<?xml version="1.0" encoding="utf-8" standalone="yes"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom" xmlns:content="http://purl.org/rss/1.0/modules/content/">
  <channel>
    <title>Math on Quasicoherent</title>
    <link>https://www.quasicoherent.blog/tags/math/</link>
    <description>Recent content in Math on Quasicoherent</description>
    <generator>Hugo -- gohugo.io</generator>
    <language>en</language>
    <copyright>Copyright &amp;copy; 2019 - Kevin shu</copyright>
    <lastBuildDate>Mon, 12 Oct 2026 00:00:00 -0400</lastBuildDate>
    <atom:link href="https://www.quasicoherent.blog/tags/math/index.xml" rel="self" type="application/rss+xml" />
    <item>
      <title>A Digestion of A Counterexample to the Generalized Lax Conjecture</title>
      <link>https://www.quasicoherent.blog/posts/generalized_lax/</link>
      <pubDate>Mon, 12 Oct 2026 00:00:00 -0400</pubDate>
      <guid>https://www.quasicoherent.blog/posts/generalized_lax/</guid>
      <description>$$&#xA;\newcommand{\R}{\mathbb{R}}&#xA;\newcommand{\C}{\mathbb{C}}&#xA;\newcommand{\S}{\mathbb{S}}&#xA;\newcommand{\T}{\intercal}&#xA;\newcommand{\adj}{\operatorname{adj}}&#xA;\newcommand{\im}{\operatorname{Im}}&#xA;\newcommand{\re}{\operatorname{Re}}&#xA;\newcommand{\op}{\mathrm{op}}&#xA;\newcommand{\proj}{\operatorname{proj}}&#xA;\renewcommand{\iota}{\operatorname{\mathfrak{I}}}&#xA;$$&lt;h1 id=&#34;preliminaries-on-hyperbolic-polynomials&#34;&gt;Preliminaries on Hyperbolic Polynomials&lt;/h1&gt;&#xA;&lt;p&gt;The generalized Lax conjecture was a central topic of study in the field of hyperbolic polynomials. Hyperbolic polynomials in general serve as a bridge between the seemingly disparate fields of algebraic geometry and convex optimization, as each hyperbolic polynomial gives rise in a natural way to a convex cone. The generalized Lax conjecture in particular gave a characterization of the possible convex cones that could arise in this way. The conjecture was disproven in October of 2026 by an internal model from OpenAI, which provided an explicit counterexample to the stated characterization. In a sense, the conjecture being false is the more interesting possibility, as it suggests that there is more richness to the theory of hyperbolic polynomials than previously thought. This post is an attempt to understand this counterexample and its generalizations.&lt;/p&gt;</description>
      <content:encoded><![CDATA[$$
\newcommand{\R}{\mathbb{R}}
\newcommand{\C}{\mathbb{C}}
\newcommand{\S}{\mathbb{S}}
\newcommand{\T}{\intercal}
\newcommand{\adj}{\operatorname{adj}}
\newcommand{\im}{\operatorname{Im}}
\newcommand{\re}{\operatorname{Re}}
\newcommand{\op}{\mathrm{op}}
\newcommand{\proj}{\operatorname{proj}}
\renewcommand{\iota}{\operatorname{\mathfrak{I}}}
$$<h1 id="preliminaries-on-hyperbolic-polynomials">Preliminaries on Hyperbolic Polynomials</h1>
<p>The generalized Lax conjecture was a central topic of study in the field of hyperbolic polynomials. Hyperbolic polynomials in general serve as a bridge between the seemingly disparate fields of algebraic geometry and convex optimization, as each hyperbolic polynomial gives rise in a natural way to a convex cone. The generalized Lax conjecture in particular gave a characterization of the possible convex cones that could arise in this way. The conjecture was disproven in October of 2026 by an internal model from OpenAI, which provided an explicit counterexample to the stated characterization. In a sense, the conjecture being false is the more interesting possibility, as it suggests that there is more richness to the theory of hyperbolic polynomials than previously thought. This post is an attempt to understand this counterexample and its generalizations.</p>
<p>A homogeneous polynomial $p \in \R[x_1, \dots, x_n]$ is said to be hyperbolic with respect to $v \in \R^n$ if $p(v) \neq 0$ and for each $x \in \R^n$, the univariate polynomial $p(x-tv)$ has only real roots. The hyperbolicity cone associated to such a polynomial is then
</p>
$$ \Lambda_p = \{x \in \R^n : p(x-tv) \text{ has only nonnegative roots}\}. $$<p>The prototypical example of a hyperbolic polynomial is the determinant of an $n\times n$ symmetric matrix, which is hyperbolic with respect to the identity matrix due to the spectral theorem, and whose hyperbolicity cone is the set of positive semidefinite matrices. The generalized Lax conjecture essentially states that every hyperbolicity cone can be expressed as the intersection between a positive semidefinite cone and a linear subspace <a href="#ref-helton2007linear" class="citation" id="cite-helton2007linear">[1]</a>. Such a cone is said to be <strong>spectrahedral</strong> <a href="#ref-netzer2012smooth" class="citation" id="cite-netzer2012smooth">[3]</a>.</p>
<p>There are two steps to disproving the generalized Lax conjecture: firstly, one must find a polynomial $p$ and show that it is hyperbolic, and secondly one must show that $\Lambda_p$ is not spectrahedral. Both of these steps are nontrivial.
While showing that a polynomial is hyperbolic is not always easy, there have been many interesting constructions of hyperbolic polynomials in the literature <a href="#ref-vinnikov2012lmi" class="citation" id="cite-vinnikov2012lmi">[2]</a>.
None of these constructions had been shown to have nonspectrahedral hyperbolicity cones in the past (though in light of the disproof of the generalized Lax conjecture, one may wish to revisit some of these earlier constructions to see if they also are counterexamples to the conjecture).
So one suspects that showing that the hyperbolicity cone is not spectrahedral will be the hard part.
