Ge Li

• Peking University

With the advent of neural language models, the performance of code generation has been significantly boosted. However, the problem of repetitions during the generation process continues to linger. Previous work has primarily focused on content repetition, which is merely a fraction of the broader repetition problem in code generation. A more preval...

Large language models (LLMs) have shown promising performance in automated code generation, especially excelling in simple tasks such as generating standalone codes. Different from simple tasks, real-world code generation usually depends on specific programming environment (e.g., code repositories). It contains complex dependencies and domain knowl...

Code search has been a critical software development activity in facilitating the efficiency of developers. It retrieves programs to satisfy the user intent from a codebase. Recently, many researchers have applied contrastive learning to learn the semantic relationships between queries and code snippets, resulting in impressive performance in code...

Recent advances in code generation models have demonstrated impressive capabilities in automating software development tasks, yet these models still struggle in real-world software engineering scenarios. Although current training methods, particularly post-training, excel at solving competitive programming problems, they fail to adequately prepare...

Repository-level code completion aims to complete code based on the long contexts of the repository. Existing studies extract long contexts from the repository as inputs and leverage Large Language Models (LLMs) to generate code. However, we reveal a severe limitation of LLMs, i.e., LLMs may ignore the information within long contexts in code compl...

Chain-of-Thought (CoT) reasoning has been demonstrated as an effective technique for improving the problem-solving capabilities of large language models (LLMs) in the context of code generation. However, existing CoT methods often exhibit a tendency toward "overthinking", where the LLM consistently applies reasoning strategies without adequately co...

Generating meaningful assert statements is one of the key challenges in automated test case generation, which requires understanding the intended functionality of the tested code. Recently, deep learning based models have shown promise in improving the performance of assert statement generation. However, the existing models only rely on the test pr...

Current advanced long-context language models offer great potential for real-world software engineering applications. However, progress in this critical domain remains hampered by a fundamental limitation: the absence of a rigorous evaluation framework for long code understanding. To gap this obstacle, we propose a long code understanding benchmark...

Large Language Models (LLMs) have shown impressive In-Context Learning (ICL) ability in code generation. LLMs take a prompt context consisting of a few demonstration examples and a new requirement as input, and output new programs without any parameter update. Existing studies have found that the performance of ICL-based code generation heavily dep...

Periodicity, as one of the most important basic characteristics, lays the foundation for facilitating structured knowledge acquisition and systematic cognitive processes within human learning paradigms. However, the potential flaws of periodicity modeling in Transformer affect the learning efficiency and establishment of underlying principles from...

Bug fixing holds significant importance in software development and maintenance. Recent research has made substantial strides in exploring the potential of large language models (LLMs) for automatically resolving software bugs. However, a noticeable gap in existing approaches lies in the oversight of collaborative facets intrinsic to bug resolution...

Code generation models have shown significant potential for automating programming tasks. However, the challenge of generating accurate and reliable code persists due to the highly complex and long-reasoning nature of the task. Even state-of-the-art models often fail in code generation due to small errors, which can drastically affect the overall f...

Bug fixing holds significant importance in software development and maintenance. Recent research has made substantial strides in exploring the potential of large language models (LLMs) for automatically resolving software bugs. However, a noticeable gap in existing approaches lies in the oversight of collaborative facets intrinsic to bug resolution...

Source code representation with deep learning techniques is an important research field. There have been many studies to learn sequential or structural information for code representation. However, existing sequence-based models and non-sequence models both have their limitations. Although researchers attempt to incorporate structural information i...

Recent statements about the impressive capabilities of large language models (LLMs) are usually supported by evaluating on open-access benchmarks. Considering the vast size and wide-ranging sources of LLMs' training data, it could explicitly or implicitly include test data, leading to LLMs being more susceptible to data contamination. However, due...

Language models (LMs) have been widely used to generate text on the Internet. The generated text is often collected into the training corpus of the next generations of LMs. Previous work has experimentally found that LMs collapse when trained on recursively generated text. This paper contributes to existing knowledge from two aspects. We present a...

