Publications | Le Chen

For a complete and up-to-date list, see my Google Scholar profile.

2026

[Preprint] ParaCodex: A Profiling-Guided Autonomous Coding Agent for Reliable Parallel Code Generation and Translation

Erel Kaplan, Tomer Bitan, Lian Ghrayeb, Le Chen, Tom Yotam, Niranjan Hasabnis, and Gal Oren

CoRR, 2026

@article{DBLP:journals/corr/abs-2601-04327,
  author = {Kaplan, Erel and Bitan, Tomer and Ghrayeb, Lian and Chen, Le and Yotam, Tom and Hasabnis, Niranjan and Oren, Gal},
  title = {ParaCodex: {A} Profiling-Guided Autonomous Coding Agent for Reliable
                    Parallel Code Generation and Translation},
  journal = {CoRR},
  volume = {abs/2601.04327},
  year = {2026},
  url = {https://doi.org/10.48550/arXiv.2601.04327},
  doi = {10.48550/ARXIV.2601.04327},
  eprinttype = {arXiv},
  eprint = {2601.04327},
  timestamp = {Wed, 11 Feb 2026 16:57:09 +0100},
  biburl = {https://dblp.org/rec/journals/corr/abs-2601-04327.bib},
  bibsource = {dblp computer science bibliography, https://dblp.org}
}

[Preprint] Latent Reasoning Guidance for Parallel Code Translation

Tomer Bitan, Erel Kaplan, Ron Bar-Yadin, Lian Ghrayeb, Le Chen, Shrey Jhaveri, and others

arXiv preprint arXiv:2606.05518, 2026

Bib

@article{bitan2026latent,
  title = {Latent Reasoning Guidance for Parallel Code Translation},
  author = {Bitan, Tomer and Kaplan, Erel and Bar-Yadin, Ron and Ghrayeb, Lian and Chen, Le and Jhaveri, Shrey and others},
  journal = {arXiv preprint arXiv:2606.05518},
  year = {2026},
  url = {https://arxiv.org/abs/2606.05518}
}

[Preprint] ParBench: A Benchmark for Reliable Evaluation of LLM Parallel Code Translation

Tom Yotam, Le Chen, Tomer Bitan, Niranjan Hasabnis, Gal Oren, and others

ResearchGate preprint, 2026

Bib

@article{chen2026parbench,
  title = {ParBench: A Benchmark for Reliable Evaluation of LLM Parallel Code Translation},
  author = {Yotam, Tom and Chen, Le and Bitan, Tomer and Hasabnis, Niranjan and Oren, Gal and others},
  journal = {ResearchGate preprint},
  year = {2026},
  url = {https://www.researchgate.net/publication/405207787_ParBench_A_Benchmark_for_Reliable_Evaluation_of_LLM_Parallel_Code_Translation}
}

[PEARC’26] Improving Usability and Productivity of PETSc with Agent-Based Workflows

Barry Smith, Hong Zhang, Junchao Zhang, Satish Balay, Le Chen, and Murat Keçeli

In Practice and Experience in Advanced Research Computing (PEARC) 2026, 2026

Bib

@inproceedings{smith2026petsc,
  title = {Improving Usability and Productivity of PETSc with Agent-Based Workflows},
  author = {Smith, Barry and Zhang, Hong and Zhang, Junchao and Balay, Satish and Chen, Le and Ke{\c{c}}eli, Murat},
  booktitle = {Practice and Experience in Advanced Research Computing (PEARC) 2026},
  year = {2026},
  address = {Minneapolis, MN, USA},
  url = {https://www.researchgate.net/publication/403500465_Improving_Usability_and_Productivity_of_PETSc_with_Agent-Based_Workflows}
}

[Preprint] An Agentic Evaluation Framework for AI-Generated Scientific Code in PETSc

Hong Zhang, Barry Smith, Satish Balay, Le Chen, Murat Keçeli, Lois Curfman McInnes, and Junchao Zhang

arXiv preprint arXiv:2603.15976, 2026

Bib

@article{zhang2026agentic,
  title = {An Agentic Evaluation Framework for AI-Generated Scientific Code in PETSc},
  author = {Zhang, Hong and Smith, Barry and Balay, Satish and Chen, Le and Ke{\c{c}}eli, Murat and McInnes, Lois Curfman and Zhang, Junchao},
  journal = {arXiv preprint arXiv:2603.15976},
  year = {2026},
  url = {https://arxiv.org/abs/2603.15976}
}

