Skip to content

Compose gauge refinement with PR150 recycling: conditional κ = 0.000472154791 - #151

Merged
Mister-Meeseeks merged 3 commits into
CrocSwap:mainfrom
SovereignSteak:research/gauge-recycling-composition
Oct 9, 2026
Merged

Mister-Meeseeks merged 3 commits into
CrocSwap:mainfrom
SovereignSteak:research/gauge-recycling-composition

Conversation

@SovereignSteak

Copy link
Copy Markdown
Contributor

Result

Conditional kappa = 472154791/10^12 = 0.000472154791, versus #150's best reported .000472026274: approximately +0.0272%.

Stacked on #150 at 40d4038760ebb6d6d3d702ce88fa3645de29f0c1; please review/merge that dependency first. Its commits are included in the comparison against main, not claimed as this contribution. The incremental change is only research/gauge-recycling/ (README, exact plan/row, verifier).

Mathematical contribution

The 187 additional gauges from our Round19 refinement (concurrent with the public #146 gauge optimization) compose with all 3,338 #150 handoffs without changing old read times or adding any new non-inherited frame transition. Restoring a first occupant's initial-value read improves the entrance/target ledger without shortening an existing donor lifetime. This is a checked compatibility property, not a claim for arbitrary gauge additions.

R=23,979, W=27,521, deficit=2,024, maximum child=60. The full histogram is rebuilt and its additional 1,012 positive-child charges are included. Coarse bit saving is 472806533/10^12.

Row Atom exponent Conditional kappa
#150 .0005 .000472026274
Composition .0005 .000472143085
#150 repriced .000473 .000472037976
Composition .000473 .000472154791

At identical atom exponent the construction contributes .000000116815; further wrapper tightening contributes .000000011706 on the composed row. The README gives the stopped recurrence and strict wrapper conditions. Bit supply remains limiting.

Verification

python3 research/bit-reuse-147/verify.py
python3 research/gauge-recycling/verify.py
# Optional direct re-check instead of transporting baseline geometry:
python3 research/gauge-recycling/verify.py --frames

Python stdlib only; run without -O.

Credits

DaysSky #150; jamesyc #124; hpst3r #147; icekylinx #144; Th0rgal #146; gupt1156 #148; SovereignSteak #122; Swapnil Jain and Zhihao Chen's word/frame tools; an664 #128. This is compatible composition and accounting, not priority for their underlying mechanisms. Prepared with AI assistance. No publication-level novelty claim.

DaysSky and others added 3 commits October 9, 2026 14:23
On the merged CrocSwap#144 bit branch with CrocSwap#147's 1,549 terminal eliminations,
run the selected deferred reads as late as the word allows and hand
3,338 dead auxiliary registers to later deferred slots whose entrance
frame contains the dead register's frame (the CrocSwap#124 birth-read operation
on the bit word). Physical roles fall from 27,317 to 23,979 and the
coarse bit saving rises to 472689442/10^12.

Adds research/bit-reuse-147 and its workflow; no existing file changes.
verify.py now runs the literal word with every data port, target and
physical register as its own variable, over F2 and over Z, and requires
the exact identity; this is complete and uses no random vectors. The
random replay is kept for the three tamper controls. Retained centres
are excluded as donors (none was used), and the proof note states the
rank-witness remark for hand-offs. No certified value changes.
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

3 participants