{
  "id": "quadratic-reciprocity",
  "title": "Quadratic Reciprocity",
  "lang": "en",
  "source_lang": "en",
  "summary": "Quadratic reciprocity relates the solvability of two prime-modulus square congruences and provides a foundation for general reciprocity laws in number theory.",
  "category": "mathematics",
  "subcategory": "algebra",
  "type": "concept",
  "aliases": [
    "Law of quadratic reciprocity",
    "quadratic reciprocity theorem"
  ],
  "generated_at": "2026-11-05",
  "model": "pro/gpt-6.1-sol",
  "url": "https://aiwiki.page/en/quadratic-reciprocity/",
  "translations": {
    "zh-Hans": {
      "url": "https://aiwiki.page/zh/quadratic-reciprocity/",
      "json": "https://aiwiki.page/zh/quadratic-reciprocity.json"
    }
  },
  "sections": [
    {
      "heading": "",
      "text": "Quadratic reciprocity is a theorem in number theory describing a symmetry, with a precisely determined sign correction, between two questions: whether an odd prime \\(p\\) is a square modulo another odd prime \\(q\\), and whether \\(q\\) is a square modulo \\(p\\). Its central assertion is that the answers agree unless both primes are congruent to \\(3\\) modulo \\(4\\), in which case they are opposite. Together with two supplementary laws concerning \\(-1\\) and \\(2\\), it provides a systematic method for deciding whether a quadratic congruence has a solution.\n\nThe theorem emerged from eighteenth-century investigations of squares and prime numbers and received its first complete proof from Carl Friedrich Gauss in 1796. Its significance extends beyond the calculation of square residues: it connects elementary modular arithmetic with quadratic fields, character sums, and local–global principles. The search for analogous laws for higher powers became one of the principal motivations for modern algebraic number theory."
    },
    {
      "heading": "Quadratic residues and the Legendre symbol",
      "text": "Let \\(p\\) be an odd prime and let \\(a\\) be an integer not divisible by \\(p\\). The integer \\(a\\) is a quadratic residue modulo \\(p\\) if there is an integer \\(x\\) satisfying\n\n\\[x^2\\equiv a\\pmod p.\\]\n\nOtherwise it is a quadratic nonresidue. The terminology depends only on the residue class of \\(a\\): adding a multiple of \\(p\\) does not change the question. For example, the nonzero squares modulo \\(7\\) are \\(1,2,4\\), since the squares of \\(1,2,3\\) are congruent to \\(1,4,2\\), and the remaining nonzero inputs give the same values.\n\nThe Legendre symbol records this distinction:\n\n\\[\\left(\\frac ap\\right)= \\begin{cases} 0,&p\\mid a,\\\\ 1,&a\\text{ is a nonzero square modulo }p,\\\\ -1,&a\\text{ is a nonsquare modulo }p. \\end{cases}\\]\n\nDespite its appearance, this notation is not a fraction. Its lower argument is an odd prime, while its upper argument may be any integer. In particular, \\(\\left(\\frac ap\\right)\\) depends on \\(a\\bmod p\\), not on the numerical size or sign of \\(a\\).\n\nThe residue classes modulo \\(p\\) form a finite field \\(\\mathbb F_p\\). Its nonzero elements form a multiplicative group, and the squaring map on this group has kernel \\(\\{1,-1\\}\\). Consequently exactly \\((p-1)/2\\) nonzero residue classes are squares. Their products and inverses are again squares, so they form a subgroup of index \\(2\\). This group-theoretic structure explains the multiplicative identity\n\n\\[\\left(\\frac{ab}{p}\\right) = \\left(\\frac ap\\right)\\left(\\frac bp\\right).\\]\n\nThus the product of two nonresidues is a residue, although neither factor is itself a square.\n\nEuler’s criterion gives an equivalent computational description:\n\n\\[a^{(p-1)/2}\\equiv\\left(\\frac ap\\right)\\pmod p\\]\n\nwhen \\(p\\nmid a\\). Indeed, Fermat’s little theorem implies that this power is a root of \\(X^2-1\\), so it is \\(1\\) or \\(-1\\). Every nonzero square gives \\(1\\); a root-counting argument for a polynomial shows that no additional inputs can do so.\n\nEuler’s criterion tests a particular numerator against a particular prime. Quadratic reciprocity adds something qualitatively different: it relates two tests with the numerator and denominator interchanged. That relationship permits repeated reductions to smaller numbers and reveals that the apparent irregularity of square residues is governed by a simple rule."