2025

[NAACL’25] AutoParLLM: GNN-guided Context Generation for Zero-Shot Code Parallelization using LLMs

Quazi Ishtiaque Mahmud, Ali TehraniJamsaz, Hung D Phan, Le Chen, Mihai Capotă, Theodore L. Willke, Nesreen K. Ahmed, and Ali Jannesari

In Proceedings of the 2025 Conference of the Nations of the Americas Chapter of the Association for Computational Linguistics: Human Language Technologies (Volume 1: Long Papers), Apr 2025

Abs DOI Bib

In-Context Learning (ICL) has been shown to be a powerful technique to augment the capabilities of LLMs for a diverse range of tasks. This work proposes AutoParLLM, a novel way to generate context using guidance from graph neural networks (GNNs) to generate efficient parallel codes. We evaluate AutoParLLM on 12 applications from two well-known benchmark suites of parallel codes: NAS Parallel Benchmark and Rodinia Benchmark. Our results show that AutoParLLM improves the state-of-the-art LLMs (e.g., GPT-4) by 19.9% in NAS and 6.48% in Rodinia benchmark in terms of CodeBERTScore for the task of parallel code generation. Moreover, AutoParLLM improves the ability of the most powerful LLM to date, GPT-4, by achieving 17% (on NAS benchmark) and 16% (on Rodinia benchmark) better speedup. In addition, we propose OMPScore for evaluating the quality of the parallel code and show its effectiveness in evaluating parallel codes.
@inproceedings{mahmud2025autoparllm, title = {{A}uto{P}ar{LLM}: {GNN}-guided Context Generation for Zero-Shot Code Parallelization using {LLM}s}, author = {Mahmud, Quazi Ishtiaque and TehraniJamsaz, Ali and Phan, Hung D and Chen, Le and Capot{\u{a}}, Mihai and Willke, Theodore L. and Ahmed, Nesreen K. and Jannesari, Ali}, editor = {Chiruzzo, Luis and Ritter, Alan and Wang, Lu}, booktitle = {Proceedings of the 2025 Conference of the Nations of the Americas Chapter of the Association for Computational Linguistics: Human Language Technologies (Volume 1: Long Papers)}, month = apr, year = {2025}, address = {Albuquerque, New Mexico}, publisher = {Association for Computational Linguistics}, url = {https://aclanthology.org/2025.naacl-long.593/}, doi = {10.18653/v1/2025.naacl-long.593}, pages = {11821--11841}, isbn = {979-8-89176-189-6} }

[IPDPS’25] PCEBench: A Multi-Dimensional Benchmark for Evaluating Large Language Models in Parallel Code Generation

Le Chen, Nesreen Ahmed, Mihai Capotă, Ted Willke, Niranjan Hasabnis, and Ali Jannesari

In 2025 IEEE International Parallel and Distributed Processing Symposium (IPDPS), Apr 2025

DOI Bib

@inproceedings{chen2025pcebench,
  author = {Chen, Le and Ahmed, Nesreen and Capotă, Mihai and Willke, Ted and Hasabnis, Niranjan and Jannesari, Ali},
  booktitle = {2025 IEEE International Parallel and Distributed Processing Symposium (IPDPS)},
  title = {PCEBench: A Multi-Dimensional Benchmark for Evaluating Large Language Models in Parallel Code Generation},
  year = {2025},
  volume = {},
  number = {},
  pages = {546-557},
  keywords = {Technological innovation;Codes;Parallel programming;Large language models;Scalability;Benchmark testing;Software systems;Multitasking;Natural language processing;Synchronization;large language model;parallel code generation;benchmark;evaluation;LLM agent},
  doi = {10.1109/IPDPS64566.2025.00055}
}

[SC-W’25] FIRST: Federated Inference Resource Scheduling Toolkit for Scientific AI Model Access

Aditya Tanikanti, Benoı̂t Côté, Yanfei Guo, Le Chen, Nicholaus Saint, Ryan Chard, Ken Raffenetti, Rajeev Thakur, Thomas D. Uram, Ian T. Foster, Michael E. Papka, and Venkatram Vishwanath

In Proceedings of the SC ’25 Workshops of the International Conference for High Performance Computing, Networking, Storage and Analysis, SC Workshops 2025, St Louis, MO, USA, November 16-21, 2025, Apr 2025

