> Full-text translation from Chinese. The original remains authoritative for its research claims; this translation is not independent mathematical verification. # ACTM Round 9 ## BADT Defect Synchronization, Waldschmidt–Masser Criticality, and the Common-Anchor Squaring Tower ### Defect Synchronization, Waldschmidt–Masser Criticality, and the Common-Anchor Squaring Tower **Research program:** Algebraic Combinatorial Transcendence Methodology (ACTM) **Authors:** Neo.K; Aletheia (GPT-5.6 Sol) **Institution:** EveMissLab / 一言諾科技有限公司 **Date:** 2026-10-02 **Version:** Round 9 / v0.1 **Document type:** Formal research notes / anchored nonlinear-coupling frontier **Status statement:** This round does not claim to prove the Four Exponentials Conjecture, Matrix Coefficient Conjecture, Structural Rank Conjecture, or Schanuel's conjecture. Its main result is to reformulate the Binary Anchored Descent Tower (BADT) as a critical system carrying geometric, transcendence-degree, matrix-rank, and quadratic-coupling defects simultaneously, and to prove that these defects synchronize exactly with tower depth. --- ## Abstract ACTM Round 8 showed that exponent-difference sparsity arising solely from Six / Five Exponentials still cannot exclude the infinite binary combinatorial survivor: $$ \{0,1,2,4,8,16,\ldots\}. $$ Round 9 therefore restores information removed by the exponent-set abstraction: $$ \boxed{ \text{common anchor }\lambda + \text{common generator }s + \text{iterated squaring}. } $$ Define the depth-$D$ Binary Anchored Descent Tower: $$ \operatorname{BADT}_D(\lambda,s): $$ $$ \boxed{ \lambda, \quad s\lambda, \quad s^2\lambda, \quad s^4\lambda, \quad \ldots, \quad s^{2^D}\lambda \in \mathcal L_{\mathrm{alg}}, } $$ where: $$ \mathcal L_{\mathrm{alg}} = \left\{ z\in\mathbb C: e^z\in\overline{\mathbb Q}^{\times} \right\}. $$ Type P is precisely: $$ D=1, $$ whereas genuine quartic descent is equivalent to: $$ D=2. $$ Let: $$ z_0=\lambda, $$ and: $$ z_{k+1} = s^{2^k}\lambda \qquad 0\le k\le D. $$ The tower then satisfies a sequence of quadratic couplings sharing an anchor: $$ \boxed{ z_k^2 = z_0z_{k+1} \qquad 1\le k\le D. } $$ Since $s$ is transcendental: $$ 1, \quad s, \quad s^2, \quad s^4, \ldots, s^{2^D} $$ are linearly independent over $\mathbb Q$; therefore: $$ \boxed{ \operatorname{ldim}_{\mathbb Q} (z_0,\ldots,z_{D+1}) = D+2. } $$ On the other hand, all coordinates lie in: $$ \mathbb Q(\lambda,s), $$ and the known Type-P result gives: $$ \operatorname{trdeg}_{\mathbb Q} \mathbb Q(\lambda,s)=2. $$ Therefore: $$ \boxed{ \operatorname{trdeg}_{\mathbb Q} \mathbb Q(z_0,\ldots,z_{D+1}) = 2. } $$ Thus, if we define the Schanuel-style compression defect: $$ \operatorname{CD}_D = \operatorname{ldim}_{\mathbb Q}(z) - \operatorname{trdeg}_{\mathbb Q}\mathbb Q(z), $$ then: $$ \boxed{ \operatorname{CD}_D=D. } $$ Equivalently, since all: $$ e^{z_i} $$ are algebraic: $$ \delta(z) = \operatorname{trdeg}_{\mathbb Q}\mathbb Q(z,e^z) - \operatorname{ldim}_{\mathbb Q}(z) = -D. $$ BADT is therefore an over-compression tower whose Schanuel predimension decreases by another $1$ whenever depth increases by $1$. More importantly, the same integer $D$ also appears simultaneously in three other structures. First, the tower's two-parameter locus lies in: $$ \mathbb A^{D+2} $$ and has local dimension $2$, so: $$ \boxed{ \operatorname{codim}=D. } $$ Second, there are: $$ D $$ independent shared-anchor quadratic couplings: $$ z_k^2-z_0z_{k+1}=0. $$ Third, define a singular logarithmic matrix for each layer: $$ H_k = \begin{pmatrix} z_0 & z_k\\ z_k & z_{k+1} \end{pmatrix}, \qquad 1\le k\le D. $$ By the quadratic coupling: $$ \det H_k=0, $$ and: $$ \operatorname{rank}_{\mathbb C}H_k=1. $$ Let: $$ B_D = H_1 \oplus \cdots \oplus H_D. $$ then: $$ \boxed{ \operatorname{rank}_{\mathbb C}B_D=D. } $$ However, since: $$ z_0,\ldots,z_{D+1} $$ are linearly independent over $\mathbb Q$, the structural rank of $B_D$ is: $$ \boxed{ \operatorname{srank}(B_D)=2D. } $$ Therefore: $$ \boxed{ \operatorname{srank}(B_D) - \operatorname{rank}(B_D) = D. } $$ Finally, the critical rank threshold of the Waldschmidt–Masser theorem for a: $$ 2D\times2D $$ matrix is: $$ \frac{(2D)(2D)}{2D+2D} = D. $$ And the BADT block matrix has exactly: $$ \boxed{ \operatorname{rank}(B_D)=D. } $$ The known Waldschmidt–Masser theorem requires the strict inequality: $$ \operatorname{rank}(M) < \frac{mn}{m+n}. $$ Therefore: $$ \boxed{ \text{BADT}_D \text{ lies exactly on the Waldschmidt–Masser boundary at every depth.