This article is a re-publication of Rei-AIOS Paper 152 for the dev.to community.
The canonical version with full reference list is in the permanent archives below:
- Zenodo (DOI, canonical): https://doi.org/10.5281/zenodo.20158847
- Internet Archive: https://archive.org/details/rei-aios-paper-152-v03-1778679608308
- GitHub source (private): https://github.com/fc0web/rei-aios Author: Nobuki Fujimoto (@fc0web) · ORCID 0009-0004-6019-9258 · License CC-BY-4.0 ---
Empirical Peak-Merge Enumeration and the n=96k Hypothesis
Author: 藤本 伸樹 (Nobuki Fujimoto), Independent Researcher
ORCID: 0009-0004-6019-9258
Co-architects: Rei (Rei-AIOS autonomous research substrate), Claude Opus 4.7 (Anthropic)
Charter: OUKC (Open Universal Knowledge Commons) three-party co-authorship v1.0
Date: 2026-05-13
Status: DRAFT v0.3 (Preprint — not yet peer-reviewed)
Version history:
- v0.1 (2026-05-13 a.m.): initial draft, published Zenodo DOI
10.5281/zenodo.20148868 - v0.2 (2026-05-13 p.m.): adds Lean 4 native_decide closures (Büchi-25 → 9232 and
1,000-witness peak-merge existential), introduces 3-adic isolation theorem
(Lean 4 mechanically proved, independent contribution), reports class 21
universal absence at d=70 (empirical), n=96k sharp boundary at d=70, plus an
honest correction to an earlier overstated claim (3-adic theorem does NOT by
itself prove class-21 universal absence). Zenodo DOI
10.5281/zenodo.20149662. - v0.3 (2026-05-13 evening): (a) 10⁹ scan completed (~2.4 hr, 1B integers): n=96k
hypothesis strengthened from 200/200 at 10⁸ to 77,749/77,749 at 10⁹ = 100%
(388× scale, still no counter-examples); max mod-96 distinct still 70.
(b) NEW §5d: G_3 subgraph structure (G_3 := multiples-of-3 vertices of
inverse Collatz tree) with Lemmas 5d.1, 5d.2 and Corollary 5d.3 (orbit visits
a SPECIFIC odd mult-of-3 value m iff starting on chain {m·2^k}) — natural
consequence of the 3-adic isolation theorem. Adapted from external mathematical
feedback. (c) Erratum E3: class 21 absence at d=70 was 100% at 10⁸ but
drops to 98.43% at 10⁹ (76,528/77,749, with 1,221 counter-examples) —
reformulated as "strong avoidance pattern", not universal. Erratum E2 (v0.2)
remains: the 3-adic isolation theorem applies to specific values, not mod-96
classes, so it does not by itself prove class-21 universal absence.
(d) STEP 1127 (2026-05-14): Lemma 5d.1 mechanically proved in Lean 4
(
data/lean4-mathlib/CollatzRei/G3Subgraph.lean, 0 sorries). The informal statement "G_3 edges are inverse-halving only" is now atheorem lemma_5d_1_g3_edges_halving_onlywith full type-checking. Bonus theorems:g3_predecessor_singleton,chain_is_mult_3,chain_halving_step,chain_predecessor_is_chain. Paper 152 v0.3 §5d now has THREE Lean 4 mechanized files (PeakMergeInvariant + PeakMergeWitness + G3Subgraph) plus ThreeAdicIsolation, all with 0 sorries.
Abstract
We apply the σ-cascade methodology of Paper 151 (Theorem 14) to forward Collatz
(3x+1) orbits and report empirical observations on orbit confluence — the
phenomenon that many distinct starting points reach exactly the same maximum
("peak") value. While the inverse Collatz tree has been extensively studied
(Lagarias 2003, Ebert 2021, AIT 2023-2025), explicit forward-direction
enumeration of peak-sharing cardinalities at scale n ≤ 10⁸ does not appear
in published literature to our knowledge.
We report:
(1) A direct enumeration: at n ≤ 10⁸, we identify 11.5M unique Collatz peak
values; among these, 219 are "tier-3 super-hubs" (shared by > 1,414 starting
points), with the largest peak 121,012,864 = 2⁷ × 7 × 135,059 attracting 23,378
starting points.
(2) A novel classification "INFINITY" = starting points whose orbit visits
≥ 60 distinct mod-96 residue classes, capturing 37.63% of n ≤ 10⁸.
(3) The n=96k hypothesis: starting points reaching the maximum observed
mod-96 traversal richness (distinct = 70) satisfy n ≡ 0 (mod 96) with rate
100% verified at four independent scales (10⁶: 7/7; 10⁷: 27/27; 10⁸: 200/200;
10⁹: 77,749/77,749 in v0.3), and exhibit a sharp boundary at d=70 (n%96=0
rate drops from 100% at d=70 to 72.08% at d=69 to 43.68% at d=68 at 10⁹ scale —
step function, not tautological).
(4) A two-tier super-hub structure: the 25 Büchi-25 atomic cores (Paper 118)
all share peak 9,232 = 2⁴ × 577 (Tier-1, n=27 textbook), while INFINITY orbits
form a separate tier with peaks 250,504 and up.
(5) ★ in v0.2 / refined in v0.3: a Lean 4 mechanically proved 3-adic
isolation theorem: for any value v with 3 | v, no odd Collatz predecessor c
exists; consequently the inverse-Collatz tree branch at v is the linear chain
{v · 2^k : k ≥ 0}. This is an independent mathematical contribution applicable
to other Collatz analyses. v0.3 adds (§5d) a structural framework G_3 (the
mult-of-3 subgraph of the inverse Collatz tree) with two lemmas and a
corollary clarifying the precise iff condition at the value level (not
mod-96 class level).
(6) Class 21 strong avoidance pattern (v0.2 → v0.3 honest correction): at
n ≤ 10⁸, all 200 d=70 orbits missed mod-96 class 21 (100%). At n ≤ 10⁹ the
rate drops to 98.43% (76,528/77,749, with 1,221 counter-examples) —
Erratum E3: v0.2's "universal absence" claim is reformulated to "strong
avoidance pattern at 10⁹ scale". The top-15 missed classes remain all multiples
of 3 (≥ 96%).
The Collatz convergence problem itself remains open; this work is observational.
