Xinyu Yang

Token-oriented inference optimizations can cut production costs and boost efficiency, transforming large model services from merely callable to fully operable.

LLM agents are shockingly vulnerable to multi-stage "trojan" attacks that inject malicious instructions into their workspace, achieving near-perfect success rates where standard prompt injection defenses fail.

Unlock scalable, high-quality singing voice synthesis by directly generating structured musical scores from audio, outperforming existing systems on multiple datasets.