arXiv subjects

Xuanming Zhang

Publications and source records attributed to Xuanming Zhang.

33 records · Page 2Linked to original sources

arXiv2025

Theoretical Proof that Auto-regressive Language Models Collapse when Real-world Data is a Finite Set

Auto-regressive language models (LMs) have been widely used to generate data in data-scarce domains to train new LMs, compensating for the scarcity of real-world data. Previous work experimentally found that LMs collapse when trained on recursively generated data. This paper presents a theoretical proof: once a corpus (such as a subset of the World Wide Web) begins to incorporate generated data and no new real-world data is added to the corpus, then no matter how small the amount of data each LM generates and contributes to the corpus, LM collapse is inevitable after sufficient time. This finding suggests that attempts to mitigate collapse by limiting the quantity of synthetic data in the corpus are fundamentally insufficient. Instead, avoiding collapse hinges on ensuring the quality of synthetic data.

cs.CL↗

arXiv2025

Quantitative Energy Prediction based on Carbon Emission Analysis by DPR Framework

This study proposes a novel analytical framework that integrates DBSCAN clustering with the Elastic Net regression model to address multifactorial problems characterized by structural complexity and multicollinearity, exemplified by carbon emissions analysis. DBSCAN is employed for unsupervised learning to objectively cluster features, while the Elastic Net is utilized for high-dimensional feature selection and complexity control. The Elastic Net is specifically chosen for its ability to balance feature selection and regularization by combining L1 (lasso) and L2 (ridge) penalties, making it particularly suited for datasets with correlated predictors. Applying this framework to energy consumption data from 46 industries in China (2000-2019) resulted in the identification of 16 categories. Emission characteristics and drivers were quantitatively assessed for each category, demonstrating the framework's capacity to identify primary emission sources and provide actionable insights. This research underscores the global applicability of the framework for analyzing complex regional challenges, such as carbon emissions, and highlights its potential to identify opportunities for emission reduction.

cs.LG↗

arXiv2025

Toward Effective PBFT Consensus Service under Software Aging in Dynamic Scenarios

The increasing application and deployment of blockchain in various services necessitates the assurance of the effectiveness of PBFT (Practical Byzantine Fault Tolerance) consensus service. However, the performance of PBFT consensus service is challenged in dynamic scenarios. The paper explores how to reduce the consensus processing time and maintenance cost of PBFT consensus service under software aging in dynamic scenarios. We first propose a PBFT system, consisting of three subsystems, one active-node subsystem, one standby-node subsystem and a repair subsystem. All the active nodes participate in the consensus and all standby nodes aim for fault-tolerance. Each aging/crashed nodes become standby nodes after completing its repairing in the repair subsystem. The nodes migrate between the active-node and standby-node subsystems in order to support the continuity of the PBFT consensus service while reducing maintenance cost. Then, we develop a Markov-chain-based analytical model for capturing the behaviors of the system and also derive the formulas for calculating the metrics, including consensus processing time, PBFT service availability, the mean number of nodes in each subsystem. Finally, we design a Multi-Objective Evolutionary Algorithm-based method for minimizing both the PBFT service response time and the PBFT system maintenance cost. We also conduct experiments for evaluation.

cs.SE↗

arXiv2024

Seeker: Towards Exception Safety Code Generation with Intermediate Language Agents Framework

In real world software development, improper or missing exception handling can severely impact the robustness and reliability of code. Exception handling mechanisms require developers to detect, capture, and manage exceptions according to high standards, but many developers struggle with these tasks, leading to fragile code. This problem is particularly evident in open-source projects and impacts the overall quality of the software ecosystem. To address this challenge, we explore the use of large language models (LLMs) to improve exception handling in code. Through extensive analysis, we identify three key issues: Insensitive Detection of Fragile Code, Inaccurate Capture of Exception Block, and Distorted Handling Solution. These problems are widespread across real world repositories, suggesting that robust exception handling practices are often overlooked or mishandled. In response, we propose Seeker, a multi-agent framework inspired by expert developer strategies for exception handling. Seeker uses agents: Scanner, Detector, Predator, Ranker, and Handler to assist LLMs in detecting, capturing, and resolving exceptions more effectively. Our work is the first systematic study on leveraging LLMs to enhance exception handling practices in real development scenarios, providing valuable insights for future improvements in code reliability.

