Repository navigation
Distinguish model-tier comparison from protocol replay savings - #18
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Compact benchmark summaries could suggest that the 0.657 ratio directly measures the TierDecay protocol, while the 24% saving is a modeled replay based on the controlled pilot.
This change names both comparisons explicitly and keeps the scope at 36 executor runs plus 3 distillations on one fixture. It preserves the pilot limitations and does not introduce a production savings claim.
Validation: the existing analysis reproduces 0.657 (95% interval 0.626–0.694); the protocol replay reproduces 23.9526% against always-frontier. Tracked benchmark data remain byte-identical. No runtime or release changes.