published imports an objective October 02, 2026

With a stack NAND on a metabolism tape, paid parts assemble the logic rungs that need an input twice, in a sixth of the worlds

DESIGN §1.3 sweep 17 moved the Logic sweep's computing off the copier: a second, 32-byte tape the soup never runs, handed on with each copy and read by the assay with a NAND that keeps its operands. The 18 emerged worlds ran another 40 000 epochs, three seeds a parent, paired with the Logic sweep's children. All 162 children finished and none died out. 9 children of 54, from 7 parents, held XOR or EQU at a tenth of their world, against none in any other arm, so the rung-4 question is shown, 9 pairs to 0 with 45 ties (p = 0.00195). The stepping-stone test against the arm paid for XOR and EQU alone is shown too, 9 to 0: the deep rungs were assembled from paid parts, Lenski's mechanism in a soup. The same tape read by the NAND that writes in place built none. Without the four piloted parents the deep tests read 8 to 0 and rest on two parents. In 14 of 15 plantings a world the deep rung was wiped from re-grew it within 500 epochs: a property of the population more than of a lineage. The ladder tops out at EQU, and the substrate imports an objective, a primitive, a hereditary channel and the semantics that let the deep rungs come, so rung 4 on Soup stays not shown.

Read from Metabolism, a second substrate that pays a cell for computing: it imports an objective and is never pooled with the fitness-free runs. Rung 4 on Soup is unaffected by it. What that means.

Question

Given a NAND, paid replicators climbed every logic rung that reads its inputs once and stopped below XOR and EQU, the two that need an input twice. The design study behind this sweep read the stall as fan-out in a two-head machine whose NAND writes in place: a value used twice needs a stored copy and re-aligned heads, and in a tape that must also copy itself only the copy loop supplies them. Meta-stack, the variant of the Metabolism substrate in DESIGN §1.4, moves the computing off the copier and changes what the NAND writes. On a tape the soup never runs, read with a NAND that keeps its operands, does a soup assemble XOR and EQU from paid parts?

Method

The parents and seeds are the Logic sweep's: the emerged worlds the from-emerged sweep continued, each run another 40 000 epochs, seeds 2001–2003, under Logic's economy and paid ladder. Every child carries a second, 32-byte metabolism tape, seeded from its cell's own tape. The soup never executes it; the logic assay reads it, and the world hands it on with each copy, mutating it at 0.00390625 a byte. Three arms:

meta-stack: every rung paid, the tape read with the stack NAND, which writes its result one cell to the left, moves its head there and leaves its operands intact.
meta-stack-deep-only: the same with only XOR and EQU paid, Lenski's control. It separates "the deep rungs are built on paid parts" from "the deep rungs are directly reachable".
meta-inplace: the tape read with Logic's NAND, which writes in place, the control on the NAND's semantics.

Each child is paired with the child of the same parent and seed in another arm. Two arms come from the Logic sweep, no new run needed: logic-none, its unpaid child, and logic-full, its paid child computing on the copied tape itself.

The reading is pre-registered, with every number fixed before a child was seeded: docs/design_record.md, 2026-10-02, "Meta-stack: does a soup assemble the deep rungs from paid parts on a metabolism tape read with a stack NAND?", DESIGN §1.3 item 17. The rules are Logic's: a settling window of 1 000 epochs, last-decile medians, and no pair with an extinct child in a test. There are six one-sided sign tests at p < 0.05:

H-deep-Ms, the rung-4 question (meta-stack against logic-none): logic_capability_deep, the number of XOR and EQU held by a tenth of the world.
H-stones-Ms (meta-stack against meta-stack-deep-only): the same count.
H-stack (meta-stack against meta-inplace): the same count. Does the NAND's semantics decide it?
H-deep-M (meta-inplace against logic-none): the same count, the tape under the in-place NAND.
H-decouple (meta-inplace against logic-full): the same count, the tape against the copied tape under the same NAND.
H-capability-M (meta-stack against logic-none): logic_capability, every rung from ECHO to EQU.

Each test is also read with the extinct pairs kept and without the four parents the design study piloted (runs 1007, 944, 2577, and 2700). Those two re-readings decide no outcome. A shown test lists any one or two parents whose pairs it cannot do without. The entry also locks three measurements made after the sweep on each deep child's stored worlds: a substitution distance, heritability by planting, and the circuit behind the deep solver.

