Software Engineering
CRUXEval
Tests whether models can predict a Python function's output or infer an input that produces a given output.
4,000items
31subjects
MITlicense
software_engineeringdomain
textmodality
item-level responses released
Saturation status: No
About this data
Recorded execution verdicts and postprocessed generations for input and output prediction on Python functions.
4,000 items · 992,000 recorded attempts. Browse item pages to see all observations. Filters and means apply to the current page.
Loading observation records…