DOI Bib

@inproceedings{Aditya2025first,
  author = {Tanikanti, Aditya and C{\{o}}t{\'{e}}, Beno{\{\i}}t and Guo, Yanfei and Chen, Le and Saint, Nicholaus and Chard, Ryan and Raffenetti, Ken and Thakur, Rajeev and Uram, Thomas D. and Foster, Ian T. and Papka, Michael E. and Vishwanath, Venkatram},
  title = {{FIRST:} Federated Inference Resource Scheduling Toolkit for Scientific
                    {AI} Model Access},
  booktitle = {Proceedings of the {SC} '25 Workshops of the International Conference
                    for High Performance Computing, Networking, Storage and Analysis,
                    {SC} Workshops 2025, St Louis, MO, USA, November 16-21, 2025},
  pages = {52--60},
  publisher = {{ACM}},
  year = {2025},
  url = {https://doi.org/10.1145/3731599.3767346},
  doi = {10.1145/3731599.3767346},
  timestamp = {Fri, 26 Dec 2025 20:53:24 +0100},
  biburl = {https://dblp.org/rec/conf/sc/TanikantiCGCSCR25.bib},
  bibsource = {dblp computer science bibliography, https://dblp.org}
}

[HPEC’25] UniPar: A Unified LLM-Based Framework for Parallel and Accelerated Code Translation in HPC

Tomer Bitan, Tal Kadosh, Erel Kaplan, Shira Meiri, Le Chen, Peter Morales, Niranjan Hasabnis, and Gal Oren

In IEEE High Performance Extreme Computing Conference, HPEC 2025, Wakefield, MA, USA, September 15-19, 2025, Apr 2025

DOI Bib

@inproceedings{bitan2025unipar,
  author = {Bitan, Tomer and Kadosh, Tal and Kaplan, Erel and Meiri, Shira and Chen, Le and Morales, Peter and Hasabnis, Niranjan and Oren, Gal},
  title = {UniPar: {A} Unified LLM-Based Framework for Parallel and Accelerated
                    Code Translation in {HPC}},
  booktitle = {{IEEE} High Performance Extreme Computing Conference, {HPEC} 2025,
                    Wakefield, MA, USA, September 15-19, 2025},
  pages = {1--9},
  publisher = {{IEEE}},
  year = {2025},
  url = {https://doi.org/10.1109/HPEC67600.2025.11196677},
  doi = {10.1109/HPEC67600.2025.11196677},
  timestamp = {Sat, 08 Nov 2025 10:18:23 +0100},
  biburl = {https://dblp.org/rec/conf/hpec/BitanKKMCMHO25.bib},
  bibsource = {dblp computer science bibliography, https://dblp.org}
}

[ICPP-W’25] AI Assistants to Enhance and Exploit the PETSc Knowledge Base

Barry Smith, Junchao Zhang, Hong Zhang, Lois Curfman McInnes, Murat Keçeli, Archit Vasan, Satish Balay, Toby Isaac, Le Chen, and Venkatram Vishwanath

Apr 2025

DOI Bib

@article{smith2025ai,
  author = {Smith, Barry and Zhang, Junchao and Zhang, Hong and McInnes, Lois Curfman and Ke{\c{c}}eli, Murat and Vasan, Archit and Balay, Satish and Isaac, Toby and Chen, Le and Vishwanath, Venkatram},
  title = {{AI} Assistants to Enhance and Exploit the PETSc Knowledge Base},
  booktitle = {Workshop Proceedings of the 54th International Conference on Parallel
                    Processing, {ICPP} Workshops 2025, San Diego, CA, USA, September 8-11,
                    2025},
  pages = {9--17},
  publisher = {{ACM}},
  year = {2025},
  url = {https://doi.org/10.1145/3750720.3757281},
  doi = {10.1145/3750720.3757281},
  timestamp = {Thu, 05 Mar 2026 17:17:13 +0100},
  biburl = {https://dblp.org/rec/conf/icppw/000200MKVBICV25.bib},
  bibsource = {dblp computer science bibliography, https://dblp.org}
}

[SC-W’25] Evaluation of Test-Time Compute Constraints on Safety and Skill Large Reasoning Models

Adarsha Balaji, Le Chen, Rajeev Thakur, Franck Cappello, and Sandeep Madireddy

In Proceedings of the SC ’25 Workshops of the International Conference for High Performance Computing, Networking, Storage and Analysis, , Apr 2025