Result

H-deep-Ms shown H-stones-Ms shown H-stack shown H-deep-M refuted H-decouple refuted H-capability-M shown With a stack NAND on a metabolism tape, paid parts assembled the logic rungs that need an input twice.

H-deep-Ms, meta-stack against logic-none: 9 pairs favour meta-stack, 0 logic-none and 45 tie, of 54 measured, p = 0.00195, shown. By parent, 7 lean to meta-stack, 11 tie throughout and 0 lean to logic-none. With the extinct pairs kept: 9 to 0 with 45 ties, p = 0.00195, shown. Without the piloted parents: 8 to 0 with 34 ties, p = 0.00391, shown.
H-stones-Ms, meta-stack against meta-stack-deep-only: 9 pairs favour meta-stack, 0 meta-stack-deep-only and 45 tie, of 54 measured, p = 0.00195, shown. By parent, 7 lean to meta-stack, 11 tie throughout and 0 lean to meta-stack-deep-only. With the extinct pairs kept: 9 to 0 with 45 ties, p = 0.00195, shown. Without the piloted parents: 8 to 0 with 34 ties, p = 0.00391, shown.
H-stack, meta-stack against meta-inplace: 9 pairs favour meta-stack, 0 meta-inplace and 45 tie, of 54 measured, p = 0.00195, shown. By parent, 7 lean to meta-stack, 11 tie throughout and 0 lean to meta-inplace. With the extinct pairs kept: 9 to 0 with 45 ties, p = 0.00195, shown. Without the piloted parents: 8 to 0 with 34 ties, p = 0.00391, shown.
H-deep-M, meta-inplace against logic-none: 0 pairs favour meta-inplace, 0 logic-none and 54 tie, of 54 measured, refuted. By parent, 0 lean to meta-inplace, 18 tie throughout and 0 lean to logic-none. With the extinct pairs kept: 0 to 0 with 54 ties, refuted. Without the piloted parents: 0 to 0 with 42 ties, refuted.
H-decouple, meta-inplace against logic-full: 0 pairs favour meta-inplace, 0 logic-full and 53 tie, of 53 measured, refuted. By parent, 0 lean to meta-inplace, 18 tie throughout and 0 lean to logic-full. With the extinct pairs kept: 0 to 0 with 54 ties, refuted. Without the piloted parents: 0 to 0 with 41 ties, refuted.
H-capability-M, meta-stack against logic-none: 54 pairs favour meta-stack, 0 logic-none and 0 tie, of 54 measured, p = 5.55e-17, shown. Every pair of all 18 parents favours meta-stack. With the extinct pairs kept: 54 to 0 with 0 ties, p = 5.55e-17, shown. Without the piloted parents: 42 to 0 with 0 ties, p = 2.27e-13, shown.

No one or two parents carry a shown test: whichever one or two are left out, the rest still read p < 0.05. Re-read as H-deep-Ms, unpiloted parents, it rests on parents run 967 and run 2771: without their pairs it falls to p ≥ 0.05. Re-read as H-stones-Ms, unpiloted parents, it rests on parents run 967 and run 2771: without their pairs it falls to p ≥ 0.05. Re-read as H-stack, unpiloted parents, it rests on parents run 967 and run 2771: without their pairs it falls to p ≥ 0.05.

9 meta-stack children of 54, from 7 parents, held XOR or EQU at a tenth of the world for 5 samples running. Every other arm reached a deep rung in 0 meta-stack-deep-only, 0 meta-inplace, 0 logic-none, and 0 logic-full children. The arms lost 0 meta-stack, 0 meta-stack-deep-only, 0 meta-inplace, 0 logic-none, and 1 logic-full children to extinction.

The deep children, read from their stored samples at render time: the first epoch each deep rung held a tenth of the world under the persistence rule.

