← Back to PeopleDLResearcherDing LeileiResearcherLead author of Φ-Bench paper evaluating large language models on infrastructure engineering tasks.1Mentions1Articles1OutletsArticlesMost recent firstMentioned InReferenced in coverageThe best AI solved only 36.5% of tasks for building AI infrastructurearXivcautiousSep 10, 2026“Lead author of Φ-Bench paper evaluating large language models on infrastructure engineering tasks.”