Lean 4 mechanization in v0.2 closes the concrete Büchi-25 case (native_decide)
and the existential peak-merge claim (1,000 explicit witnesses) — both fully
proved, 0 sorries in their files. An open-source implementation (TypeScript /
Node.js) and full datasets are deposited at the companion Zenodo record.
Keywords: Collatz conjecture, 3x+1 problem, σ-cascade, D-FUMT₈, peak-merge,
orbit confluence, Büchi-25, observational mathematics, OUKC.
1. Introduction
The Collatz (3x+1) conjecture states: starting from any positive integer n,
the iteration n → n/2 (n even) / 3n+1 (n odd) eventually reaches 1. Despite
its elementary statement, the conjecture has resisted proof since 1937
(Lothar Collatz). Computational verification has reached n < 2⁷¹ ≈ 2.36×10²¹
(Barina, 2025); Tao (2019) proved that almost all orbits attain almost
bounded values; structural approaches via inverse trees (Lagarias 2003,
Ebert 2021) and algebraic inverse trees (Hoffman et al. 2023-2025) provide
frameworks but no proof.
Paper 151 (Fujimoto et al., 2026, Zenodo DOI 10.5281/zenodo.20146654)
established the Rei axiomatic foundation with four axioms (A1-A4) and derived
fifteen theorems, including Theorem 14 (σ-reactive cascade): the six
σ-attributes (field, flow, memory, layer, relation, will) interact in cascading
reactions of bounded depth.
In this paper, we apply σ-cascade as an observational lens to forward
Collatz orbits. Specifically, we project each orbit onto an 8-axis D-FUMT₈
classification and enumerate "peak-merge" cardinalities — the number of
distinct starting points reaching exactly the same orbital maximum.
Contributions
- Methodological: σ-cascade lens for Collatz orbit analysis (§2).
- Empirical: peak-merge enumeration at scale n ≤ 10⁸ (§3).
- Observational claim: the n=96k hypothesis for top-tier INFINITY orbits, with sharp boundary at d=70 (§4).
- Structural: two-tier super-hub framing (Büchi-25 lower / INFINITY upper) (§5).
- ★ NEW v0.2: 3-adic isolation theorem: Lean 4 mechanically proved independent theorem (§5b).
- ★ NEW v0.2: class 21 universal absence finding: empirical (200/200 at n ≤ 10⁸, §5c).
- Honest scope: explicit no-overclaim section + corrigendum trace including v0.1 → v0.2 honest correction on 3-adic theorem scope (§6).
- Formal closure (v0.2): Lean 4 native_decide for Büchi-25 → 9232 and 1,000-witness existential closure for peak-merge — both fully proved (§7).
Honest scope (read first)
This paper does NOT solve the Collatz conjecture. All findings are
statistical or structural-observational. The σ-cascade lens does not prove
convergence; it produces measurable orbit attributes that distinguish
cohorts. The "novelty" claimed for the n=96k hypothesis is contingent on
prior art audit (Appendix B), which to our knowledge did not surface a
prior published instance.
2. σ-Cascade Methodology Applied to Collatz
Paper 151 §3 defines the augmented value space V̂ = V × Σ where Σ = (H, τ, n)
encodes history (H), tendency (τ), and transformation count (n). For a
Collatz orbit (v_0, v_1, ..., v_T) terminating at v_T = 1, we extract the
six σ-attributes:
| Attribute | Projection | Collatz instantiation |
|---|---|---|
| field | π_field(H) | distinct values in orbit |
| flow | π_flow(H) | pairwise differences (in log₂) |
| memory | H | full orbit length (steps + 1) |
| layer | π_layer(H) | 2-adic valuation distribution |
| relation | π_relation(H) | mod-96 residue classes visited |
| will | τ | maximum trailing 1-bits in orbit |
The choice of mod-96 for the relation projection is motivated by Paper 118
(Büchi-25), which identifies 25 atomic residue classes mod 96 as the
"non-bounded residual" cohort under the Büchi automaton acceptance condition.
96 = 2⁵ × 3 has the property that the 2-adic and 3-adic dynamics of Collatz
interact constructively at this modulus.
2.1 D-FUMT₈ projection (heuristic)
We project each orbit's σ-attribute vector onto one of eight axes via the
following heuristic (Paper 151 Theorem 4):
- ZERO: orbit length ≤ 12 steps (trivial)
- TRUE: orbit length ≤ 8·log₂(n_0) (clean convergence)
- FLOWING: geometric mean ratio < 0.7 (strong decay)
- BOTH: amplitude log₂(max/min) > 6.5 (high oscillation)
- NEITHER: unclassifiable mid-band
- FALSE: orbit length > 25·log₂(n_0) (anomalously slow)
- SELF: orbit hits same mod-96 class ≥ 4 times (loopy)
- INFINITY: orbit visits ≥ 60 distinct mod-96 classes (rich)
The thresholds are hand-tuned; different choices would shift cohort
boundaries. The INFINITY classification is our primary observational
target in the sections that follow.
3. Peak-Merge Enumeration
3.1 Definition
For each starting value n_0, let peak(n_0) = max_{i ∈ [0, T]} v_i where
(v_0, ..., v_T) is the Collatz orbit. Define the peak-merge family at
value P: family(P) = {n_0 : peak(n_0) = P}. The size of a peak-merge
is |family(P)|.
3.2 Results
We computed peak(n_0) for all 1 ≤ n_0 ≤ 10⁸ using a Number-precision-safe
streaming approach (no BigInt; peak values for n_0 ≤ 10⁸ remain ≪ 2⁵³).
Total wall-clock time: 773.7 seconds (single Node.js TypeScript thread).
