Bo Liu

Optimization performance varies significantly by workload, challenging the notion that larger models are always superior in coding tasks.

KOAL's innovative use of clinical context and expert knowledge allows for a more nuanced and accurate prediction of prostate cancer grading, outperforming traditional methods.

Membership inference attacks can now be effectively executed across multiple generative modalities with a single, unified framework.

ActWorld bridges the navigation-interaction gap in interactive world models, enabling rich object interactions that were previously unattainable.