Skip to content

data: rename seven tier5 stems to mirror DistQLDPC (reproduced GH-225/GH-227 CC lower bounds) - #18

Merged
guluchen merged 1 commit into
mainfrom
data/gh225-lb-stems
Nov 5, 2026
Merged

guluchen merged 1 commit into
mainfrom
data/gh225-lb-stems

Conversation

@guluchen

@guluchen guluchen commented Nov 5, 2026

Copy link
Copy Markdown
Owner

Mirrors guluchen/DistQLDPC#230: renames seven tier5 stems whose CC lower bounds were reproduced. These are the same matrices; only the Gx/Gz header comments change, and all 80 matrix files are byte-identical to the DistQLDPC branch.

Old New
TN_324_4_22to25 TN_324_4_23to25
TN_496_2_27to30 TN_496_2_28to30
TN_540_4_23to34 TN_540_4_24to34
TN_648_14_21to28 TN_648_14_22to28
TN_684_2_22to29 TN_684_2_24to29
COLOR_1027_1_28to37 COLOR_1027_1_31to37
COLOR_2791_1_26to61 COLOR_2791_1_28to61
  • The lower bounds come from DistQLDPC's optional CC engine, mostly with the erasure rule. They were first seen in the GH-225 / GH-227 Tier 5 runs ([Mac agent] CC engine: erasure bound as a default rule (on top of GH-217 candidate A) DistQLDPC#225, #227).
  • Each was reached by at least two runs with different search trees: relabelled qubits and rows, a different row order, or a different binary.
  • This is the same engine, not an independent second method.
  • Upper bounds are unchanged.
  • Evidence and per-run times: DistQLDPC#230 and DistQLDPC optimization/experiments/GH-225/lb-repro/.

Changes:

  • git mv of the matrix files;
  • data/README.md (tier table and a dated note);
  • NOTICE (stem list, rename paragraph, color-code list);
  • tests/test_repository_paths.py.

The layout tests in tests/test_repository_paths.py pass locally.

🤖 Generated with Claude Code

… bounds

TN_324_4_23to25 (was TN_324_4_22to25), TN_496_2_28to30 (was TN_496_2_27to30),
TN_540_4_24to34 (was TN_540_4_23to34), TN_648_14_22to28 (was TN_648_14_21to28),
TN_684_2_24to29 (was TN_684_2_22to29), COLOR_1027_1_31to37 (was COLOR_1027_1_28to37),
COLOR_2791_1_28to61 (was COLOR_2791_1_26to61). Same matrices; only the Gx/Gz header
comments change. Each lower bound was reached by at least two CC runs with different
search trees (relabelled qubits and rows, a different row order, or a different binary);
same engine, not an independent method; upper bounds unchanged.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
@guluchen
guluchen merged commit c5eef91 into main Nov 5, 2026
3 checks passed
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.

1 participant