Context Many studies have moved toward saliva and peripheral blood sampling for studying cortisol, even in relation to disorders of the brain. However, the degree to which peripheral cortisol reflects central cortisol levels has yet to be comprehensively described. Data describing the effect that biological characteristics such as age and sex have...

Large Language Models (LLMs) have been widely used in code completion, and researchers are focusing on scaling up LLMs to improve their accuracy. However, larger LLMs will increase the response time of code completion and decrease the developers’ productivity. In this paper, we propose a lightweight and effective LLM for code completion named aiXco...

Large language models (LLMs) have achieved impressive performance in code generation recently, offering programmers revolutionary assistance in software development. However, due to the auto-regressive nature of LLMs, they are susceptible to error accumulation during code generation. Once an error is produced, LLMs can merely continue to generate t...

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 tw...

Large Language Models (LLMs) have been widely used in code completion, and researchers are focusing on scaling up LLMs to improve their accuracy. However, larger LLMs will increase the response time of code completion and decrease the developers' productivity. In this paper, we propose a lightweight and effective LLM for code completion named aiXco...

In recent years, Large language model-powered Automated Program Repair (LAPR) techniques have achieved state-of-the-art bug-fixing performance and have been pervasively applied and studied in both industry and academia. Nonetheless, LLMs were proved to be highly sensitive to input prompts, with slight differences in the expressions of semantically...

Code generation models have shown significant potential for programming tasks. However, existing training methods like supervised fine-tuning face key limitations: they do not effectively teach models to prioritize correct over incorrect solutions in ambiguous situations, nor do they effectively optimize the runtime efficiency of the generated code...

Equivalent Representations (ERs) of code are textual representations that preserve the same semantics as the code itself, e.g., natural language comments and pseudocode. ERs play a critical role in software development and maintenance. However, how to automatically generate ERs of code remains an open challenge. In this paper, we propose a self-ref...

Large Language Models (LLMs) have shown impressive abilities in code generation, but they may generate erroneous programs. Reading a program takes ten times longer than writing it. Showing these erroneous programs to developers will waste developers' energies and introduce security risks to software. To address the above limitations, we propose Hon...

Despite the remarkable success achieved by neural networks, particularly those represented by MLP and Transformer, we reveal that they exhibit potential flaws in the modeling and reasoning of periodicity, i.e., they tend to memorize the periodic data rather than genuinely understanding the underlying principles of periodicity. However, periodicity...

A proper code evaluation metric (CEM) profoundly impacts the evolution of code generation, which is an important research field in NLP and software engineering. Prevailing match-based CEMs (e.g., BLEU, Accuracy, and CodeBLEU) suffer from two significant drawbacks. 1. They primarily measure the surface differences between codes without considering t...

Symbolic execution is a key technology in software testing, which generates test cases by collecting symbolic path constraints and then solving constraints with SMT solvers. Symbolic execution has been proven helpful in generating high-coverage test cases, but its limitations, e.g., the difficulties in solving path constraints, prevent it from broa...

Large Language Models (LLMs) have shown impressive abilities in code generation. Chain-of-Thought (CoT) prompting is the state-of-the-art approach to utilizing LLMs. CoT prompting asks LLMs first to generate CoTs ( i.e., intermediate natural language reasoning steps) and then output the code. However, the accuracy of CoT prompting still can not sat...

Large Language Models (LLMs), such as GPT-4, StarCoder, and CodeLlama, are transforming the way developers approach programming by automatically generating code based on given natural language descriptions. Despite advancements, generating syntactically and semantically correct code remains challenging, especially for complex programming tasks. Typ...

Large language models (LLMs) have behaved well in generating unit tests for Java projects. However, the performance for covering the complex focal methods within the projects is poor. Complex methods comprise many conditions and loops, requiring the test cases to be various enough to cover all lines and branches. However, existing test generation m...

In the past decade, thanks to the powerfulness of deep-learning techniques, we have witnessed a whole new era of automated code generation. To sort out developments, we have conducted a comprehensive review of solutions to deep learning-based code generation. In this survey, we generally formalize the pipeline and procedure of code generation and c...

