Picture for Zheyuan Chen

Zheyuan Chen

Llamas on the Web: Memory-Efficient, Performance-Portable, and Multi-Precision LLM Inference with WebGPU

Add code
May 20, 2026
Viaarxiv icon