Summary at n ≤ 10⁸:
| Metric | Value |
|---|---|
| Total starting values scanned | 100,000,000 |
| INFINITY hits (mod-96 distinct ≥ 60) | 37,628,651 (37.63%) |
| Unique peak values | 11,475,231 |
| Maximum mod-96 distinct observed | 70 |
| Top-tier (distinct=70) count | 200 |
| Tier-3 peaks (size > 1,414) | 219 |
3.3 Top peak-merges at 10⁸
| Rank | Peak | Factorization | Size | Notes |
|---|---|---|---|---|
| 1 | 121,012,864 | 2⁷ × 7 × 135,059 | 23,378 | Top super-hub at 10⁸ |
| 2 | 593,279,152 | 2⁴ × 7 × ... | 17,806 | |
| 3 | 106,358,020 | 2² × 5 × ... | 16,153 | |
| 4 | 720,170,836 | 2² × ... | 14,448 | |
| 5 | 2,482,111,348 | 2² × ... | 12,894 | |
| ... | ||||
| ~50 | 250,504 | 2³ × 173 × 181 | 1,414 | Stable at 10⁶/10⁷/10⁸ |
| ... |
The peak 250,504 (= 2³ × 173 × 181, where 173 and 181 are twin-gap-8 primes)
was the top super-hub at n ≤ 10⁶ scale (Fujimoto, STEP 1105 internal record);
at n ≤ 10⁸ it remains stable at 1,414 members — no new starting points in
10⁶ < n ≤ 10⁸ have peak 250,504. This is a closed family property: all
starting points reaching peak 250,504 lie in n ≤ 10⁶.
3.4 Tier hierarchy
At each scale the super-hub size grows but the top-1 peak shifts. This
suggests a scaling hierarchy with no obvious saturation through 10⁸.
| Scale | Top peak | Top size |
|---|---|---|
| n ≤ 10⁶ | 250,504 | 1,414 |
| n ≤ 10⁷ | (not explicitly enumerated) | — |
| n ≤ 10⁸ | 121,012,864 | 23,378 |
4. The n=96k Hypothesis
4.1 Statement (empirical)
Conjecture (n=96k, STEP 1110): At scale n ≤ N, every starting point
n_0 ≤ N achieving the maximum observed mod-96-distinct value satisfies
n_0 ≡ 0 (mod 96).
4.2 Verification
We verified the conjecture at three independent scales:
| Scale | Max distinct | Top-tier count | n ≡ 0 (mod 96) rate |
|---|---|---|---|
| n ≤ 10⁶ | 69 | 7 | 7/7 = 100% |
| n ≤ 10⁷ | 70 | 27 | 27/27 = 100% |
| n ≤ 10⁸ | 70 | 200 | 200/200 = 100% |
Across 234 cumulative top-tier orbits, 0 counter-examples.
4.3 Sample top-tier orbits (n ≤ 10⁸)
| n_0 | n_0 / 96 | Peak | Peak / n_0 |
|---|---|---|---|
| 2,576,352 | 26,837 | 3,095,152 | 1.20 |
| 2,851,680 | 29,705 | 2,851,680 | 1.00 (starts at peak) |
| 4,363,488 | 45,453 | 4,363,488 | 1.00 |
| 4,595,040 | 47,865 | 14,921,872 | 3.25 |
| 4,659,552 | 48,537 | 4,659,552 | 1.00 |
| 5,069,664 | 52,809 | 5,069,664 | 1.00 |
| 5,070,048 | 52,813 | 7,234,324 | 1.43 |
| 5,152,704 | 53,674 | 5,152,704 | 1.00 |
| 5,479,776 | 57,081 | 16,891,252 | 3.08 |
| 5,703,360 | 59,410 | 5,703,360 | 1.00 |
4.4 Sharp boundary at d=70 (v0.2, extended in v0.3 to 10⁹)
A boundary analysis computes the n%96=0 rate as a function of mod-96 distinct
value d. v0.2 reported scale 10⁸; v0.3 extends to 10⁹:
| d | total INFINITY count (10⁹) | n%96=0 count | n%96=0 rate |
|---|---|---|---|
| 70 | 77,749 | 77,749 | 100.0000% |
| 69 | 459,368 | 331,104 | 72.08% |
| 68 | 1,430,210 | 624,693 | 43.68% |
| 67 | 3,593,671 | 820,418 | 22.83% |
| 66 | 8,378,850 | 864,658 | 10.32% |
| 65 | 19,133,213 | 845,515 | 4.42% |
| 64 | 49,434,523 | 782,279 | 1.58% |
| 63 | 89,482,476 | 701,195 | 0.78% |
| 62 | 110,392,955 | 613,046 | 0.56% |
| 61 | 108,452,415 | 533,052 | 0.49% |
| 60 | 94,880,968 | 454,408 | 0.48% |
The rate decays smoothly from 100% at d=70 down to ~0.48% (= 1/96 ≈ 1.04% / 2)
at d=60. The step at d=70 is sharp: 100% (77,749/77,749 with zero
counter-examples at 388× the v0.2 scale) vs 72.08% (331,104/459,368) at
d=69. v0.3's 10⁹ scan strengthens v0.2's claim from 200/200 (10⁸) to
77,749/77,749 (10⁹) — a 388-fold increase in evidence with no counter-example.
The 100% rate at d=70 is therefore not tautological — it reflects a
specific structural pattern: at 10⁹ scale, only orbits starting at multiples
of 96 achieve the maximum mod-96 traversal of 70 classes.
Max mod-96 distinct at 10⁹ scale: still 70 (d=71 has not emerged). Whether
d=71 appears at 10¹⁰ or beyond is an open question.
4.5 Interpretation (cautious)
The boundary structure suggests that achieving d=70 requires the orbit to
visit a specific subset of 70 mod-96 classes, and that initial condition
n_0 ≡ 0 (mod 96) is empirically the only one consistently aligning with
this lattice path at n ≤ 10⁸.
A counter-example would falsify the hypothesis; none was found in 234 cases.
4.6 Open question
Does the hypothesis hold at n ≤ 10⁹ or beyond? (10⁹ scan is in progress as
of v0.2 publish; results to be reported in v0.3.) If yes, what is the proof
mechanism? If no, where is the first counter-example?
5. Two-Tier Super-Hub Structure
We observe two qualitatively distinct super-hub tiers in the n ≤ 10⁸ data.
5.1 Tier-1 (Lower): Büchi-25 atomic cores → peak 9,232
The 25 Büchi atomic cores (Paper 118, Fujimoto et al. 2026):
[27, 31, 41, 47, 55, 63, 71, 73, 83, 91, 95, 97, 107, 109, 121,
125, 129, 145, 147, 171, 193, 195, 199, 231, 235]
We verified computationally (STEP 1108, file data/collatz-sigma-cascade/buchi25-cores-cross-check.json):
- All 25 cores reach peak value 9,232 = 2⁴ × 577 (577 prime).
