3. Preparatory Lemmas
Lemma 1 (Lemma of irreducibility). For all odd x, is the closure of by actions A, C, and V. Each of its non-x element has a uniquely ordered, finite, and non-cyclic decomposition into actions A, C, and V from x, and the only possible cycle is from x to itself.
Proof. so as x is odd. As residue classes in base 3 are periodic by V, it always reaches numbers where actions C and A are applicable. If x is in , then is in and in . Since , we have that for any x in , , and since , we also have for any x in .
For any x in , we could generalize action A to output the sequence for all because all elements of such sequence also converge to , but we already have them counted in all iterations of action V over . The same goes for action C: it is generalizable with for all with x in , but we will already have those numbers counted in any iteration of action V on .
To prove that each element of has a unique non-cyclic decomposition into actions A, C, and V, we now define a one-to-one correspondence between any and either action. For any odd y, either is divisible by exactly 2, by exactly 4, or by at least 8. If is divisible by 2, then ; if it is divisible by 4, then ; and if it is divisible by 8 or more, there are always finite k and z such that and .
Any odd x in leads to , so the only possible cycle is x to itself. For example, 1 is the only fixed point of C so the cycle in is . □
Remark 1. We already have that no composition of C, and V may be cyclical as both actions are strictly increasing, but proving that no composition of A, C, and V be cyclical, as they combine coprimes 2 and 3, would be of at least the same complexity as proving the Furstenbergconjecture, of which the entire difficulty lies in that “base 2 and base 3 representations share no known common structures” (Furstenberg 1960 [3] and 1967 [4], and Shmerkin 2021 [5]). We already have, however, that, in any base, - 1.
always ends with the representation of theMersenne number, e.g., by
- 2.
always ends with the representation of theMersenne number divided by 3, e.g., by
- 3.
always ends with the representation of numbere.g., by
The chaoticity ofcompositions (that is, any legal word of the form, whereis representing exactly one of either operation and x is any odd number) and hence the difficulty of proving the non-existence of cycles directly come from the fact that A is allowed to delete digits and that both A and C contain divisions by 3. However, the beauty of chaotic cryptology is that, whenever a simple system appears to be intractably chaotic, it comes with the participation trophy of at least offering a strong (and straightforward) industrial chaotic cipher.
Lemma 2 (Lemma of prolificity). Let x be odd with no odd y such that . There is always a unique odd z such that .
Proof. Let x be in , then so is . is in and so is . , , and . Thus, . For example,
Let x be in ; then, is in and . For example
Let x be in ; then, is in and , with not in . For example, . Note (in the example, ) always has a V-sequence of its own so will in turn output another sequence.
□
Remark 2. For any odd x, outputs a sequence (and if x is in , a single additional odd number), in which V-closure is an infinite sequence of sequences (since no element of any sequence can be written with y being odd). Then, is an infinite sequence of sequences and is an infinite k-sequence (a sequence of sequences k times) of sequences. is a polynomial of n of degree 1 (it is exactly equal to ), and the amount of numbers up to extra binary length n from x in each is a polynomial of n of degree . This algebraic property is the reason we call Lemma 2 that of "prolificity". Its first immediate consequence is that , being an infinite sequence of positive-valued polynomials of n of incremented degree and with a positive coefficient for each term of maximal degree, is now necessarily at least of the form for some constants , and c.
Lemma 3 (Lemma of homogeneity). All residue classes in base 3 are equally frequent in when n and k tend toward infinity.
Proof. with mod 3 and , so
Let x end in ; then, for all k, ends in and ends in .
Let x end in ; then, for all k, ends in .
If x ends in , is in . If x ends in , is in . If x ends in , is in .
If x ends in , is in . If x ends in , is in . If x ends in , is in .
□
Remark 3. Concordant with Lemma 3, we can verify that actions D and S already preserve the periodicity of residue classes in base 3: if x is in , is in ; if x is in , is in ; and if x is in , is in . Furthermore, if x is in , is in ; if x is in , is in ; and if x is in , is in . The Lemma of homogeneity is an essential result to construct the following “Lemma of density”, which in turn is the most critical preparatory result for the fundamental Lemma 1.
Lemma 4 (Lemma of density)
. As , every , is associated with a single sequence , where a is a pseudorandom variable, of which the average is less than 2, and Proof. By Lemma 2 for any in some , there are always y and z such that . Let us review all the possible cases.
For x in , and . If x began with , and ; otherwise, and .
For x in , and . If x began with , and ; otherwise, and
For x in , and . If x began with , and ; otherwise, and
For simplification, we assume the two cases of
x beginning with either
or
are at worst equifrequent but
(which is always outputing smaller numbers) will in fact be more frequent because for any
a beginning in
,
always begins in
, whereas exactly one third of all odd
a beginning in
(for example, those beginning in
) are such that
also begins with
(See Figure 27 in Rahn et al., 2021 [
2]).
Thus, assuming the limiting conditions of Lemma 3, we can now associate a unique disjoint pair to each . From all cases we reviewed above, we have that is a pseudorandom variable, in which the pseudorandomness entirely comes from the Furstenberg principle that “base 2 and base 3 representations share no known common structure”, and its actual (and again, unknown) structure is in fact at the heart of the Bocart proof of work (we may call the larger family “base-change ciphers”). However, what we now know is that the mean of over all k is less than 2 because all of the possible cases are {1; 2; 0; 3; −1; 4} if x began with and {2; 3; 1; 4; 0; 5} if x began with , and within each prefix starting condition ( or ), Lemma 3 guarantees their being equifrequent.
