arXiv subjects

Qi Guo

Publications and source records attributed to Qi Guo.

At least 127 records · Page 7Linked to original sources

arXiv2025

QiMeng-TensorOp: Automatically Generating High-Performance Tensor Operators with Hardware Primitives

Computation-intensive tensor operators constitute over 90\% of the computations in Large Language Models (LLMs) and Deep Neural Networks.Automatically and efficiently generating high-performance tensor operators with hardware primitives is crucial for diverse and ever-evolving hardware architectures like RISC-V, ARM, and GPUs, as manually optimized implementation takes at least months and lacks portability.LLMs excel at generating high-level language codes, but they struggle to fully comprehend hardware characteristics and produce high-performance tensor operators. We introduce a tensor-operator auto-generation framework with a one-line user prompt (QiMeng-TensorOp), which enables LLMs to automatically exploit hardware characteristics to generate tensor operators with hardware primitives, and tune parameters for optimal performance across diverse hardware. Experimental results on various hardware platforms, SOTA LLMs, and typical tensor operators demonstrate that QiMeng-TensorOp effectively unleashes the computing capability of various hardware platforms, and automatically generates tensor operators of superior performance. Compared with vanilla LLMs, QiMeng-TensorOp achieves up to $1291 \times$ performance improvement. Even compared with human experts, QiMeng-TensorOp could reach $251 \%$ of OpenBLAS on RISC-V CPUs, and $124 \%$ of cuBLAS on NVIDIA GPUs. Additionally, QiMeng-TensorOp also significantly reduces development costs by $200 \times$ compared with human experts.

cs.LG↗

arXiv2025

QiMeng-CPU-v2: Automated Superscalar Processor Design by Learning Data Dependencies

Automated processor design, which can significantly reduce human efforts and accelerate design cycles, has received considerable attention. While recent advancements have automatically designed single-cycle processors that execute one instruction per cycle, their performance cannot compete with modern superscalar processors that execute multiple instructions per cycle. Previous methods fail on superscalar processor design because they cannot address inter-instruction data dependencies, leading to inefficient sequential instruction execution. This paper proposes a novel approach to automatically designing superscalar processors using a hardware-friendly model called the Stateful Binary Speculation Diagram (State-BSD). We observe that processor parallelism can be enhanced through on-the-fly inter-instruction dependent data predictors, reusing the processor's internal states to learn the data dependency. To meet the challenge of both hardware-resource limitation and design functional correctness, State-BSD consists of two components: 1) a lightweight state-selector trained by the simulated annealing method to detect the most reusable processor states and store them in a small buffer; and 2) a highly precise state-speculator trained by the BSD expansion method to predict the inter-instruction dependent data using the selected states. It is the first work to achieve the automated superscalar processor design, i.e. QiMeng-CPU-v2, which improves the performance by about $380\times$ than the state-of-the-art automated design and is comparable to human-designed superscalar processors such as ARM Cortex A53.

cs.AR↗

arXiv2025

QiMeng-Xpiler: Transcompiling Tensor Programs for Deep Learning Systems with a Neural-Symbolic Approach

Heterogeneous deep learning systems (DLS) such as GPUs and ASICs have been widely deployed in industrial data centers, which requires to develop multiple low-level tensor programs for different platforms. An attractive solution to relieve the programming burden is to transcompile the legacy code of one platform to others. However, current transcompilation techniques struggle with either tremendous manual efforts or functional incorrectness, rendering "Write Once, Run Anywhere" of tensor programs an open question. We propose a novel transcompiler, i.e., QiMeng-Xpiler, for automatically translating tensor programs across DLS via both large language models (LLMs) and symbolic program synthesis, i.e., neural-symbolic synthesis. The key insight is leveraging the powerful code generation ability of LLM to make costly search-based symbolic synthesis computationally tractable. Concretely, we propose multiple LLM-assisted compilation passes via pre-defined meta-prompts for program transformation. During each program transformation, efficient symbolic program synthesis is employed to repair incorrect code snippets with a limited scale. To attain high performance, we propose a hierarchical auto-tuning approach to systematically explore both the parameters and sequences of transformation passes. Experiments on 4 DLS with distinct programming interfaces, i.e., Intel DL Boost with VNNI, NVIDIA GPU with CUDA, AMD MI with HIP, and Cambricon MLU with BANG, demonstrate that QiMeng-Xpiler correctly translates different tensor programs at the accuracy of 95% on average, and the performance of translated programs achieves up to 2.0x over vendor-provided manually-optimized libraries. As a result, the programming productivity of DLS is improved by up to 96.0x via transcompiling legacy tensor programs.