The proof given by OpenAI actually reduces the issue of showing that the cone to the well known fact that there exist positive operators which are not completely positive, a distinction that has been studied in the context of quantum mechanics for many years <a href="#ref-paulsen2002completely" class="citation" id="cite-paulsen2002completely">[4]</a>.</p>
<p>A positive operator is a linear map $\Phi :\S^n \rightarrow \S^m$ that sends positive semidefinite matrices to positive semidefinite matrices, and a completely positive such operator is one that takes the form of a `Kraus operator&rsquo;, which is to say it is of the form
</p>
$$
    \Phi(X) = A^{\intercal}(I\otimes X)A,
$$<p>
where $A : \R^m \rightarrow \R^n \times \R^k$ for some $k$.</p>
<p>It is noteworthy that there is an equivalence between such linear operators and <em>biquadratic forms</em>, meaning polynomials that are separately homogeneous of degree 2 in two separate blocks of variables, so that a positive operator corresponds precisely a nonnegative biquadratic form, and a completely positive one corresponds to a sum-of-squares biquadratic form. In this way, the distinction between positive operators and completely positive operators is related to the distinction between sums-of-squares polynomials and nonnegative polynomials.</p>
<p>The most general construction of a hyperbolicity cone that can be derived from the OpenAI paper shows that for any nonnegative <em>triquadratic form</em>, which can be associated to a quadratic family of positive maps, there is an associated hyperbolicity cone. The OpenAI paper further shows that if this triquadratic form is constructed from a <em>extremal nonnegative biquadratic form that is not sums-of-squares</em>, then the resulting hyperbolicity cone is not spectrahedral.</p>
<p>Another paper in the same OpenAI release shows that there is a hyperbolic polynomial whose hyperbolicity cone is not even a spectrahedral shadow. We will not discuss this example here, except to note that the relevant hyperbolicity cone that serves as a counterexample to that construction also arises from the same construction applied to a certain nonnegative triquadratic form. Also, the OpenAI paper that produced the counter example to the generalized Lax conjecture applied some additional ideas not mentioned here that reduce the number of variables and the size of the counter example.</p>
<p>This document is constructed so that the proofs of each major result immediately follows the statement, but it is hidden underneath a dropdown menu. This is meant to give the reader the ability to go over the argument at a high level, and then pass through the details afterwards.</p>
<h1 id="the-construction-generalized-schur-complements-and-positive-operators">The Construction: Generalized Schur Complements and Positive Operators</h1>
<p>The hyperbolicity cones that are ultimately shown to be nonspectrahedral arise from a natural and fairly general construction. To motivate the construction, we will recall the Schur complement characterization of the positive semidefinite cone:
</p>
$$
\S^n_+ = \overline{\{\begin{pmatrix}X & Y \\ Y^{\T} & Z\end{pmatrix} : X \succ 0,\ Z \succ Y^{\T}X^{-1}Y\}}.
$$<p>
We will note a few features of the map $(X,Y) \mapsto Y^{\T}X^{-1}Y$:</p>
<ol>
<li>It is linear in $X^{-1}$.</li>
<li>It is quadratic in $Y$.</li>
<li>For any $Y$ and any $X \succeq 0$, $Y^{\T}X^{-1}Y \succeq 0$.</li>
</ol>
<p>We will say that a linear map $\Phi : \S^n \rightarrow \S^m$ is <strong>positive</strong> if for every $X \succeq 0$, $\Phi(X) \succeq 0$. We will then define a <strong>quadratic family of positive maps (QFPM)</strong> to be a polynomial function $\Phi : \S^n \times \R^k \rightarrow \S^m$ so that $\Phi(X,Y)$ is linear in $X$, quadratic in $Y$, and for every $Y$, the linear map $\Phi(\cdot, Y)$ is positive.</p>
<p>We then come to the construction of the relevant hyperbolicity cone.</p>
<h4 id="theorem-1">Theorem 1</h4>
<blockquote>
<p>Let $\Phi(X,Y)$ be a QFPM. Then the following is a hyperbolicity cone,
</p>
$$
      H_{\Phi} = \overline{\{(X,Y,Z) \in \S^m \times \S^n \times \R^k : X \succ 0,\ Z \succ \Phi(X^{-1},Y)\}},
  $$<p>
for the polynomial
</p>
$$
     p_{\Phi}(X,Y,Z) = \det\big(\det(X)Z - \Phi(\adj(X), Y)\big),
  $$<p>
which is hyperbolic with respect to $(I,0,I)$.
Here $\adj(X) = \det(X) X^{-1}$ for invertible $X$ is the adjugate of $X$.</p>
</blockquote>
<details class="proof-toggle">
<summary style="color:rgb(157, 157, 230)">Click to expand the proof</summary>
<div class="proof-body">
<p>We will use repeatedly in this proof that if $X$ is nonsingular, then
</p>
$$ p_{\Phi}(X,Y,Z) = \det(X)^m\det\big(Z - \Phi(X^{-1}, Y)\big). $$<p>The proof breaks into two parts: firstly that $p_{\Phi}$ is indeed hyperbolic, and secondly that its hyperbolicity cone is $H_{\Phi}$. We will show this in two lemmas:</p>
<h4 id="lemma">Lemma</h4>
<blockquote>
<p>For any QFPM $\Phi$, the polynomial
</p>
$$
     p_{\Phi}(X,Y,Z) = \det\big(\det(X)Z - \Phi(\adj(X), Y)\big),
  $$<p>
is hyperbolic with respect to $(I,0,I)$.</p>
</blockquote>
<details class="proof-toggle">
<summary style="color:rgb(157, 157, 230)">Click to expand the proof</summary>
<div class="proof-body">
<p>We consider $p_{\Phi}(X+tI,Y,Z+tI)$ where $\im t > 0$ and $(X,Y,Z) \in \S^n \times \R^k \times \S^m$, and we want to show that this is nonzero. Assume for contradiction that $p_{\Phi}(X+tI,Y,Z+tI) = 0$ with $\im t > 0$.</p>
<p>Note that if $\det(X+tI) = 0$, then $t$ would be real because it would be the negative of an eigenvalue of a symmetric matrix. Thus, we may assume that $\det(X+tI) \neq 0$, and therefore that
</p>
$$
    p_{\Phi}(X+tI,Y,Z+tI) = \det(X+tI)^{m} \det\big(Z+tI - \Phi((X+tI)^{-1}, Y)\big) = 0.
