Hailong Jiang

Hailong Jiang

he/him

Tenure Track Assistant Professor in Computer Science at Youngstown State University. Research focuses on HPC systems, compiler technologies, and the intersection of program analysis and large language models (LLMs).

New Research Content Added - Summary

New Research Content Added - Summary

Overview

I have successfully added your latest research on LLM Understanding of Intermediate Representations to your personal homepage. This complements your existing SC’24 Doctoral Showcase work and demonstrates the breadth of your research in AI-driven code intelligence.

What Has Been Added

1. New Publication Entry

  • File: _publications/2025-02-07-llm-ir-understanding.md
  • Title: “Can Large Language Models Understand Intermediate Representations?”
  • Venue: arXiv preprint arXiv:2502.06854
  • Date: February 7, 2025
  • Link: https://arxiv.org/abs/2502.06854

2. Updated Portfolio Project

  • File: _portfolio/portfolio-2.html
  • Title: “LLM Understanding of Intermediate Representations”
  • Focus: Pioneering empirical study of LLM capabilities in IR comprehension
  • Models Tested: GPT-4, GPT-3, Gemma 2, LLaMA 3.1, Code Llama
  • Tasks Evaluated: CFG reconstruction, decompilation, summarization, execution reasoning

3. Enhanced Homepage Content

  • New Section: “🔬 Latest Research: LLM Understanding of IRs”
  • Research Highlights: Four critical IR-related tasks
  • Direct Link: To your arXiv paper
  • Updated Skills: Added IR analysis to compiler technologies

4. Updated CV Page

  • New Publication Category: “Preprints” section
  • Research Experience: Added LLM IR understanding research
  • Skills Enhancement: IR analysis added to compiler technologies
  • Last Updated: February 2025

5. New Project Image

  • File: images/llm-ir-understanding.jpg
  • Purpose: LLM IR Understanding research project visualization
  • Status: Placeholder created (ready for your actual content)

Research Impact Highlighted

Key Contributions

  • Pioneering Study: First comprehensive evaluation of LLMs in IR understanding
  • Multi-Model Analysis: Testing across 5 major LLM architectures
  • Task-Specific Evaluation: Four critical compiler-level tasks
  • Practical Insights: Recommendations for IR-specific LLM enhancements

Technical Significance

  • Compiler Design: AI-assisted compiler development
  • Program Analysis: Automation of low-level code analysis
  • LLM Training: Specialized training for IR comprehension
  • Software Engineering: Integration of AI in development tools

Website Now Showcases

Complete Research Portfolio

  1. HAPPA Platform - HPC resilience analysis with LLM integration
  2. LLM IR Understanding - Compiler-level AI capabilities research
  3. SC’24 Doctoral Showcase - Premier conference presentation
  4. National Lab Experience - Argonne and Los Alamos collaborations

Research Breadth

  • High-Performance Computing: Resilience and fault tolerance
  • Artificial Intelligence: Large Language Models and deep learning
  • Compiler Technologies: LLVM, CUDA, OpenMP, IR analysis
  • System Programming: Low-level code understanding and optimization

What You Need to Do Next

1. Add Real Images

  • LLM IR Project: Replace images/llm-ir-understanding.jpg with actual research diagrams
  • HAPPA Platform: Replace images/hpc-resilience.jpg with platform screenshots
  • Profile Photo: Replace images/profile.jpg with your actual photo

2. Update Publications

  • Replace placeholder publications with your actual IEEE conference papers
  • Add your Springer journal articles
  • Include paper PDFs in the files/ directory

3. Customize Content

  • Update talks and teaching experience with your real activities
  • Modify project descriptions to match your actual implementations
  • Add any additional research projects or collaborations

Current Status

✅ New Research Added: LLM IR Understanding study
✅ Publication Updated: arXiv paper linked
✅ Portfolio Enhanced: Second major project documented
✅ Homepage Updated: Latest research prominently featured
✅ CV Enhanced: New publication category and research experience
✅ Images Ready: Placeholders created for your content

How to View Updates

Your enhanced website is now running at:

  • Local: http://localhost:4000
  • New Content: LLM IR Understanding research prominently featured
  • Updated Navigation: All new content accessible through existing menus

Your personal homepage now showcases a comprehensive research portfolio spanning HPC resilience, LLM integration, and compiler-level AI research! 🚀