cs.CL↗

arXiv2025

FCGHunter: Towards Evaluating Robustness of Graph-Based Android Malware Detection

Graph-based detection methods leveraging Function Call Graphs (FCGs) have shown promise for Android malware detection (AMD) due to their semantic insights. However, the deployment of malware detectors in dynamic and hostile environments raises significant concerns about their robustness. While recent approaches evaluate the robustness of FCG-based detectors using adversarial attacks, their effectiveness is constrained by the vast perturbation space, particularly across diverse models and features. To address these challenges, we introduce FCGHunter, a novel robustness testing framework for FCG-based AMD systems. Specifically, FCGHunter employs innovative techniques to enhance exploration and exploitation within this huge search space. Initially, it identifies critical areas within the FCG related to malware behaviors to narrow down the perturbation space. We then develop a dependency-aware crossover and mutation method to enhance the validity and diversity of perturbations, generating diverse FCGs. Furthermore, FCGHunter leverages multi-objective feedback to select perturbed FCGs, significantly improving the search process with interpretation-based feature change feedback. Extensive evaluations across 40 scenarios demonstrate that FCGHunter achieves an average attack success rate of 87.9%, significantly outperforming baselines by at least 44.7%. Notably, FCGHunter achieves a 100% success rate on robust models (e.g., AdaBoost with MalScan), where baselines achieve only 11% or are inapplicable.

cs.CR↗

arXiv2025

Constraints on the galaxy formation models during epoch of reionization with high redshift observations

We use high resolution N-body dark matter simulations and L-Galaxies semi-analytical galaxy formation models to explore the high-$z$ galaxy properties and estimate the budget of ionizing photons. The parameters within L-Galaxies are obtained using a Markov Chain Monte Carlo (MCMC) method with high-$z$ galaxy observations from JWST and other telescopes. We consider two versions of L-Galaxies with and without dust correction on galaxy UV luminosities. With the best-fit parameters, both L-Galaxies 2015 and L-Galaxies 2020 reproduce well observations of UV luminosity functions, stellar mass functions, star formation rate densities and ionizing photon emission efficiency. With the assumption of escape fraction of $20\%$, all models produce more ionizing photons than the number of Hydrogen atoms in the Universe at $z>6$. The inclusion of dust correction within MCMC results in higher star formation efficiency, which predicts $\sim 50\%$ more ionizing photons, with better consistency between the predicted stellar mass functions and observations.

astro-ph.CO↗

arXiv2025

Focal Split: Untethered Snapshot Depth from Differential Defocus

We introduce Focal Split, a handheld, snapshot depth camera with fully onboard power and computing based on depth-from-differential-defocus (DfDD). Focal Split is passive, avoiding power consumption of light sources. Its achromatic optical system simultaneously forms two differentially defocused images of the scene, which can be independently captured using two photosensors in a snapshot. The data processing is based on the DfDD theory, which efficiently computes a depth and a confidence value for each pixel with only 500 floating point operations (FLOPs) per pixel from the camera measurements. We demonstrate a Focal Split prototype, which comprises a handheld custom camera system connected to a Raspberry Pi 5 for real-time data processing. The system consumes 4.9 W and is powered on a 5 V, 10,000 mAh battery. The prototype can measure objects with distances from 0.4 m to 1.2 m, outputting 480$\times$360 sparse depth maps at 2.1 frames per second (FPS) using unoptimized Python scripts. Focal Split is DIY friendly. A comprehensive guide to building your own Focal Split depth camera, code, and additional data can be found at https://focal-split.qiguo.org.

cs.CV↗

arXiv2025

Wavefront Estimation From a Single Measurement: Uniqueness and Algorithms

Wavefront estimation is an essential component of adaptive optics where the goal is to recover the underlying phase from its Fourier magnitude. While this may sound identical to classical phase retrieval, wavefront estimation faces more strict requirements regarding uniqueness as adaptive optics systems need a unique phase to compensate for the distorted wavefront. Existing real-time wavefront estimation methodologies are dominated by sensing via specialized optical hardware due to their high speed, but they often have a low spatial resolution. A computational method that can perform both fast and accurate wavefront estimation with a single measurement can improve resolution and bring new applications such as real-time passive wavefront estimation, opening the door to a new generation of medical and defense applications. In this paper, we tackle the wavefront estimation problem by observing that the non-uniqueness is related to the geometry of the pupil shape. By analyzing the source of ambiguities and breaking the symmetry, we present a joint optics-algorithm approach by co-designing the shape of the pupil and the reconstruction neural network. Using our proposed lightweight neural network, we demonstrate wavefront estimation of a phase of size $128\times 128$ at $5,200$ frames per second on a CPU computer, achieving an average Strehl ratio up to $0.98$ in the noiseless case. We additionally test our method on real measurements using a spatial light modulator. Code is available at https://pages.github.itap.purdue.edu/StanleyChanGroup/wavefront-estimation/.

