Before submitting
Area
apps/desktop
Problem or use case
I feel like there is a UI mismatch between the reasoning level picker in the text box saying e.g. 'Extra High - Normal' where "Normal" is ambiguous and then turning Fast Mode to "On" changes the indicator to 'Extra High - Fast'. In this case, it's difficult to understand what Normal means at first glance and it feels odd to click 'On' and see the indicator change to 'Fast', and clicking 'Off' and see it changing to 'Normal'.
Proposed solution
I propose two different solutions:
- Changing the options in the reasoning level picker to the same as in the textbox ('Normal' and 'Fast').
- Simply not indicating the speed of the model when that speed is 'Normal' and then adding some icon or text to clarify that the model is in fast mode when that is the case.
Why this matters
This creates confusion because people might think 'Normal' means: "This reasoning level is the default / normal one" or maybe not be sure whether 'Normal' refers to the reasoning level or other qualities of the model.
Smallest useful scope
The smallest change would be to change the options in the reasoning level dropdown to 'Normal' and 'Fast'.
Alternatives considered
No response
Risks or tradeoffs
No response
Examples or references
No response
Contribution
Before submitting
Area
apps/desktop
Problem or use case
I feel like there is a UI mismatch between the reasoning level picker in the text box saying e.g. 'Extra High - Normal' where "Normal" is ambiguous and then turning Fast Mode to "On" changes the indicator to 'Extra High - Fast'. In this case, it's difficult to understand what Normal means at first glance and it feels odd to click 'On' and see the indicator change to 'Fast', and clicking 'Off' and see it changing to 'Normal'.
Proposed solution
I propose two different solutions:
Why this matters
This creates confusion because people might think 'Normal' means: "This reasoning level is the default / normal one" or maybe not be sure whether 'Normal' refers to the reasoning level or other qualities of the model.
Smallest useful scope
The smallest change would be to change the options in the reasoning level dropdown to 'Normal' and 'Fast'.
Alternatives considered
No response
Risks or tradeoffs
No response
Examples or references
No response
Contribution