A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2023; you can also visit the original URL.
The file type is application/pdf
.
Filters
Boosting Automated Patch Correctness Prediction via Pre-trained Language Model
[article]
2023
arXiv
pre-print
In this paper, we propose APPT, a pre-trained model-based automated patch correctness assessment technique, which treats the source code as token sequences without extra overhead to design hand-crafted ...
Our additional investigation on 49,694 real-world patches shows that APPT achieves the optimum performance (exceeding 99% in five common metrics for assessing patch classification techniques) compared ...
We also highlight this direction of integrating code-aware features (e.g., code edits and AST representation) with pre-trained models for patch correctness assessment. ...
arXiv:2301.12453v1
fatcat:e4dljj5b5vctpn2oa6knaiaqsa
CCRep: Learning Code Change Representations via Pre-Trained Code Model and Query Back
[article]
2023
arXiv
pre-print
To evaluate CCRep and demonstrate its applicability to diverse code-change-related tasks, we apply it to three tasks: commit message generation, patch correctness assessment, and just-in-time defect prediction ...
Specifically, CCRep regards a code change as the combination of its before-change and after-change code, leverages a pre-trained code model to obtain high-quality contextual embeddings of code, and uses ...
ACKNOWLEDGMENTS This research/project is supported by the National Natural Science Foundation of China (No. 62202420) and the Fundamental Research Funds for the Central Universities (No. 226-2022-00064 ...
arXiv:2302.03924v1
fatcat:xmza6cs4k5fu3icke3kouppuku
The Best of Both Worlds: Combining Learned Embeddings with Engineered Features for Accurate Prediction of Correct Patches
[article]
2022
arXiv
pre-print
Our empirical work investigates different representation learning approaches for code changes to derive embeddings that are amenable to similarity computations of patch correctness identification, and ...
assess the possibility of accurate classification of correct patch by combining learned embeddings with engineered features. ...
This finding suggests we use the embedding model built for code changes (e.g., CC2Vec) for the objective of having a high recall in identifying correct patches. ...
arXiv:2203.08912v2
fatcat:sowqqxbmtjb7rmakcnumow4fsu
The Remarkable Role of Similarity in Redundancy-based Program Repair
[article]
2019
arXiv
pre-print
We show that with similarity analysis, at least 90% of search space can be ignored to find the correct patch. ...
Code similarity is capable of ranking the correct repair ingredient first in 4 - 33 % of the considered cases. ...
By combining the best context-less ranking and the best context-aware ranking, we achieve better results than context-less ranking only. ...
arXiv:1811.05703v3
fatcat:sgf5eeeaizgplamogknn3ztzdi
PatchZero: Zero-Shot Automatic Patch Correctness Assessment
[article]
2024
arXiv
pre-print
To mitigate the issue, in this study, we propose \toolname, the patch correctness assessment by adopting a large language model for code. ...
Specifically, for patches generated by a new or unseen APR tool, \toolname does not need labeled patches of this new or unseen APR tool for training but directly queries the large language model for code ...
[33] proposed Cache, a patch correctness assessment technique that learns a context-aware code change embedding, considering program structures. ...
arXiv:2303.00202v3
fatcat:sq7wz7vvmjg4nb5j3eel23z5h4
A Survey of Learning-based Automated Program Repair
[article]
2023
arXiv
pre-print
We highlight several practical guidelines on applying DL techniques for future APR studies, such as exploring explainable patch generation and utilizing code features. ...
We illustrate the general workflow of learning-based APR techniques and detail the crucial components, including fault localization, patch generation, patch ranking, patch validation, and patch correctness ...
ACKNOWLEDGMENTS The authors would like to thank the anonymous reviewers for their insightful comments. ...
arXiv:2301.03270v3
fatcat:dm3hgnvj2bhe5nmlxfxt3epjdm
Automated Description Generation for Software Patches
[article]
2024
arXiv
pre-print
Patch descriptions provide detailed accounts of changes, aiding comprehension and collaboration among developers. ...
Moreover, the translation model in PATCHEXPLAINER is designed with an awareness of description similarity. ...
description
Semantic Context Extraction For a patch, code changes are introduced to fix a bug. ...
arXiv:2402.03805v1
fatcat:kliz3ie53fc33iq26bxjrdfane
Attention: Not Just Another Dataset for Patch-Correctness Checking
[article]
2023
arXiv
pre-print
To date, various Patch-Correctness Checking (PCC) techniques have been proposed to address this important issue. ...
