Add files via upload

Author: DeepMathLLM

preliminary-archive-results/001-Alexander polynomial.md

@@ -0,0 +1,59 @@
+# Mathematical Object Origin Archive | Alexander Polynomial
+
+## 1. Archive Information
+
+- Standard Name: Alexander polynomial
+- Mathematical Field: Algebraic Topology (knot theory)
+- Abstract: The Alexander polynomial is a Laurent polynomial invariant of an oriented knot, defined up to multiplication by a unit \(\pm t^n\). It arose from the problem of deciding when two knot diagrams represent different knot types: it converts part of the topology of the knot complement into a polynomial that can be calculated from finite algebraic data. It supplies effective obstructions to knot equivalence, although it is not a complete classifier.
+
+## 2. Core Record
+
+### Precise Description
+
+Let \(K\subset S^3\) be an oriented knot and let
+\[
+X_K=S^3\setminus \operatorname{int}\nu(K)
+\]
+be its exterior. The canonical abelianization
+\[
+\pi_1(X_K)\longrightarrow H_1(X_K;\mathbb Z)\cong \mathbb Z
+\]
+determines an infinite cyclic cover \(\widetilde X_K\to X_K\). Choosing the positive deck transformation gives
+\[
+A_K=H_1(\widetilde X_K;\mathbb Z)
+\]
+the structure of a finitely presented module over the Laurent polynomial ring \(\Lambda=\mathbb Z[t,t^{-1}]\). This is the Alexander module.
+
+For a finite presentation \(\Lambda^m\xrightarrow{P}\Lambda^n\to A_K\to0\), the zeroth Fitting ideal \(\operatorname{Fitt}_0(A_K)\) is generated by all \(n\times n\) minors of \(P\) (and is zero if \(m<n\)). This ideal is independent of the chosen finite presentation. The Alexander polynomial \(\Delta_K(t)\) is the order of the torsion \(\Lambda\)-module \(A_K\): equivalently, it is a greatest common divisor of the generators of \(\operatorname{Fitt}_0(A_K)\), defined up to a unit \(\pm t^r\). For a knot, a Seifert matrix \(V\) gives the square presentation matrix \(tV-V^{T}\), so one may compute
+\[
+\Delta_K(t)\doteq \det(tV-V^{T}),
+\]
+where \(f\doteq g\) means \(f=\pm t^r g\) for some \(r\in\mathbb Z\). Thus the unnormalized invariant is an equivalence class in \(\mathbb Z[t,t^{-1}]\) modulo its units. It can be normalized by imposing symmetry and \(\Delta_K(1)=1\) [2,3].
+
+### Mathematical Context and Formation
+
+The motivating problem is the knot-equivalence problem: given two embeddings \(S^1\hookrightarrow S^3\), or two planar diagrams encoding them, determine whether an ambient isotopy carries one knot to the other. A diagram is finite, but its combinatorics is not itself invariant: Reidemeister moves can change the number and arrangement of crossings without changing the knot. The complement group is an invariant and admits a presentation from a diagram, but comparing finitely presented nonabelian groups is generally difficult; ordinary abelianization is too coarse because every knot exterior has first homology \(\mathbb Z\). Hence neither raw diagram data nor \(H_1(X_K;\mathbb Z)\) provides a practical discriminator for many knots.
+
+Alexander's 1928 construction attached matrices with polynomial entries to knot and link diagrams and extracted polynomial data invariant under the allowed diagram changes [1]. In modern topological terms, the decisive refinement is to use the same map to \(\mathbb Z\) that makes ordinary first homology uninformative, but retain the associated infinite cyclic cover. Its deck transformation acts on homology, and the indeterminate \(t\) records that action. Consequently \(H_1(\widetilde X_K;\mathbb Z)\) retains how one-cycles in the cover are translated from sheet to sheet, information erased when one keeps only \(H_1(X_K;\mathbb Z)\). A diagrammatic Wirtinger presentation, followed by Fox differentiation, or a Seifert surface and its Seifert pairing, converts this \(\Lambda\)-module into a finite matrix; the invariant Fitting ideal and its gcd then compress the module to one polynomial [2,3]. This module-and-order description is a modern structural interpretation of the diagrammatic invariant, not a claim that Alexander originally formulated it in precisely these terms.
+
+### Essential Role
+
+The polynomial makes a definite part of the equivalence problem tractable. Ambient isotopy induces the corresponding module isomorphism and therefore preserves \(\Delta_K(t)\) up to \(\pm t^r\). Thus, if two computed polynomials are inequivalent under multiplication by a unit, the knots cannot be ambient-isotopic. For example, the unknot has \(\Delta(t)=1\), whereas a trefoil has \(\Delta(t)\doteq t^2-t+1\); the polynomial therefore certifies that the trefoil is not the unknot [2,3].
+
+The features accomplishing this are specific: the cyclic cover preserves meridional winding through the \(t\)-action; finite presentation makes that information computable from a diagram or Seifert surface; and the zeroth Fitting ideal generated by the full-size minors is invariant under changes of finite presentation. The individual minors need not remain fixed or become associates; rather, the invariant ideal they collectively generate, followed by its gcd/order, produces the polynomial up to a unit. The object thereby replaces a difficult direct search for an isotopy with an algebraic obstruction obtained by polynomial arithmetic. It also introduces the structural viewpoint that knot invariants can be extracted from modules over a group ring rather than from ordinary homology alone.
+
+Its contribution is obstructive rather than classificatory. Equal Alexander polynomials do not imply equivalent knots, and some nontrivial knots have \(\Delta_K(t)=1\). The polynomial overcomes the inadequacy of ordinary abelianization only partially: taking the order discards information contained in the full Alexander module [2,3].
+
+## 3. Notes
+
+- The Alexander polynomial of a link with more than one component requires additional conventions; multivariable Alexander polynomials retain one variable for each independent meridional direction. This archive concerns the one-variable polynomial of a knot.
+- Depending on conventions, a presentation may use \(V-tV^T\), \(tV-V^T\), or substitutions such as \(t\mapsto t^{-1}\); these yield the same invariant up to the stated unit ambiguity.
+- The Alexander module and the Alexander polynomial are related but distinct objects: the latter is a determinantal summary of the former.
+
+## 4. Sources
+
+[1] J. W. Alexander, “Topological Invariants of Knots and Links,” *Transactions of the American Mathematical Society* 30 (1928), 275–306. https://doi.org/10.1090/S0002-9947-1928-1501429-1
+
+[2] R. H. Crowell and R. H. Fox, *Introduction to Knot Theory*, Graduate Texts in Mathematics 57, Springer, 1977, especially the development of the Alexander module and polynomial.
+
+[3] Dale Rolfsen, *Knots and Links*, Publish or Perish, 1976; AMS Chelsea reprint, 2003, Chapter 7.

