Jing Liang

Assigning reward credit to specific tokens, rather than uniformly, dramatically improves multi-reward RL for LLMs.

Sparse rewards in multi-objective RL are no longer a bottleneck; a simple density correction unlocks faster, more robust learning.

Censored sales data can be transformed into a powerful tool for optimizing pricing and inventory decisions, leading to significant profit improvements.

Decoupling semantic reasoning from real-time control allows robots to master complex tasks with unprecedented stability and generalization.

Training updates that improve performance in LLMs can actually degrade inference quality—unless you use the new Monotonic Inference Policy Update framework.

Forget scaling laws: verifiable RL can teach smaller models to abstain and clarify unanswerable questions as well as much larger models.