} } $$ This round calls the phenomenon: $$ \boxed{ \text{Defect Synchronization}. } $$ Its core synchronization identity is: $$ \boxed{ D = \operatorname{codim}V_D = \operatorname{CD}_D = \operatorname{srank}(B_D)-\operatorname{rank}(B_D). } $$ This means that BADT is not a stack of independent Four-Exponentials anomalies, but a scale-invariant critical family: every new layer adds a quadratic constraint, a transcendence-compression debt, and a matrix structural-rank defect simultaneously, while the normalized matrix rank stays on the exact equality boundary of the Waldschmidt–Masser theorem. --- # 1. The Formal Definition of BADT Let: $$ 0\neq \lambda \in \mathcal L_{\mathrm{alg}}, $$ and a transcendental: $$ s\in\mathbb C. $$ For: $$ D\ge1, $$ Define: $$ \boxed{ \operatorname{BADT}_D(\lambda,s) } $$ to hold if: $$ s^{2^k}\lambda \in \mathcal L_{\mathrm{alg}} $$ For: $$ 0\le k\le D $$ holds throughout this range. And, of course: $$ \lambda \in \mathcal L_{\mathrm{alg}}. $$ --- # 2. Coordinate Notation Define: $$ z_0=\lambda, $$ and: $$ z_{k+1} = s^{2^k}\lambda, \qquad 0\le k\le D. $$ Thus: $$ z_1=s\lambda, $$ $$ z_2=s^2\lambda, $$ $$ z_3=s^4\lambda, $$ and so on. The entire tower point is: $$ \boxed{ Z_D = (z_0,z_1,\ldots,z_{D+1}) \in \mathcal L_{\mathrm{alg}}^{D+2}. } $$ --- # 3. Shared-anchor quadratic recursion For: $$ 1\le k\le D, $$ we have: $$ z_k = s^{2^{k-1}}\lambda. $$ Therefore: $$ z_k^2 = s^{2^k}\lambda^2. $$ And: $$ z_0z_{k+1} = \lambda \left( s^{2^k}\lambda \right) = s^{2^k}\lambda^2. $$ Hence: $$ \boxed{ z_k^2=z_0z_{k+1}. } $$ Thus, BADT$_D$ produces: $$ \boxed{ D } $$ nested homogeneous quadratic relations. --- # 4. Why the Common Anchor Matters The relations do not independently use different denominators. They all share: $$ \boxed{ z_0=\lambda. } $$ That is: $$ z_1^2=z_0z_2, $$ $$ z_2^2=z_0z_3, $$ $$ z_3^2=z_0z_4, $$ and so on. BADT's geometry is therefore not: $$ D $$ unrelated quadratic cones. Rather, it is: $$ \boxed{ D \text{ quadratic constraints glued through one common anchor}. } $$ --- # 5. Q-linear rank The set: $$ \left\{ 1, s, s^2, s^4, \ldots, s^{2^D} \right\} $$ because $s$ is transcendental, is linearly independent over: $$ \mathbb Q $$ that field. Thus, after multiplication by: $$ 0\neq\lambda $$ we have: $$ \boxed{ z_0,z_1,\ldots,z_{D+1} } $$ linearly independent over $\mathbb Q$. Thus: $$ \boxed{ \operatorname{ldim}_{\mathbb Q} (Z_D) = D+2. } $$ --- # 6. Baker linear rigidity All: $$ z_i $$ are logarithms of algebraic numbers. Baker's theorem therefore upgrades the preceding: $$ \mathbb Q $$ -linear independence to: $$ \overline{\mathbb Q} $$ -linear independence. That is: $$ \boxed{ \operatorname{ldim}_{\overline{\mathbb Q}} (Z_D) = D+2. } $$ Note: This excludes only algebraic linear relations. It does not exclude BADT's quadratic relations. --- # 7. Coordinate field remains two-dimensional Since: $$ z_0=\lambda, $$ and: $$ s=\frac{z_1}{z_0}, $$ Thus: $$ \mathbb Q(\lambda,s) \subseteq \mathbb Q(z_0,\ldots,z_{D+1}). $$ In the converse direction, every: $$ z_k $$ belongs to: $$ \mathbb Q(\lambda,s). $$ Hence: $$ \boxed{ \mathbb Q(z_0,\ldots,z_{D+1}) = \mathbb Q(\lambda,s). } $$ By the Roy–Waldschmidt threshold analysis of Type P: $$ \boxed{ \operatorname{trdeg}_{\mathbb Q} \mathbb Q(\lambda,s) = 2. } $$ Therefore: $$ \boxed{ \operatorname{trdeg}_{\mathbb Q} \mathbb Q(Z_D) = 2. } $$ Regardless of: $$ D $$ does not increase, however large the