Child	Parent	Seed	First XOR	First EQU
run 4845	run 944	2003	21 870	21 940
run 4862	run 967	2002	21 760	21 700
run 4863	run 967	2003	33 530	22 690
run 4871	run 991	2002	23 180	23 220
run 4890	run 1029	2003	28 290	22 880
run 4952	run 2654	2002	28 790	28 220
run 4961	run 2676	2002	21 630	21 700
run 4979	run 2771	2002	—	41 480
run 4980	run 2771	2003	—	22 720

Every count above is read from the two sweeps' stored samples at render time, through the same reading the sweep page draws.

The pre-registered reading final imports an objective

Every child of every qualifying parent has finished, here and in the Logic sweep. Each child is paired with the child of the same parent and seed in another arm: the Logic sweep's unpaid child and its paid child on the woven tape are the twins from that sweep. All are read over their own samples past a settling window of 1,000 epochs. A child is extinct where its last-decile median share of replicators on the replicating tape is below 10%, and a pair with an extinct child is in no test. Each test is a one-sided sign test at p < 0.05, re-read with the extinct pairs kept and without the parents the design study piloted (runs 1007, 944, 2577, and 2700). A deep rung shown here reads as assembled from paid parts only where H-stones-Ms is shown too, and as a crossing of several steps only where its substitution distance, measured after the sweep, is at least two. This substrate imports an objective, a primitive, a hereditary channel and a choice of the primitive's semantics made because it lets the deep rungs come, so rung 4 on Soup stays not shown whatever it reads.

Children per arm, read so far. A deep rung is XOR or EQU held by a tenth of the world for 5 samples running; a stepping stone is OR, ANDN or NOR held first.

Arm	Children	Finished	Settled relapses	Extinct	Capability measured	Reached a deep rung	Through a stepping stone	Complexity survivors	Rises
meta-stack	54	54	0	0	54	9	9	54	3
meta-stack-deep-only	54	54	0	0	54	0	0	54	2
meta-inplace	54	54	0	0	54	0	0	54	3
logic-none	54	54	0	0	54	0	0	54	1
logic-full	54	54	6	1	53	0	0	48	2

Hypotheses

H-deep-Ms, meta-stack against logic-none: shown

54 pairs measured on both sides: 9 favour meta-stack, 0 logic-none, 45 tie; p = 0.00195. With the extinct pairs kept: shown, 9 to 0, p = 0.00195. Without the piloted parents: shown, 8 to 0, p = 0.00391.

Per-parent agreement

Parent	Favour meta-stack	Tie
run 944	1	2
run 950	0	3
run 967	2	1
run 991	1	2
run 1007	0	3
run 1029	1	2
run 1087	0	3
run 1089	0	3
run 1103	0	3
run 2568	0	3
run 2577	0	3
run 2590	0	3
run 2654	1	2
run 2676	1	2
run 2700	0	3
run 2771	2	1
run 2802	0	3
run 2862	0	3

H-stones-Ms, meta-stack against meta-stack-deep-only: shown

54 pairs measured on both sides: 9 favour meta-stack, 0 meta-stack-deep-only, 45 tie; p = 0.00195. With the extinct pairs kept: shown, 9 to 0, p = 0.00195. Without the piloted parents: shown, 8 to 0, p = 0.00391.

Per-parent agreement

Parent	Favour meta-stack	Tie
run 944	1	2
run 950	0	3
run 967	2	1
run 991	1	2
run 1007	0	3
run 1029	1	2
run 1087	0	3
run 1089	0	3
run 1103	0	3
run 2568	0	3
run 2577	0	3
run 2590	0	3
run 2654	1	2
run 2676	1	2
run 2700	0	3
run 2771	2	1
run 2802	0	3
run 2862	0	3

H-stack, meta-stack against meta-inplace: shown

54 pairs measured on both sides: 9 favour meta-stack, 0 meta-inplace, 45 tie; p = 0.00195. With the extinct pairs kept: shown, 9 to 0, p = 0.00195. Without the piloted parents: shown, 8 to 0, p = 0.00391.

