I am Haoxiang Sun, an undergraduate student in Computer Science and Engineering at Sichuan University (2023.09 - expected 2027.06), currently with GPA 3.79/4.0.

Since 2024.12, I have been a research intern at DICALab (Data Intelligence and Computing Art Laboratory), supervised by Dr. Tao Wang. My current work focuses on:

• Multimodal Large Language Models (MLLMs / VLLMs)
• Transferring VLLMs to real-world applications
• Inference acceleration and efficient decoding

Updated on April 2026.

🔥 News

• 2026.04:  I will join Annu Zhineng (Chengdu) as an external on-site intern.
• 2026.03:  I will join Peking University (Shenzhen Graduate School) as a master’s student in Fall 2027.
• 2026.03:  From Structure to Synergy: A Survey of Vision-Language Perception Paradigm Evolution in Multimodal Large Language Models accepted by Information Fusion.
• 2026.02:  Dr. Seg: Revisiting GRPO Training for Visual Large Language Models through Perception-Oriented Design accepted by CVPR 2026.
• 2024.12:  Joined DICALab as a research intern.

🧩 Open Source Contributions

• 2026.01 - now, Contributor @ vLLM-speculators

💻 Internships

• 2024.12 - now, Research Intern, DICALab (Data Intelligence and Computing Art Laboratory), Chengdu, China.
  Supervised by Dr. Tao Wang; focusing on multimodal LLMs for visual perception.

• 2026.04 - now, Research Intern, Annu Zhineng, Chengdu, China.
  Focusing on deploying world action model.

🛠 Technical Skills

• Languages: Mandarin (native), English (fluent)
• Programming Languages: C/C++, Python
• Frameworks: PyTorch, vLLM, VeRL