Abs DOI Bib

Test-time compute scaling has demonstrated the ability to improve the performance of reasoning language models by generating longer chain-of-thought (CoT) sequences. However, this increase in performance comes with a significant increase in computation cost. In this work, we investigate two compute constraint strategies: (1) reasoning length constraint and (2) model quantization, and study their impact on the safety performance of reasoning models. Specifically, we explore two approaches to apply compute constraints to reasoning models: (1) fine-tuning reasoning models using a length-controlled policy optimization (LCPO) based reinforcement learning method to satisfy a user-defined CoT reasoning length, and (2) applying quantization to maximize the generation of CoT sequences within a user-defined compute constraint. Furthermore, we study the trade-off between the computational efficiency and the safety of the model.
@inproceedings{balaji2025evaluation, author = {Balaji, Adarsha and Chen, Le and Thakur, Rajeev and Cappello, Franck and Madireddy, Sandeep}, title = {Evaluation of Test-Time Compute Constraints on Safety and Skill Large Reasoning Models}, year = {2025}, isbn = {9798400718717}, publisher = {Association for Computing Machinery}, address = {New York, NY, USA}, url = {https://doi.org/10.1145/3731599.3767406}, doi = {10.1145/3731599.3767406}, booktitle = {Proceedings of the SC '25 Workshops of the International Conference for High Performance Computing, Networking, Storage and Analysis}, pages = {534–539}, numpages = {6}, keywords = {Large Language Models, Foundation Models, Reasoning, Safety Analysis, Artificial Intelligence}, location = { }, series = {SC Workshops '25} }

[Preprint] Beyond Code Pairs: Dialogue-Based Data Generation for LLM Code Translation

Le Chen, Nan Xu, Wenhao Chen, Bin Lei, Pei-Hung Lin, Dunzhi Zhou, Rajeev Thakur, Caiwen Ding, and others

arXiv preprint arXiv:2512.03086, Apr 2025

Bib

@article{chen2025beyond,
  title = {Beyond Code Pairs: Dialogue-Based Data Generation for LLM Code Translation},
  author = {Chen, Le and Xu, Nan and Chen, Wenhao and Lei, Bin and Lin, Pei-Hung and Zhou, Dunzhi and Thakur, Rajeev and Ding, Caiwen and others},
  journal = {arXiv preprint arXiv:2512.03086},
  year = {2025},
  url = {https://arxiv.org/abs/2512.03086}
}

[Preprint] Who Gets the Reward, Who Gets the Blame? Evaluation-Aligned Training Signals for Multi-LLM Agents

Chih-Hsuan Yang, Tanwi Mallick, Le Chen, Krishnan Raghavan, Aaron Wells, Amal Gueroudji, Ian T. Foster, and others

arXiv preprint arXiv:2511.10687, Apr 2025

Bib

@article{yang2025reward,
  title = {Who Gets the Reward, Who Gets the Blame? Evaluation-Aligned Training Signals for Multi-LLM Agents},
  author = {Yang, Chih-Hsuan and Mallick, Tanwi and Chen, Le and Raghavan, Krishnan and Wells, Aaron and Gueroudji, Amal and Foster, Ian T. and others},
  journal = {arXiv preprint arXiv:2511.10687},
  year = {2025},
  url = {https://arxiv.org/abs/2511.10687}
}

[Preprint] OMPILOT: Harnessing Transformer Models for Auto Parallelization to Shared Memory Computing Paradigms

Arijit Bhattacharjee, Ali TehraniJamsaz, Le Chen, Niranjan Hasabnis, Mihai Capotă, and others

arXiv preprint arXiv:2511.03866, Apr 2025

Bib

@article{bhattacharjee2025ompilot,
  title = {OMPILOT: Harnessing Transformer Models for Auto Parallelization to Shared Memory Computing Paradigms},
  author = {Bhattacharjee, Arijit and TehraniJamsaz, Ali and Chen, Le and Hasabnis, Niranjan and Capot{\u{a}}, Mihai and others},
  journal = {arXiv preprint arXiv:2511.03866},
  year = {2025},
  url = {https://arxiv.org/abs/2511.03866}
}