preliminary-archive-results/002-Bakry–Émery curvature-dimension condition.md

@@ -0,0 +1,82 @@
+# Mathematical Object Origin Archive | Bakry–Émery curvature-dimension condition
+
+## 1. Archive Information
+
+- Standard Name: Bakry–Émery curvature-dimension condition
+- Mathematical Field: Probability Theory and Stochastic Processes; geometric analysis
+- Abstract: The Bakry–Émery curvature-dimension condition, denoted \(CD(\rho,N)\), is a pointwise inequality for the first and second carré du champ forms of a diffusion generator. It arose as a locally checkable criterion for the concrete problem of proving hypercontractivity of Markov diffusion semigroups, replacing a difficult global operator-norm question by an infinitesimal inequality that abstracts the curvature and dimension terms in the Bochner formula.
+
+## 2. Core Record
+
+### Precise Description
+
+Let \(L\) be the generator of a Markov diffusion semigroup \((P_t)_{t\ge 0}\), acting on a suitable algebra \(\mathcal A\) of test functions and satisfying the diffusion chain rule. Its carré du champ is
+\[
+\Gamma(f,g)=\frac12\bigl(L(fg)-fLg-gLf\bigr),
+\qquad \Gamma(f)=\Gamma(f,f),
+\]
+and its iterated carré du champ is
+\[
+\Gamma_2(f,g)=\frac12\bigl(L\Gamma(f,g)-\Gamma(f,Lg)-\Gamma(g,Lf)\bigr),
+\qquad \Gamma_2(f)=\Gamma_2(f,f).
+\]
+For \(\rho\in\mathbb R\) and \(N\in[1,\infty]\), the generator satisfies the Bakry–Émery curvature-dimension condition \(CD(\rho,N)\) when, pointwise for every admissible \(f\),
+\[
+\Gamma_2(f)\ge \rho\,\Gamma(f)+\frac1N(Lf)^2,
+\]
+where \(1/\infty=0\). Thus \(CD(\rho,\infty)\) is the \(\Gamma_2\)-criterion \(\Gamma_2\ge \rho\Gamma\) [1,2].
+
+For the Laplace–Beltrami generator \(L=\Delta\) on an \(n\)-dimensional Riemannian manifold,
+\[
+\Gamma(f)=|\nabla f|^2,
+\qquad
+\Gamma_2(f)=\|\operatorname{Hess}f\|^2+\operatorname{Ric}(\nabla f,\nabla f).
+\]
+Consequently, the lower Ricci bound \(\operatorname{Ric}\ge \rho g\), together with \(\|\operatorname{Hess}f\|^2\ge (\Delta f)^2/n\), gives \(CD(\rho,n)\). This model explains the two terms in the abstract condition: \(\rho\) acts as a curvature lower bound and \(N\) as an upper dimension parameter [2].
+
+### Mathematical Context and Formation
+
+The motivating problem was to decide when a diffusion Markov semigroup is hypercontractive: given an invariant probability measure \(\mu\) and \(p>1\), one seeks times \(t>0\) and exponents \(q(t)>p\) for which
+\[
+\|P_t f\|_{L^{q(t)}(\mu)}\le \|f\|_{L^p(\mu)}.
+\]
+This is a global, nonlinear norm-improvement property. Gross's logarithmic Sobolev theory had related hypercontractivity to a logarithmic Sobolev inequality, but that reformulation still left a global integral inequality to be verified for each diffusion [3]. Direct estimates of transition kernels or of changing \(L^p\)-norms were strongly dependent on the particular process and did not supply a uniform criterion expressed in the generator itself.
+
+Bakry and Émery's formation of the condition used the observation that the generator already contains an algebraic substitute for a squared gradient, namely \(\Gamma\). Differentiating this energy under the semigroup introduces exactly the second-order expression \(\Gamma_2\). On a manifold, the Bochner identity decomposes that expression into a nonnegative Hessian term and a Ricci-curvature term; the trace inequality for the Hessian contributes the dimension-dependent quantity \((Lf)^2/N\). The resulting inequality