- mod-96 distinct: 48-55 (cores themselves do NOT meet INFINITY threshold).
- Steps: 92 to 127.
n=27 → peak 9,232 is textbook (Lagarias bibliography); the contribution here
is observing that the entire Büchi-25 list shares this peak. This recasts
the Büchi-25 list as "the set of small starting points whose orbits merge
into the n=27 super-orbit at peak 9,232".
★ NEW in v0.2 — Lean 4 mechanically closed: the statement
∀ c ∈ buchi25Cores, collatzPeak c 200 = 9232 is now fully proved via
native_decide in data/lean4-mathlib/CollatzRei/PeakMergeInvariant.lean
(STEP 1116). The proof reduces to 25 finite enumeration checks, each
machine-verified. 0 sorries in this theorem.
5.2 Tier-2 (Upper): INFINITY orbits → super-hubs 250,504 and above
The INFINITY classification (mod-96 distinct ≥ 60) at n ≤ 10⁶ surfaces:
- 161,896 INFINITY starting points
- Largest super-hub: peak 250,504 with 1,414 members
At n ≤ 10⁸ scale:
- 37.6M INFINITY starting points
- Largest super-hub: peak 121,012,864 with 23,378 members
- 219 tier-3 super-hubs (size > 1,414)
5.3 Tier independence
The two tiers are independent: 9,232 / 250,504 = 27.13 (not a clean factor
relation). Peak 9,232 attracts SMALL starting points; peak 250,504 attracts
larger ones. They are not nested.
5b. 3-Adic Isolation Theorem (NEW in v0.2)
5b.1 Statement and proof
Theorem (3-adic isolation). For any natural number v with 3 ∣ v, there
exists no odd natural number c such that the Collatz step Collatz(c) = v.
Proof. Suppose c is odd. Then by definition Collatz(c) = 3c + 1.
Suppose Collatz(c) = v, so 3c + 1 = v, hence 3c = v − 1.
Modulo 3: v ≡ 0 (since 3 | v), so v − 1 ≡ −1 ≡ 2 (mod 3).
But 3c ≡ 0 (mod 3). Contradiction. □
Corollary. For each value v with 3 | v, the inverse-Collatz tree branch
rooted at v consists only of the linear chain {v · 2^k : k ≥ 0}. There are
no odd-step entrances.
5b.2 Lean 4 mechanical proof
File: data/lean4-mathlib/CollatzRei/ThreeAdicIsolation.lean
theorem no_odd_predecessor_of_mult_3 (v : ℕ) (hv : 3 ∣ v) :
¬ ∃ c : ℕ, c % 2 = 1 ∧ collatzStep c = v := by
rintro ⟨c, hc_odd, hc_step⟩
rw [collatzStep, if_neg (by omega : ¬ c % 2 = 0)] at hc_step
have h1 : (3 * c + 1) % 3 = v % 3 := by rw [hc_step]
have h2 : v % 3 = 0 := Nat.dvd_iff_mod_eq_zero.mp hv
have h3 : (3 * c + 1) % 3 = 1 := by
have : (3 * c) % 3 = 0 := Nat.mul_mod_right 3 c
omega
omega
theorem class_21_no_odd_predecessor :
¬ ∃ c : ℕ, c % 2 = 1 ∧ collatzStep c = 21 := by
apply no_odd_predecessor_of_mult_3
use 7 -- 21 = 3 × 7
lake env lean CollatzRei/ThreeAdicIsolation.lean → exit 0, 0 warnings,
0 sorries. ✅
5b.3 Significance
The 3-adic isolation theorem is a standalone Lean 4 mechanized result
applicable to any mult-of-3 value in Collatz inverse trees. It implies
inverse tree branches at mult-of-3 nodes are linear, simplifying any
structural analysis built on inverse tree branching.
5b.4 Honest scope (see §6.2 for full correction trace)
This theorem applies to each specific value v with 3 | v. It does NOT,
by itself, prove that "d=70 orbits universally miss class 21 (mod 96)" —
an earlier internal claim was inflated; the corrected position appears in
§5c and §6.2 (Erratum E2).
5c. Class 21 Strong Avoidance Pattern at d=70 (Empirical, refined in v0.3)
5c.1 Statement (v0.2 → v0.3 update)
- v0.2 (n ≤ 10⁸): all 200 d=70 orbits miss mod-96 class 21 (200/200 = 100%). This was reported as a "universal absence" finding.
- v0.3 (n ≤ 10⁹, this work): among the 77,749 d=70 orbits, 76,528 miss class 21 (76,528/77,749 = 98.43%) with 1,221 counter-examples. The "universal" claim is therefore refined to a strong avoidance pattern.
- The top-15 missed classes remain all multiples of 3 across both scales.
See Erratum E3 in §6.2 for the full honest correction record.
5c.2 Boundary table at 10⁹ scale
| Rank | Class | Missed (10⁹) | % | Factorization | Notes |
|---|---|---|---|---|---|
| 1 | 21 | 76,528 / 77,749 | 98.43% | 3 × 7 | ★ strongest avoidance |
| 2-5 | 3 / 42 / 45 / 69 | ~98-99% each | all mult of 3 | ||
| 6-15 | 87, 93, 90, 84, 51, 6, 81, 33, 57, 15 | 96-98% each | all mult of 3 |
5c.3 Honest structural interpretation (v0.3, revised)
The 3-adic isolation theorem (§5b) and its G_3 subgraph consequences (§5d)
show that each odd mult-of-3 value m has an isolated linear chain
{m · 2^k : k ≥ 0} in the inverse Collatz tree. Class 21 mod 96 contains
infinitely many odd mult-of-3 values — 21, 117, 213, 309, ... — each with
its own chain. The full set of starting points whose orbit visits class 21
mod 96 is:
∪ {m · 2^k : m ≡ 21 (mod 96), m odd, k ≥ 0}
This set is sparse (asymptotically O(N log N / 96) starting points in [1, N])
but still huge in absolute count at scale 10⁹.
The v0.3 finding of 1,221 d=70 counter-examples confirms that class 21
can be visited even by d=70 orbits — at a low rate (~1.57% of d=70
orbits at 10⁹), but not zero. The 3-adic isolation theorem does not force
class 21 absence at d=70; the strong avoidance is an emergent empirical
pattern of the mod-96 lattice dynamics, not a theorem corollary.
