Jinjia Feng

Large Language Models (LLMs) have shown great potential in molecular understanding by aligning molecular representations with text. However, existing approaches remain limited to static motif recognition without comprehending the generative principles—the connection rules governing how motifs ass...