eess.SP↗

arXiv2025

Exploring cosmological constraints of the weak gravitational lensing and galaxy clustering joint analysis in the CSST photometric survey

We explore the joint weak lensing and galaxy clustering analysis from the photometric survey operated by the China Space Station Telescope (CSST), and study the strength of the cosmological constraints. We employ a high-resolution JiuTian-1G simulation to construct a partial-sky light cone to $z=3$ covering 100 deg$^2$, and obtain the CSST galaxy mock samples based on an improved semi-analytical model. We perform a multi-lens-plane algorithm to generate corresponding synthetic weak lensing maps and catalogs. Then we generate the mock data based on these catalogs considering the instrumental and observational effects of the CSST, and use the Markov Chain Monte Carlo (MCMC) method to perform the constraints. The covariance matrix includes non-Gaussian contributions and super-sample covariance terms, and the systematics from intrinsic alignments, galaxy bias, photometric redshift uncertainties, shear calibration, and non-linear effects are considered in the analysis. We find that the constraint result is comparable to that from Stage III surveys, and it can be significantly improved further in the full CSST survey with 17500 deg$^2$. This indicates the CSST photometric survey is powerful for exploring the Universe.

astro-ph.CO↗

arXiv2025

FlexMem: High-Parallel Near-Memory Architecture for Flexible Dataflow in Fully Homomorphic Encryption

Fully Homomorphic Encryption (FHE) imposes substantial memory bandwidth demands, presenting significant challenges for efficient hardware acceleration. Near-memory Processing (NMP) has emerged as a promising architectural solution to alleviate the memory bottleneck. However, the irregular memory access patterns and flexible dataflows inherent to FHE limit the effectiveness of existing NMP accelerators, which fail to fully utilize the available near-memory bandwidth. In this work, we propose FlexMem, a near-memory accelerator featuring high-parallel computational units with varying memory access strides and interconnect topologies to effectively handle irregular memory access patterns. Furthermore, we design polynomial and ciphertext-level dataflows to efficiently utilize near-memory bandwidth under varying degrees of polynomial parallelism and enhance parallel performance. Experimental results demonstrate that FlexMem achieves 1.12 times of performance improvement over state-of-the-art near-memory architectures, with 95.7% of near-memory bandwidth utilization.

cs.AR↗

arXiv2025

A Survey on Evaluating Large Language Models in Code Generation Tasks

This paper provides a comprehensive review of the current methods and metrics used to evaluate the performance of Large Language Models (LLMs) in code generation tasks. With the rapid growth in demand for automated software development, LLMs have demonstrated significant potential in the field of code generation. The paper begins by reviewing the historical development of LLMs and their applications in code generation. Next, it details various methods and metrics for assessing the code generation capabilities of LLMs, including code correctness, efficiency, readability, and evaluation methods based on expert review and user experience. The paper also evaluates the widely used benchmark datasets, identifying their limitations and proposing directions for future improvements. Specifically, the paper analyzes the performance of code generation models across different tasks by combining multiple evaluation metrics, such as code compilation/interpretation success rates, unit test pass rates, and performance and efficiency metrics, to comprehensively assess the practical application of LLMs in code generation. Finally, the paper discusses the challenges faced in evaluating LLMs in code generation, particularly how to ensure the comprehensiveness and accuracy of evaluation methods and how to adapt to the evolving practices of software development. These analyses and discussions provide valuable insights for further optimizing and improving the application of LLMs in code generation tasks.

cs.SE↗

arXiv2025

Enhancing the coherence time of a neutral atom by an optical quartic trap

The coherence time of an optically trapped neutral atom is a crucial parameter for quantum technologies. We found that optical dipole traps with higher-order spatial forms inherently offer lower decoherence rates compared to those with lower-order spatial forms. We formulated the decoherence rate caused by the variance of the differential energy shift and photon jumping rate. Then, we constructed blue-detuned harmonic and quartic optical dipole traps, and experimentally investigated the coherence time of a trapped single cesium atom. The experimental results qualitatively verified our theory. Our approach provides a novel method to enhance the coherence time of optically trapped neutral atoms.