cs.CL↗

arXiv2024

EvoCodeBench: An Evolving Code Generation Benchmark with Domain-Specific Evaluations

How to evaluate Large Language Models (LLMs) in code generation remains an open question. Existing benchmarks have two limitations - data leakage and lack of domain-specific evaluation. The former hurts the fairness of benchmarks, and the latter hinders practitioners from selecting superior LLMs for specific programming domains. To address these two limitations, we propose a new benchmark - EvoCodeBench, which has the following advances: (1) Evolving data. EvoCodeBench will be dynamically updated every period (e.g., 6 months) to avoid data leakage. This paper releases the first version - EvoCodeBench-2403, containing 275 samples from 25 repositories. (2) A domain taxonomy and domain labels. Based on the statistics of open-source communities, we design a programming domain taxonomy consisting of 10 popular domains. Based on the taxonomy, we annotate each sample in EvoCodeBench with a domain label. (3) Domain-specific evaluations. Besides the Pass@k, we compute the Domain-Specific Improvement (DSI) and define LLMs' comfort and strange domains. These evaluations help practitioners select superior LLMs in specific domains and discover the shortcomings of existing LLMs. We evaluate 8 popular LLMs (e.g., gpt-4, DeepSeek Coder) on EvoCodeBench and summarize some insights. EvoCodeBench reveals the actual abilities of these LLMs in real-world repositories. For example, the highest Pass@1 of gpt-4 on EvoCodeBench-2403 is only 20.74%. Besides, we evaluate LLMs in different domains and discover their comfort and strange domains. For example, gpt-4 performs best in most domains but falls behind others in the Internet domain. StarCoder 2-15B unexpectedly performs well in the Database domain and even outperforms 33B LLMs. EvoCodeBench has been released.

cs.CL↗

arXiv2024

DECOR: Improving Coherence in L2 English Writing with a Novel Benchmark for Incoherence Detection, Reasoning, and Rewriting

Coherence in writing, an aspect that second-language (L2) English learners often struggle with, is crucial in assessing L2 English writing. Existing automated writing evaluation systems primarily use basic surface linguistic features to detect coherence in writing. However, little effort has been made to correct the detected incoherence, which could significantly benefit L2 language learners seeking to improve their writing. To bridge this gap, we introduce DECOR, a novel benchmark that includes expert annotations for detecting incoherence in L2 English writing, identifying the underlying reasons, and rewriting the incoherent sentences. To our knowledge, DECOR is the first coherence assessment dataset specifically designed for improving L2 English writing, featuring pairs of original incoherent sentences alongside their expert-rewritten counterparts. Additionally, we fine-tuned models to automatically detect and rewrite incoherence in student essays. We find that incorporating specific reasons for incoherence during fine-tuning consistently improves the quality of the rewrites, achieving a result that is favored in both automatic and human evaluations.

cs.CL↗

arXiv2024

VarBench: Robust Language Model Benchmarking Through Dynamic Variable Perturbation

As large language models achieve impressive scores on traditional benchmarks, an increasing number of researchers are becoming concerned about benchmark data leakage during pre-training, commonly known as the data contamination problem. To ensure fair evaluation, recent benchmarks release only the training and validation sets, keeping the test set labels closed-source. They require anyone wishing to evaluate his language model to submit the model's predictions for centralized processing and then publish the model's result on their leaderboard. However, this submission process is inefficient and prevents effective error analysis. To address this issue, we propose to variabilize benchmarks and evaluate language models dynamically. Specifically, we extract variables from each test case and define a value range for each variable. For each evaluation, we sample new values from these value ranges to create unique test cases, thus ensuring a fresh evaluation each time. We applied this variable perturbation method to four datasets: GSM8K, ARC, CommonsenseQA, and TruthfulQA, which cover mathematical generation and multiple-choice tasks. Our experimental results demonstrate that this approach provides a more accurate assessment of the true capabilities of language models, effectively mitigating the contamination problem.