Per-parent agreement

Parent	Favour meta-stack	Tie
run 944	1	2
run 950	0	3
run 967	2	1
run 991	1	2
run 1007	0	3
run 1029	1	2
run 1087	0	3
run 1089	0	3
run 1103	0	3
run 2568	0	3
run 2577	0	3
run 2590	0	3
run 2654	1	2
run 2676	1	2
run 2700	0	3
run 2771	2	1
run 2802	0	3
run 2862	0	3

H-deep-M, meta-inplace against logic-none: refuted

54 pairs measured on both sides: 0 favour meta-inplace, 0 logic-none, 54 tie. With the extinct pairs kept: refuted, 0 to 0. Without the piloted parents: refuted, 0 to 0.

Per-parent agreement

Parent	Favour meta-inplace	Favour logic-none	Tie	Not both measured
run 944	0	0	3	0
run 950	0	0	3	0
run 967	0	0	3	0
run 991	0	0	3	0
run 1007	0	0	3	0
run 1029	0	0	3	0
run 1087	0	0	3	0
run 1089	0	0	3	0
run 1103	0	0	3	0
run 2568	0	0	3	0
run 2577	0	0	3	0
run 2590	0	0	3	0
run 2654	0	0	3	0
run 2676	0	0	3	0
run 2700	0	0	3	0
run 2771	0	0	3	0
run 2802	0	0	3	0
run 2862	0	0	3	0

H-decouple, meta-inplace against logic-full: refuted

53 pairs measured on both sides: 0 favour meta-inplace, 0 logic-full, 53 tie. With the extinct pairs kept: refuted, 0 to 0. Without the piloted parents: refuted, 0 to 0.

Per-parent agreement

Parent	Tie	Not both measured
run 944	3	0
run 950	3	0
run 967	3	0
run 991	3	0
run 1007	3	0
run 1029	3	0
run 1087	3	0
run 1089	3	0
run 1103	2	1
run 2568	3	0
run 2577	3	0
run 2590	3	0
run 2654	3	0
run 2676	3	0
run 2700	3	0
run 2771	3	0
run 2802	3	0
run 2862	3	0

H-capability-M, meta-stack against logic-none: shown

54 pairs measured on both sides: 54 favour meta-stack, 0 logic-none, 0 tie; p = 5.55e-17. With the extinct pairs kept: shown, 54 to 0, p = 5.55e-17. Without the piloted parents: shown, 42 to 0, p = 2.27e-13.

Per-parent agreement

Parent	Favour meta-stack	Favour logic-none	Tie	Not both measured
run 944	3	0	0	0
run 950	3	0	0	0
run 967	3	0	0	0
run 991	3	0	0	0
run 1007	3	0	0	0
run 1029	3	0	0	0
run 1087	3	0	0	0
run 1089	3	0	0	0
run 1103	3	0	0	0
run 2568	3	0	0	0
run 2577	3	0	0	0
run 2590	3	0	0	0
run 2654	3	0	0	0
run 2676	3	0	0	0
run 2700	3	0	0	0
run 2771	3	0	0	0
run 2802	3	0	0	0
run 2862	3	0	0	0