+\(\Gamma_2\ge \rho\Gamma+(Lf)^2/N\) therefore retains precisely the pieces of the geometric computation needed for semigroup estimates while making no reference to coordinates, a Riemannian metric, or even a finite-dimensional state space. In their work on hypercontractive diffusions, Bakry and Émery identified strong positivity of the iterated carré du champ as the sufficient condition that could be checked locally and then propagated by the semigroup [1].
+
+### Essential Role
+
+The condition made the infinitesimal-to-global step in the hypercontractivity problem tractable. For example, under \(CD(\rho,\infty)\), apply the condition to the interpolation
+\[
+\Phi(s)=P_s\!\left(\Gamma(P_{t-s}f)\right),\qquad 0\le s\le t.
+\]
+A direct differentiation gives
+\[
+\Phi'(s)=2P_s\!\left(\Gamma_2(P_{t-s}f)\right)
+\ge 2\rho\,\Phi(s).
+\]
+Integration yields the gradient estimate
+\[
+\Gamma(P_t f)\le e^{-2\rho t}P_t\Gamma(f).
+\]
+Thus the particular form of \(\Gamma_2\)—the derivative of carré-du-champ energy along the semigroup—and its lower bound by \(\Gamma\) produce a closed differential inequality to which Grönwall's lemma applies. With an invariant probability measure and the usual symmetry, ergodicity, domain, and regularity assumptions, the same semigroup interpolation and diffusion chain rule give, for \(\rho>0\),
+\[
+\operatorname{Ent}_\mu(f^2)
+\le \frac{2}{\rho}\int \Gamma(f)\,d\mu,
+\]
+and hence the desired hypercontractivity through the logarithmic Sobolev equivalence [1,3].
+
+This was the direct contribution: a global \(L^p\)-to-\(L^q\) property could be proved from a pointwise inequality involving only \(L\) and its first two carré du champ forms. The curvature term supplies exponential control, while the finite-\(N\) term records the extra coercivity coming from dimension and permits dimension-sensitive refinements. Conceptually, the condition also showed that the analytic content of a Ricci lower bound can be encoded at the level of a Markov generator, allowing the hypercontractivity argument to survive beyond the original Riemannian examples [1,2].
+
+## 3. Notes
+
+The condition is a property of a diffusion generator, not the same object as the Bakry–Émery Ricci tensor on a weighted manifold. For a weighted Laplacian, that tensor appears in the Bochner formula and can be used to verify the generator condition. The notation \(CD(\rho,N)\) here refers to the carré-du-champ formulation; later metric-measure curvature-dimension conditions use related notation but are defined by different structures. Geometric comparison theorems, concentration bounds, and nonsmooth extensions are important later developments, not the motivating role recorded here.
+
+## 4. Sources
+
+[1] Dominique Bakry and Michel Émery, “Diffusions hypercontractives,” *Séminaire de probabilités de Strasbourg* 19, Lecture Notes in Mathematics 1123 (1985), 177–206. https://www.numdam.org/article/SPS_1985__19__177_0.pdf
+
+[2] Dominique Bakry, “The Geometry of Markov Diffusion Generators,” *Annales de la Faculté des sciences de Toulouse: Mathématiques*, Série 6, 9(2) (2000), 305–366. https://www.numdam.org/item/?id=AFST_2000_6_9_2_305_0
+
+[3] Leonard Gross, “Logarithmic Sobolev Inequalities,” *American Journal of Mathematics* 97(4) (1975), 1061–1083.