cs.CL↗

arXiv2024

ProLex: A Benchmark for Language Proficiency-oriented Lexical Substitution

Lexical Substitution discovers appropriate substitutes for a given target word in a context sentence. However, the task fails to consider substitutes that are of equal or higher proficiency than the target, an aspect that could be beneficial for language learners looking to improve their writing. To bridge this gap, we propose a new task, language proficiency-oriented lexical substitution. We also introduce ProLex, a novel benchmark designed to assess systems' ability to generate not only appropriate substitutes but also substitutes that demonstrate better language proficiency. Besides the benchmark, we propose models that can automatically perform the new task. We show that our best model, a Llama2-13B model fine-tuned with task-specific synthetic data, outperforms ChatGPT by an average of 3.2% in F-score and achieves comparable results with GPT-4 on ProLex.

cs.CL↗

arXiv2024

DevEval: A Manually-Annotated Code Generation Benchmark Aligned with Real-World Code Repositories

How to evaluate the coding abilities of Large Language Models (LLMs) remains an open question. We find that existing benchmarks are poorly aligned with real-world code repositories and are insufficient to evaluate the coding abilities of LLMs. To address the knowledge gap, we propose a new benchmark named DevEval, which has three advances. (1) DevEval aligns with real-world repositories in multiple dimensions, e.g., code distributions and dependency distributions. (2) DevEval is annotated by 13 developers and contains comprehensive annotations (e.g., requirements, original repositories, reference code, and reference dependencies). (3) DevEval comprises 1,874 testing samples from 117 repositories, covering 10 popular domains (e.g., Internet, Database). Based on DevEval, we propose repository-level code generation and evaluate 8 popular LLMs on DevEval (e.g., gpt-4, gpt-3.5, StarCoder 2, DeepSeek Coder, CodeLLaMa). Our experiments reveal these LLMs' coding abilities in real-world code repositories. For example, in our experiments, the highest Pass@1 of gpt-4-turbo is only 53.04%. We also analyze LLMs' failed cases and summarize their shortcomings. We hope DevEval can facilitate the development of LLMs in real code repositories. DevEval, prompts, and LLMs' predictions have been released.

cs.CL↗

arXiv2024

EvoCodeBench: An Evolving Code Generation Benchmark Aligned with Real-World Code Repositories

How to evaluate Large Language Models (LLMs) in code generation is an open question. Existing benchmarks demonstrate poor alignment with real-world code repositories and are insufficient to evaluate the coding abilities of LLMs. This paper proposes a new benchmark - EvoCodeBench to address the preceding problems, which has three primary advances. (1) EvoCodeBench aligns with real-world repositories in multiple dimensions, e.g., code distributions and dependency distributions. (2) EvoCodeBench offers comprehensive annotations (e.g., requirements, reference code, and reference dependencies), and robust evaluation metrics (e.g., Pass@k and Recall@k). (3) EvoCodeBench is an evolving benchmark to avoid data leakage. We build an automatic pipeline to update EvoCodeBench from the latest repositories. We release the first version - EvoCodeBench-2403, containing 275 samples from 25 real-world repositories. Based on EvoCodeBench, we propose repository-level code generation and evaluate 10 popular LLMs (e.g., gpt-4, gpt-3.5, DeepSeek Coder, StarCoder 2, CodeLLaMa, Gemma, and Qwen 1.5). Our experiments reveal the coding abilities of these LLMs in real-world repositories. For example, the highest Pass@1 of gpt-4 only is 20.73% in our experiments. We also analyze failed cases and summarize the shortcomings of existing LLMs in EvoCodeBench. We release EvoCodeBench, all prompts, and LLMs' completions for further community analysis.

cs.CL↗

arXiv2024

Parameter Estimation from Single Patient, Single Time-Point Sequencing Data of Recurrent Tumors