The pairs

Parent	Seed	Meta-stack child	Capability	Deep capability	Meta-stack-deep-only child	Meta-inplace child	Capability	Logic-none child	Capability	Logic-full child	Capability
run 944	2001	run 4843	4	0	run 4846	run 4849	4	run 4687	0	run 4681	2
run 944	2002	run 4844	4	0	run 4847	run 4850	4	run 4688	0	run 4682	4
run 944	2003	run 4845	4	2	run 4848	run 4851	4	run 4689	0	run 4683	2
run 950	2001	run 4852	5	0	run 4855	run 4858	5	run 4696	0	run 4690	5
run 950	2002	run 4853	4	0	run 4856	run 4859	4	run 4697	0	run 4691	5
run 950	2003	run 4854	5	0	run 4857	run 4860	4	run 4698	0	run 4692	3
run 967	2001	run 4861	4	0	run 4864	run 4867	4	run 4705	0	run 4699	4
run 967	2002	run 4862	7	2	run 4865	run 4868	4	run 4706	0	run 4700	3
run 967	2003	run 4863	5	2	run 4866	run 4869	4	run 4707	0	run 4701	2
run 991	2001	run 4870	4	0	run 4873	run 4876	4	run 4714	0	run 4708	3
run 991	2002	run 4871	5	2	run 4874	run 4877	4	run 4715	0	run 4709	1
run 991	2003	run 4872	4	0	run 4875	run 4878	4	run 4716	0	run 4710	2
run 1007	2001	run 4879	5	0	run 4882	run 4885	4	run 4723	0	run 4717	4
run 1007	2002	run 4880	3	0	run 4883	run 4886	4	run 4724	0	run 4718	3
run 1007	2003	run 4881	4	0	run 4884	run 4887	6	run 4725	0	run 4719	4
run 1029	2001	run 4888	3	0	run 4891	run 4894	4	run 4732	0	run 4726	4
run 1029	2002	run 4889	4	0	run 4892	run 4895	3	run 4733	0	run 4727	4
run 1029	2003	run 4890	7	2	run 4893	run 4896	4	run 4734	0	run 4728	4
run 1087	2001	run 4897	4	0	run 4900	run 4903	5	run 4741	0	run 4735	4
run 1087	2002	run 4898	6	0	run 4901	run 4904	4	run 4742	0	run 4736	3
run 1087	2003	run 4899	5	0	run 4902	run 4905	4	run 4743	0	run 4737	4
run 1089	2001	run 4906	4	0	run 4909	run 4912	3	run 4750	0	run 4744	5
run 1089	2002	run 4907	5	0	run 4910	run 4913	4	run 4751	0	run 4745	3
run 1089	2003	run 4908	4	0	run 4911	run 4914	4	run 4752	0	run 4746	3
run 1103	2001	run 4915	4	0	run 4918	run 4921	4	run 4759	0	run 4753	4
run 1103	2002	run 4916	5	0	run 4919	run 4922	4	run 4760	0	run 4754	4
run 1103	2003	run 4917	3	0	run 4920	run 4923	4	run 4761	0	run 4755 extinct	3
run 2568	2001	run 4924	4	0	run 4927	run 4930	3	run 4768	1	run 4762	4
run 2568	2002	run 4925	5	0	run 4928	run 4931	4	run 4769	0	run 4763	4
run 2568	2003	run 4926	7	0	run 4929	run 4932	4	run 4770	0	run 4764	4
run 2577	2001	run 4933	4	0	run 4936	run 4939	4	run 4777	0	run 4771	4
run 2577	2002	run 4934	4	0	run 4937	run 4940	4	run 4778	0	run 4772	2
run 2577	2003	run 4935	5	0	run 4938	run 4941	4	run 4779	0	run 4773	2
run 2590	2001	run 4942	4	0	run 4945	run 4948	3	run 4786	0	run 4780	4
run 2590	2002	run 4943	5	0	run 4946	run 4949	4	run 4787	0	run 4781	6
run 2590	2003	run 4944	4	0	run 4947	run 4950	3	run 4788	0	run 4782	4
run 2654	2001	run 4951	5	0	run 4954	run 4957	4	run 4795	0	run 4789	4
run 2654	2002	run 4952	4	2	run 4955	run 4958	3	run 4796	0	run 4790	3
run 2654	2003	run 4953	5	0	run 4956	run 4959	4	run 4797	0	run 4791	3
run 2676	2001	run 4960	4	0	run 4963	run 4966	4	run 4804	0	run 4798	2
run 2676	2002	run 4961	5	2	run 4964	run 4967	4	run 4805	0	run 4799	4
run 2676	2003	run 4962	4	0	run 4965	run 4968	4	run 4806	0	run 4800	4
run 2700	2001	run 4969	5	0	run 4972	run 4975	4	run 4813	0	run 4807	2
run 2700	2002	run 4970	4	0	run 4973	run 4976	4	run 4814	0	run 4808	3
run 2700	2003	run 4971	4	0	run 4974	run 4977	4	run 4815	0	run 4809	1
run 2771	2001	run 4978	3	0	run 4981	run 4984	4	run 4822	0	run 4816	4
run 2771	2002	run 4979	8	1	run 4982	run 4985	4	run 4823	0	run 4817	4
run 2771	2003	run 4980	6	1	run 4983	run 4986	3	run 4824	0	run 4818	4
run 2802	2001	run 4987	4	0	run 4990	run 4993	4	run 4831	0	run 4825	7
run 2802	2002	run 4988	4	0	run 4991	run 4994	3	run 4832	0	run 4826	4
run 2802	2003	run 4989	4	0	run 4992	run 4995	4	run 4833	0	run 4827	4
run 2862	2001	run 4996	5	0	run 4999	run 5002	4	run 4840	0	run 4834	4
run 2862	2002	run 4997	4	0	run 5000	run 5003	4	run 4841	1	run 4835	4
run 2862	2003	run 4998	6	0	run 5001	run 5004	5	run 4842	0	run 4836	3

The ladder, descriptive only: how many children's worlds held each rung at a share of 10% for 5 samples running, settling window included, and the median first epoch they did.