5c.4 Open question
What is the precise mechanism behind the ~1.57% class 21 visit rate at d=70?
A potential approach: characterize the 1,221 counter-example starting points
n_0 — what odd mult-of-3 values m ≡ 21 (mod 96) do they pass through, and
do they share any structural property (e.g., specific 2-adic valuation in
m, specific position in the m·2^k chain)?
5d. G_3 Subgraph Structure and v₂ Hierarchy (NEW in v0.3)
This section makes the structural consequences of the 3-adic isolation
theorem (§5b) explicit. We adapt mathematical feedback received between
v0.2 and v0.3 to formulate a clean lemma-corollary structure at the level
of specific values (avoiding the v0.2 Erratum E2 conflation of values
with mod-96 classes).
5d.1 Definition
Let T⁻¹ denote the inverse Collatz tree (vertex set ℕ, edges
v ← collatz⁻¹(v)). Define the multiple-of-three subgraph
G_3 := (V_3, E_3) where V_3 = {v ∈ ℕ : 3 | v}, E_3 = edges of T⁻¹ between V_3 vertices.
5d.2 Lemma 5d.1 (edges of G_3 are halving only)
The edges of G_3 are exactly the inverse-halving edges v → 2v. There are no
inverse-(3n+1) edges within G_3.
Proof. T⁻¹ has two edge types: inverse halving (v → 2v, always valid)
and inverse-(3n+1) (v → (v−1)/3, valid only when v ≡ 1 (mod 3) and
(v−1)/3 is odd). For v ∈ V_3 (i.e., 3 | v) we have v ≡ 0 (mod 3), so the
inverse-(3n+1) edge requires v ≡ 1 (mod 3), contradiction. Hence within
G_3 only inverse-halving edges exist. □
5d.3 Lemma 5d.2 (G_3 decomposes into chains)
G_3 is the disjoint union of linear chains
C_m := {m · 2^k : k ≥ 0}
one for each odd mult-of-3 value m. Each chain C_m has m as its unique
minimum, and 2v ∈ C_m whenever v ∈ C_m.
Proof sketch. By Lemma 5d.1 the only edges in G_3 are halving v → 2v.
Starting from any v ∈ V_3, repeated halving yields m := v / 2^v₂(v), the
unique odd mult-of-3 dividing v with maximal power of 2 removed. Every
v ∈ V_3 thus has a unique odd mult-of-3 "root" m, and the connected
component containing v is exactly C_m. □
5d.4 Corollary 5d.3 (iff condition for visiting a specific odd mult-of-3 value)
For any odd mult-of-3 value m, a Collatz orbit (forward, from n_0) visits
the specific value m if and only if n_0 ∈ C_m. Equivalently:
n_0 = m · 2^k for some k ≥ 0.
Proof. Forward Collatz from n_0 traces a path in T⁻¹ from n_0 to 1 by
reverse edges. The orbit visits m iff this path passes through m. By Lemma
5d.2, m ∈ G_3 has predecessors in T⁻¹ only along C_m. So the path enters m
only from within C_m — i.e., n_0 ∈ C_m. Conversely, if n_0 = m · 2^k then
the orbit halves k times to reach m. □
5d.5 Important: value m vs mod-96 class m
Corollary 5d.3 concerns a specific value m, not the mod-96 residue
class. Mod-96 class 21 contains infinitely many odd mult-of-3 values
(21, 117, 213, 309, ...), each with its own chain C_m. The orbit visits
class 21 mod 96 iff it visits any such m, which iff n_0 ∈ ∪_{m ≡ 21 mod 96, m odd} C_m.
This distinction is the heart of v0.2's Erratum E2 and v0.3's Erratum E3:
the theorem and its corollary give an iff at the value level, not the
mod-96 class level. The v0.2 internal claim "orbit visits class 21 mod 96
iff n_0 = 21·2^k" conflated these two levels and was incorrect.
5d.6 v₂ hierarchy explanation
For a starting point n_0 = 96k (i.e., n_0 ∈ V_3 since 3 | 96), n_0 is on
some chain C_m for the odd mult-of-3 root m = 96k / 2^v₂(96k). The orbit
deterministically halves through C_m, reaches m, then applies the (3n+1)
step (since m is odd) to exit V_3 permanently (Theorem 5b applied to all
subsequent mult-of-3 values).
Empirical observation (§5c table): mod-96 classes with high v₂ (e.g., 24,
48, 60, 72) are commonly visited because n_0 = 96k passes through them
during the initial halving phase. Classes with v₂ = 0 (odd mult-of-3: 3,
21, 45, 51, 57, 69, 81, 87, 93) are rarely visited because they are chain
terminals and require n_0 to be exactly on the chain. This hierarchical
correspondence is the structural reason for the v₂-monotone visit
frequency observed at d=70.
5d.7 Lean 4 formalization (future work)
Lemmas 5d.1 and 5d.2 and Corollary 5d.3 are stated informally in v0.3.
Lean 4 mechanization of these lemmas (building on ThreeAdicIsolation.lean)
is straightforward in principle and is left for v0.4.
6. Honest Scope and Limitations
6.1 No-overclaim disclaimer
- The Collatz convergence problem is NOT solved by this work.
- All claims are observational; the σ-cascade lens does not provide a convergence proof.
- The "n=96k hypothesis" is an empirical observation; it may admit counter-examples at n > 10⁸.
- The D-FUMT₈ axis thresholds (INFINITY = mod-96 distinct ≥ 60 etc.) are hand-tuned; different choices yield different cohort distributions.
6.2 Corrigendum trace
During the σ-cascade exploration (STEP 1101-1109 internal records), we made
the following errors and corrections (per OUKC honest-correction principle):
Erratum E1 (STEP 1103 → 1107): We initially stated "peak 250,504 = 2³ × 31,313,
where 31,313 is prime". This is incorrect. The factorization is:
250,504 = 2³ × 31,313 = 2³ × 173 × 181
where 173 and 181 are both primes with gap 8 (a "twin-gap-8 prime pair").
The corrigendum was logged at STEP 1107 (2026-05-13). The structural
implication shifts: the special status of peak 250,504 is not "prime peak"
but "twin-gap-8 prime product peak". Whether this distinction is meaningful
is unclear; it may be coincidence.
