Research: present-head flag 9 hosts basis head only after source drop (no new kappa)

[VictorLiwentao]

Change and scope

One-flag follow-up under priority revision 6, card p104-side-correction. #3349 shows the basis head meets the target final on only 4 of 208 matched flags (present-head class (0,1,9,43)). #3392 pins one of the other 204 (absent-head flag 2). This package pins one of those four: flag 9 (X4 → Y23, basis head 196).

On the published 4,058-event word:

• X4 starts at [196] and drops 196 at transport event 500
• every annotated Y23 frame before the terminal transport avoids 196
• the head appears on Y23 only in the after-frame of event 3489 (gap 2989)
• no event names both endpoints; neither frame is restored

The published channel therefore cannot chronologically dirty-restore that source line into the matched target along co-hosted basis-head geometry, even though the target final contains the head.

Scoped obstruction. No new kappa. Selected κ stays 0.001672660197076918. Coefficient delivery (flag-0 cascade) is outside this package.

Result

Package: research/p104-present-head-flag9-audit/.
Tip: acc2fa45374ce34c2d6293b862c3dcafb853ffa6.
Content tip: ae18b8db189f777bef159530d5395ebfe024f7b1.
Base: 6f100d4bb5083f8c35d96fd95a539811c62f245c.

Validation

python3.12 -I -B research/p104-present-head-flag9-audit/verify.py
python3.12 -I -O -Werror -B research/p104-present-head-flag9-audit/verify.py
python3.12 -I -O -Werror -B research/p104-present-head-flag9-audit/independent.py
python3.12 -I -O -Werror -B research/p104-present-head-flag9-audit/test_package.py
python3.12 -B scripts/research_package.py --package research/p104-present-head-flag9-audit

Expected:

PASS_P104_PRESENT_HEAD_FLAG9 flag=9 head=196 preterminal_hosting=0 drop_event=500 arrive_event=3489 no_new_kappa
PASS_P104_PRESENT_HEAD_FLAG9_INDEPENDENT flag=9 head=196 preterminal_hosting=0 drop_event=500 arrive_event=3489 no_new_kappa
PASS_P104_PRESENT_HEAD_FLAG9_TESTS groups=6
PASS_REGISTERED_RESEARCH_PACKAGE research/p104-present-head-flag9-audit outcome=completed

Default verify is read-only. Admission uses need() and still fires under -O.

Overlap and credit

Closest priors: #3349 for the 4/208 present-head census; #3392 for the complementary absent-head obstruction. Does not reopen that census, the flag-2 path, #3441/#3436/#3421/#3411, deferred V45/rho=0, or the merged readout series through #2883. Soft-overlap checked against open #3363/#3365/#3371/#3372 (coefficient delivery).

Credit Nick Petalas / #2981, jwr404 / #3242, #3349, and #3392. Prepared by Wentao Li (VictorLiwentao, @Is__Victor) with substantial AI assistance, under Apache-2.0.

[VictorLiwentao]

Limits note (follow-up from flag-1 span work). On this published word the event-500 after-frame [8, 64, 132, 260] still spans basis head 196 via 196 = 132 ^ 64, so the literal “source drop” at event 500 is a change of generators rather than a GF(2) span exit. The package claim that remains solid is the target-side late hosting: Y23 first hosts 196 only at terminal after-frame 3489, with preterminal_hosting=0. See also #3451 for the flag-1 package that certifies the span reading explicitly (source_span_drops=0).