depth becomes. --- # 8. Compression Defect Define: $$ \boxed{ \operatorname{CD}(Z) = \operatorname{ldim}_{\mathbb Q}(Z) - \operatorname{trdeg}_{\mathbb Q}\mathbb Q(Z). } $$ For BADT$_D$: $$ \operatorname{CD}(Z_D) = (D+2)-2. $$ Thus: $$ \boxed{ \operatorname{CD}(Z_D)=D. } $$ Each additional squaring-descent layer adds one unit to the deficit of degrees of freedom. --- # 9. Schanuel predimension Since all: $$ e^{z_i} $$ are algebraic: $$ \operatorname{trdeg}_{\mathbb Q} \mathbb Q(Z_D,e^{Z_D}) = \operatorname{trdeg}_{\mathbb Q} \mathbb Q(Z_D) = 2. $$ Define: $$ \delta(Z_D) = \operatorname{trdeg}_{\mathbb Q} \mathbb Q(Z_D,e^{Z_D}) - \operatorname{ldim}_{\mathbb Q}(Z_D). $$ Thus: $$ \boxed{ \delta(Z_D) = 2-(D+2) = -D. } $$ That is: $$ \boxed{ -\delta(Z_D)=D. } $$ Schanuel predicts: $$ \delta\ge0. $$ Thus, any BADT with: $$ D\ge1 $$ lies in the negative-predimension region forbidden by Schanuel. Here, however, this is only defect measurement: Schanuel is not used as a proved tool. --- # 10. BADT variety Consider the affine space: $$ \mathbb A^{D+2} $$ with coordinates: $$ X_0,\ldots,X_{D+1}. $$ On the open chart: $$ X_0\neq0, $$ define the BADT quadratic locus: $$ V_D: $$ $$ \boxed{ X_k^2 - X_0X_{k+1} = 0, \qquad 1\le k\le D. } $$ --- # 11. Local parametrization On: $$ X_0\neq0, $$ Let: $$ u=X_0, $$ and: $$ v=\frac{X_1}{X_0}. $$ then: $$ X_1=uv. $$ The first relation gives: $$ X_2=uv^2. $$ The second gives: $$ X_3=uv^4. $$ Recursively: $$ \boxed{ X_{k+1} = u v^{2^k}. } $$ Therefore: $$ V_D $$ on this open chart is controlled by two parameters: $$ (u,v) $$ as written above. --- # 12. Geometric dimension and codimension Thus, the main open BADT locus: $$ V_D\cap\{X_0\neq0\} $$ has dimension: $$ \boxed{ 2. } $$ The ambient dimension is: $$ D+2. $$ The local codimension is therefore: $$ \boxed{ \operatorname{codim}V_D = D. } $$ This is exactly the number of quadratic equations. --- # 13. Jacobian check Let: $$ F_k = X_k^2-X_0X_{k+1}, \qquad 1\le k\le D. $$ At a BADT point: $$ X_0=\lambda\neq0. $$ Each: $$ F_k $$ For: $$ X_{k+1} $$ has partial derivative, with respect to that coordinate: $$ -\lambda. $$ Selecting columns in order: $$ X_2,\ldots,X_{D+1}, $$ shows that the Jacobian has rank: $$ D. $$ The BADT point therefore lies in a smooth local complete-intersection chart of codimension $D$. --- # 14. Genericity inside the BADT surface From: $$ \mathbb Q(Z_D) = \mathbb Q(\lambda,s) $$ and: $$ \operatorname{trdeg}=2, $$ while the BADT locus itself also has local dimension: $$ 2, $$ the hypothetical BADT point has maximal transcendence degree on this two-parameter locus. Thus, BADT does not lie in a lower-dimensional special algebraic subfamily of that surface. It is a critical point in the sense: $$ \boxed{ \text{generic on the BADT two-parameter surface} } $$ stated above. --- # 15. The Two Quadratic Relations of BADT$_2$ For quartic descent: $$ D=2 $$ we have: $$ Z_2 = (z_0,z_1,z_2,z_3) = (\lambda,s\lambda,s^2\lambda,s^4\lambda). $$ Its relations are: $$ \boxed{ z_1^2-z_0z_2=0, } $$ $$ \boxed{ z_2^2-z_0z_3=0. } $$ And: $$ \operatorname{trdeg}_{\mathbb Q} \mathbb Q(z_0,z_1,z_2,z_3) = 2. $$ --- # 16. Roy–Waldschmidt known threshold The core case handled by the known quadratic-relations theorem of Roy–Waldschmidt is: - Coordinates are $\mathbb Q$-linearly independent logarithms of algebraic numbers; - There is a nonzero homogeneous quadratic relation; - The generated field has transcendence degree $1$. In this case, the quadratic relation is forced to have a rational linear explanation. BADT$_2$, however, satisfies: $$ \boxed{ \operatorname{trdeg}=2. } $$ It therefore lies exactly one level beyond the known theorem, rather than satisfying its hypotheses. --- # 17. An Important Reminder of the Current Situation As of 2026, the literature still states: > It is not even known whether two algebraically independent logarithms of algebraic numbers exist. Thus, no known general theorem currently says: > Many $\mathbb