Kun Wang

Visual token pruning creates a silent vulnerability vector: lightweight perturbations can manipulate token selection to blind compressed vision-language models while leaving uncompressed inference completely intact.

MAOL outperforms existing methods in fine-grained defect grading by effectively integrating morphological cues and adaptive thresholds, achieving robust performance even with noisy data.

Forget-Retain Alignment Gap reveals that the structure of weight updates, not just their distance, is key to preventing LLMs from relearning forgotten information.

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

TINA+ uncovers that current concept erasure methods often fail to eliminate visual knowledge, revealing a hidden layer of understanding in diffusion models.

Mint-Agent models not only outperform existing benchmarks in financial reasoning but also integrate long-horizon execution with auditable evidence trails, setting a new standard for financial intelligence.