$$<p>
We conclude that $Z + tI - \Phi((X+tI)^{-1}, Y)$ is singular. Let $w \in \C^m$ be in the kernel of this matrix, and let us consider
</p>
$$
\begin{aligned}
    2\im \big(w^\dagger (Z + tI - \Phi((X+tI)^{-1}, Y)) w \big)
    &= \frac{1}{i} \big(w^\dagger (Z + tI - \Phi((X+tI)^{-1}, Y)) w - w^\T \overline{(Z + tI - \Phi((X+tI)^{-1}, Y))}\, \overline{w} \big) \\
    &= \frac{1}{i} \big(w^\dagger (Z + tI - \Phi((X+tI)^{-1}, Y)) w - w^\dagger (Z + tI - \Phi((X+tI)^{-1}, Y))^{\dagger} w \big) \\
    &= w^\dagger\, \iota\big(Z + tI - \Phi((X+tI)^{-1}, Y)\big)\, w,
\end{aligned}
$$<p>
where
</p>
$$
    \iota(X) = \frac{1}{i}(X - X^{\dagger}).
$$<p>We see that
</p>
$$
    \iota\big(Z + tI - \Phi((X+tI)^{-1}, Y)\big) = \im t\, I - \Phi(\iota((X+tI)^{-1}), Y),
$$<p>
where we use the fact that $\Phi(X,Y)$ is linear in $X$ with real coefficients, and that $X$ is symmetric.</p>
<p>Our final claim before concluding that $p_{\Phi}$ is hyperbolic is that $\iota((X+tI)^{-1})$ is negative definite. This finishes the argument because the fact that $\Phi(\cdot, Y)$ is positive would also imply that
</p>
$$
    0 \preceq \im t\, I \preceq \im t\, I - \Phi(\iota((X+tI)^{-1}), Y),
$$<p>
and therefore that $\im\, w^\dagger \iota(Z + tI - \Phi((X+tI)^{-1}, Y)) w < 0$, contradicting the fact that $w$ is in the kernel of this matrix.</p>
<p>The fact that $\iota((X+tI)^{-1})$ is negative definite can be checked eigenvalue by eigenvalue of this matrix; every eigenvalue of $(X+tI)^{-1}$ is of the form $\frac{1}{\lambda + t}$ where $\lambda$ is an eigenvalue of $X$, and thus the eigenvalues of $\iota((X+tI)^{-1})$ are of the form
</p>
$$
    \im \frac{1}{(\lambda + \re t) + i\,\im t} =
    \im \frac{\overline{t}}{|t|^2} = - \frac{\im t}{|t|^2} < 0.
$$
<p>$\Box$</p>
</div>
</details>
<h4 id="lemma-1">Lemma</h4>
<blockquote>
<p>For any QFPM $\Phi$, the hyperbolicity cone of the polynomial
</p>
$$
     p_{\Phi}(X,Y,Z) = \det\big(\det(X)Z - \Phi(\adj(X), Y)\big),
  $$<p>
with respect to $(I,0,I)$ is
</p>
$$
      H_{\Phi} = \overline{\{(X,Y,Z) \in \S^m \times \S^n \times \R^k : X \succ 0,\ Z \succ \Phi(X^{-1},Y)\}},
  $$</blockquote>
<details class="proof-toggle">
<summary style="color:rgb(157, 157, 230)">Click to expand the proof</summary>
<div class="proof-body">
<p>We will use the characterization that the interior of the hyperbolicity cone of $p_{\Phi}$ is the connected component of the complement of the zero set of $p_{\Phi}$ containing $(I,0,I)$.</p>
<p>Note that because
</p>
$$
    \{(X,Y,Z) \in \S^m \times \S^n \times \R^k : X \succ 0,\ Z \succ \Phi(X^{-1},Y)\}
$$<p>
is open, it is the interior of $H_{\Phi}$. Thus, it is clear that $(I,0,I) \in H_{\Phi}^o$.
It remains to show that that $H_{\Phi}^o$ is connected, $p_{\Phi}$ does not vanish on $H_{\Phi}^o$ and that $p_{\Phi}$ does vanish everywhere on the boundary of $H_{\Phi}$.</p>
<p>To see that $H_{\Phi}^o$ is connected and indeed path connected, note that for any $(X,Y,Z) \in H_{\Phi}^o$ and $1 \ge s \ge 0$, $(X,sY,Z) \in H_{\Phi}^o$. Then we note that $(t X + (1-t)I, 0, t Z + (1-t)I) \in H_{\Phi}^o$ for any $1 \ge t \ge 0$, and so we are done.</p>
<p>It is also clear that $p_{\Phi}$ does not vanish on $H_{\Phi}^o$. If $(X,Y,Z) \in H_{\Phi}^o$, then
</p>
$$
    p_{\Phi}(X,Y,Z) = \det(X)^{m} \det(Z - \Phi(X^{-1}, Y)) > 0.
$$<p>It remains to show that $p_{\Phi}(X,Y,Z) = 0$ for any $(X,Y,Z) \in \partial H_{\Phi}$. For this, we note that if $(X,Y,Z) \in \partial H_{\Phi}$, then either $X$ is singular or $Z - \Phi(X^{-1}, Y)$ is singular. If $X$ is nonsingular, and $Z - \Phi(X^{-1}, Y)$ is singular, then it is clear from our above factorization that $p_{\Phi}(X,Y,Z) = 0$.
If $X$ is singular, then
</p>
$$
    p_{\Phi}(X,Y,Z) = \det\big(\det(X)Z - \Phi(\adj(X), Y)\big) = \det\big(\Phi(\adj(X),Y)\big).
$$<p>
If $X$ is of rank less than $n-1$, then $\adj(X) = 0$, and we can see that $p_{\Phi}(X,Y,Z) = 0$.
So we must finally consider the case that $X$ is singular, of rank $n-1$, and $X \in H_{\Phi}$. Under these assumptions, we must show that $\Phi(\adj(X),Y)$ is singular, so that $p_{\Phi}(X,Y,Z) = 0$.</p>
<p>Suppose that $\Phi(\adj(X),Y)$ is not singular, and consider a sequence $\{(X_i, Y_i, Z_i)\}_{i=0}^{\infty}$ where each $(X_i, Y_i, Z_i) \in H_{\Phi}^o$ and $\lim_{i\rightarrow \infty}(X_i, Y_i, Z_i) = (X,Y,Z)$.
We can consider the spectral decomposition of each $X_i$,
</p>
$$X_i = \sum_{k=1}^n \lambda_{ik} v_{ik}v_{ik}^{\T},$$<p>
where $0<\lambda_{i1}\le \lambda_{i2} \le \lambda_{i3}\dots\le \lambda_{in}$ are the eigenvalues of $X_i$ and each $v_{ik}$ is a unit eigenvector associated to the eigenvalue $\lambda_{ik}$.
