Rongxiang Wang

About Me

I am a Ph.D. candidate in Computer Science at the University of Virginia, advised by Professor Felix Xiaozhu Lin, and I expect to graduate in May 2027. My research is on on-device machine learning systems: making LLMs and speech models (ASR and TTS) run efficiently and reliably on consumer GPUs and personal devices. Recent work includes WhisperFlow (MobiSys 2025) and Turbocharge (MobiCom 2024) for real-time speech understanding, SpQt for dynamic sparsity in quantized LLM inference, and LACI for reliable long-form text-to-speech.

I have interned at Argmax (Summer 2026, long-form TTS), ByteDance (Spring 2026, LLM memory systems), and Zoom (2025, on-device LLM inference). Before UVA, I received my B.Eng. in Automation from Tsinghua University.

News

• Sep 2026: Our Argmax paper, Taming Long-form Text-to-Speech, is on arXiv.
• May 2026: Started a summer internship at Argmax in Palo Alto, working on long-form TTS.
• Mar 2026: Started a research internship at ByteDance in San Jose, working on LLM memory systems.
• Nov 2025: Enabling Dynamic Sparsity in Quantized LLM Inference (SpQt), from my Zoom internship, is on arXiv.
• Oct 2025: Proto appeared at SOSP 2025.
• May 2025: WhisperFlow appeared at MobiSys 2025.

Publications

Powered by Jekyll and Minimal Light theme.