Jiaqing Liang

PlanRAG transforms exploratory reasoning by leveraging logical query trees, achieving superior performance in complex query resolution.

ADaPT enables a single model to flexibly navigate the efficiency-performance trade-off, achieving significant cost savings without sacrificing reasoning quality.

Naive RL in recommender systems suffers from biased gradients that favor longer paths, but ProRL fixes this with a novel reward centering and advantage estimation scheme.

LLM agent performance hinges on maximizing decision-relevant information density within context, not just context length, and GenericAgent proves it.