Different from the flow semantics of natural languages, programming languages are inherently rigid in structure and grammar. Existing fine-tuning methodologies for code vulnerability detection generally treat code as long text sequences, stripping away structural elements such as newlines ('/n') and whitespace. However, this approach inadvertently...

Code translation tools, namely transpilers, are developed for automatic source-to-source translation. Latest learning-based transpilers have shown impressive enhancement against rule-based counterparts in both translation accuracy and readability, owing to their task-specific pre-training on extensive monolingual corpora. Nevertheless, their curren...

Code completion aims to enhance programming productivity by predicting potential code based on the current programming context. Recently, pre-trained language models (LMs) have become prominent in this field. Various approaches have been proposed to fine-tune LMs using supervised fine-tuning (SFT) techniques for code completion. However, the inhere...

Large Language Models (LLMs) have shown great success in code generation. LLMs take as the input a prompt and output the code. How to make prompts ( i.e., Prompting Techniques ) is a key question. Existing prompting techniques are designed for natural language generation and have low accuracy in code generation. In this paper, we propose a new prom...

Although large language models (LLMs) have demonstrated impressive ability in code generation, they are still struggling to address the complicated intent provided by humans. It is widely acknowledged that humans typically employ planning to decompose complex problems and schedule solution steps prior to implementation. To this end, we introduce pl...

Although Large Language Models (LLMs) have demonstrated remarkable code-generation ability, they still struggle with complex tasks. In real-world software development, humans usually tackle complex tasks through collaborative teamwork, a strategy that significantly controls development complexity and enhances software quality. Inspired by this, we...

Large Language Models (LLMs) have strong capabilities in code comprehension, but fine-tuning costs and semantic alignment issues limit their project-specific optimization; conversely, code models such CodeBERT are easy to fine-tune, but it is often difficult to learn vulnerability semantics from complex code languages. To address these challenges,...

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) DevE...

Bug localization, which is used to help programmers identify the location of bugs in source code, is an essential task in software development. Researchers have already made efforts to harness the powerful deep learning (DL) techniques to automate it. However, training bug localization model is usually challenging because it requires a large quanti...

Recently, Large Language Models (LLMs) have shown impressive abilities in code generation. However, existing LLMs' decoding strategies are designed for Natural Language (NL) generation, overlooking the differences between NL and programming languages (PL). Due to this oversight, a better decoding strategy for code generation remains an open questio...

In the software engineering (SE) community, deep learning (DL) has recently been applied to many source code processing tasks, achieving state-of-the-art results. Due to the poor interpretability of DL models, their security vulnerabilities require scrutiny. Recently, researchers have identified an emergent security threat to DL models, namely, poi...

Background and objectives: Moderate-to-severe traumatic brain injuries (TBI) have been reported to increase the risk of Alzheimer disease (AD). Whether mild TBI (mTBI) in veterans confers a similar increased risk of AD is less known. This study investigated early AD changes using CSF biomarkers in veterans with blast mTBI. Methods: This was a cr...

Software developers frequently use code completion tools to accelerate software development by suggesting the following code elements. Researchers usually employ AutoRegressive (AR) decoders to complete code sequences in a left-to-right, token-by-token fashion. To improve the accuracy and efficiency of code completion, we argue that tokens within a...

Background While epidemiologic evidence links higher levels of exposure to fine particulate matter (PM2.5) to decreased cognitive function, fewer studies have investigated links with traffic-related air pollution (TRAP), and none have examined ultrafine particles (UFP, ≤100 nm) and late-life dementia incidence. Objective To evaluate associations b...

Background Depression has been repeatedly linked to increased risk for dementia. While there is some research to suggest that an individual’s trajectory of depression symptoms over time may inform dementia risk above and beyond depression symptoms at a single time point, this has yet to be demonstrated in a dataset specifically designed to capture...

Background Traumatic brain injury (TBI) has been reported to increase risk of Alzheimer’s dementia (AD). It remains unknown whether mild TBI (mTBI) caused by blast in Veterans confers a similar increase of risk. This study investigated early AD changes defined by cerebrospinal fluid (CSF) AD biomarkers in Veterans with blast mTBI. Method Fifty‐one...

