Repository navigation
Fix power precedence and bounded refusals in MathGate - #1
Merged
Merged
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
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
MathGate previously evaluated
-2^2as4, associated chained powers incorrectly, and could raise instead of returning a refusal for large factorials or deeply nested input. This change applies conventional power precedence and returns sealed refusals for documented computation, output-size and parsing limits.The public 240-case corpus is unchanged: 192 exact answers and 48 expected refusals, with zero failures or certificate drift. Eighteen unit tests pass, including new regressions for the discovered failures. The README now identifies this corpus as a public development suite rather than an independent evaluation.
Validation:
python -m unittest discover -s tests,python -m mathgate --bench, demo, andgit diff --checkpassed. AI-assisted implementation and review.