Yong-Qi Tong

Frontier LLMs fail to achieve even 65% strict success on real-world business tasks, revealing that current benchmarks mask critical execution invalidity and inefficiency.

Timing the entry of preference dimensions can lead to substantial performance gains in multi-preference alignment for LLMs.

Models trained with ACA-RL not only outperform on missing-premise tasks but also redefine how we evaluate reasoning under uncertainty in NLP.

Fair behavior comparisons in agent interactions can dramatically improve performance, reducing response times from nearly 5 seconds to just over 1 second.