Latest

We announce the release of our new pretraining corpus and appeal outcome prediction benchmark.

We describe barriers to responsibly deploying natural language understanding tools with existing technology, and ways to overcome them.

Announcing a submission of a Freedom of Information Act public records request for Medicare data, prepared in collaboration with the Data Liberation Project.

We analyze data from the CMS Medicare Part C and D Qualified Independent Contractor appeal database.

We look into demographic trends in two California databases documenting outcomes of external appeals of health insurance denials.