quant-ph↗

arXiv2025

2D watershed void clustering for probing the cosmic large-scale structure

Cosmic void has been proven to be an effective cosmological probe of the large-scale structure (LSS). However, since voids are usually identified in spectroscopic galaxy surveys, they are generally limited to low number density and redshift. We propose to utilize the clustering of two-dimensional (2D) voids identified using Voronoi tessellation and watershed algorithm without any shape assumption to explore the LSS. We generate mock galaxy and void catalogs for the next-generation Stage IV photometric surveys in $z = 0.8-2.0$ from simulations, develop the 2D void identification method, and construct the theoretical model to fit the 2D watershed void and galaxy angular power spectra. We find that our method can accurately extract the cosmological information, and the constraint accuracies of some cosmological parameters from the 2D watershed void clustering are even comparable to the galaxy angular clustering case, which can be further improved by as large as $\sim30\%$ in the void and galaxy joint constraints. This indicates that the 2D void clustering is a good complement to galaxy angular clustering measurements, especially for the forthcoming Stage IV surveys that detect high-redshift universe.

astro-ph.CO↗

arXiv2025

Intention is All You Need: Refining Your Code from Your Intention

Code refinement aims to enhance existing code by addressing issues, refactoring, and optimizing to improve quality and meet specific requirements. As software projects scale in size and complexity, the traditional iterative exchange between reviewers and developers becomes increasingly burdensome. While recent deep learning techniques have been explored to accelerate this process, their performance remains limited, primarily due to challenges in accurately understanding reviewers' intents. This paper proposes an intention-based code refinement technique that enhances the conventional comment-to-code process by explicitly extracting reviewer intentions from the comments. Our approach consists of two key phases: Intention Extraction and Intention Guided Revision Generation. Intention Extraction categorizes comments using predefined templates, while Intention Guided Revision Generation employs large language models (LLMs) to generate revised code based on these defined intentions. Three categories with eight subcategories are designed for comment transformation, which is followed by a hybrid approach that combines rule-based and LLM-based classifiers for accurate classification. Extensive experiments with five LLMs (GPT4o, GPT3.5, DeepSeekV2, DeepSeek7B, CodeQwen7B) under different prompting settings demonstrate that our approach achieves 79% accuracy in intention extraction and up to 66% in code refinement generation. Our results highlight the potential of our approach in enhancing data quality and improving the efficiency of code refinement.

cs.SE↗

arXiv2025

DiffZOO: A Purely Query-Based Black-Box Attack for Red-teaming Text-to-Image Generative Model via Zeroth Order Optimization

Current text-to-image (T2I) synthesis diffusion models raise misuse concerns, particularly in creating prohibited or not-safe-for-work (NSFW) images. To address this, various safety mechanisms and red teaming attack methods are proposed to enhance or expose the T2I model's capability to generate unsuitable content. However, many red teaming attack methods assume knowledge of the text encoders, limiting their practical usage. In this work, we rethink the case of \textit{purely black-box} attacks without prior knowledge of the T2l model. To overcome the unavailability of gradients and the inability to optimize attacks within a discrete prompt space, we propose DiffZOO which applies Zeroth Order Optimization to procure gradient approximations and harnesses both C-PRV and D-PRV to enhance attack prompts within the discrete prompt domain. We evaluated our method across multiple safety mechanisms of the T2I diffusion model and online servers. Experiments on multiple state-of-the-art safety mechanisms show that DiffZOO attains an 8.5% higher average attack success rate than previous works, hence its promise as a practical red teaming tool for T2l models.

cs.CR↗

arXiv2025

CDIO: Cross-Domain Inference Optimization with Resource Preference Prediction for Edge-Cloud Collaboration

Currently, massive video tasks are processed by edge-cloud collaboration. However, the diversity of task requirements and the dynamics of resources pose great challenges to efficient inference, resulting in many wasted resources. In this paper, we present CDIO, a cross-domain inference optimization framework designed for edge-cloud collaboration. For diverse input tasks, CDIO can predict resource preference types by analyzing spatial complexity and processing requirements of the task. Subsequently, a cross-domain collaborative optimization algorithm is employed to guide resource allocation in the edge-cloud system. By ensuring that each task is matched with the ideal servers, the edge-cloud system can achieve higher efficiency inference. The evaluation results on public datasets demonstrate that CDIO can effectively meet the accuracy and delay requirements for task processing. Compared to state-of-the-art edge-cloud solutions, CDIO achieves a computing and bandwidth consumption reduction of 20%-40%. And it can reduce energy consumption by more than 40%.

cs.MM↗

arXiv2025

A Neural Network Model for the Cosmic Dispersion Measure in the CAMELS Simulations

The probability distribution, $p(\mathrm{DM})$ of cosmic dispersion measures (DM) measured in fast radio bursts (FRBs) encodes information about both cosmology and galaxy feedback. In this work, we study the effect of feedback parameters in the $p(\mathrm{DM})$ calculated from the full Latin Hypercube of parameters sampled by the CAMELS hydrodynamical simulation suite, building a neural network (NN) model that performs well in emulating the effect of feedback on $p(\mathrm{DM})$ at arbitrary redshifts at $z\leq1$. Using this NN model, we further study the parameter $F\equiv σ_{\rm DM} \, z^{1/2}$, which is commonly used to summarize the scatter on $p(\mathrm{DM})$. We find that $F$ does not depend monotonically on every feedback parameter; instead each feedback mechanism jointly influences the final feedback strength in non-trivial ways. Even the largest values of $F$ that we find in our entire parameter space are small compared to the current constraints from observed FRB DMs by Baptista et al. 2024, pointing at the limitations of the CAMELS suite due to the small simulation box sizes. In the future, with larger box-sizes from CAMELS-like suites, similar models can be used to constrain the parameters governing galaxy feedback in the increasing observational samples of FRBs.

astro-ph.CO↗

arXiv2025

DWCL: Dual-Weighted Contrastive Learning for Multi-View Clustering

Multi-view contrastive clustering (MVCC) has gained significant attention for generating consistent clustering structures from multiple views through contrastive learning. However, most existing MVCC methods create cross-views by combining any two views, leading to a high volume of unreliable pairs. Furthermore, these approaches often overlook discrepancies in multi-view representations, resulting in representation degeneration. To address these challenges, we introduce a novel model called Dual-Weighted Contrastive Learning (DWCL) for Multi-View Clustering. Specifically, to reduce the impact of unreliable cross-views, we introduce an innovative Best-Other (B-O) contrastive mechanism that enhances the representation of individual views at a low computational cost. Furthermore, we develop a dual weighting strategy that combines a view quality weight, reflecting the quality of each view, with a view discrepancy weight. This approach effectively mitigates representation degeneration by downplaying cross-views that are both low in quality and high in discrepancy. We theoretically validate the efficiency of the B-O contrastive mechanism and the effectiveness of the dual weighting strategy. Extensive experiments demonstrate that DWCL outperforms previous methods across eight multi-view datasets, showcasing superior performance and robustness in MVCC. Specifically, our method achieves absolute accuracy improvements of 5.4\% and 5.6\% compared to state-of-the-art methods on the Caltech6V7 and MSRCv1 datasets, respectively.

cs.CV↗

arXiv2025

AGON: Automated Design Framework for Customizing Processors from ISA Documents

Customized processors are attractive solutions for vast domain-specific applications due to their high energy efficiency. However, designing a processor in traditional flows is time-consuming and expensive. To address this, researchers have explored methods including the use of agile development tools like Chisel or SpinalHDL, high-level synthesis (HLS) from programming languages like C or SystemC, and more recently, leveraging large language models (LLMs) to generate hardware description language (HDL) code from natural language descriptions. However, each method has limitations in terms of expressiveness, correctness, and performance, leading to a persistent contradiction between the level of automation and the effectiveness of the design. Overall, how to automatically design highly efficient and practical processors with minimal human effort remains a challenge. In this paper, we propose AGON, a novel framework designed to leverage LLMs for the efficient design of out-of-order (OoO) customized processors with minimal human effort. Central to AGON is the nano-operator function (nOP function) based Intermediate Representation (IR), which bridges high-level descriptions and hardware implementations while decoupling functionality from performance optimization, thereby providing an automatic design framework that is expressive and efficient, has correctness guarantees, and enables PPA (Power, Performance, and Area) optimization. Experimental results show that superior to previous LLM-assisted automatic design flows, AGON facilitates designing a series of customized OoO processors that achieve on average 2.35 $\times$ speedup compared with BOOM, a general-purpose CPU designed by experts, with minimal design effort.

cs.AR↗