Because the eigenvalues of a symmetric matrix are continuous functions of that matrix,
</p>
$$\lim_{i\to\infty} \lambda_{ik}=\lambda_k(X)$$<p>
for each $k$. In particular, because $X$ is of rank $n-1$, for all $i$ large enough, and all $k > 1$, $\lambda_{ik} > \epsilon$ and $\lambda_{i1} < \epsilon$ for some $\epsilon > 0$. We therefore conclude that the projection onto the eigenspace of the minimum eigenvalue of each $X_i$, $v_{i1}v_{i1}^{\T}$, is unique and therefore continuous as a function of $X$, and so we have that $\lim_{i\rightarrow \infty} v_{i1}v_{i1}^{\T} = vv^{\T}$, where $v$ is the unit vector in the kernel of $X$. This implies that
</p>
$$
\Phi(X_i^{-1}, Y) = \Phi\Big(X_i^{-1}-\tfrac{1}{\lambda_{i1}}v_{i1}v_{i1}^{\T},Y\Big) + \Phi\Big(\tfrac{1}{\lambda_{i1}}v_{i1}v_{i1}^{\T},Y\Big) \succeq \tfrac{1}{\lambda_{i1}}\Phi(v_{i1}v_{i1}^{\T},Y).
$$<p>We have by continuity of $\Phi$ that $\lim_{i\rightarrow \infty}\Phi(v_{i1}v_{i1}^{\T},Y)= \Phi(vv^{\T},Y) \neq 0$, and so, for $i$ large enough, $\|\Phi(v_{i1}v_{i1}^{\T},Y)\|_{\op} > \delta$ for some $\delta > 0$. This implies that for $i$ large enough,
</p>
$$\|Z_i\|_{\op} \ge \|\Phi(v_{i1}v_{i1}^{\T},Y)\|_{\op} \ge \frac{\delta}{\lambda_{i1}}.$$<p>Since
</p>
$$\lim_{i\to\infty} \lambda_{i1} = 0,$$<p>
we see that $\|Z_i\|_{\op}$ is unbounded, and so the limit of the $Z_i$ does not exist. This contradiction implies that $\Phi(\adj(X),Y)$ is singular, and we are done.</p>
<p>$\Box$</p>
</div>
</details>
<p>Combining these two lemmas gives the theorem.</p>
<p>$\Box$</p>
</div>
</details>
<h4 id="remark">Remark</h4>
<blockquote>
<p>This definition is remarkably close to that of a Siegel cone for the PSD cone, which amounts to the special case of this construction when $n = 1$. Siegel cones are of interest because every homogeneous cone can be constructed by applying the Siegel cone operation repeatedly a one dimensional cone <a href="#ref-vinberg1963homogeneous" class="citation" id="cite-vinberg1963homogeneous">[5]</a>.</p>
</blockquote>
<h1 id="spectrahedral-cones-and-completely-positive-maps">Spectrahedral Cones and Completely Positive Maps</h1>
<p>There are in fact many hyperbolicity cones that result from this construction. We now want to find one such cone which is not spectrahedral. Or, to put it another way, we want to find <em>necessary</em> conditions that $\Phi$ needs to satisfy in order for this cone to be spectrahedral, and then show that some $\Phi$ yields a hyperbolicity cone that does not satisfy those conditions.</p>
<p>One <strong>sufficient</strong> condition for the cone to be spectrahedral is for $\Phi$ to be completely positive in the following sense: there exist some positive operator $E : \S^n \rightarrow \S^{n'}$ and linear map $A : \R^k \rightarrow \R^{n' \times m}$ so that
</p>
$$
    \Phi(X,Y) = A(Y)^{\T} E(X) A(Y),
$$<p>
in which case
</p>
$$
    H_{\Phi} = \{(X,Y,Z) : \begin{pmatrix} E(X) & A(Y)^{\T} \\ A(Y) & Z\end{pmatrix} \succeq 0\}.
$$<p>The first thing that the OpenAI paper <a href="#ref-openai2026lax" class="citation" id="cite-openai2026lax">[8]</a> did to show that there is a nonspectrahedral hyperbolicity cone is to show a partial converse to this result.</p>
<p>Recall that a positive operator is unital if $\Phi(I) = I$.</p>
<h4 id="theorem-2">Theorem 2</h4>
<blockquote>
<p>Let $\Phi(X,Y)$ be a QFPM, and suppose that $H_{\Phi}$ is spectrahedral. Then there exist unital positive operators $E : \S^n \rightarrow \S^{n'}$ and $G: \S^m \rightarrow \S^{m'}$, together with a linear map $F : \R^k \rightarrow \R^{n' \times m'}$ so that
</p>
$$
    H_{\Phi} = \left\{(X,Y,Z) \in \S^n \times \R^k \times \S^m : \begin{pmatrix} E(X) & F(Y) \\ F(Y)^{\T} & G(Z) \end{pmatrix} \succeq 0\right\}.
$$</blockquote>
<details class="proof-toggle">
<summary style="color:rgb(157, 157, 230)">Click to expand the proof</summary>
<div class="proof-body">
<p>Since $H_{\Phi}$ is spectrahedral, there exists some natural number $d$ and a linear map $L : \S^n \times \R^k \times \S^m \rightarrow \S^d$ so that $(X,Y,Z) \in H_{\Phi} \Leftrightarrow L(X,Y,Z) \succeq 0$. We may assume by applying a compression transformation if necessary that there is no common kernel for the image of $L$, and by applying a congruence transformation, we may assume that $L(I,0,I)$ is positive definite.</p>
<p>We may write $L(X,Y,Z) = L_1(X) + L_2(Y) + L_3(Z)$, where $L_1 : \S^n \rightarrow \S^d$, $L_2 : \R^k \rightarrow \S^d$, and $L_3 : \S^m \rightarrow \S^d$.</p>
<p>Now, we may perform a block decomposition on $\S^d$: let $U = \bigcap_{X \in \S^n} \ker(L_1(X))$ (which may in principle be $0$), and we can perform a congruence transformation sending $U$ to the span of the last $\dim(U)$ coordinate vectors, so that in block matrix form,
</p>
$$
    L_1(X) = \begin{pmatrix} E(X) & 0 \\ 0 & 0\end{pmatrix},
$$<p>
for some positive map $E : \S^n \rightarrow \S^{d'}$. We can similarly block decompose $L_2$ and $L_3$ to obtain
</p>
$$
    L_2(Y) = \begin{pmatrix} L_2^{11}(Y) &  F(Y) \\  F(Y)^{\T} & L_2^{22}(Y)\end{pmatrix},
$$<p>
and
</p>
$$
    L_3(Z) = \begin{pmatrix} L_3^{11}(Z) &  L_3^{12}(Z) \\  L_3^{12}(Z)^{\T} & G(Z)\end{pmatrix}.
