Fix overflow in historical scoring model point count summation #3616
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
In adb0afc we started raising bucket weights to the power four in the historical model. This improved our model's accuracy greatly, but resulted in a much larger `total_valid_points_tracked`. In the same commit we converted `total_valid_points_tracked` to a float, but retained the 64-bit integer math to build it out of integer bucket values. Sadly, 64 bits are not enough to sum 1024 bucket pairs of 16-bit integers multiplied together and then squared (we need 16*4 + 10 = 74 bits to avoid overflow). Thus, here we replace the summation with 128-bit integers.
wpaulino
approved these changes
Feb 24, 2025
|
Landing as trivial |
|
Backported in #3616 |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
In adb0afc we started raising bucket weights to the power four in the historical model. This improved our model's accuracy greatly, but resulted in a much larger
total_valid_points_tracked. In the same commit we convertedtotal_valid_points_trackedto a float, but retained the 64-bit integer math to build it out of integer bucket values.Sadly, 64 bits are not enough to sum 1024 bucket pairs of 16-bit integers multiplied together and then squared (we need 16*4 + 10 = 74 bits to avoid overflow). Thus, here we replace the summation with 128-bit integers.