    },
    {
      "heading": "The reciprocity law and its supplements",
      "text": "For distinct odd primes \\(p\\) and \\(q\\), quadratic reciprocity states\n\n\\[\\boxed{ \\left(\\frac pq\\right)\\left(\\frac qp\\right) = (-1)^{\\frac{p-1}{2}\\frac{q-1}{2}} }.\\]\n\nThe exponent is odd precisely when both \\(p\\) and \\(q\\) are congruent to \\(3\\pmod4\\). Equivalently,\n\n\\[\\left(\\frac pq\\right)= \\begin{cases} \\left(\\frac qp\\right), & p\\equiv1\\pmod4\\text{ or }q\\equiv1\\pmod4,\\\\[2mm] -\\left(\\frac qp\\right), & p\\equiv q\\equiv3\\pmod4. \\end{cases}\\]\n\nThe restriction to distinct primes excludes a trivial zero case: if \\(p=q\\), both symbols vanish and the displayed sign formula does not apply.\n\nThe first supplementary law determines the status of \\(-1\\):\n\n\\[\\left(\\frac{-1}{p}\\right)=(-1)^{(p-1)/2}.\\]\n\nThus \\(-1\\) is a square modulo \\(p\\) exactly when \\(p\\equiv1\\pmod4\\). For example, \\(5^2\\equiv-1\\pmod{13}\\), whereas no square is congruent to \\(-1\\pmod7\\). This supplement also determines the effect of replacing a numerator by its negative.\n\nThe second supplementary law handles the even prime:\n\n\\[\\left(\\frac2p\\right)=(-1)^{(p^2-1)/8}.\\]\n\nThe exponent is an integer for every odd \\(p\\). In terms of residue classes,\n\n\\[\\left(\\frac2p\\right)= \\begin{cases} 1,&p\\equiv1,7\\pmod8,\\\\ -1,&p\\equiv3,5\\pmod8. \\end{cases}\\]\n\nFor instance, \\(2\\) is a square modulo \\(17\\), since \\(6^2\\equiv2\\pmod{17}\\), but it is not a square modulo \\(13\\).\n\nSmall examples illustrate the essential sign distinction. The primes \\(3\\) and \\(11\\) are both \\(3\\pmod4\\). Here \\(3\\) is a square modulo \\(11\\), since \\(5^2\\equiv3\\), but \\(11\\equiv2\\pmod3\\) is not a square modulo \\(3\\). By contrast, \\(5\\equiv1\\pmod4\\), so\n\n\\[\\left(\\frac5{11}\\right)=\\left(\\frac{11}5\\right)=1.\\]\n\nIndeed, \\(4^2\\equiv5\\pmod{11}\\).\n\nA frequent consequence is a classification of primes for which a fixed small integer is a square. Applying reciprocity to \\(3\\), and excluding \\(p=3\\), gives\n\n\\[\\left(\\frac3p\\right)=1 \\quad\\Longleftrightarrow\\quad p\\equiv1\\text{ or }11\\pmod{12}.\\]\n\nSimilarly,\n\n\\[\\left(\\frac5p\\right)=1 \\quad\\Longleftrightarrow\\quad p\\equiv1\\text{ or }4\\pmod5\\]\n\nfor odd primes \\(p\\ne5\\). The first classification requires both a reciprocity sign and reduction modulo \\(3\\); the second has no sign correction because \\(5\\equiv1\\pmod4\\)."