2024

[Preprint] The landscape and challenges of HPC research and LLMs

Le Chen, Nesreen K Ahmed, Akash Dutta, Arijit Bhattacharjee, Sixing Yu, Quazi Ishtiaque Mahmud, Waqwoya Abebe, Hung Phan, Aishwarya Sarkar, Branden Butler, and others

arXiv preprint arXiv:2402.02018, Apr 2024

Bib

@article{chen2024landscape,
  url = {https://arxiv.org/abs/2402.02018},
  title = {The landscape and challenges of HPC research and LLMs},
  author = {Chen, Le and Ahmed, Nesreen K and Dutta, Akash and Bhattacharjee, Arijit and Yu, Sixing and Mahmud, Quazi Ishtiaque and Abebe, Waqwoya and Phan, Hung and Sarkar, Aishwarya and Butler, Branden and others},
  journal = {arXiv preprint arXiv:2402.02018},
  year = {2024}
}

[EuroPar’24] OMPGPT: A Generative Pre-trained Transformer Model for OpenMP

Le Chen, Arijit Bhattacharjee, Nesreen Ahmed, Niranjan Hasabnis, Gal Oren, Vy Vo, and Ali Jannesari

In Euro-Par 2024: Parallel Processing: 30th European Conference on Parallel and Distributed Processing, Madrid, Spain, August 26–30, 2024, Proceedings, Part I, Madrid, Spain, Apr 2024

Abs DOI Bib

Large language models (LLMs)such as ChatGPT have significantly advanced the field of Natural Language Processing (NLP). This trend led to the development of code-based large language models such as StarCoder, WizardCoder, and CodeLlama, which are trained extensively on vast repositories of code and programming languages. While the generic abilities of these code LLMs are helpful for many programmers in tasks like code generation, the area of high-performance computing (HPC) has a narrower set of requirements that make a smaller and more domain-specific model a smarter choice. This paper presents OMPGPT, a novel domain-specific model meticulously designed to harness the inherent strengths of language models for OpenMP pragma generation. Furthermore, we leverage prompt engineering techniques from the NLP domain to create Chain-of-OMP, an innovative strategy designed to enhance OMPGPT’s effectiveness. Our extensive evaluations demonstrate that OMPGPT outperforms existing large language models specialized in OpenMP tasks and maintains a notably smaller size, aligning it more closely with the typical hardware constraints of HPC environments. We consider our contribution as a pivotal bridge, connecting the advantage of language models with the specific demands of HPC tasks.
@inproceedings{chen2024ompgpt, author = {Chen, Le and Bhattacharjee, Arijit and Ahmed, Nesreen and Hasabnis, Niranjan and Oren, Gal and Vo, Vy and Jannesari, Ali}, title = {OMPGPT: A Generative Pre-trained Transformer Model for OpenMP}, year = {2024}, isbn = {978-3-031-69576-6}, publisher = {Springer-Verlag}, address = {Berlin, Heidelberg}, url = {https://doi.org/10.1007/978-3-031-69577-3_9}, doi = {10.1007/978-3-031-69577-3_9}, booktitle = {Euro-Par 2024: Parallel Processing: 30th European Conference on Parallel and Distributed Processing, Madrid, Spain, August 26–30, 2024, Proceedings, Part I}, pages = {121–134}, numpages = {14}, keywords = {Large Language model, OpenMP, HPC}, location = {Madrid, Spain} }

[NeurIPS’24] CodeRosetta: Pushing the Boundaries of Unsupervised Code Translation for Parallel Programming

Ali TehraniJamsaz, Arijit Bhattacharjee, Le Chen, Nesreen K. Ahmed, Amir Yazdanbakhsh, and Ali Jannesari

In Advances in Neural Information Processing Systems 38: Annual Conference on Neural Information Processing Systems 2024, NeurIPS 2024, Vancouver, BC, Canada, December 10 - 15, 2024, Apr 2024

Bib

@inproceedings{tehrani2024coderosetta,
  author = {TehraniJamsaz, Ali and Bhattacharjee, Arijit and Chen, Le and Ahmed, Nesreen K. and Yazdanbakhsh, Amir and Jannesari, Ali},
  editor = {Globersons, Amir and Mackey, Lester and Belgrave, Danielle and Fan, Angela and Paquet, Ulrich and Tomczak, Jakub M. and Zhang, Cheng},
  title = {CodeRosetta: Pushing the Boundaries of Unsupervised Code Translation
                    for Parallel Programming},
  booktitle = {Advances in Neural Information Processing Systems 38: Annual Conference
                    on Neural Information Processing Systems 2024, NeurIPS 2024, Vancouver,
                    BC, Canada, December 10 - 15, 2024},
  year = {2024},
  url = {http://papers.nips.cc/paper\_files/paper/2024/hash/b6edb87876bec4ac2260bffa083cb992-Abstract-Conference.html},
  timestamp = {Thu, 13 Feb 2025 16:56:44 +0100},
  biburl = {https://dblp.org/rec/conf/nips/TehraniJamsazBC24.bib},
  bibsource = {dblp computer science bibliography, https://dblp.org}
}

[Preprint] Fortran2cpp: Automating fortran-to-c++ migration using llms via multi-turn dialogue and dual-agent integration

Le Chen, Bin Lei, Dunzhi Zhou, Pei-Hung Lin, Chunhua Liao, Caiwen Ding, and Ali Jannesari

arXiv e-prints, Apr 2024

Bib

@article{chen2024fortran2cpp,
  url = {https://arxiv.org/abs/2412.19770},
  title = {Fortran2cpp: Automating fortran-to-c++ migration using llms via multi-turn dialogue and dual-agent integration},
  author = {Chen, Le and Lei, Bin and Zhou, Dunzhi and Lin, Pei-Hung and Liao, Chunhua and Ding, Caiwen and Jannesari, Ali},
  journal = {arXiv e-prints},
  pages = {arXiv--2412},
  year = {2024}
}

2023

[SC-W’23] HPC-GPT: Integrating Large Language Model for High-Performance Computing

Xianzhong Ding, Le Chen, Murali Emani, Chunhua Liao, Pei-Hung Lin, Tristan Vanderbruggen, Zhen Xie, Alberto Cerpa, and Wan Du

In Proceedings of the SC ’23 Workshops of the International Conference on High Performance Computing, Network, Storage, and Analysis, Denver, CO, USA, Apr 2023

Abs DOI Bib

Large Language Models (LLMs), including the LLaMA model, have exhibited their efficacy across various general-domain natural language processing (NLP) tasks. However, their performance in high-performance computing (HPC) domain tasks has been less than optimal due to the specialized expertise required to interpret the model’s responses. In response to this challenge, we propose HPC-GPT, a novel LLaMA-based model that has been supervised fine-tuning using generated QA (Question-Answer) instances for the HPC domain. To evaluate its effectiveness, we concentrate on two HPC tasks: managing AI models and datasets for HPC, and data race detection. By employing HPC-GPT, we demonstrate comparable performance with existing methods on both tasks, exemplifying its excellence in HPC-related scenarios. Our experiments on open-source benchmarks yield extensive results, underscoring HPC-GPT’s potential to bridge the performance gap between LLMs and HPC-specific tasks. With HPC-GPT, we aim to pave the way for LLMs to excel in HPC domains, simplifying the utilization of language models in complex computing applications.
@inproceedings{ding2023hpc, author = {Ding, Xianzhong and Chen, Le and Emani, Murali and Liao, Chunhua and Lin, Pei-Hung and Vanderbruggen, Tristan and Xie, Zhen and Cerpa, Alberto and Du, Wan}, title = {HPC-GPT: Integrating Large Language Model for High-Performance Computing}, year = {2023}, isbn = {9798400707858}, publisher = {Association for Computing Machinery}, address = {New York, NY, USA}, url = {https://doi.org/10.1145/3624062.3624172}, doi = {10.1145/3624062.3624172}, booktitle = {Proceedings of the SC '23 Workshops of the International Conference on High Performance Computing, Network, Storage, and Analysis}, pages = {951–960}, numpages = {10}, keywords = {Data Race Detection, High-performance Computing, Large Language Model, Neural Network., OpenMP}, location = {Denver, CO, USA}, series = {SC-W '23} }
[IWOMP’23] LM4HPC: Towards Effective Language Model Application in High-Performance Computing

Le Chen, Pei-Hung Lin, Tristan Vanderbruggen, Chunhua Liao, Murali Emani, and Bronis Supinski

In OpenMP: Advanced Task-Based, Device and Compiler Programming: 19th International Workshop on OpenMP, IWOMP 2023, Bristol, UK, September 13–15, 2023, Proceedings, Bristol, United Kingdom, Apr 2023