However, they are only evaluated on very limited datasets as the APR tools used for generating such patches can only explore a small subset of the search space of possible patches, posing serious threats ...
AVR is a better metric for simulating actual efforts of selecting one correct patch from plausible patches. ...
arXiv:2207.06590v2
fatcat:vxiz53qezrhtbbaytg25tbwaqy
Vision Transformer Inspired Automated Vulnerability Repair
2024
ACM Transactions on Software Engineering and Methodology
In addition, we incorporate our VM into encoders' self-attention to learn embeddings that emphasize the vulnerable areas of a program. ...
In this article, we propose a novel vulnerability repair framework inspired by the Vision Transformer based approaches for object detection in the computer vision domain. ...
code areas in a vulnerable function for producing better repairs. ...
doi:10.1145/3632746
fatcat:q3cksaseybf4xknq34qm5no2d4
A Syntax-Guided Edit Decoder for Neural Program Repair
[article]
2022
arXiv
pre-print
architecture to ensure the syntactic correctness of the patched program and accurate generation; 3) Recoder generates placeholders that could be instantiated as project-specific identifiers later. ...
This result suggests that Recoder has better generalizability than existing APR approaches. ...
Threats to internal validity. mainly lie in our manual assessment of patch correctness. ...
arXiv:2106.08253v6
fatcat:afjebribknb6vcqjb5k6fhhh7u
Neuron-level LLM Patching for Code Generation
[article]
2024
arXiv
pre-print
It can correct a neural model by patching 1 or 2 neurons. As the pioneer work on neuron-level model editing of generative models, we formalize the editing process and introduce the involved concepts. ...
In this paper, we propose a novel and effective model editing approach, MENT, to patch LLMs in coding tasks. MENT is effective, efficient, and reliable. ...
For example, Code LLMs are not version-aware and may require updates to handle breaking changes of dependencies [12] . ...
arXiv:2312.05356v3
fatcat:ofkg5t6xmbahxe3yivqbu2l32i
ContrastRepair: Enhancing Conversation-Based Automated Program Repair via Contrastive Test Case Pairs
[article]
2024
arXiv
pre-print
Automated Program Repair (APR) aims to automatically generate patches for rectifying software bugs. ...
Our key insight is to minimize the difference between the generated passing test and the given failing test, which can better isolate the root causes of bugs. ...
Consequently, the two aforementioned metrics used to assess code similarity are not applicable to our context. ...
arXiv:2403.01971v2
fatcat:hvgdgfol6va7ph5mlo7t2x4ewa
Neural Program Repair with Execution-based Backpropagation
[article]
2021
arXiv
pre-print
Neural machine translation (NMT) architectures have achieved promising results for automatic program repair. ...
Yet, they have the limitation of generating low-quality patches(e.g., not compilable patches). ...
ACKNOWLEDGMENTS We thank the anonymous reviewers for the insightful feedback. ...
arXiv:2105.04123v1
fatcat:fbvylcfwgra7lnrv55iojum654
Automated patch assessment for program repair at scale
2021
Empirical Software Engineering
AbstractIn this paper, we do automatic correctness assessment for patches generated by program repair systems. ...
We build a curated dataset of 638 patches for Defects4J generated by 14 state-of-the-art repair systems, we evaluate automated patch assessment on this dataset. ...
For example, in the context of Java code, Yu et al. (2018) studied the overfitting on Defects4J. ...
doi:10.1007/s10664-020-09920-w
fatcat:zic7uge77fgxthma3xrz2ht6ti
Learning Transformation-Aware Embeddings for Image Forensics
[article]
2020
arXiv
pre-print
Further experimentation validates the proposed approach in the context of image provenance analysis. ...
Our approach learns transformation-aware descriptors using weak supervision via composited transformations and a rank-based quadruplet loss. ...
Code for the full pipeline to perform provenance graph construction using transformation-aware learned embeddings will be released to the community, upon the acceptance of this paper for publication. ...
arXiv:2001.04547v1
fatcat:4vz2jdn62bfcze5w3qagdpytra
« Previous
Showing results 1 — 15 out of 12,880 results