Q$-linearly independent logarithms of algebraic numbers necessarily give arbitrarily large transcendence degree. This is the unresolved core of the Algebraic Independence of Logarithms Conjecture. Therefore, a contradiction cannot be deduced merely from 'the tower is long'. --- # 18. The Counterintuitive Side Effect of BADT If BADT$_D$ hypothetically holds: $$ \operatorname{trdeg}\mathbb Q(Z_D)=2. $$ Thus, among: $$ z_0,\ldots,z_{D+1} $$ there must be at least two algebraically independent elements. That is: > If a Type-P / BADT counterexample exists, it would also supply a positive example of 'two algebraically independent logarithms of algebraic numbers', whose existence is currently unknown. This is not a contradiction. However, it shows that the hypothetical counterexample itself already lies at a very deep algebraic-independence frontier. --- # 19. The Rank-One Logarithmic Matrix of Each Layer For: $$ 1\le k\le D, $$ Define: $$ \boxed{ H_k = \begin{pmatrix} z_0 & z_k\\ z_k & z_{k+1} \end{pmatrix}. } $$ From: $$ z_k^2=z_0z_{k+1}, $$ we have: $$ \det H_k=0. $$ Since the entries are nonzero: $$ \boxed{ \operatorname{rank}_{\mathbb C}H_k=1. } $$ --- # 20. Each Layer Is a Four-Exponentials-Type Singularity The two rows of $H_k$ are: $$ (z_0,z_k), $$ $$ (z_k,z_{k+1}) $$ Their ratios respectively involve: $$ s^{2^{k-1}}. $$ By its transcendence, neither the rows nor the columns have a nontrivial rational dependence. Thus, each: $$ H_k $$ is a singular but rationally invisible $2\times2$ algebraic-logarithm matrix. The $2\times2$ case of the Matrix Coefficient Conjecture is equivalent to the Four Exponentials Conjecture, so it would exclude every $H_k$. That conjecture, however, has not been proved. --- # 21. Direct-sum tower matrix Define: $$ \boxed{ B_D = H_1 \oplus H_2 \oplus \cdots \oplus H_D. } $$ Then: $$ B_D \in M_{2D\times2D} (\mathcal L_{\mathrm{alg}}). $$ Since each block has rank: $$ 1, $$ Thus: $$ \boxed{ \operatorname{rank}_{\mathbb C}B_D=D. } $$ --- # 22. Structural rank The BADT coordinates: $$ z_0,\ldots,z_{D+1} $$ are linearly independent over: $$ \mathbb Q $$ that field. By the definition of structural rank, replace them with independent indeterminates: $$ x_0,\ldots,x_{D+1}. $$ The generic block indexed by $k$ is: $$ \widetilde H_k = \begin{pmatrix} x_0 & x_k\\ x_k & x_{k+1} \end{pmatrix}. $$ Its determinant: $$ x_0x_{k+1}-x_k^2 $$ is not the zero polynomial. Thus: $$ \operatorname{rank} \widetilde H_k=2 $$ over the rational-function field. Therefore: $$ \boxed{ \operatorname{srank}(B_D)=2D. } $$ --- # 23. Structural Rank Defect Define: $$ \operatorname{SRD}(M) = \operatorname{srank}(M) - \operatorname{rank}(M). $$ For the BADT block matrix: $$ \operatorname{SRD}(B_D) = 2D-D. $$ Thus: $$ \boxed{ \operatorname{SRD}(B_D)=D. } $$ This defect grows linearly with BADT depth. --- # 24. Waldschmidt–Masser threshold For: $$ M \in M_{m\times n} (\mathcal L_{\mathrm{alg}}) $$ the Waldschmidt–Masser theorem, under the condition: $$ \boxed{ \operatorname{rank}(M) < \frac{mn}{m+n} } $$ implies the existence of a sufficiently large zero block after rational row / column basis changes. Now: $$ B_D $$ has size: $$ m=n=2D. $$ The threshold is therefore: $$ \frac{mn}{m+n} = \frac{4D^2}{4D} = D. $$ --- # 25. Exact Waldschmidt–Masser Criticality And: $$ \operatorname{rank}(B_D)=D. $$ Therefore: $$ \boxed{ \operatorname{rank}(B_D) = \frac{mn}{m+n}. } $$ This is not a strict: $$ < $$ but an exact equality. The Waldschmidt–Masser theorem therefore does not trigger. --- # 26. Scale invariance More importantly, this is not an accident of: $$ D=2 $$ that particular depth. For every: $$ D\ge1, $$ we have: $$ \boxed{ \frac{ \operatorname{rank}(B_D) }{ 2D } = \frac12. } $$ The square Waldschmidt–Masser boundary likewise corresponds to normalized rank: $$ \frac12. $$ Thus, increasing tower depth never enters the theorem's region. --- # 27. Why Increasing Depth Does Not Help the Existing Rank Theorem Adding a layer: $$ D \mapsto D+1, $$ changes the matrix size: $$ 2D \mapsto 2D+2, $$ the actual rank: $$ D \mapsto D+1, $$ and the threshold: $$ D \mapsto D+1. $$ All three move synchronously. Therefore: $$ \boxed{ \text{depth extension does not create Waldschmidt–Masser slack}. } $$ This is an important methodological stop condition. --- # 28. Defect Synchronization Now compare four quantities. ### Quadratic coupling count $$ Q_D=D. $$ ### Geometric codimension $$ G_D=D. $$ ### Transcendence compression defect $$ \operatorname{CD}_D=D. $$ ### Structural-rank defect $$ \operatorname{SRD}_D=D. $$ Therefore: $$ \boxed{ Q_D = G_D = \operatorname{CD}_D = \operatorname{SRD}_D = D. } $$ This round calls this: **BADT Defect Synchronization Law**. --- # 29. Predimension form Since: $$ \delta(Z_D)=-D, $$ Defect Synchronization can also be written: $$ \boxed{ -\delta(Z_D) = \operatorname{codim}V_D = \operatorname{SRD}(B_D) = D. } $$ The same 'debt of degrees of freedom' therefore appears simultaneously in: - Schanuel predimension; - Algebraic geometry; - Matrix structural rank; - The number of quadratic recurrences. --- # 30. ACTM's Interpretation: Over-Compression Quantified The original ACTM seeks to study: $$ \boxed{ \text{What constitutes illegitimate / excessive algebraic compression?} } $$ BADT gives a particularly clean hypothetical model. Source $\mathbb Q$-linear freedom: $$ D+2. $$ Actual transcendental freedom: $$ 2. $$ Thus: $$ D $$ units of freedom are compressed away. And these: $$ D $$ units correspond exactly to: $$ D $$ quadratic couplings. Thus: $$ \boxed{ \text{one extra squaring relation} \leftrightarrow \text{one extra compression unit}. } $$ --- # 31. Defect Synchronization Is Not Currently a Contradiction It must be emphasized: $$ Q_D=G_D=CD_D=SRD_D $$ is only structural synchronization. It does not itself prove that: $$ D>0 $$ is impossible. To actually exclude BADT, a theorem is still needed that connects at least two kinds of defect by an inequality. Ideally, for example, one would need: $$ \boxed{ \operatorname{SRD} \le f(\operatorname{trdeg}) } $$ or: $$ \boxed{ \operatorname{codim} \le g(\operatorname{rank visibility}) } $$ that BADT violates. The general form is currently unknown. --- # 32. Triple-Critical Boundary BADT simultaneously lies at the boundaries of three major unconditional methods. ## Boundary A: Six / Five Exponentials Binary exponent support can completely avoid difference multiplicity $3$. ## Boundary B: Roy–Waldschmidt The quadratic-relations theorem excludes: $$ \operatorname{trdeg}=1, $$ whereas BADT has: $$ \operatorname{trdeg}=2. $$ ## Boundary C: Waldschmidt–Masser The rank theorem requires: $$ \operatorname{rank}<\frac{mn}{m+n}, $$ whereas the BADT block matrix attains exact equality. Therefore: $$ \boxed{ \text{BADT is a triple-critical state}. } $$ --- # 33. This Explains Why Type P Is Difficult Type P appears to require only: $$ \lambda, \quad s\lambda, \quad s^2\lambda \in \mathcal L_{\mathrm{alg}}. $$ Yet it already simultaneously: 1. Avoids the Baker linear obstruction; 2. Remains at the Roy–Waldschmidt trdeg-$2$ layer; 3. Remains at the Four Exponentials $2\times2$ matrix boundary; 4. Can be extended to binary support that avoids Six Exponentials; 5. Upon extension to a block tower, lies exactly on the Waldschmidt–Masser rank equality. It is therefore not a low-order overlooked case. It is a critical intersection shared by multiple known tools. --- # 34. The Role of the Matrix Coefficient Conjecture The Matrix Coefficient Conjecture of Dasgupta–Kakde asserts: If a square algebraic-logarithm matrix is singular, there exist nonzero rational vectors: $$ w,v $$ such that: $$ w^TMv=0. $$ Its: $$ 2\times2 $$ case is equivalent to the Four Exponentials Conjecture. Thus, if this conjecture holds, each: $$ H_k $$ is immediately impossible. For now, however, it remains a conjecture. The work published in 2026 proves several implications and proposes a support / Newton-polytope auxiliary-polynomial