[guluchen]

PI-approved stem renames (2026-10-29) for the six tier5 TN codes whose CC lower bounds were raised by GH-152 (#152). Do not merge before the PI's confirmation.

Old	New	lb	ub
TN_288_2_9to24	TN_288_2_22to24	9 -> 22	24 (unchanged)
TN_324_4_9to25	TN_324_4_22to25	9 -> 22	25 (unchanged)
TN_432_8_9to29	TN_432_8_23to29	9 -> 23	29 (unchanged)
TN_496_2_17to30	TN_496_2_23to30	17 -> 23	30 (unchanged)
TN_540_4_9to34	TN_540_4_23to34	9 -> 23	34 (unchanged)
TN_576_2_9to37	TN_576_2_24to37	9 -> 24	37 (unchanged)

Evidence. CC lower bounds from the optional CC engine with -cc-aut=on (Tanner-graph automorphism candidates for the orbit seeds and the XZ-dual map; every candidate passes the existing GF(2) checks), 6 h on yfclab2, binary dfd0717. Reproduction: same binary and labels, different search order (-cc-rules=excl,pack,orbf,parity,leaf,overlap,mrv, 12 h): all six reproduced, every level EMPTY, node counts differ at every level (same algorithm, not a second method). Record: optimization/experiments/GH-152/REPRODUCTION.md on experiment/gh-152-tanner-aut and the comments on #152.

Changes (pattern of #149):

• git mv of Hx/Hz/Gx/Gz for each stem; only the Gx/Gz header comment changes.
• data/matrices/README.md: tier5 list, six rename-history rows citing [Mac agent] CC engine: Tanner-graph automorphism discovery for orbit seeds and the XZ-dual map (open TN codes) #152 ([Mac agent] CC engine: Tanner-graph automorphism discovery for orbit seeds and the XZ-dual map (open TN codes) #152), CC lower bound, -cc-aut=on and the reproduction; open items updated.
• NOTICE: a 2026-10-29 rename paragraph.
• optimization/baselines/SERVER.md and main_server_reference.json: name columns (as Adopt GH-146 Stern finder as default (both engines); tier5 stems get verified upper bounds #149 did).
• The [Mac agent] Stern/Dumer collision step in the verified upper-bound finder (large open codes) #146 upper-bound witnesses for these six stems are copied under the new stems to optimization/experiments/GH-152/witnesses/ (VERIFY.txt: 6/6 VALID with the pure-Python verify_witnesses.py). The 2026-10-23 copies in GH-146/witnesses/renamed/ are left as they were (history).
• No old stem remains outside rename-history entries (data/matrices/README.md, NOTICE) and historical experiment records under optimization/experiments/ ([Mac agent] Stern/Dumer collision step in the verified upper-bound finder (large open codes) #146, [Mac agent] MaxSAT path: stabilizer-overlap dominance constraints in every CSS-half probe #148), which are unchanged. No tests, scripts, docs or web files referenced these stems.

Checks (Mac, no timing runs):

• Build: OK.
• make test-cc-flags: 376 checks, 0 failed.
• scripts/smoke_test.sh: passed; scripts/test_partition_soft_literals.py: passed.
• Tier table vs disk: 51 stems listed = 51 on disk, all four files present.
• Bare stems resolve: bin/distqldpc TN_576_2_24to37 -engine=cc -cpu-lim=5 (lb 15, TIMEOUT as expected) and bin/distqldpc TN_496_2_23to30 -cpu-lim=3.

Note: #155 (-cc-aut=on as the CC default) also edits NOTICE; a trivial merge conflict is possible depending on merge order.

QDistSAT mirror: guluchen/QDistSAT#14

🤖 Generated with Claude Code