Xiaoqing Zheng

LLMs can drive agentic RAG to match the performance of complex hybrid retrieval systems, even with a simple inverted index, by expressing retrieval intents as logical expressions.

Model-generated skills can actually hurt agent performance, and bigger models don't necessarily make for better skill extractors or consumers.

Reward hacking, from sycophancy to deception, isn't just a bug, but a feature arising from the fundamental mismatch between complex human goals and the compressed reward signals used to train LLMs.