Repository navigation
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Compare three fixed PiPER hold rules after identical evidence-gated test lifts. The frozen 36-attempt study records native five-second grip retention of 2/12, 4/12 and 4/12 for measured, issued and Cartesian targets; grip plus TCP stationarity is 2/12, 0/12 and 0/12. The Cartesian grip count includes its terminal fallback. No controller is promoted and no learning or hardware qualification is claimed.
Includes complete native records, independent command/predicate/pose audit, CPU measurements, failure cases and source-only reproduction tools. The existing wheel, stable main and physical interfaces are unchanged.
Validation: 77 core tests; 13 independent-audit tests; bounded packaging and publisher tests; all36 native attempts valid; 720,000 native rows and 36,000 commands independently checked; 12/12 shared acquisition prefixes exactly matched; real browser report verification.
Research: https://research.dvidia.org/papers/piper-retention-v02/
Evidence: https://huggingface.co/datasets/Dvidia/piper-retention-v02
Runtime source: 6ab5e34
Frozen protocol: a6b16cce45f364995787baf456d6c3e0b858c8572402957c777c784beeaf4565
Independent audit: 710516172db5e7851ed88a37278d8696271860b84643b456fcb2b3e64b2bb5ec
This draft PR targets the previous probe-evidence feature branch. It is an experimental comparison, not a stable release promotion.