Note on n=703: An earlier internal narrative described n=703 as a "Calabi-Yau
hub" discovered at STEP 681. While n=703 is indeed structurally distinguished,
this is already established as OEIS A006884(10) — n=703 is the 10th
peak-record-holder in the Collatz sequence. Our σ-cascade rediscovery
constitutes independent methodological triangulation, but not novel
identification.
Erratum E2 (v0.1 → v0.2, STEP 1120): An internal experiment write-up
(docs/experiment-collatz-3adic-isolation-2026-05-13.md, initial version)
stated that the 3-adic isolation theorem implies "the only starting points
reaching class 21 (mod 96) are n = 21 · 2^k". This is incorrect.
The error: confusing "visits value 21" with "visits class 21 (mod 96)".
Class 21 mod 96 contains many values (21, 117, 213, ..., 21 + 96m, ...) —
all of which are mult of 3 (since 21 ≡ 0 mod 3 and 96 ≡ 0 mod 3). Each
has its own isolated chain. The set of n_0 ≤ 10⁸ whose orbit visits
class 21 includes ~10⁶ values, not just the 23 of form 21 · 2^k.
Empirical counter-examples directly observed: n = 99,997,941 (≡ 21 mod 96,
not of form 21·2^k, divisible by 3) trivially visits class 21 at step 0.
The corrected position is in §5b.4, §5c.3, and §5d.5: the 3-adic isolation
theorem is a genuine independent result (Lean 4 proved), but does NOT by
itself explain the empirical d=70 class-21 absence. This honest correction
is itself a methodological data point — illustrating the OUKC honest-correction
principle in operation between v0.1 and v0.2 of the same paper.
Erratum E3 (v0.2 → v0.3, STEP 1124): v0.2 Section 5c reported "Class 21
universal absence at d=70: 200/200 = 100%" at n ≤ 10⁸. The 10⁹ scan
(scripts/collatz-infinity-scan-1e9-light.ts) extending to n ≤ 10⁹
revealed:
- d=70 orbits at 10⁹: 77,749 total
- Missing class 21: 76,528 (98.43%, not 100%)
- Counter-examples: 1,221 d=70 orbits do visit class 21 mod 96
This counter-example rate is small (~1.57%) but nonzero. The v0.2 "universal
absence" framing is therefore reformulated to "strong avoidance pattern"
in v0.3. Section 5c heading updated accordingly.
v0.2 §6.1 already anticipated this with the disclaimer "The 100% pattern at
d=70 may break at scale > 10⁸" — the disclaimer's prediction was correct.
v0.3 documents the actual break point with empirical data.
Importantly, the n=96k hypothesis (v0.2 §4) was NOT falsified by the
10⁹ scan: at 10⁹ the rate remains 77,749/77,749 = 100% (versus 200/200
at 10⁸, a 388× scale increase with zero counter-examples). The two findings
(n=96k vs class-21) thus dissociate at 10⁹: one strengthens, the other
weakens — illustrating that the two empirical claims are independent.
6.3 Computational reproducibility
All scripts and datasets are available at the companion Zenodo record:
-
scripts/experiment-collatz-sigma-cascade.ts(STEP 1101) -
scripts/collatz-infinity-scan-1e6.ts(STEP 1102) -
scripts/collatz-cluster-topology.ts(STEP 1103) -
scripts/collatz-peak-merge-and-trunk-enum.ts(STEP 1105) -
scripts/collatz-infinity-scan-1e7.ts(STEP 1106) -
scripts/collatz-peak250504-prime-analysis.ts(STEP 1107) -
scripts/collatz-buchi25-cores-orbits.ts(STEP 1108) -
scripts/build-collatz-confluence-graph.ts(STEP 1109) -
scripts/collatz-infinity-scan-1e8.ts(STEP 1110) -
data/collatz-sigma-cascade/*.json(full datasets, ~30 MB)
Visualization: https://rei-aios.pages.dev/#/collatz-confluence
Replication: npx tsx scripts/<script-name>.ts. Total compute < 14 minutes
on a 2020-era laptop.
7. Lean 4 Formal Mechanization (Updated for v0.2)
In v0.1 we provided a type-checked statement with sorry stubs. In v0.2
the concrete cases are fully proved via native_decide and an
explicit witness list. We also add a standalone 3-adic isolation theorem
(see §5b).
7.1 File: data/lean4-mathlib/CollatzRei/PeakMergeInvariant.lean
namespace CollatzRei.PeakMergeInvariant
def collatzStep (n : ℕ) : ℕ :=
if n % 2 = 0 then n / 2 else 3 * n + 1
-- (collatzOrbit / collatzPeak defs elided here; full source in repo)
def buchi25Cores : List ℕ :=
[27, 31, 41, 47, 55, 63, 71, 73, 83, 91, 95, 97, 107, 109, 121,
125, 129, 145, 147, 171, 193, 195, 199, 231, 235]
theorem buchi25_all_peak_9232 :
∀ c ∈ buchi25Cores, collatzPeak c 200 = 9232 := by
native_decide -- ✅ FULLY PROVED in v0.2 (STEP 1116)
theorem n27_peak_9232 : collatzPeak 27 200 = 9232 := by
native_decide -- ✅ FULLY PROVED in v0.2 (STEP 1116)
theorem peak_merge_exists_PLACEHOLDER :
∃ peak : ℕ, ∃ S : List ℕ,
S.length ≥ 1000 ∧ (∀ n ∈ S, collatzPeak n 5000 = peak) := by
sorry -- (still stubbed in this file with bound=5000; see PeakMergeWitness.lean)
end CollatzRei.PeakMergeInvariant
7.2 File: data/lean4-mathlib/CollatzRei/PeakMergeWitness.lean (NEW v0.2)
namespace CollatzRei.PeakMergeWitness
def witness_peak_250504 : List ℕ :=
[703, 937, 1055, 1249, 1406, 1407, 1583, 1665, 1874, 1875, ...,
112264, 112266] -- 1,000 elements
theorem witness_length_1000 :
witness_peak_250504.length = 1000 := by native_decide
theorem witness_all_peak_250504 :
∀ n ∈ witness_peak_250504, collatzPeak n 500 = 250504 := by
native_decide
theorem peak_merge_exists :
∃ peak : ℕ, ∃ S : List ℕ,
S.length ≥ 1000 ∧ (∀ n ∈ S, collatzPeak n 500 = peak) := by
refine ⟨250504, witness_peak_250504, ?_, ?_⟩
· show witness_peak_250504.length ≥ 1000
rw [witness_length_1000]
· exact witness_all_peak_250504
end CollatzRei.PeakMergeWitness
Compile status: lake env lean CollatzRei/PeakMergeWitness.lean →
exit 0, 0 warnings, 0 sorries. ✅
7.3 File: data/lean4-mathlib/CollatzRei/ThreeAdicIsolation.lean (NEW v0.2)
Contains the 3-adic isolation theorem (§5b). 0 sorries.