$$<p>We now claim that
</p>
$$(X,Y,Z) \in H_{\Phi} \iff \begin{pmatrix} E(X) & F(Y) \\ F(Y)^{\T} & G(Z) \end{pmatrix} \succeq 0.$$<p>One of the key observations about this generalized Schur complement construction is that there is a natural action of $(\R^{\times})^2$ on $H_{\Phi}$, i.e.</p>
<h4 id="lemma">Lemma</h4>
<blockquote>
<p>For $t,s \neq 0$, $(X,Y,Z) \in H_{\Phi}$ if and only if $(t^2X,stY, s^2 Z) \in H_{\Phi}$.</p>
</blockquote>
<details class="proof-toggle">
<summary style="color:rgb(157, 157, 230)">Click to expand the proof</summary>
<div class="proof-body">
<p>We will show this for $(X,Y,Z) \in H_{\Phi}^o$ and then claim that it therefore holds for $H_{\Phi}$ by continuity.</p>
<p>Note that
</p>
\[
    s^2Z \succ \Phi((t^2X)^{-1}, st Y) \Leftrightarrow 
    s^2Z \succ \frac{(s t)^2}{t^2}\Phi(X^{-1}, Y) \Leftrightarrow 
    Z \succ \Phi(X^{-1}, Y).
\]
<p>$\Box$</p>
</div>
</details>
<p>Using this observation, we can show the following:</p>
<h4 id="lemma-1">Lemma</h4>
<blockquote>
<p>Both $L_2^{22}(Y)$ and $L_2^{11}(Y)$ are identically $0$.</p>
</blockquote>
<details class="proof-toggle">
<summary style="color:rgb(157, 157, 230)">Click to expand the proof</summary>
<div class="proof-body">
<p>Fix some $Y \in \R^k$, and let $Z_0 = \Phi(I,Y)$. Note that for any $s$, $(I, sY, s^2Z_0), (I,-sY,s^2Z_0) \in H_{\Phi}$, and therefore we have that for any $s > 0$,
</p>
$$
    s^2 L_3^{22}(Z_0) + s L_2^{22}(Y) \succeq 0, \qquad s^2 L_3^{22}(Z_0) - s L_2^{22}(Y) \succeq 0.
$$<p>
Dividing by $s$ and taking the limit as $s$ approaches $0$ shows that $L_2^{22}(Y) = 0$ for any $Y \in \R^k$. A similar argument noting that $(s^2 I, s Y, Z_0), (s^2 I, -s Y, Z_0) \in H_{\Phi}$ for $s > 0$ shows that $L_2^{11}(Y) = 0$ for all $Y$.</p>
<p>$\Box$</p>
</div>
</details>
<p>From this, we can also conclude that $E(I)$ and $G(I)$ are positive definite; if they were not, $L(I,0,I)$ would have a nontrivial kernel.</p>
<p>We now show one direction of the equivalence.</p>
<h4 id="lemma-2">Lemma</h4>
<blockquote>
<p>If $(X,Y,Z) \in H_{\Phi}$, then $\begin{pmatrix} E(X) & F(Y) \\ F(Y)^{\T} & G(Z) \end{pmatrix} \succeq 0.$</p>
</blockquote>
<details class="proof-toggle">
<summary style="color:rgb(157, 157, 230)">Click to expand the proof</summary>
<div class="proof-body">
<p>Let $D_s = \begin{pmatrix} I & 0 \\ 0 & \frac{1}{s}I \end{pmatrix}$. For any $(X,Y,Z) \in H_{\Phi}$ and $s$, we have that $(X,sY,s^2Z) \in H_{\Phi}$, so
</p>
$$
    D_s\big(L_1(X) + s L_2(Y) + s^2 L_3(Z)\big)D_s = \begin{pmatrix} E(X) + s^2 L_3^{11}(Z) & F(Y) + s L_3^{12}(Z) \\ F(Y)^{\T} + s L_3^{12}(Z)^{\T} & G(Z)\end{pmatrix}  \succeq 0.
$$<p>
Taking the limit as $s \rightarrow 0$ shows that $\begin{pmatrix} E(X) & F(Y) \\ F(Y)^{\T} & G(Z) \end{pmatrix} \succeq 0$.</p>
<p>$\Box$</p>
</div>
</details>
<h4 id="lemma-3">Lemma</h4>
<blockquote>
<p>If $(X,Y,Z) \not \in H_{\Phi}$, then $\begin{pmatrix} E(X) & F(Y) \\ F(Y)^{\T} & G(Z) \end{pmatrix} \not \succeq 0.$</p>
</blockquote>
<details class="proof-toggle">
<summary style="color:rgb(157, 157, 230)">Click to expand the proof</summary>
<div class="proof-body">
<p>Since $(X,Y,Z) \not\in H_{\Phi}$, $L(X,Y,Z) \not\succeq 0$. We will assume for the sake of contradiction that $\begin{pmatrix} E(X) & F(Y) \\ F(Y)^{\T} & G(Z) \end{pmatrix} \succeq 0.$ Because $(X,Y,Z) \not\in H_{\Phi}$, we have $(X + \delta I, Y, Z +\delta I) \not\in H_{\Phi}$ for $\delta$ small enough, yet we would in fact have that
</p>
$$\begin{pmatrix} E(X) & F(Y) \\ F(Y)^{\T} & G(Z) \end{pmatrix} + \delta \begin{pmatrix} E(I) & 0 \\ 0 & G(I) \end{pmatrix} \succ 0.$$<p>
So, by replacing $(X,Y,Z)$ by $(X+\delta I, Y, Z +\delta I)$ if necessary, we may assume that $\begin{pmatrix} E(X) & F(Y) \\ F(Y)^{\T} & G(Z) \end{pmatrix}$ is positive definite, yet $(X,Y,Z) \not\in H_{\Phi}$.</p>
<p>Now, we simply note that for every $s \neq 0$, $(X, sY, s^2 Z) \not\in H_{\Phi}$, and therefore,
</p>
$$
    D_s\big(L_1(X) + s L_2(Y) + s^2 L_3(Z)\big)D_s = \begin{pmatrix} E(X) + s^2 L_3^{11}(Z) & F(Y) + s L_3^{12}(Z) \\ F(Y)^{\T} + s L_3^{12}(Z)^{\T} & G(Z)\end{pmatrix} \not\succeq 0,
$$<p>
and in particular, the limit as $s\rightarrow 0$ of this matrix cannot be positive definite, a contradiction.</p>
<p>Finally, we can achieve the fact that $E$ and $G$ are unital by applying a congruence transformation and using the fact that $E(I)$ and $G(I)$ are positive definite.</p>
<p>$\Box$</p>
</div>
</details>
<p>We conclude that
</p>
$$
    H_{\Phi} = \{(X,Y,Z) \in \S^n \times \R^k \times \S^m : \begin{pmatrix} E(X) & F(Y) \\ F(Y)^{\T} & G(Z)\end{pmatrix} \succeq 0\}. 