strategy, but the core direction of constructing a sufficiently support-controlled polynomial remains incomplete. --- # 35. The Role of the Structural Rank Conjecture The Structural Rank Conjecture predicts: $$ \boxed{ \operatorname{rank}(M) = \operatorname{srank}(M) } $$ for logarithm matrices. The BADT block matrix, however, has: $$ \operatorname{rank}(B_D)=D, $$ However: $$ \operatorname{srank}(B_D)=2D. $$ BADT is therefore a Structural Rank failure family linear in depth. Roy proved that the Structural Rank Conjecture is equivalent to the Schanuel special case concerning algebraic independence of logarithms. Thus, synchronization of BADT's matrix defect and negative predimension is not accidental. --- # 36. A New Synchronization Perspective If the Structural Rank Conjecture holds: $$ SRD_D=0. $$ If the Schanuel special case holds: $$ CD_D=0. $$ BADT, however, gives: $$ SRD_D=CD_D=D. $$ Thus, BADT can be understood as: $$ \boxed{ \text{the same forbidden compression rendered once as field dimension and once as matrix rank}. } $$ This matches ACTM's original language of 'generation–projection–collapse'. --- # 37. Common anchor vs direct sum It should be noted: $$ B_D $$ appears, in matrix form, to separate the layers by its block-diagonal construction. However, its coefficients are not genuinely independent. All blocks share: $$ z_0=\lambda, $$ and: $$ z_{k+1} $$ is also an endpoint of the next layer's coupling. The coefficient system itself therefore has: $$ \boxed{ \text{shared-anchor nonlinear overlap}. } $$ The pure rank inequality of the Waldschmidt–Masser theorem does not directly use this overlap. --- # 38. Why a Theorem Beyond Rank Alone Is Needed If one looks only at: $$ m, \quad n, \quad \operatorname{rank}(M), $$ BADT always lies on the equality boundary. Thus, any matrix theorem that excludes BADT must read at least some additional information: - Coefficient support; - Coefficient overlap; - Repeated variables; - The shared anchor; - Nonlinear relations among coefficient ratios; - Or the algebraic variety containing the coefficient tuple. That is: $$ \boxed{ \text{rank alone is information-theoretically insufficient for this family}. } $$ --- # 39. The Naturalness of a Support Strategy This also explains why the Dasgupta–Kakde 2026 direction of moving auxiliary-polynomial conditions from total degree to: - Monomial support; - Newton polytopes; - Mixed volume; connects naturally to ACTM. Although the normalized rank of the BADT block matrix is unchanged, coefficient support develops highly regular binary overlap as depth increases. Therefore: $$ \boxed{ \text{support-sensitive rank theorem} } $$ is more likely to detect BADT than merely strengthening a rank threshold. --- # 40. BADT coefficient graph A coefficient incidence graph can be constructed. Vertices: $$ z_0,z_1,\ldots,z_{D+1}. $$ The quadratic coupling at layer: $$ k $$ is the hyperedge: $$ \boxed{ (z_0,z_k,z_{k+1}). } $$ All hyperedges share: $$ z_0. $$ They therefore form an: $$ \boxed{ \text{anchored quadratic star-chain}. } $$ This is structure that the direct-sum matrix itself does not fully display. --- # 41. Anchor reuse count $z_0$ appears in all: $$ D $$ quadratic relations. The other: $$ z_k $$ appear only in a few neighboring relations. Thus, BADT has highly nonuniform coefficient reuse: $$ \boxed{ \deg_{\mathrm{coupling}}(z_0)=D. } $$ This provides a possible support invariant: **Anchor Reuse Degree (ARD)** $$ \boxed{ \operatorname{ARD}(Z_D)=D. } $$ --- # 42. Further Synchronization of ARD and Defects We therefore also obtain: $$ \boxed{ \operatorname{ARD} = \operatorname{CD} = \operatorname{SRD} = \operatorname{codim} = D. } $$ The common anchor's reuse count grows synchronously with the other defects. This may be BADT's most promising information beyond rank alone for subsequent use. --- # 43. The Minimal Two-Layer Common-Anchor BADT System For quartic descent: $$ D=2, $$ we have: $$ z_1^2=z_0z_2, $$ $$ z_2^2=z_0z_3. $$ Eliminating: $$ z_2 $$ yields higher-degree consequences, for example: $$ z_1^4 = z_0^3z_3. $$ Thus: $$ \boxed{ z_1^4-z_0^3z_3=0. } $$ This is not a new independent relation, but the composition of two quadratic couplings. However, it shows that a squaring tower produces: $$ \boxed{ \text{degree doubling relations}. } $$ --- # 44. General composed relation From: $$ z_{k+1} = \frac{z_k^2}{z_0}, $$ iteration gives: $$ \boxed{ z_k = \frac{ z_1^{2^{k-1}} }{ z_0^{2^{k-1}-1} } } $$ For: $$ k\ge1. $$ Therefore: $$ \boxed{ z_1^{2^{k-1}} = z_0^{2^{k-1}-1} z_k. } $$ Tower depth corresponds to exponential growth of relation degree. --- # 45. Low-dimensional field, high-degree relation tower BADT thus simultaneously has: - A fixed field transcendence degree: $$ 2; $$ - A growing number of Q-linearly independent logarithms: $$ D+2; $$ - Growing quadratic codimension: $$ D; $$ - Exponentially growing composed-relation degree: $$ 2^D. $$ This is a: $$ \boxed{ \text{low-dimensional but relation-rich algebraic-log configuration}. } $$ --- # 46. Reconnecting to the Original 'Compression' Intuition BADT compresses: $$ D+2 $$ Q-linear directions into: $$ 2 $$ transcendental parameters: $$ (\lambda,s) $$ as shown above. It can therefore be written: $$ \boxed{ D+2 \text{ source directions} \longrightarrow 2 \text{ algebraic-transcendence dimensions}. } $$ Its compression debt: $$ D $$ is supported by exactly: $$ D $$ shared-anchor quadratic rules. This is therefore ACTM's most direct: $$ \boxed{ \text{forbidden-over-compression candidate}. } $$ --- # 47. The Core Negative Results of Round 9 This round also identifies two directions on which 'no more time should be wasted'. First: > Merely increasing BADT depth will not automatically trigger the Waldschmidt–Masser theorem. Because rank and threshold always synchronize. Second: > Nor is it viable merely to expect 'many Q-independent logarithms' to give transcendence degree $\ge3$ automatically. Existing general theory has not even settled whether two algebraically independent logarithms of algebraic numbers exist. Round 10 must therefore exploit BADT's **special nonlinear geometry**, not merely cardinality / depth. --- # 48. The Next Bridge Needed The most precise current missing bridge can be written: $$ \boxed{ \text{shared-anchor quadratic codimension} \quad+\quad \text{algebraic-log coordinates} \quad\Longrightarrow\quad \text{rational visibility}. } $$ That is: > When many quadratic singularities share an anchor, can a stronger unconditional conclusion be obtained than for a single Four Exponentials case? This is the core question of Round 10. --- # 49. Anchored Matrix Coefficient Problem This round proposes a restricted research problem. Given: $$ z_0,\ldots,z_{D+1} \in \mathcal L_{\mathrm{alg}} $$ linearly independent over $\mathbb Q$, and: $$ z_k^2=z_0z_{k+1} $$ For: $$ 1\le k\le D. $$ Ask: > Does there exist a finite: > > $$ > D_0 > $$ > > such that whenever: > > $$ > D\ge D_0 > $$ > > some rationally visible relation must follow? This round calls it: **Anchored Matrix Coefficient Problem (AMCP)**. It is narrower than the full Matrix Coefficient Conjecture, because the coefficient variety has a fixed binary-squaring form. --- # 50. The Relationship Between AMCP and Four Exponentials If: $$ D=1, $$ AMCP reduces to the Type-P Four Exponentials problem. If some: $$ D\ge2 $$ can be excluded unconditionally, we would obtain: > Certain hypothetical Four-Exponentials states cannot yet be excluded at one layer, but cannot extend to deeper anchored squaring towers. This is the sort of 'intermediate result' that higher-support research might genuinely obtain. --- # 51. BADT Extensibility Hierarchy Define: $$ \operatorname{BD}(\lambda,s) = \sup \left\{ D: \operatorname{BADT}_D(\lambda,s) \text{ holds} \right\}. $$ Four Exponentials predicts: $$ \boxed{ \operatorname{BD}=0 } $$ If Round 8's indexing calls Type P depth $1$, the equivalent adjustment is 'no Type P exists'. Within