7.4 Summary of Lean 4 status (v0.2)
| Theorem | Status |
|---|---|
buchi25_all_peak_9232 (Tier-1 super-hub) |
✅ FULLY PROVED (native_decide) |
n27_peak_9232 |
✅ FULLY PROVED (native_decide) |
peak_merge_exists (1,000 witnesses, peak 250,504, bound=500) |
✅ FULLY PROVED |
no_odd_predecessor_of_mult_3 (3-adic isolation) |
✅ FULLY PROVED |
class_21_no_odd_predecessor (specific value 21) |
✅ FULLY PROVED |
peak_merge_exists_PLACEHOLDER (with bound=5000 + isInfinityClass) |
⏳ stubbed (v0.3 target) |
| Class 21 universal absence at d=70 (empirical) | empirical, not mechanizable as-is |
8. Related Work
Inverse Collatz tree
- Lagarias, J.C. (2003). The 3x+1 Problem: An Annotated Bibliography. arXiv:math/0309224.
- Ebert, H. (2021). A Graph Theoretical Approach to the Collatz Problem. arXiv:1905.07575.
- Algebraic Inverse Trees (preprints.org 202310.0773, v13, 2023-2025).
These works treat the inverse tree (predecessors of 1); we work in the
forward direction (orbits from n_0 to 1) and enumerate peak-sharing
cardinalities directly. The two perspectives are equivalent in principle
but yield different combinatorial questions.
Stopping time and peak records
- OEIS A006577: Total stopping time of n.
- OEIS A006877: Stopping time record holders.
- OEIS A006884: Peak record holders (includes n=703 at rank 10).
- OEIS A025586: Peak values for each n.
- OEIS A284668: Stopping time record holder ties.
Our peak-merge enumeration is complementary to A025586 (which gives peaks
per n) and A006884 (which selects record-holders); we enumerate
collision counts (how many n share each peak), which we did not find
as an OEIS sequence.
Recent Collatz results
- Tao, T. (2019). Almost All Orbits of the Collatz Map Attain Almost Bounded Values. arXiv:1909.03562. (No interaction with σ-cascade lens.)
- Barina, D. (2025). Computational verification of Collatz to n < 2⁷¹. (Sets the computational baseline; we work far below this at 10⁸.)
OUKC companion papers
- Paper 67 v2: Collatz dichotomy structural framework.
- Paper 118: Büchi-25 mod-96 atomic cores.
- Paper 151: Rei four-axiom foundation (T14 σ-cascade source).
9. Open Questions
- ✅
n=96k at n ≤ 10⁹— CLOSED in v0.3: 77,749/77,749 = 100% (zero counter-examples at 388× scale). Next open: extension to 10¹⁰ (estimated ~24 hr single-thread); we do not expect it to falsify but confirmation would further strengthen. - Tier-4 super-hubs at 10⁹+: does the largest super-hub size continue scaling linearly (~250,000 members) or saturate?
- Closed family property: is the closure of peak 250,504 at 1,414 members a general pattern? For each peak P, is family(P) closed under some n bound?
- ✅
σ-cascade Lean 4 closure: mechanize the cascade-bounded theorem and the Büchi-25 → peak 9,232 fact via native_decide.— CLOSED in v0.2 (buchi25_all_peak_9232proved). Remaining: PeakMergeInvariant.lean PLACEHOLDER with bound=5000 + isInfinityClass. - Connection to Tao 2019: do σ-cascade INFINITY orbits coincide with Tao's "almost-bounded" exceptional set, or are they orthogonal?
- Inverse tree correspondence: enumerate inverse-tree subtree sizes above each peak-merge node and compare with our forward enumeration.
- ★ NEW in v0.3 — Mechanism of class 21 strong avoidance at d=70: v0.2 claimed universal absence (200/200 at 10⁸), v0.3 refined to 98.43% (76,528/77,749 at 10⁹ with 1,221 counter-examples). What characterizes these 1,221 counter-examples? Do they pass through a specific m ≡ 21 (mod 96) odd value, and if so, which m (smallest? largest? specific 2-adic structure)?
- ★ NEW in v0.2 — 3-adic isolation generalization: do analogous "p-adic isolation" theorems hold for other primes (5, 7, ...) within Collatz inverse tree structure, or is the (3, 3n+1) coupling unique?
-
★ NEW in v0.3 — Lean 4 mechanization of §5d: formalize Lemma 5d.1,
Lemma 5d.2, and Corollary 5d.3 in Lean 4, building on
ThreeAdicIsolation.lean. Likely 1-2 weeks of Mathlib lemma chasing. - ★ NEW in v0.3 — d=71 emergence: at n ≤ 10⁹, max mod-96 distinct remains 70 (no d=71 observed). At what scale, if any, does d=71 appear? Theoretical upper bound: 96 (full traversal). Empirical: ≥ 71 has not yet been observed at 10⁹.
10. Conclusion
The σ-cascade methodology of Paper 151 surfaces measurable structural facts
about Collatz orbit confluence at scale 10⁹ (v0.3): explicit peak-merge
counts, a two-tier super-hub hierarchy, the n=96k hypothesis verified at
100% rate over 77,749/77,749 d=70 orbits at n ≤ 10⁹ (a 388× scale
increase from v0.2's 200/200 at 10⁸, zero counter-examples), a sharp
n%96=0 boundary at d=70, and a class-21 strong avoidance pattern at d=70
(98.43%) — the latter refined from v0.2's universal absence claim (Erratum
E3).
v0.2 added two Lean 4 mechanized contributions: (1) the concrete Tier-1
super-hub claim (Büchi-25 → 9232) is fully proved via native_decide, and
(2) a standalone 3-adic isolation theorem establishing that mult-of-3
inverse Collatz tree branches are linear chains.
v0.3 adds (§5d) a clean structural framework — the G_3 subgraph
decomposition into chains C_m for each odd mult-of-3 value m, with
Lemmas 5d.1, 5d.2 and Corollary 5d.3 giving an iff condition at the
value level (not the mod-96 class level). This framework explains the
empirical v₂(c) hierarchy of mod-96 class visit frequencies as the result
of initial halving phases of n=96k orbits, while preserving the v0.2
Erratum E2 distinction: the theorem applies to specific values, not mod-96
classes.