$$
<p>$\Box$</p>
</div>
</details>
<p>Thus, if $H_{\Phi}$ is spectrahedral, then it must look like the construction applied to a completely positive $\Phi$.
In particular, this suggests that if a $\Phi$ which is not completely positive is passed through this construction, then $H_{\Phi}$ will not be spectrahedral. While we will not show that this is the case, we will show that if $\Phi$ is a positive map that is not completely positive and also `extremal&rsquo; in a certain sense, then $H_{\Phi}$ is not spectrahedral.</p>
<p>It may not seem obvious that there exists such a positive QFPM that is not completely positive. However, there is a fairly general class of constructions that will supply us with these QFPMs. For this, it is helpful to pass to an equivalent characterization of QFPMs: nonnegative triquadratic forms.</p>
<h1 id="qfpms-and-nonnegative-triquadratic-forms">QFPMs and Nonnegative Triquadratic Forms</h1>
<h2 id="preliminaries-on-positive-operators-and-nonnegative-biquadratic-forms">Preliminaries on Positive Operators and Nonnegative Biquadratic Forms</h2>
<p>Associated to a linear map $\Phi : \S^n \rightarrow \S^m$ is the following biquadratic form $B : \R^m \times \R^n \rightarrow \R$:
</p>
$$
    B(u,v) = u^{\T}\Phi(vv^{\T})u.
$$<p>
By biquadratic, we mean that $B(\lambda u, \mu v) = \lambda^2 \mu^2 B(u,v)$ for any $\lambda, \mu \in \R$ and $u,v \in \R^n \times \R^m$.
In fact, by dimension counting, we can see that for any biquadratic form $B$, there is some $\Phi$ so that $B(u,v) = u^{\T}\Phi(vv^{\T})u$. Thus, a biquadratic form is equivalent to a linear map $\Phi : \S^n \rightarrow \S^m$.</p>
<p>One reason to study $B(u,v)$ is that it is pointwise nonnegative if and only if $\Phi$ is a positive operator.<!--TODO: cite-->
This follows directly from the fact that the extreme rays of $\S^m$ are spanned by rank $1$ matrices.</p>
<p>One sufficient condition for $B(u,v)$ to be nonnegative is that it is a sum-of-squares of lower degree polynomials (which will necessarily be bilinear forms).
A linear map associated to a sums-of-squares biquadratic form is said to be <strong>completely positive</strong>. Note that $\Phi$ being completely positive is equivalent to the statement that the map $\Phi$ is of the form $A^{\T}(I\otimes X)A$ for some $A$.</p>
<p>An equivalent characterization is that a linear map $\Phi$ is completely positive if and only if for every $k$, the linear map $I_k \otimes \Phi$ is positive. <!--TODO: cite--> Such maps have been studied intently in quantum mechanics because they correspond to quantum communication channels that are respect entanglement relations, while more general positive maps do not. In particular, it is known that there are positive maps that are not completely positive, or in otherwise nonnegative biquadratic forms that are not sums-of-squares whenever $n,m > 2$.</p>
<p>An important example is the <em>Choi–Lam</em> form <a href="#ref-choi1975positive" class="citation" id="cite-choi1975positive">[6]</a>, defined as
</p>
$$
    B(u,v) = \sum_{i=1}^3 u_i^2v_i^2 + \sum_{i=1}^3 u_i^2 v_{i+1}^2 - \sum_{i = 1}^3 \sum_{j=1}^3 u_iu_{i+1}v_i v_{i+1}.
$$<p>
This is an example of a nonnegative biquadratic form that is not a sum of squares. It is worth noting that this particular nonnegative biquadratic form is <em>extreme</em>, meaning that it cannot be written in a nontrivial way as a sum of other nonnegative biquadratic forms <a href="#ref-choi1977extremal" class="citation" id="cite-choi1977extremal">[7]</a>. This is the key fact that the ultimate nonspectrahedrality argument relies on. However, it is also worth noting that any extremal nonnegative biquadratic form will suffice for this argument. Indeed, it is possible to deduce the existence of an extremal nonnegative biquadratic form that is not a sums-of-squares just from the fact that there exists any biquadratic form which is nonnegative but not sums-of-squares and general convex geometry.</p>
<h2 id="triquadratic-forms-and-qfpms">Triquadratic forms and QFPMs</h2>
<p>In analogy with the above relationship between linear maps between symmetric matrices and biquadratic forms, we can associate to each QFPM a <em>triquadratic</em> polynomial
</p>
$$
    T(u,v,Y) = u^{\T}\Phi(vv^{\T}, Y)u.
$$<p>
Again, there is a bijective relationship between triquadratic forms and quadratic families of linear maps between spaces of symmetric matrices. This polynomial is pointwise nonnegative exactly when $\Phi$ is a QFPM, and $\Phi$ is completely positive if and only if $T$ is a sum of squares.</p>
<p>There is a way to turn a general <strong>biquadratic</strong> form into a triquadratic one that preserves nonnegativity and the property of being a sum of squares. Given a biquadratic form $B : \R^{\ell} \times \R^k \rightarrow \R$, let $f : \R^n \times \R^m \rightarrow \R^{\ell}$ be a bilinear map. Then $B(f(u,v), y)$ is a nonnegative triquadratic form.</p>
<p>For the purposes of the construction below, we will consider the case in which $n = m$ and the bilinear  $f(u,v) = u_0 v' - v_0 u'$, where $u = (u_0, u')$ and $v = (v_0, v')$.