In this study, we develop consistent estimators for key parameters that govern the dynamics of tumor cell populations when subjected to pharmacological treatments. While these treatments often lead to an initial reduction in the abundance of drug-sensitive cells, a population of drug-resistant cells frequently emerges over time, resulting in cancer recurrence. Samples from recurrent tumors present as an invaluable data source that can offer crucial insights into the ability of cancer cells to adapt and withstand treatment interventions. To effectively utilize the data obtained from recurrent tumors, we derive several large number limit theorems, specifically focusing on the metrics that quantify the clonal diversity of cancer cell populations at the time of cancer recurrence. These theorems then serve as the foundation for constructing our estimators. A distinguishing feature of our approach is that our estimators only require a single time-point sequencing data from a single tumor, thereby enhancing the practicality of our approach and enabling the understanding of cancer recurrence at the individual level.

stat.AP↗

arXiv2024

Limit theorems for the site frequency spectrum of neutral mutations in an exponentially growing population

The site frequency spectrum (SFS) is a widely used summary statistic of genomic data. Motivated by recent evidence for the role of neutral evolution in cancer, we investigate the SFS of neutral mutations in an exponentially growing population. Using branching process techniques, we establish (first-order) almost sure convergence results for the SFS of a Galton-Watson process, evaluated either at a fixed time or at the stochastic time at which the population first reaches a certain size. We finally use our results to construct consistent estimators for the extinction probability and the effective mutation rate of a birth-death process.

math.PR↗

arXiv2024

DevEval: Evaluating Code Generation in Practical Software Projects

How to evaluate Large Language Models (LLMs) in code generation is an open question. Many benchmarks have been proposed but are inconsistent with practical software projects, e.g., unreal program distributions, insufficient dependencies, and small-scale project contexts. Thus, the capabilities of LLMs in practical projects are still unclear. In this paper, we propose a new benchmark named DevEval, aligned with Developers' experiences in practical projects. DevEval is collected through a rigorous pipeline, containing 2,690 samples from 119 practical projects and covering 10 domains. Compared to previous benchmarks, DevEval aligns to practical projects in multiple dimensions, e.g., real program distributions, sufficient dependencies, and enough-scale project contexts. We assess five popular LLMs on DevEval (e.g., gpt-4, gpt-3.5-turbo, CodeLLaMa, and StarCoder) and reveal their actual abilities in code generation. For instance, the highest Pass@1 of gpt-3.5-turbo only is 42 in our experiments. We also discuss the challenges and future directions of code generation in practical projects. We open-source DevEval and hope it can facilitate the development of code generation in practical projects.

cs.SE↗

arXiv2021

From Symbols to Embeddings: A Tale of Two Representations in Computational Social Science

Computational Social Science (CSS), aiming at utilizing computational methods to address social science problems, is a recent emerging and fast-developing field. The study of CSS is data-driven and significantly benefits from the availability of online user-generated contents and social networks, which contain rich text and network data for investigation. However, these large-scale and multi-modal data also present researchers with a great challenge: how to represent data effectively to mine the meanings we want in CSS? To explore the answer, we give a thorough review of data representations in CSS for both text and network. Specifically, we summarize existing representations into two schemes, namely symbol-based and embedding-based representations, and introduce a series of typical methods for each scheme. Afterwards, we present the applications of the above representations based on the investigation of more than 400 research articles from 6 top venues involved with CSS. From the statistics of these applications, we unearth the strength of each kind of representations and discover the tendency that embedding-based representations are emerging and obtaining increasing attention over the last decade. Finally, we discuss several key challenges and open issues for future directions. This survey aims to provide a deeper understanding and more advisable applications of data representations for CSS researchers.

cs.CY↗

arXiv2021

A High-fidelity, Machine-learning Enhanced Queueing Network Simulation Model for Hospital Ultrasound Operations

We collaborate with a large teaching hospital in Shenzhen, China and build a high-fidelity simulation model for its ultrasound center to predict key performance metrics, including the distributions of queue length, waiting time and sojourn time, with high accuracy. The key challenge to build an accurate simulation model is to understanding the complicated patient routing at the ultrasound center. To address the issue, we propose a novel two-level routing component to the queueing network model. We apply machine learning tools to calibrate the key components of the queueing model from data with enhanced accuracy.

cs.LG↗