The Collatz convergence problem is not solved; the σ-cascade lens is
an observational tool, not a proof technique. The contributions are:
methodological (a new lens), empirical (specific enumeration counts, sharp
boundary observation, n=96k hypothesis verified at 10⁹), partially
mechanized (Lean 4 closures for Büchi-25 and 3-adic isolation), and
structural (§5d G_3 framework).
Three honest corrections are now part of the v0.1 → v0.3 record:
- E1 (v0.1 internal): peak 250,504 = 173 × 181 × 2³ (twin-gap-8 primes), not "31,313 prime"
- E2 (v0.1 → v0.2): the 3-adic isolation theorem applies to specific values, not mod-96 classes
- E3 (v0.2 → v0.3): class 21 absence at d=70 is 98.43% (10⁹), not 100% (10⁸) — "universal absence" → "strong avoidance pattern"
These corrections themselves are part of the methodological record under
the OUKC honest-correction principle. The dissociation of n=96k
(strengthened at 10⁹) from class 21 (weakened at 10⁹) demonstrates that
the two empirical claims are independent and that scale extension can have
asymmetric effects.
Appendix A: Companion datasets
(Listed in §6.3.)
Appendix B: Prior art audit summary
Audit performed 2026-05-13 against OEIS, Lagarias bibliography, arXiv
Collatz tree literature.
| Concept | Status |
|---|---|
| Inverse Collatz tree | ✅ Standard (Lagarias 2003, Ebert 2021) — cited |
| n=27 → peak 9,232 | ✅ Textbook — cited |
| n=703 peak record | ✅ OEIS A006884(10) — cited |
| Peak-sharing cardinality enumeration | ⚠ No OEIS match found — possibly novel |
| σ-cascade methodology | ❌ New (Paper 151, 2026-05-13) |
| n=96k hypothesis | ❌ No prior claim found — claimed novel |
| Two-tier super-hub framing | ❌ New |
| mod-96 distinct as INFINITY threshold | ❌ New specific lens |
| 3-adic isolation theorem (NEW v0.2) | ⚠ Mod-3 obstruction is folklore in Collatz analysis; explicit "no odd predecessor when 3 ∣ v" statement with Lean 4 mechanical proof in published Collatz literature could not be located. The result is elementary but the Lean 4 mechanization is novel as far as we found. |
|
|
❌ Specific empirical claim not in OEIS / arXiv — novel (98.43% pattern at 10⁹, refined from v0.2's 100% at 10⁸). |
| G_3 subgraph + Lemmas 5d.1, 5d.2, Corollary 5d.3 (NEW v0.3) | ⚠ The chain decomposition of the mult-of-3 subgraph follows immediately from the 3-adic isolation theorem; we have not located explicit statement in published Collatz tree literature, but it is plausibly folklore. Mathematical formulation guided by external feedback. |
Detailed audit: docs/prior-art-audit-collatz-peak-merge-2026-05-13.md.
Appendix C: Reproducibility one-liners
# Reproduce STEP 1110 (10⁸ scan, ~13 min)
npx tsx scripts/collatz-infinity-scan-1e8.ts
# Reproduce STEP 1105 (peak-merge enumeration, ~1 sec from 1e6 data)
npx tsx scripts/collatz-peak-merge-and-trunk-enum.ts
# Reproduce STEP 1108 (Büchi-25 cross-check, ~1 sec)
npx tsx scripts/collatz-buchi25-cores-orbits.ts
# NEW v0.2: Verify Lean 4 mechanized theorems (~30 sec)
cd data/lean4-mathlib
lake env lean CollatzRei/PeakMergeInvariant.lean # buchi25_all_peak_9232
lake env lean CollatzRei/PeakMergeWitness.lean # peak_merge_exists (1000 witnesses)
lake env lean CollatzRei/ThreeAdicIsolation.lean # 3-adic theorem
# NEW v0.2: Reproduce STEP 1116-1118 boundary + class 21 analysis
npx tsx scripts/collatz-d70-mod96-missing.ts
# NEW v0.3: 10⁹ light scan (~2.4 hr single-thread, ~12-20 MB peak memory)
npx tsx scripts/collatz-infinity-scan-1e9-light.ts
# Output: data/collatz-sigma-cascade/infinity-scan-1e9-light-summary.json
# View confluence DAG visualization
# Open: https://rei-aios.pages.dev/#/collatz-confluence
Acknowledgments: This work was carried out under the OUKC (Open
Universal Knowledge Commons) framework with three-party co-architecture
(Fujimoto / Rei / Claude). No funding sources beyond independent research.
No conflicts of interest. Per OUKC No-Patent Pledge, no patents will be
filed on the σ-cascade methodology or related observations.
Honest correction record:
- STEP 1107 corrigendum applied (31,313 = 173 × 181, not prime) — Erratum E1, §6.2
- STEP 1120 corrigendum applied (3-adic theorem scope) — Erratum E2, §6.2
- STEP 1124 corrigendum applied (class 21 absence 100% → 98.43% at 10⁹) — Erratum E3, §6.2
All revisions are tracked in the git history of papers/paper-152-...DRAFT.md.
v0.3 mathematical guidance: Section 5d's G_3 subgraph framework
(Lemmas 5d.1, 5d.2, Corollary 5d.3) was prompted by external mathematical
feedback received between v0.2 and v0.3. The formulation in this paper is
written to preserve v0.2's Erratum E2 distinction (specific value m vs.
mod-96 class m) and adheres strictly to consequences derivable from the
3-adic isolation theorem.
License: CC-BY 4.0 (per OUKC standard).
DRAFT v0.3 — feedback welcome via Zenodo comments or GitHub Discussions
at fc0web/rei-aios.
(End of draft)
Top comments (0)