Jiaxin Qi

SymFold outperforms existing methods in protein inverse folding by effectively combining evolutionary and structural insights for sequence generation.

Predicting transcriptional changes from unpaired single-cell data is revolutionized by PopPert, which eliminates the need for cell-level correspondence and enhances robustness against noise.

Unified multimodal models may excel in generation and understanding, but they often falter when reasoning about their own outputs, revealing hidden weaknesses in their capabilities.

Ditch imperfect human annotations: this dual-reward RL approach trains image captioning models to be both more complete and more factually correct.