You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
While investigating the presence of medical myths in RTX-KG2 (particularly those caused by retracted papers, see RTXteam/RTX-KG2#386), I ran the query "what causes autism?" on ARAX:
The top result is "Vaccines". The second result is "Measles-Mumps-Rubella Vaccine". While RTX-KG2 is responsibly for passing along these edges, Steve recommended that I post an issue for ARAX as well. These two results are widely considered not just misinformation, but actively harmful. Is it possible that we could consider, for major medical myths like this, at least ranking harmful answers lower or blacklisting them entirely? Or is this dependent on RTX-KG2 to filter out these edges (which came largely from SemMedDB misparses)?
The text was updated successfully, but these errors were encountered:
While investigating the presence of medical myths in RTX-KG2 (particularly those caused by retracted papers, see RTXteam/RTX-KG2#386), I ran the query "what causes autism?" on ARAX:
https://arax.rtx.ai/?r=243414
The top result is "Vaccines". The second result is "Measles-Mumps-Rubella Vaccine". While RTX-KG2 is responsibly for passing along these edges, Steve recommended that I post an issue for ARAX as well. These two results are widely considered not just misinformation, but actively harmful. Is it possible that we could consider, for major medical myths like this, at least ranking harmful answers lower or blacklisting them entirely? Or is this dependent on RTX-KG2 to filter out these edges (which came largely from SemMedDB misparses)?
The text was updated successfully, but these errors were encountered: