I ran a comparison between Asta and MiroThinker using the same prompt on each requesting a qualitative report on the metrology (measurement) of AI frameworks and applications. MiroThinker was able to understand and respond better to qualitative aspects of the prompt such as asking for enumeration of first- , second- , and third-order concepts.
Asta report:
https://asta.allen.ai/chat/bcc5a999-9fb3-4068-be88-4ffee961c46f
MiroThinker report:
https://dr.miromind.ai/chat/e074cdc1-ab88-41c3-b261-7bbc727d4ada