Background The novel diffusion‐based Neurite Orientation and Dispersion Imaging (NODDI) model may provide complementary information to Diffusion Tensor Imaging (DTI) to increase sensitivity in characterizing tissue microarchitecture. We examined if one or a combination of both approaches can improve our understanding of at‐risk tissue in and around...

Background The NINDS criteria define Traumatic Encephalopathy Syndrome (TES) as the clinical manifestation of the neuropathologically diagnosed chronic traumatic encephalopathy (CTE). The core clinical features of TES include both cognitive and neuropsychiatric symptoms, the latter termed neurobehavioral dysregulation (NBD) . We explored potential...

Background The Hachinski ischemic score (HIS) scale is used to exclude as much as possible, individuals with dominant cerebrovascular pathology from those with dominant AD pathology. White matter hyperintensities (WMH) are the gold‐standard for evaluating cerebrovascular pathology and linked to worsening cognitive impairment. HIS ≤ 4 are used to mi...

Background The Hachinski Ischemia Scale (HIS) differentiates AD and vascular dementia and is used to include individuals with minimal vascular risk factors in ADNI i.e., HIS ≤ 4. HIS score includes history of hypertension (HMHYPERT), a known risk for mild cognitive impairment (MCI) and vascular dementia. Cerebrovascular disease affects performance...

Large Language Models (LLMs) have shown great success in code generation. LLMs take as the input a prompt and output the code. A key question is how to make prompts (i.e., Prompting Techniques). Existing prompting techniques are designed for natural language generation and have low accuracy in code generation. In this paper, we propose a new prompt...

Large Language Models (LLMs) (e.g., ChatGPT) have shown impressive performance in code generation. LLMs take prompts as inputs, and Chain-of-Thought (CoT) prompting is the state-of-the-art prompting technique. CoT prompting asks LLMs first to generate CoTs (i.e., intermediate natural language reasoning steps) and then output the code. However, CoT...

Code generation focuses on automatically converting natural language (NL) utterances into code snippets. Sequence-to-tree (Seq2Tree) approaches are proposed for code generation with the aim of ensuring grammatical correctness of the generated code. These approaches generate subsequent Abstract Syntax Tree (AST) nodes based on the preceding predicti...

Recently, Large Language Models (LLMs) have shown impressive results in code generation. However, existing decoding strategies are designed for Natural Language (NL) generation, overlooking the differences between NL and programming languages (PL). Due to this oversight, a better decoding strategy for code generation remains an open question. In th...

Bug fixing holds significant importance in software development and maintenance. Recent research has made notable progress in exploring the potential of large language models (LLMs) for automatic bug fixing. However, existing studies often overlook the collaborative nature of bug resolution, treating it as a single-stage process. To overcome this l...

Existing studies show that code summaries help developers understand and maintain source code. Unfortunately, these summaries are often missing or outdated in software projects. Code summarization aims to generate natural language descriptions automatically for source code. Code summaries are highly structured and have repetitive patterns. Besides...

Developers introduce code clones to improve programming productivity. Many existing studies have achieved impressive performance in monolingual code clone detection. However, during software development, more and more developers write semantically equivalent programs with different languages to support different platforms and help developers transl...

Large language models (LLMs) have showcased remarkable potential across various tasks by conditioning on prompts. However, the quality of different human-written prompts leads to substantial discrepancies in LLMs' performance, and improving prompts usually necessitates considerable human effort and expertise. To this end, this paper proposes Prompt...

Despite their ability to aid developers in detecting potential defects early in the software development life cycle, static analysis tools often suffer from precision issues (i.e., high false positive rates of reported alarms). To improve the availability of these tools, many automated warning identification techniques have been proposed to assist...

Bug localization is a key software development task, where a developer locates the portion of the source code that must be modified based on the bug report. It is label-intensive and time-consuming due to the increasing size and complexity of the modern software. Effectively automating this task can greatly reduce costs by cutting down the develope...

