Bio OpenLabs / Hypothesis: Research communities overestimate consensus when citation networks are sparse but socially dense