-
Notifications
You must be signed in to change notification settings - Fork 34
Calculate the mode correctly, return the 'maximum' value in case of multiple modes.#96
Calculate the mode correctly, return the 'maximum' value in case of multiple modes. #96Megh-Thakkar wants to merge 2 commits intomain from
Conversation
Megh-Thakkar
commented
Aug 25, 2025
Since we are only creating dummy reports for compositional tasks, the mode validation correctly follows this now: https://github.com/ServiceNow/WorkArena/blob/main/src/browsergym/workarena/tasks/compositional/dash_do_base.py#L346, https://github.com/ServiceNow/WorkArena/blob/main/src/browsergym/workarena/tasks/compositional/dash_do_base.py#L477
@aldro61
aldro61
left a comment
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
@Megh-Thakkar for the cheat function, it's ok to use the max value as tiebreaker and just use the largest of the modes. However, for the validation function, we must accept any value that is a valid mode. You should move it to its one condition case, just like max/min and make sure you allow any number that is a valid mode.
#94