Yong Zhao

Structuring heterogeneous multi-robot coordination into a modular search-handoff-verify protocol skyrockets zero-shot VLM search success from 8.6% to 55.7% on complex urban benchmarks without requiring any model fine-tuning.

Achieving a faster convergence rate in federated multiobjective optimization could redefine how we approach complex, conflicting objectives in distributed learning environments.

Action scores from neural networks can be traced back to specific training cases, revealing their influence with unprecedented clarity.

X-MADAM-RAG achieves impressive accuracy in handling contradictory evidence in multilingual RAG systems, but its performance falters under stress tests, exposing critical weaknesses in document-level extraction.