In particular, there is a QFPM $\Phi$ that is associated to the triquadratic form obtained by composing the Choi-Lam form with this bilinear map, and this is precisely the $\Phi$ that the OpenAI paper shows leads to a nonspectrahedral hyperbolicity cone.</p>
<p>Note that more general bilinear maps $f$ and indeed triquadratic maps $T$ that do not arise in this way could be used to obtain more general classes of nonspectrahedral hyperbolicity cones, but we will focus on these cases for simplicity.</p>
<h1 id="a-nonspectrahedral-hyperbolicity-cone">A Nonspectrahedral Hyperbolicity Cone</h1>
<h4 id="theorem-3">Theorem 3</h4>
<blockquote>
<p>Let $B : \R^{m-1} \times \R^{k}\rightarrow \R$ be an extremal nonnegative biquadratic form that is not a sum of squares. Let $u = \begin{pmatrix}u_0\\u'\end{pmatrix}$ and $v = \begin{pmatrix}v_0\\v'\end{pmatrix}$, and let $\Phi$ be the QFPM associated to the nonnegative triquadratic form
</p>
$$
      T(u,v,Y) = B(u_0 v' - v_0 u', Y).
  $$<p>
Then the convex cone $H_{\Phi}$ is not spectrahedral.</p>
</blockquote>
<details class="proof-toggle">
<summary style="color:rgb(157, 157, 230)">Click to expand the proof</summary>
<div class="proof-body">
<p>We first record the consequence of extremality of $B$ that will be used at the end of the proof.</p>
<h4 id="lemma">Lemma</h4>
<blockquote>
<p>If $\ell(z, Y)$ is any bilinear form satisfying
</p>
$$
    \ell(z, Y)^2 \leq B(z, Y)
    \qquad
    \text{for all }(z, Y)\in \R^{m-1}\times \R^{k},
$$<p>
then $\ell=0$.</p>
</blockquote>
<details class="proof-toggle">
<summary style="color:rgb(157, 157, 230)">Click to expand the proof</summary>
<div class="proof-body">
$$
    B=\ell^2+(B-\ell^2)
$$<p><p><p>
expresses $B$ as a sum of two nonnegative biquadratic forms. Since $B$ spans an extreme ray of the cone of nonnegative biquadratic forms, there is some $\alpha\in[0,1]$ such that</p></p>
</p>
$$
    \ell^2=\alpha B.
$$<p>
If $\alpha>0$, then
</p>
$$
    B=\alpha^{-1}\ell^2,
$$<p>
so $B$ is a sum of squares, contrary to assumption. Hence $\alpha=0$, and therefore $\ell=0$. Thus
</p>
$$
\begin{equation}
    \label{eq:weak-extremality}
    \ell^2\leq B
    \quad\Longrightarrow\quad
    \ell=0.
\end{equation}
$$
<p>$\Box$</p>
</div>
</details>
<p>Now suppose, for contradiction, that $H_\Phi$ is spectrahedral. We will show the following:</p>
<h4 id="lemma-1">Lemma</h4>
<blockquote>
<p>If $Y \in \R^k$, $v \in \R^n$ and $u \in \R^m$ are unit vectors, then
</p>
$$
\begin{equation}
    \label{eq:B-norm}
    T(u, v, Y)
    =
    \left\|P(v)F(Y)R(u)\right\|_{\op}^2,
\end{equation}
$$<p>
where $P(v)$ denotes the orthogonal projection onto $\ker E(vv^{\intercal})$ and $R(u)$ denotes the orthogonal projection onto $\ker G(uu^{\intercal})$.</p>
</blockquote>
<details class="proof-toggle">
<summary style="color:rgb(157, 157, 230)">Click to expand the proof</summary>
<div class="proof-body">
<p>Theorem 2 implies that there exist positive unital maps $E: \S^n \rightarrow \S^{n'}$, $G : \S^m \rightarrow \S^{m'}$ and a linear map $F : \R^k \rightarrow \R^{n' \times m'}$ so that
</p>
$$
\begin{equation}
    \label{eq:block-equivalence}
    Z\succeq\Phi(X^{-1}, Y)
    \Longleftrightarrow
    \begin{pmatrix}
        E(X) & F(Y)\\
        F(Y)^{\T} & G(Z)
    \end{pmatrix} \succeq 0.
\end{equation}
$$<p>
Since $E$ is positive and unital, $E(X)\succ 0$ whenever $X\succ 0$.
Thus, taking the Schur complement of $E(X)$, we obtain the equivalent
formulation
</p>
$$
\begin{equation}
    \label{eq:schur-equivalence}
    Z\succeq\Phi(X^{-1}, Y)
    \quad\Longleftrightarrow\quad
    G(Z)\succeq F(Y)^{\T} E(X)^{-1}F(Y).
\end{equation}
$$<p>We now consider what happens when we allow $X$ and $Z$ to both degenerate to rank one matrices.
Fix unit vectors $v\in\R^n$ and $u\in\R^m$ and fix $Y\in\R^{k}$. For $t\geq1$, define
</p>
$$
    X_t
    :=
    vv^{\T}+t(I-vv^{\T}),
    \qquad
    Z_t
    :=
    uu^{\T}+t(I-uu^{\T}).
$$<p>We see that for $t \ge 1$,
</p>
$$
    X_t\succeq I,
    \qquad
    Z_t\succeq I.
$$<p>
Positivity and unitality of $E$ and $G$ therefore imply
</p>
$$
    E(X_t)\succeq I,
    \qquad
    G(Z_t)\succeq I,
$$<p>
so both matrices are positive definite.</p>
<p>Define $r_t$ for each $t$ to be the minimum value so that $r_t Z_t \succeq \Phi(X_t^{-1}, Y)$, and
note that there are two equivalent ways to compute the value of $r_t$ for a given $t$: one coming from the definition of $H_{\Phi}$ and one from its purported spectrahedral characterization.