Abs DOI Bib

In recent years, language models (LMs), such as GPT-4, have been widely used in multiple domains, including natural language processing, visualization, and so on. However, applying them for analyzing and optimizing high-performance computing (HPC) software is still challenging due to the lack of HPC-specific support. In this paper, we design the LM4HPC framework to facilitate the research and development of HPC software analyses and optimizations using LMs. Tailored for supporting HPC datasets, AI models, and pipelines, our framework is built on top of a range of components from different levels of the machine learning software stack, with Hugging Face-compatible APIs. Using three representative tasks, we evaluated the prototype of our framework. The results show that LM4HPC can help users quickly evaluate a set of state-of-the-art models and generate insightful leaderboards.
@inproceedings{chen2023lm4hpc, author = {Chen, Le and Lin, Pei-Hung and Vanderbruggen, Tristan and Liao, Chunhua and Emani, Murali and de Supinski, Bronis}, title = {LM4HPC: Towards Effective Language Model Application in High-Performance Computing}, year = {2023}, isbn = {978-3-031-40743-7}, publisher = {Springer-Verlag}, address = {Berlin, Heidelberg}, url = {https://doi.org/10.1007/978-3-031-40744-4_2}, doi = {10.1007/978-3-031-40744-4_2}, booktitle = {OpenMP: Advanced Task-Based, Device and Compiler Programming: 19th International Workshop on OpenMP, IWOMP 2023, Bristol, UK, September 13–15, 2023, Proceedings}, pages = {18–33}, numpages = {16}, keywords = {High-performance computing, Programming language processing, Language model}, location = {Bristol, United Kingdom} }
[SC-W’23] Data Race Detection Using Large Language Models

Le Chen, Xianzhong Ding, Murali Emani, Tristan Vanderbruggen, Pei-Hung Lin, and Chunhua Liao

In Proceedings of the SC ’23 Workshops of the International Conference on High Performance Computing, Network, Storage, and Analysis, Denver, CO, USA, Apr 2023

Abs DOI Bib

Large language models (LLMs) are demonstrating significant promise as an alternate strategy to facilitate analyses and optimizations of high-performance computing programs, circumventing the need for resource-intensive manual tool creation. In this paper, we explore a novel LLM-based data race detection approach combining prompting engineering and fine-tuning techniques. We create a dedicated dataset named DRB-ML, which is derived from DataRaceBench, with fine-grain labels showing the presence of data race pairs and their associated variables, line numbers, and read/write information. DRB-ML is then used to evaluate representative LLMs and fine-tune open-source ones. Our experiment shows that LLMs can be a viable approach to data race detection. However, they still cannot compete with traditional data race detection tools when we need detailed information about variable pairs causing data races.
@inproceedings{chen2023data, author = {Chen, Le and Ding, Xianzhong and Emani, Murali and Vanderbruggen, Tristan and Lin, Pei-Hung and Liao, Chunhua}, title = {Data Race Detection Using Large Language Models}, year = {2023}, isbn = {9798400707858}, publisher = {Association for Computing Machinery}, address = {New York, NY, USA}, url = {https://doi.org/10.1145/3624062.3624088}, doi = {10.1145/3624062.3624088}, booktitle = {Proceedings of the SC '23 Workshops of the International Conference on High Performance Computing, Network, Storage, and Analysis}, pages = {215–223}, numpages = {9}, keywords = {OpenMP, data race detection, large language model}, location = {Denver, CO, USA}, series = {SC-W '23} }

[MLSys’23] Learning to Parallelize with OpenMP by Augmented Heterogeneous AST Representation

Le Chen, Quazi Ishtiaque Mahmud, Hung Phan, Nesreen K. Ahmed, and Ali Jannesari

In Proceedings of the Sixth Conference on Machine Learning and Systems, MLSys 2023, Miami, FL, USA, June 4-8, 2023, Apr 2023

Bib

@inproceedings{chen2023learning,
  author = {Chen, Le and Mahmud, Quazi Ishtiaque and Phan, Hung and Ahmed, Nesreen K. and Jannesari, Ali},
  editor = {Song, Dawn and Carbin, Michael and Chen, Tianqi},
  title = {Learning to Parallelize with OpenMP by Augmented Heterogeneous {AST}
                    Representation},
  booktitle = {Proceedings of the Sixth Conference on Machine Learning and Systems,
                    MLSys 2023, Miami, FL, USA, June 4-8, 2023},
  publisher = {mlsys.org},
  year = {2023},
  url = {https://proceedings.mlsys.org/paper\_files/paper/2023/hash/8ee477d6175a03d7098fa23641a2d298-Abstract-mlsys2023.html},
  timestamp = {Fri, 28 Jun 2024 15:58:54 +0200},
  biburl = {https://dblp.org/rec/conf/mlsys/ChenMPAJ23.bib},
  bibsource = {dblp computer science bibliography, https://dblp.org}
}

[NeurIPS’23] Perfograph: A numerical aware program graph representation for performance optimization and program analysis