    },
    {
      "heading": "Historical development",
      "text": "Problems about the representation of primes by quadratic expressions supplied much of the background to quadratic reciprocity. Pierre de Fermat studied assertions such as the representation of primes congruent to \\(1\\pmod4\\) as sums of two squares. Such results naturally lead to questions about when \\(-1\\), \\(2\\), or another fixed integer is a square modulo a prime divisor of a quadratic expression. Before reciprocity was formulated as a single theorem, many of its individual consequences appeared as separate patterns.\n\nLeonhard Euler investigated these patterns extensively during the eighteenth century. He recognized relationships between the prime divisors of quadratic forms and the congruence classes of the primes involved, and formulated versions of the reciprocity phenomenon. The modern symbolic expression, however, should not be projected unchanged onto these early investigations: their language was often that of divisors of forms rather than a uniform calculus of residue symbols.\n\nAdrien-Marie Legendre gave an explicit formulation in a memoir published in 1785 and developed the subject further in his Essai sur la théorie des nombres of 1798. The notation now called the Legendre symbol made the statement especially compact. His attempted proofs did not establish the theorem in full generality; they relied on assertions about the existence of primes with specified residue properties that had not been proved in the required generality.\n\nGauss found a complete proof in 1796, at the age of eighteen, and published it in Disquisitiones Arithmeticae in 1801. His treatment made reciprocity a central organizing theorem of arithmetic rather than an isolated observation. The proof used a carefully structured induction with auxiliary results about quadratic residues. Gauss subsequently returned to the theorem repeatedly and produced several proofs based on different ideas, including methods associated with products of residues and roots of unity.\n\nThe multiplication of proofs was mathematically productive. A result whose elementary statement concealed several different structures could be approached through counting, permutations, cyclotomic expressions, and quadratic forms. During the nineteenth century, these approaches helped identify the concepts needed for reciprocity laws in broader settings. Gauss’s work on biquadratic residues, for example, led naturally to arithmetic involving complex integers rather than ordinary integers alone.\n\nThe historical importance of quadratic reciprocity therefore lies both in its resolution and in the problems it generated. The effort to generalize it contributed to the development of algebraic integers, ideals, and field extensions. Later reciprocity theories preserved its basic theme—connecting residue information at different primes—while replacing the elementary sign rule with more general algebraic structures."
    },
    {
      "heading": "An elementary proof through Gauss’s lemma",
      "text": "One of the most transparent proofs combines Gauss’s lemma for quadratic residues with a lattice-point count. Let \\(p\\) be an odd prime, \\(a\\) coprime to \\(p\\), and \\(m=(p-1)/2\\). Consider\n\n\\[a,2a,\\ldots,ma\\]\n\nand reduce each term to its least positive residue modulo \\(p\\). Let \\(N\\) count how many of these residues exceed \\(p/2\\). Gauss’s lemma asserts\n\n\\[\\left(\\frac ap\\right)=(-1)^N.\\]\n\nTo see why, replace every residue above \\(p/2\\) by its negative modulo \\(p\\), producing integers between \\(1\\) and \\(m\\). Their absolute values are distinct: an equality would imply \\(ai\\equiv\\pm aj\\pmod p\\), which forces \\(i=j\\) within the given range. They therefore constitute a permutation of \\(1,\\ldots,m\\). Multiplying gives\n\n\\[a^m m!\\equiv(-1)^N m!\\pmod p.\\]\n\nCancellation and Euler’s criterion prove the lemma.\n\nFor odd positive \\(a\\), the same lemma yields the useful parity formula\n\n\\[\\left(\\frac ap\\right) = (-1)^{\\displaystyle\\sum_{i=1}^{(p-1)/2} \\left\\lfloor\\frac{ai}{p}\\right\\rfloor}.\\]\n\nHere \\(\\lfloor t\\rfloor\\) is the greatest integer not exceeding \\(t\\). To verify the parity relation, write \\(ai=pk_i+r_i\\), with \\(1\\le r_i<p\\), and let \\(t_i\\) be the absolute least residue. Modulo \\(2\\), replacing \\(r_i\\) by \\(t_i\\) changes the sum by one for each of the \\(N\\) replacements, because \\(p\\) is odd. Since the \\(t_i\\) permute \\(1,\\ldots,m\\) and \\(a\\) is odd, summing the equations gives \\(\\sum k_i\\equiv N\\pmod2\\).\n\nNow take distinct odd primes \\(p,q\\), and put\n\n\\[m=\\frac{p-1}{2},\\qquad n=\\frac{q-1}{2}.\\]\n\nIn the rectangle of integer points\n\n\\[1\\le i\\le m,\\qquad1\\le j\\le n,\\]\n\nno point lies on the line \\(q i=p j\\). Such an equality would force \\(p\\mid i\\), which is impossible in this range. For each \\(i\\), the number of points with \\(j<qi/p\\) is \\(\\lfloor qi/p\\rfloor\\). Similarly, summing \\(\\lfloor pj/q\\rfloor\\) counts the points on the other side. Hence\n\n\\[\\sum_{i=1}^{m}\\left\\lfloor\\frac{qi}{p}\\right\\rfloor+ \\sum_{j=1}^{n}\\left\\lfloor\\frac{pj}{q}\\right\\rfloor =mn.\\]\n\nApplying the parity formula to the two sums proves reciprocity.\n\nThe supplements fit the same framework. For \\(a=-1\\), all \\(m\\) residues lie above \\(p/2\\). For \\(a=2\\), the relevant count is\n\n\\[N=\\frac{p-1}{2}-\\left\\lfloor\\frac p4\\right\\rfloor.\\]\n\nIts parity, evaluated for \\(p\\equiv1,3,5,7\\pmod8\\), gives the supplementary law for \\(2\\). Thus a single residue-counting principle proves all three laws."
    },
    {
      "heading": "Computation, Jacobi symbols, and quadratic equations",
      "text": "Quadratic reciprocity turns residue testing into a reduction procedure resembling the Euclidean algorithm. First reduce the numerator modulo the denominator. Remove factors of \\(2\\) using the second supplement, then interchange odd numerator and denominator, introducing a minus sign only when both are \\(3\\pmod4\\). Repetition eventually reaches a numerator of \\(1\\), or detects a common factor.\n\nFor a direct example,\n\n\\[\\left(\\frac{37}{101}\\right) = \\left(\\frac{101}{37}\\right) = \\left(\\frac{27}{37}\\right) = \\left(\\frac3{37}\\right)^3.\\]\n\nNo negative sign appears in the first interchange because \\(37\\equiv1\\pmod4\\). A further application gives\n\n\\[\\left(\\frac3{37}\\right) = \\left(\\frac{37}3\\right) = \\left(\\frac13\\right)=1.\\]\n\nThus \\(37\\) is a square modulo \\(101\\), without searching through the possible roots.\n\nFor efficient general computation, the Jacobi symbol extends the notation to every positive odd denominator. If\n\n\\[n=\\prod_i p_i^{e_i},\\]\n\ndefine\n\n\\[\\left(\\frac an\\right)= \\prod_i\\left(\\frac a{p_i}\\right)^{e_i},\\]\n\nwith value \\(1\\) when \\(n=1\\). Its multiplicativity, reciprocity law, and supplementary laws allow its evaluation without first factoring \\(n\\). In particular, for positive odd coprime \\(m,n\\),\n\n\\[\\left(\\frac mn\\right)\\left(\\frac nm\\right) = (-1)^{(m-1)(n-1)/4}.\\]\n\nAn iterative algorithm therefore needs only remainders, removal of powers of \\(2\\), swaps, and sign changes.\n\nThe distinction between Legendre and Jacobi symbols is essential. A Jacobi value of \\(1\\) need not indicate an actual square modulo a composite denominator. For example,\n\n\\[\\left(\\frac2{15}\\right) = \\left(\\frac23\\right)\\left(\\frac25\\right)=1,\\]\n\nbut \\(x^2\\equiv2\\pmod{15}\\) has no solution because it already has none modulo \\(3\\). This distinction underlies the Solovay–Strassen primality test, which checks the prime-modulus identity relating Euler’s criterion to a Jacobi symbol; composite integers can fail that identity.\n\nResidue testing also decides the solvability of a general quadratic equation modulo an odd prime. If \\(p\\nmid a\\), then\n\n\\[ax^2+bx+c\\equiv0\\pmod p\\]\n\nis equivalent to\n\n\\[(2ax+b)^2\\equiv b^2-4ac\\pmod p.\\]\n\nThe discriminant therefore determines whether there are zero, one, or two roots. Reciprocity decides whether a nonzero discriminant is a square; it does not itself construct its square root. Algorithms such as the Tonelli–Shanks algorithm perform that separate task. For composite moduli, prime-power analysis and the Chinese remainder theorem replace the single prime-field calculation."
    },
    {
      "heading": "Quadratic fields and representation by forms",
      "text": "A central algebraic interpretation concerns the splitting of primes in a quadratic number field\n\n\\[K=\\mathbb Q(\\sqrt d),\\]\n\nwhere \\(d\\ne1\\) is squarefree. This is a degree-two extension of the rational numbers. Its ring of algebraic integers is\n\n\\[\\mathcal O_K= \\begin{cases} \\mathbb Z[(1+\\sqrt d)/2],&d\\equiv1\\pmod4,\\\\ \\mathbb Z[\\sqrt d],&d\\not\\equiv1\\pmod4. \\end{cases}\\]\n\nThe field discriminant is correspondingly \\(D=d\\) or \\(D=4d\\).\n\nFor an odd prime \\(p\\nmid D\\), the prime ideal factorization of \\(p\\mathcal O_K\\) has two possibilities. If \\(\\left(\\frac dp\\right)=1\\), then\n\n\\[p\\mathcal O_K=\\mathfrak p_1\\mathfrak p_2\\]\n\nwith two distinct prime ideals, each having residue field \\(\\mathbb F_p\\); the prime is said to split. If \\(\\left(\\frac dp\\right)=-1\\), then \\(p\\mathcal O_K\\) remains prime, and its residue field has \\(p^2\\) elements; the prime is inert. Primes dividing \\(D\\) ramify instead. Thus a question about the congruence \\(x^2\\equiv d\\pmod p\\) becomes a statement about prime factorization in another number system.\n\nThe Kronecker symbol, an extension of the Jacobi symbol that also handles even denominators, packages the splitting information as\n\n\\[\\chi_D(n)=\\left(\\frac Dn\\right).\\]\n\nFor a quadratic field discriminant \\(D\\), this is a real Dirichlet character of modulus \\(|D|\\). Quadratic reciprocity explains why the behavior of varying primes in a fixed quadratic field can be expressed by congruence classes modulo a fixed integer. For \\(K=\\mathbb Q(\\sqrt5)\\), for example, odd primes other than \\(5\\) split precisely when they are \\(1\\) or \\(4\\pmod5\\).\n\nThe connection with quadratic forms is related but more refined. A binary form\n\n\\[Ax^2+Bxy+Cy^2\\]\n\nhas discriminant \\(B^2-4AC\\). If an odd prime \\(p\\nmid A\\) is represented by the form, reduction modulo \\(p\\) gives\n\n\\[(2Ax+By)^2\\equiv(B^2-4AC)y^2\\pmod p.\\]\n\nFor a prime representation, \\(y\\) cannot vanish modulo \\(p\\) under these assumptions, so the discriminant must be a square modulo \\(p\\).\n\nSuch a residue condition is not generally sufficient for representation by a specified form: distinct equivalence classes of forms can share a discriminant. In the special case of \\(x^2+y^2\\), however, the splitting condition in \\(\\mathbb Q(i)\\) does give the familiar theorem that an odd prime is a sum of two integer squares exactly when it is \\(1\\pmod4\\). For more general discriminants, the distinction between splitting and representation records additional ideal-class information that quadratic reciprocity alone does not determine."
    },
    {
      "heading": "Character sums and analytic interpretations",
      "text": "For an odd prime \\(p\\), the function\n\n\\[\\chi_p(a)=\\left(\\frac ap\\right)\\]\n\nis multiplicative on the nonzero residue classes and takes values in \\(\\{1,-1\\}\\). Extending it by \\(\\chi_p(0)=0\\) makes it convenient for finite sums. This character viewpoint connects reciprocity with harmonic analysis on finite groups: additive oscillations and multiplicative residue information can be combined in a single expression.\n\nThe basic quadratic Gauss sum is\n\n\\[\\tau_p=\\sum_{a=0}^{p-1} \\chi_p(a)e^{2\\pi i a/p}.\\]\n\nIts coefficients are arithmetic signs, while its exponential factors are roots of unity in the complex numbers. A fundamental identity is\n\n\\[\\tau_p^2=\\chi_p(-1)p.\\]\n\nIn particular, \\(|\\tau_p|=\\sqrt p\\), with \\(\\tau_p\\) real for \\(p\\equiv1\\pmod4\\) and purely imaginary for \\(p\\equiv3\\pmod4\\). The identity follows by combining the absolute-value calculation \\(\\tau_p\\overline{\\tau_p}=p\\) with\n\n\\[\\overline{\\tau_p}=\\chi_p(-1)\\tau_p.\\]\n\nThese sums also explain how quadratic characters behave under a finite Fourier transform. If \\(b\\not\\equiv0\\pmod p\\), a change of variable gives\n\n\\[\\sum_{a=0}^{p-1}\\chi_p(a)e^{2\\pi i ab/p} = \\chi_p(b)\\tau_p.\\]\n\nThus the transformed character is a scalar multiple of itself. The nontrivial square-root factor in this identity is an analytic manifestation of the same sign structure that occurs in elementary reciprocity.\n\nReciprocity can be derived by comparing Gauss sums associated with different primes inside fields generated by roots of unity. The comparison requires more than the squared identity alone: it must track how automorphisms act on the roots of unity and therefore on the sums. This approach places the elementary theorem within Galois theory, where a residue symbol describes the action of an arithmetic automorphism on an algebraic quantity.\n\nQuadratic characters also define Dirichlet L-functions,\n\n\\[L(s,\\chi_D)=\\sum_{n=1}^{\\infty}\\frac{\\chi_D(n)}{n^s}, \\qquad \\operatorname{Re}(s)>1.\\]\n\nTheir Euler products organize primes according to the splitting values \\(1\\), \\(-1\\), and \\(0\\). Reciprocity supplies the fixed-modulus character that makes this analytic treatment possible. Through Dirichlet’s theorem on primes in arithmetic progressions, splitting and inert primes in a quadratic field each have density \\(1/2\\) among all primes, after the finitely many ramified primes are excluded. Reciprocity itself does not prove that distribution theorem; it identifies the congruence classes to which the theorem applies."
    },
    {
      "heading": "Local–global formulations and higher reciprocity",
      "text": "A symmetric formulation of quadratic reciprocity uses completions of the rational numbers. These include the real field and the fields of p-adic numbers \\(\\mathbb Q_p\\), one for each prime \\(p\\). For nonzero rational \\(a,b\\), the Hilbert symbol \\((a,b)_v\\) at a place \\(v\\) is \\(1\\) if\n\n\\[z^2=ax^2+by^2\\]\n\nhas a nonzero solution over the corresponding completion, and is \\(-1\\) otherwise. It records a local solvability condition for a quadratic form.\n\nHilbert reciprocity states\n\n\\[\\prod_v(a,b)_v=1.\\]\n\nOnly finitely many factors differ from \\(1\\). For distinct odd primes \\(p,q\\), the nontrivial contributions for \\((p,q)\\) can occur only at \\(p,q,\\) and \\(2\\); the real contribution is \\(1\\) because both numbers are positive. More explicitly,\n\n\\[(p,q)_p=\\left(\\frac qp\\right),\\qquad (p,q)_q=\\left(\\frac pq\\right),\\]\n\nwhile\n\n\\[(p,q)_2=(-1)^{(p-1)(q-1)/4}.\\]\n\nThe product formula therefore gives exactly the classical reciprocity law. In this language, its apparently exceptional sign is the contribution from the prime \\(2\\).\n\nThis formulation is related to, but distinct from, the Hasse–Minkowski theorem, which states that a quadratic form over the rationals has a nontrivial rational zero if and only if it has one over the reals and over every \\(\\mathbb Q_p\\). The product formula constrains the collection of local symbols; the Hasse–Minkowski theorem turns a complete set of local solvability conditions into a global rational solution. Both express the special effectiveness of local methods for quadratic equations.\n\nFor higher powers, the analogous questions become more intricate. Cubic and quartic reciprocity laws concern third- and fourth-power residues. Their natural settings involve rings containing the corresponding roots of unity, rather than only ordinary integers. Cubic reciprocity is naturally expressed using Eisenstein integers, while quartic reciprocity uses Gaussian integers. Choices of suitable prime generators and congruence normalizations replace the simple distinction between primes congruent to \\(1\\) or \\(3\\pmod4\\).\n\nThe broad abelian generalization is class field theory. Its Artin reciprocity law relates ideal-theoretic arithmetic in a number field to the automorphism groups of its abelian extensions. For a quadratic extension of the rationals, the automorphism group has two elements, and the Frobenius automorphism of an unramified prime is encoded by the quadratic character \\(\\chi_D(p)\\). Classical quadratic reciprocity is therefore the degree-two instance of a framework in which residue symbols describe how primes act in field extensions. Extensions with nonabelian automorphism groups require structures beyond the abelian reciprocity map."
    }
  ],
  "links": [
    "number-theory",
    "prime-number",
    "carl-friedrich-gauss",
    "modular-arithmetic",
    "integer",
    "fraction",
    "finite-field",
    "group-theory",
    "fermats-little-theorem",
    "polynomial",
    "pierre-de-fermat",
    "leonhard-euler",
    "permutation",
    "euclidean-algorithm",
    "quadratic-equation",
    "discriminant",
    "chinese-remainder-theorem",
    "field-extension",
    "rational-number",
    "ideal-ring-theory",
    "kronecker-delta",
    "quadratic-form",
    "harmonic-analysis",
    "complex-number",
    "fourier-transform",
    "galois-theory"
  ],
  "sources": [],
  "license": "https://creativecommons.org/licenses/by/4.0/"
}