Hence, to each
, there is a unique
, in which V-closure contains at least
. As
contains exactly
elements with up to
n more binary digits than
x; if we take the limiting condition of
, each of them must be associated with at least
k additional (disjoint) elements to have
Those additional elements are always found in . □
Remark 4. Lemma 2 established that any was growing exponentially with n. Lemma 4 now further states that, as n goes to infinity, fits each and is associated with a unique collection with particular density conditions. As we will see in Theorem 1, this now provides enough information to pinpoint the general formula for .
5. Discussion
While the previous demonstration was one of discrete algebraic topology, following an interesting request from the reviewers, we outline here another simpler graph-theoretical demonstration of the ACV-closure of any odd being space-filling. In fact, one of the reasons we ventured to call Theorem 1 a “Fundamental theorem” is that there may be a great diversity of ways to demonstrate it.
We already know that the closure of any odd by actions S and G up to the additional binary length n, which in this case is strictly equivalent to all the odd numbers of length lesser or equal to and beginning with all the first characters of x minus the last one, is in bijection with the powerset of the set with n elements and, therefore, has exactly elements. Importantly, for the sake of this demonstration, we will only consider , but if , then we would consider until it is either in or , and with , we just use the fact that, then, .
There are already strings of k characters out of the alphabet , but no odd number exists such that it can remain whole under both A and C so cannot be of the form . For any string to belong to , the alphabet is endowed with a grammar interdicting certain strings and determined by both the residue class of odd numbers in base 3 and the total binary characters added or subtracted by actions A, C, and V. Let us begin by determining the latter:
so if x began with or otherwise.
so if x began with or otherwise.
so for all y in
By Lemma 3, for all y in , and may be in any base 3 residue class with equal frequency.
Additionally, note that prefixes
and
are not equally frequent with action · 3. Indeed, if for all
y beginning with
,
always begin with
, exactly one third of all odd numbers beginning in
—precisely, those greater than the
(for those with an odd number of digits in base 2) or than the
(for those with an even number of digits) of the same length as theirs—are such that
will also begin with
, thus favoring this outcome after either actions
A or
C with a limit frequency of
when
(Figure 27 in Rahn et al., 2021 [
2]). Action
V on the other hand leaves the base 2 prefix intact.
Exactly one out of three times, each action (V, A and C) will only be allowed to be followed by action V only, which also appends two more base 2 digits. Otherwise, action C appends digit on average, and action A removes digits on average, meaning that has 1 more digits than y if y was on the right of the odd numbers beginning with (for example, and ).
If in any ACV string we note
,
, and
as the total number of characters
V,
A, and
C, respectively, then all strings in
must verify the following topological invariant:
By Lemma 3, the following limit is also verified when
,
, and
are counted on all of
:
and since
V is defined for all base 3 residue classes in
while there is no other operation defined for residue class
, we also have
As
A and
C are equifrequent, we may represent “
A or
C exclusively and with equal frequency” with
X, and when
,
will equal the number of paths in the binary tree where
of all nodes (each representing a word) have a
V and
X vertices and
only a
V one. This is strictly equivalent to the binary tree where each node has exactly
chances of being allowed to branch non-exclusively in either
V or
X with equal frequency (thus branching to just
X with a total frequency of exactly
) and
to branch non-exclusively and equifrequently in
or
. In this tree, we now count all of the existing words, verifying
If
was only closed by two distinct actions, both defined equifrequently on all elements and with the binary length characteristics of V (for example
and
), it would be in bijection with the powerset of the set with
elements. However, the tree we defined is at least equivalent to
adding two binary digits and to
adding two exactly half of the times and subtracting
the other half. Thus counting all occurrences of
and only the pathways with a net-positive binary length extension, we now have that
is at least as large as the powerset of the set with
elements. Therefore,
so:
Remark 5. This alternative approach to proving the space-fillingness of Collatz feathers is particularly fertile conceptually. Indeed, the ACV-closure of any can now, by the development of Equation (19) and although it uses three distinct actions so would start with a maximum of branches, be interpreted as with being the Hausdorff dimension of the triadic Cantor set (the set of all real numbers that can be written without any digit 1 in base 3). Practitioners of the Furstenberg conjecture know that Cantor sets are in turn a larger family containing sets left invariant by either mod 1 or mod 1 but never by both ([4]). is also the inverse of the Hausdorff dimension of the Sierpinski triangle, a natural figure in the evaluation of the 3-adic distance, which gives us some insight into the geometry of when . With this in mind, it would be particularly interesting to further study the structure (and iterated morphisms) of the attractors of the , and Juggler sequences with the deliberate intent of constructing ("pathological") objects of an intermediate cardinality between and as such constructions, though never achieved, are already known not to contradict ZFC [7,8]. All in all, we really believe that the Collatz conjecture, the Furstenberg conjecture, and the Continuum Hypothesis should be much more tightly related in the literature. 6. Conclusions
Theorem 1, for the particular case of
when
, already implies that “almost all Collatz orbits converge to 1”, which is a result strictly stronger (and shorter) than Tao 2019 [
9], where it was only obtained that
almost all Collatz orbits attained
almost bounded values, leaving not only the possibility of cycles but also the limitation of almost boundedness. Such is not the case here: on the one side, all the elements in
converge to 1 so none other than 1 belongs to a cycle, and on the other side, the elements of
are proven to cover
almost all of
. Additionally, pushed to infinity, Collatz attractors are therefore space-filling as their cardinality is of the form
.
However, Theorem 1 does not yet prove that all Collatz orbits converge to 1. Rather, it proves that
vanishes when
n tends toward infinity (which by the way is consistent with the computational findings presented in Figure 15 in Rahn et al., 2021 [
2]). Hence,
can never reach a formula of the form
for some integer (possibly negative) constant
. At worst, it may only reach
with
. On the other side though, no Collatz attractor can vanish. Thus, in conclusion, by Lemma 1, we have
However, for all odd
x of binary length
m, we always have: