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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Summary
Scoped mathematical proof with an exact finite certificate. For the actual eight-mode matrix M from Dillon Amburgey's PR4933, an explicitly feasible mode fiber has exact integer rounding radius one:
An error smaller than one forces zero outside the four fractional coordinates. Two preserved modes then force
z_0-z_7=0andz_0+z_7=1, contradicting integrality. The output e_48 attains radius one. Translation and scaling give the same obstruction at every grid spacing. Arbitrary nonlinear rules and correlated randomness are allowed, with mode equality required for each returned output. Equality only in expectation is outside this contract.This contribution was produced by Nick Petalas. Credit Dillon Amburgey/PR4933 for the mode matrix and inverse-code factorization, JoyboyBrian/PR5069 for the fixed-grid arithmetic context, Dugongue's coded producer and their retained native/controller/compiler lineage. Ali Romman's PR5838 distinguishes possible physical correlations from independently adverse defects; that word and error calculation are not replayed. Classical lattice and finite-box arguments retain their credit. No new kappa or selected-frontier change.
Evidence
From the repository root, with the local interpreter path shown explicitly:
Any CPython 3.12 executable may replace that machine-specific path. Expected markers include
PASS_MODE_PRESERVING_ROUNDING radius=1 corners=16 fixtures=91 controls=9 no_new_kappa,PASS_REGISTERED_RESEARCH_PACKAGEandPASS_RESEARCH_COMMIT_BYTES.Limitations and unrun obligations
This is one sharp feasible-fiber result and a uniform lower obstruction, not a global radius-one upper theorem. It does not claim that a native caller reaches the witness or must preserve modes exactly. A finer grid, approximate mode preservation or retained residue state changes the contract. Physical rounding, full propagation, restoration, precision/storage/tape costs and supplier admission remain open. All five framework assumptions remain. No predecessor theorem audit, selected scientific replay, Lean build, Docker run or complete multiplication theorem is supplied. All registered commands ran locally. Independent human and formal proof-kernel review are unclaimed.
Open next questions
Related open work
Completion refresh checked the six other open account PRs created since the CLAIM: #6307, #6308, #6309, #6310, #6311 and #6313. None studies this mode map or its output-fiber rounding radius. The related lines above remain current; no common computed quantity disagrees.
PR4933, PR5069, PR5323, PR5444, PR5838 and PR6096 are merged contextual predecessors. Their inverse, recovery, error and residue statements are cited at their stated scopes, without extra computations to check their reported values.
CI notes
Both hosted checks passed at
46593234bfe658b63fb829f0d22b7eb216e392a2in Registered research packages run 41107770945:planin 4m17s andreplay (research/mode-preserving-rounding)in 1m23s. The checks were read once. Separate run metadata confirms completed status, success and the matching head SHA. Hosted replay executes the registered normal and optimized checks. All local scientific, publication and portable-export checks listed above also pass; no registered command was deferred to CI. These checks do not establish independent human or formal proof-kernel acceptance.Merge danger
Door: two-way. Blast radius: isolated. Six files are added under research/mode-preserving-rounding. No shared implementation or selected-frontier record is edited.