Lin-Hao Luo

This work formalizes a Wiki Graph schema that seamlessly bridges fine-grained structures with dense contexts, maintaining explicit topologies alongside continuous semantics and engineer an infrastructural NCCL boundary exchange protocol that hoists static partition indices and leverages fixed-shape GPU-to-GPU collectives, bypassing CPU serialization and memory copy overheads.

System Intelligence emerges as a game-changer, enabling LLM agents to collaborate effectively across complex tasks by organizing their interactions through dynamic graph structures.

EASy achieves superior performance-efficiency trade-offs by intelligently coordinating heterogeneous executors based on their capabilities and costs, reshaping agentic system design.

TriAlign achieves a remarkable balance between universal truth consistency and personalized responses, addressing a critical gap in LLM alignment.

Forget costly training or reward models: MATO unlocks personalized LLM alignment by optimizing objective weights *during* generation, offering unprecedented control and adaptability.

Forget bolting vision onto language models – truly powerful multimodal AI demands rethinking architectures from the ground up.

Multi-agent systems can dynamically adapt their communication topology to changing conditions, leading to more robust and accurate performance on complex tasks.