wang Shaoli
https://arxiv.org/abs/2506.18778 https://arxiv.org/pdf/2506…

Likelihood Ratio test for Poisson graph
Directed acyclic graphs are widely used to describe the causal effects among random variables, and the inference of those causal effects has become an popular topic in statistics and machine learning, and has wide applications in neuroinformatics, bioinformatics and so on. However, most studies focus on the estimation or inference of the directional relations among continuous random variables, those among discrete random variables have not gained much attentions. In this article we focus on the…

Principal stratification with recurrent events truncated by a terminal event: A nested Bayesian nonparametric approach
Yuki Ohnishi, Michael O. Harhay, Fan Li
https://arxiv.org/abs/2506.19015

Principal stratification with recurrent events truncated by a terminal event: A nested Bayesian nonparametric approach
Recurrent events often serve as key endpoints in clinical studies but may be prematurely truncated by terminal events such as death, creating selection bias and complicating causal inference. To address this challenge, we propose novel causal estimands within the principal stratification framework, introducing a refined ``always-survivor'' stratum that defines survival until the final recurrent event rather than a fixed time point, yielding more stable and interpretable causal contrasts. We dev…

Arctic Inference with Shift Parallelism: Fast and Efficient Open Source Inference System for Enterprise AI
Samyam Rajbhandari, Mert Hidayetoglu, Aurick Qiao, Ye Wang, Juncheng Yang, Jeff Rasley, Michael Wyatt, Yuxiong He
https://arxiv.org/abs/2507.11830

Arctic Inference with Shift Parallelism: Fast and Efficient Open Source Inference System for Enterprise AI
Inference is now the dominant AI workload, yet existing systems force trade-offs between latency, throughput, and cost. Arctic Inference, an open-source vLLM plugin from Snowflake AI Research, introduces Shift Parallelism, a dynamic parallelism strategy that adapts to real-world traffic while integrating speculative decoding, SwiftKV compute reduction, and optimized embedding inference. It achieves up to 3.4 times faster request completion, 1.75 times faster generation, and 1.6M tokens/sec per …