Generating meaningful assert statements is one of the key challenges in automated test case generation, which requires understanding the intended functionality of the tested code. Recently, deep learning-based models have shown promise in improving the performance of assert statement generation. However, existing models only rely on the test prefix...

Developers often perform repetitive code editing activities (up to 70%) for various reasons ( e.g., code refactoring) during software development. Many deep learning (DL) models have been proposed to automate code editing by learning from the code editing history. Among DL-based models, pre-trained code editing models have achieved the state-of-the...

Automated program repair is a crucial task for improving the efficiency of software developers. Recently, neural-based techniques have demonstrated significant promise in generating correct patches for buggy code snippets. However, most existing approaches arbitrarily treat the buggy context without any analysis to capture the semantic relationship...

Over the past few years, the software engineering (SE) community has widely employed deep learning (DL) techniques in many source code processing tasks. Similar to other domains like computer vision and natural language processing (NLP), the state‐of‐the‐art DL techniques for source code processing can still suffer from adversarial vulnerability, w...

Automatically generating source code from natural language descriptions has been a growing field of research in recent years. However, current large-scale code generation models often encounter difficulties when selecting appropriate APIs for specific contexts. These models may generate APIs that do not meet requirements or refer to non-existent AP...

Large language models (LLMs) have demonstrated an impressive ability to generate codes on competitive programming tasks. However, with limited sample numbers, LLMs still suffer from poor accuracy. Inspired by the process of human programming, we propose a generate-and-edit approach that utilizes execution results of the generated code from LLMs to...

Code comment generation aims at generating natural language descriptions for a code snippet to facilitate developers' program comprehension activities. Despite being studied for a long time, a bottleneck for existing approaches is that given a code snippet, they can only generate one comment while developers usually need to know information from di...

Code generation is widely regarded as a key technique for elevating the automation and ultimate quality of software development. Nevertheless, existing code generation approaches usually concentrate on a single stage of the software development process (i.e., the coding stage) and do not take into consideration other stages that are crucial in redu...

In-context learning (ICL) with pre-trained language models (PTLMs) has shown great success in code generation. ICL does not require training. PTLMs take as the input a prompt consisting of a few requirement-code examples and a new requirement, and output a new program. However, existing studies simply reuse ICL techniques for natural language gener...

Although large language models have demonstrated impressive ability in code generation, they are still struggling to address the complicated intent provided by humans. It is widely acknowledged that humans typically employ planning to decompose complex problems and schedule the solution steps prior to implementation. Thus we introduce planning into...

Recently, deep learning techniques have shown great success in automatic code generation. Inspired by the code reuse, some researchers propose copy-based approaches that can copy the content from similar code snippets to obtain better performance. Practically, human developers recognize the content in the similar code that is relevant to their need...

A proper code evaluation metric (CEM) profoundly impacts the evolution of code generation, which is an important research field in NLP and software engineering. Prevailing CEMs can be categorized into match-based CEMs (e.g., BLEU, Accuracy, and CodeBLEU) and execution-based CEMs (e.g., AvgPassRatio and Pass@k), but both of them suffer from some iss...

Large-scale pre-trained language models have achieved impressive results on a wide range of downstream tasks recently. However, fine-tuning an extremely large-scale pre-trained language model on limited target datasets is often plagued by overfitting and representation degradation. In this paper, we propose a Dynamic Parameter Selection (DPS) algor...

In the process of code generation, it is essential to guarantee the generated code satisfies grammar constraints of programming language (PL). However, neglecting grammar constraints is a fatal drawback of commonly used sequence-based code generation. In this paper, we devise a pushdown automaton (PDA)-based methodology to address this problem, exp...

In the software engineering community, deep learning (DL) has recently been applied to many source code processing tasks. Due to the poor interpretability of DL models, their security vulnerabilities require scrutiny. Recently, researchers have identified an emergent security threat, namely poison attack. The attackers aim to inject insidious backd...

Developers often perform repetitive code editing activities for various reasons (e.g., code refactor) during software development. Many deep learning models are applied to automate code editing by learning from the code editing history. Recently, pre-trained code editing models have achieved the state-of-the-art (SOTA) results. Pre-trained models a...