Ali TehraniJamsaz, Quazi Ishtiaque Mahmud, Le Chen, Nesreen K Ahmed, and Ali Jannesari

Advances in Neural Information Processing Systems, Apr 2023

Bib

@article{tehranijamsaz2023perfograph,
  title = {Perfograph: A numerical aware program graph representation for performance optimization and program analysis},
  author = {TehraniJamsaz, Ali and Mahmud, Quazi Ishtiaque and Chen, Le and Ahmed, Nesreen K and Jannesari, Ali},
  journal = {Advances in Neural Information Processing Systems},
  volume = {36},
  pages = {57783--57794},
  year = {2023}
}

[HPEC’23] Creating a dataset for high-performance computing code translation using LLMS: a bridge between OpenMP Fortran and C++

Bin Lei, Caiwen Ding, Le Chen, Pei-Hung Lin, and Chunhua Liao

In 2023 IEEE High Performance Extreme Computing Conference (HPEC), Apr 2023

Bib

@inproceedings{lei2023creating,
  title = {Creating a dataset for high-performance computing code translation using LLMS: a bridge between OpenMP Fortran and C++},
  author = {Lei, Bin and Ding, Caiwen and Chen, Le and Lin, Pei-Hung and Liao, Chunhua},
  booktitle = {2023 IEEE High Performance Extreme Computing Conference (HPEC)},
  pages = {1--7},
  year = {2023},
  organization = {IEEE}
}

[Preprint] Compcodevet: A compiler-guided validation and enhancement approach for code dataset

Le Chen, Arijit Bhattacharjee, Nesreen K Ahmed, Niranjan Hasabnis, Gal Oren, Bin Lei, and Ali Jannesari

arXiv preprint arXiv:2311.06505, Apr 2023

Bib

@article{chen2023compcodevet,
  url = {https://arxiv.org/abs/2311.06505},
  title = {Compcodevet: A compiler-guided validation and enhancement approach for code dataset},
  author = {Chen, Le and Bhattacharjee, Arijit and Ahmed, Nesreen K and Hasabnis, Niranjan and Oren, Gal and Lei, Bin and Jannesari, Ali},
  journal = {arXiv preprint arXiv:2311.06505},
  year = {2023}
}

[Preprint] Flashvideo: A framework for swift inference in text-to-video generation

Bin Lei, Caiwen Ding, and others

arXiv preprint arXiv:2401.00869, Apr 2023

Bib

@article{lei2023flashvideo,
  url = {https://arxiv.org/abs/2401.00869},
  title = {Flashvideo: A framework for swift inference in text-to-video generation},
  author = {Lei, Bin and Ding, Caiwen and others},
  journal = {arXiv preprint arXiv:2401.00869},
  year = {2023}
}

[Preprint] Dataracebench v1. 4.1 and dataracebench-ml v0. 1: Benchmark suites for data race detection

Le Chen, Wenhao Wu, Stephen F Siegel, Pei-Hung Lin, and Chunhua Liao

arXiv preprint arXiv:2308.08473, Apr 2023

Bib

@article{chen2023dataracebench,
  url = {https://arxiv.org/abs/2308.08473},
  title = {Dataracebench v1. 4.1 and dataracebench-ml v0. 1: Benchmark suites for data race detection},
  author = {Chen, Le and Wu, Wenhao and Siegel, Stephen F and Lin, Pei-Hung and Liao, Chunhua},
  journal = {arXiv preprint arXiv:2308.08473},
  year = {2023}
}

2022

[IPDPSW’22] Multi-View Learning for Parallelism Discovery of Sequential Programs

Le Chen, Quazi Ishtiaque Mahmud, and Ali Jannesari

In 2022 IEEE International Parallel and Distributed Processing Symposium Workshops (IPDPSW), Apr 2022

DOI Bib

@inproceedings{chen2022multi,
  author = {Chen, Le and Mahmud, Quazi Ishtiaque and Jannesari, Ali},
  booktitle = {2022 IEEE International Parallel and Distributed Processing Symposium Workshops (IPDPSW)},
  title = {Multi-View Learning for Parallelism Discovery of Sequential Programs},
  year = {2022},
  volume = {},
  number = {},
  pages = {295-303},
  keywords = {Training;Runtime;Dynamics;Semantics;Machine learning;Static analysis;Parallel processing;machine learning;artificial intelligence;parallel program language},
  doi = {10.1109/IPDPSW55747.2022.00059}
}