3 ms·
Neat idea! You might be able to deduplicate the list by converting the decisions to embeddings and running k-means clustering.
by rickcarlino 2y ago
Neat idea! You might be able to deduplicate the list by converting the decisions to embeddings and running k-means clustering.
- thsnyder 2y agoI like this idea! I could definitely see running an ML algo to cluster the Reasonings by keywords and then hide the shorter ones, with the assumption that the longer explanations are more in-depth.