Yikun Wang

Achieving state-of-the-art performance in document parsing, NaviDC-OCR tackles geometric distortions and structural reasoning challenges that plague existing models.

Tailoring pretraining data processing to individual examples can lead to significant performance improvements in LLMs while cutting down on compute costs.

ETAS redefines how we program agent systems by integrating semantic elements that enhance authorization and audit capabilities without sacrificing programming clarity.

Context rot leads LLMs to falter under lengthy inputs, but targeted management and rejection strategies can restore their performance.

Achieve superior 3D scene reconstruction from aerial images with significantly reduced transmission overhead by directly optimizing communication for rendering quality.