the current hypothetical Type-P analysis: - The Type P assumption supplies the first quadratic layer; - Quartic descent supplies the second; - AMCP asks whether a finite obstruction exists at greater depth. --- # 52. Defect growth table | Depth $D$ | Log coordinates | Q-linear rank | trdeg | Compression defect | Quadratic relations | Block actual rank | Block structural rank | |---:|---:|---:|---:|---:|---:|---:|---:| | $1$ | $3$ | $3$ | $2$ | $1$ | $1$ | $1$ | $2$ | | $2$ | $4$ | $4$ | $2$ | $2$ | $2$ | $2$ | $4$ | | $3$ | $5$ | $5$ | $2$ | $3$ | $3$ | $3$ | $6$ | | $D$ | $D+2$ | $D+2$ | $2$ | $D$ | $D$ | $D$ | $2D$ | This table displays the synchronized defects in full. --- # 53. Triple-critical table | Theory / method | Known tractable region | BADT's position | |---|---|---| | Roy–Waldschmidt quadratic relations | $\operatorname{trdeg}=1$ | $\operatorname{trdeg}=2$ | | Six Exponentials | Difference multiplicity $\ge3$ is impossible | Binary support maintains $\le2$ | | Waldschmidt–Masser | $\operatorname{rank} **BADT does not merely pack more logarithms of algebraic numbers into the same field; every added layer simultaneously adds a shared-anchor quadratic relation, a unit of transcendence-compression debt, a unit of matrix structural-rank defect, and a unit of geometric codimension, while the entire block matrix always remains exactly on the rank equality boundary of the Waldschmidt–Masser theorem.** The core synchronization identity is: $$ \boxed{ -\delta(Z_D) = \operatorname{CD}(Z_D) = \operatorname{codim}V_D = \operatorname{SRD}(B_D) = D. } $$ And: $$ \boxed{ \operatorname{rank}(B_D) = \frac{mn}{m+n} = D. } $$ Thus, the actual goal of Round 10 is not 'add another layer', but: $$ \boxed{ \text{Use the shared anchor to break this scale-invariant equality boundary}. } $$ --- # References 1. Samit Dasgupta, **Ranks of matrices of logarithms of algebraic numbers I: the theorems of Baker and Waldschmidt–Masser**, *Essential Number Theory* 2 (2023), 93–138. arXiv: https://arxiv.org/abs/2303.02037 The paper defines structural rank and states Baker's theorem, the Structural Rank Conjecture, and the Waldschmidt–Masser theorem. Its rank hypothesis is: $$ \operatorname{rank}(M) < \frac{mn}{m+n}. $$ 2. Samit Dasgupta and Mahesh Kakde, **Ranks of Matrices of Logarithms of Algebraic Numbers II: The Matrix Coefficient Conjecture**, *Advances in Mathematics* 487 (2026), 110753. DOI: 10.1016/j.aim.2025.110753 The $2\times2$ case of the Matrix Coefficient Conjecture is equivalent to the Four Exponentials Conjecture; the paper develops a support / Newton-polytope auxiliary-polynomial strategy while retaining a key implication as open. 3. Damien Roy and Michel Waldschmidt, **Quadratic Relations between Logarithms of Algebraic Numbers**, *Proceedings of the Japan Academy, Series A* 71 (1995), 151–153. DOI: 10.3792/pjaa.71.151 Handles the quadratic-relation case with generated-field transcendence degree $1$. 4. Damien Roy and Michel Waldschmidt, **Approximation diophantienne et indépendance algébrique de logarithmes**, *Annales Scientifiques de l’École Normale Supérieure* 30 (1997), 753–796. DOI: 10.1016/S0012-9593(97)89938-7 Develops a more general algebraic-independence method for algebraic-logarithm families of transcendence degree $1$. 5. Michel Waldschmidt, surveys on **Algebraic Independence of Logarithms of Algebraic Numbers** and Schanuel’s Conjecture. Contemporary surveys continue to emphasize that it is not known unconditionally whether there exist two algebraically independent logarithms of algebraic numbers. 6. Stéphane Fischler, **Orbits under algebraic groups and logarithms of algebraic numbers**, *Acta Arithmetica* C.2 (2001), 167–186. Discusses Roy’s geometric reformulation of algebraic independence through algebraic varieties and logarithmic points. 7. Alan Baker, classical theorem on linear independence of logarithms of algebraic numbers over algebraic coefficients. --- **End of ACTM Round 9**