Rung	Meta-stack: children	Meta-stack: median first epoch	Meta-stack-deep-only: children	Meta-stack-deep-only: median first epoch	Meta-inplace: children	Meta-inplace: median first epoch	Logic-none: children	Logic-none: median first epoch	Logic-full: children	Logic-full: median first epoch
echo	54	20,130	54	25,310	54	20,120	19	22,630	54	20,690
not	54	21,040	0	—	54	20,800	0	—	52	23,660
nand	43	21,500	0	—	21	21,250	0	—	37	25,680
and	17	22,200	0	—	0	—	0	—	14	28,550
orn	54	21,170	0	—	54	21,120	0	—	44	26,420
or	51	21,760	0	—	54	21,370	0	—	23	27,890
andn	49	21,340	0	—	38	21,650	0	—	32	29,810
nor	51	21,920	0	—	54	21,540	0	—	19	33,810
xor	7	23,180	0	—	0	—	0	—	0	—
equ	9	22,720	0	—	0	—	0	—	0	—

The deep children, measured offline

The pre-registration locks three readings for each deep child, made after the sweep on its stored worlds with the study's landscape tools, research/landscape/. They were made on 2026-10-02 at main 3cd9abb from the lab's stored worlds, read through SELECTs only. They are offline measurements, not this page's code, so the table below is typed from their summary (one row per planting), whose README gives the scripts as run. The first epochs in it equal the live table's above.

Per deep child: its tape cap (max_tape_len), its first deep rung, the locked substitution distance d, the nearest deep mutant of its dominant tape (not locked), heritability by planting, and the circuit behind its first deep solver, traced on x = 0x5a, y = 0x33.

Child	Parent, seed	Cap	First deep rung	d	Nearest deep mutant	Heritability	Circuit
4845	944, 2003	128	XOR 21 870 (EQU 21 940)	25	2	unresolved	2 NANDs a lap; XOR in 4 NANDs, EQU a lap later in 6; fans out x, y, t1
4862	967, 2002	128	EQU 21 700 (XOR 21 760)	31	none within 3	unresolved	4 NANDs a lap; XOR in 6, EQU in 8; fans out x, y, t1, t2
4863	967, 2003	128	EQU 22 690 (XOR 33 530)	25	3	unresolved	3 NANDs a lap; EQU in 6; XOR only later, a 12-NAND chain ending in a double negation; fans out x, y, t2, t3
4871	991, 2002	128	XOR 23 180 (EQU 23 220)	22	2	unresolved	4 NANDs a lap; XOR in 6, EQU in 12; fans out x, y, t1, t3
4890	1029, 2003	256	EQU 22 880 (XOR 28 290)	17	already an EQU solver	unresolved	2 NANDs a lap after a copy prologue; EQU in 7, one a NOT; fans out x, y, t2, t3. Its later XOR solver: 1 NAND a lap, XOR in 5
4952	2654, 2002	128	EQU 28 220 (XOR 28 790)	30	3	unresolved	4845's circuit: 2 NANDs a lap, XOR in 4, EQU in 6
4961	2676, 2002	128	XOR 21 630 (EQU 21 700)	26	none within 3	unresolved	a prologue that emits x, then 2 NANDs a lap; XOR in 4, EQU in 6; fans out x, y, t3
4979	2771, 2002	256	EQU 41 480 (no XOR)	20	already an EQU solver	not heritable	1 NAND a lap after a 2-NAND prologue; EQU in 5, built on NOT x; fans out y, NOT x, t2
4980	2771, 2003	256	EQU 22 720 (no XOR)	24	3	unresolved	1 NAND a lap; EQU in 5; fans out x, y, t1, t2

The substitution distance says nothing here. The entry counts a deep rung as a multi-step crossing where d, the distance from the world's dominant metabolism tape when the last lower rung reached the line to the first deep solver after it, is at least 2. All 9 read d ≥ 2, from 17 to 31, so all 9 are "multi-step" by the locked rule. The rule cannot tell here. The worlds hold 9 634 to 14 375 distinct metabolism tapes among 16 384 cells, and the dominant one sits on only 9 to 21 cells. Drift puts it a median 22 to 30 of 32 bytes from a cell of its own world. So the test could only fail where the dominant tape is the deep solver itself. Read with the study's own tool instead, every mutant within 3 substitutions of the dominant tape (not locked):

no child has a credited path within 3, one that gains a paid unit at every step;
the nearest deep mutant is 2 substitutions away in 4845 and 4871, 3 in 4863, 4952 and 4980, and beyond 3 in 4862 and 4961;
in 4890 and 4979 the dominant tape was already an EQU solver at the line, below a tenth of the world, so no rung was crossed between the two tapes the distance compares.

Heritability: 0 heritable, 8 unresolved, 1 not heritable (4979). Each first deep solver was planted back into its own end world, every other carrier of the rung wiped, beside a control with the same wipe and no planting. In 14 of the 15 locked plantings, every seed's control re-grew the deep rung to a tenth of the world within 500 epochs, level with the planted world, which the entry reads as unresolved. In 4979 neither world reached a tenth. In 10 of the 16 plantings, a labelled extension for 4890's XOR included, the tape the wipe put in place, the world's commonest tape not credited the rung, is one substitution from it. So in these worlds the deep rung is a property of the population: re-derived within hundreds of epochs, often from abundant lower-rung tapes one substitution below it, not a single lineage's invention. That is what the design study's §7.6 describes: "The soup then evolves loop-shaped ORN/ANDN/OR solvers that sit one substitution below XOR/EQU, and takes that step."

The circuits. Every deep solver is a loop that stacks NANDs leftward, lap after lap, and reads earlier laps' results. Each is credited its rung on all 18 traced cases and fans out an input. They compute XOR in 4 to 6 NANDs, but for 4863's, a 12-NAND chain ending in a double negation, and EQU in 5 to 12.

What it means

Given a tape apart from the copier, read by a NAND that keeps its operands, and paid for every rung, a soup derived from emerged BFF worlds assembled XOR and EQU in a sixth of its worlds. Paid for those two alone it built nothing deep, and with the NAND that writes in place it built nothing deep either. That is Lenski et al.'s 2003 stepping-stone mechanism, complex features from rewarded simpler ones, in a population that began as a soup. It also turns the design study's pilot reading of the Logic stall into a measurement: with the copier out of the path, the in-place NAND still built no deep rung (H-deep-M refuted), and the stack NAND did (H-stack shown).

It holds on a machine built for it. The substrate imports four things: an objective, the paid tasks; a primitive, the NAND byte; a hereditary channel, since the world, not the program, copies the metabolism tape; and the primitive's semantics, the stack NAND, chosen among seven because it was the one under which the pilot's deep rungs came.

What is not claimed

Nothing about rung 4 on Soup. Whatever rose here was paid for by tasks the programme chose, computed by a byte and a tape the programme added. It is never pooled with the fitness-free runs, and rung 4 on Soup stays "not shown".
Not open-ended. The ladder tops out at EQU. A child that holds it has nowhere further to be paid for.
Depth is what the loop's later laps compute. As the topless study puts it (§2.3–2.4), it is "the circuit complexity of the function the soup's lap structure happens to compute", a real computation, not a count of accumulated steps.
Not a lineage's invention. No deep rung read heritable, and the substitution distance does not separate a crossing from drift in these worlds.
Two parents, under one re-reading. The deep tests hold whichever one or two parents are left out of all 18. Without the piloted parents they read 8 to 0, and that re-reading rests on parents 967 and 2771, which gave 4 of the 9 deep children.
Children are correlated. Three children share each parent's world, four parents were piloted, and two arms are the Logic sweep's own children.

What comes next

The ladder here has a near top. The topless study asks whether depth keeps rising when it has none: nested three- and four-input logic ladders, each rung paid by its exact minimal NAND count, read for whether the deepest rung held keeps rising late in long runs. Its minimal-NAND tables are computed; the engine slice that adds the ladders is in review. A pre-registration of its own comes after.

Evidence

Transition, flagged, census and arm are used here in the senses set out in the glossary.

Sweep: Meta-stack

Transition epoch vs treatment

No finished runs yet.

Transitions and replicator counts

One row per finished run the detector flagged or that counted at least one replicator. "Transition epoch" is the detector's event; "Peak replicators" is the largest census the run reached, with the epoch it first reached it in brackets; "—" means the run never had a value.

No finished run of this sweep has transitioned or counted a replicator yet.

Runs

Every run of the sweep, flagged or not, with its seed — the denominator of the rate above.

Seed	Treatment	Status	Priority	Epochs done	Transition epoch	Census
2002	logic×32×isa×0.00391×own_tape×stack×8×xor×2048×initiator×0×1024×65536	finished	40	60,000	— confirmed	22 census only	Run
2003	logic×32×isa×0.00391×own_tape×stack×8×xor×2048×initiator×0×1024×65536	finished	40	60,000	— confirmed	23 census only	Run
2001	logic×32×isa×0.00391×own_tape×in_place×8×2048×initiator×0×1024×65536	finished	40	60,000	— confirmed	337 census only	Run
2002	logic×32×isa×0.00391×own_tape×in_place×8×2048×initiator×0×1024×65536	finished	40	60,000	— confirmed	39 census only	Run
2003	logic×32×isa×0.00391×own_tape×in_place×8×2048×initiator×0×1024×65536	finished	40	60,000	— confirmed	431 census only	Run
2001	logic×32×isa×0.00391×own_tape×stack×8×2048×initiator×0×1024×65536	finished	40	60,000	— confirmed	—	Run
2002	logic×32×isa×0.00391×own_tape×stack×8×2048×initiator×0×1024×65536	finished	40	60,000	— confirmed	28 census only	Run
2003	logic×32×isa×0.00391×own_tape×stack×8×2048×initiator×0×1024×65536	finished	40	60,000	— confirmed	31 census only	Run
2001	logic×32×isa×0.00391×own_tape×stack×8×xor×2048×initiator×0×1024×65536	finished	40	60,000	— confirmed	28 census only	Run
2002	logic×32×isa×0.00391×own_tape×stack×8×xor×2048×initiator×0×1024×65536	finished	40	60,000	— confirmed	9 census only	Run
2003	logic×32×isa×0.00391×own_tape×stack×8×xor×2048×initiator×0×1024×65536	finished	40	60,000	— confirmed	14 census only	Run
2001	logic×32×isa×0.00391×own_tape×in_place×8×2048×initiator×0×1024×65536	finished	40	60,000	— confirmed	37 census only	Run
2002	logic×32×isa×0.00391×own_tape×in_place×8×2048×initiator×0×1024×65536	finished	40	60,000	— confirmed	—	Run
2003	logic×32×isa×0.00391×own_tape×in_place×8×2048×initiator×0×1024×65536	finished	40	60,000	— confirmed	32 census only	Run
2001	logic×32×isa×0.00391×own_tape×stack×8×2048×initiator×0×1024×65536	finished	40	60,000	— confirmed	19 census only	Run
2002	logic×32×isa×0.00391×own_tape×stack×8×2048×initiator×0×1024×65536	finished	40	60,000	— confirmed	103 census only	Run
2003	logic×32×isa×0.00391×own_tape×stack×8×2048×initiator×0×1024×65536	finished	40	60,000	— confirmed	18 census only	Run
2001	logic×32×isa×0.00391×own_tape×stack×8×xor×2048×initiator×0×1024×65536	finished	40	60,000	— confirmed	26 census only	Run
2002	logic×32×isa×0.00391×own_tape×stack×8×xor×2048×initiator×0×1024×65536	finished	40	60,000	— confirmed	12 census only	Run
2003	logic×32×isa×0.00391×own_tape×stack×8×xor×2048×initiator×0×1024×65536	finished	40	60,000	— confirmed	15 census only	Run