</p>
$$
\begin{equation}
    \label{eq:threshold-equivalence}
    r_t Z_t\succeq\Phi(X_t^{-1}, Y)
    \quad\Longleftrightarrow\quad
    r_tG(Z_t)
    \succeq
    F(Y)^{\T} E(X_t)^{-1}F(Y).
\end{equation}
$$<p>We want to compute the value of $\lim_{t\rightarrow \infty} r_t$ using these two equivalent characterizations.
Considering the characterization of $r_t$ coming from the definition of $H_{\Phi}$, we can obtain the following:</p>
<h4 id="lemma">Lemma</h4>
<blockquote>
$$\lim_{t\rightarrow \infty} r_t = T(u,v,Y).$$</blockquote>
<details class="proof-toggle">
<summary style="color:rgb(157, 157, 230)">Click to expand the proof</summary>
<div class="proof-body">
$$
    M_t
    :=
    Z_t^{-1/2}\Phi(X_t^{-1}, Y)Z_t^{-1/2}.
$$<p><p><p>
Since $\Phi(\cdot, Y)$ is positive and $X_t^{-1}\succeq0$, we have
$M_t\succeq0$.
Congruence by $Z_t^{-1/2}$ gives</p></p>
</p>
$$
    r_tZ_t\succeq\Phi(X_t^{-1}, Y)
    \quad\Longleftrightarrow\quad
    r_tI\succeq M_t,
$$<p>
so $r_t = \lambda_{\max}(M_t)$.</p>
<p>The matrices $X_t$ and $Z_t$ were chosen
precisely so that their inverses collapse onto the rank-one directions
$v$ and $u$. Indeed,
</p>
$$
    X_t^{-1}
    =
    vv^{\T}+t^{-1}(I-vv^{\T}),
$$<p>
and
</p>
$$
    Z_t^{-1/2}
    =
    uu^{\T}+t^{-1/2}(I-uu^{\T}).
$$<p>
Consequently
</p>
$$
    X_t^{-1}\longrightarrow vv^{\T},
    \qquad
    Z_t^{-1/2}\longrightarrow uu^{\T}.
$$<p>
By linearity and continuity of $\Phi(\cdot, Y)$,
</p>
$$
\begin{align*}
    M_t
    &=
    Z_t^{-1/2}
    \Phi(X_t^{-1}, Y)
    Z_t^{-1/2}
    \\
    &\longrightarrow
    uu^{\T}\Phi(vv^{\T}, Y)uu^{\T} = T(u,v,Y) uu^{\T}.
\end{align*}
$$<p>Continuity of the largest eigenvalue yields
</p>
$$
\begin{equation}
    \label{eq:left-rank-one-limit}
    \lambda_{\max}(M_t)
    \longrightarrow
    T(u,v,Y).
\end{equation}
$$
<p>$\Box$</p>
</div>
</details>
<p>On the other hand, we will compute $\lim_{t\rightarrow \infty} r_t$ using the purported spectrahedral definition of $H_{\Phi}$.</p>
<h4 id="lemma-1">Lemma</h4>
<blockquote>
$$\lim_{t\rightarrow \infty} r_t = \left\|P(v)F(Y)R(u)\right\|_{\op}^2,$$<p> where $P(v)$ is the orthogonal projection onto the kernel of $E(vv^{\intercal})$ and $R(u)$ is the orthogonal projection onto the kernel of $G(uu^{\intercal})$.</p>
</blockquote>
<details class="proof-toggle">
<summary style="color:rgb(157, 157, 230)">Click to expand the proof</summary>
<div class="proof-body">
<p>Starting with the characterization
</p>
$$ r_tG(Z_t) \succeq F(Y)^{\T} E(X_t)^{-1}F(Y),$$<p>
we apply a congruence transformation by $G(Z_t)^{-1/2}$ to get
</p>
$$
\begin{align*}
    r_tG(Z_t)
    &\succeq
    F(Y)^{\T} E(X_t)^{-1}F(Y)
    \\
    &\Longleftrightarrow\quad
    r_tI
    \succeq
    G(Z_t)^{-1/2}
    F(Y)^{\T} E(X_t)^{-1}F(Y)
    G(Z_t)^{-1/2}
    \\
    &\Longleftrightarrow\quad
    r_tI\succeq C_t^{\T} C_t,
\end{align*}
$$<p>
where
</p>
$$
    C_t
    :=
    E(X_t)^{-1/2}F(Y)G(Z_t)^{-1/2}.
$$<p>The largest eigenvalue of $C_t^{\T} C_t$ is $\|C_t\|_{\op}^2$, so
</p>
$$
    r_t = \|C_t\|_{\op}^2.
$$<p>We will now take $t\rightarrow \infty$.
We will make use of the following fact: for any fixed positive semidefinite matrix $H$,
</p>
$$
\begin{equation}
    \label{eq:kernel-projection-limit}
    \lim_{t\rightarrow \infty}
    [I+(t-1)H]^{-1/2}
    = \proj_{\ker H},
\end{equation}
$$<p>
where $\proj_{\ker H}$ denotes the orthogonal projection onto the kernel of $H$.</p>
<p>Applying this to $E(X_t)$ and $G(Z_t)$ and using unitality yields
</p>
$$
    \lim_{t\rightarrow \infty} E(X_t)^{-1/2} = P(v) \qquad 
    \lim_{t\rightarrow \infty} G(Z_t)^{-1/2} = R(u),
$$<p>
where $P(v)$ and $R(u)$ are the orthogonal projections onto the kernels of $E(vv^{\intercal})$ and $G(uu^{\intercal})$ respectively.</p>
<p>$\Box$</p>
</div>
</details>
<p>By combining these characterizations of $\lim_{t\rightarrow \infty}r_t$, we arrive at the lemma.</p>
<p>$\Box$</p>
</div>
</details>
<p>With this lemma in mind, we now linearize \eqref{eq:B-norm} along the diagonal $u=v$. Fix
</p>
$$
    v=e_0=
    \begin{pmatrix}
        1\\0
    \end{pmatrix}\in\R\times\R^{m-1}.
$$<p>
For $z\in\R^{m-1}$, choose a smooth curve $u(t)$ on the unit sphere such that
</p>
$$
    u(0)=e_0,
    \qquad
    \dot u(0)=
    \begin{pmatrix}
        0\\z
    \end{pmatrix}.
$$<p>
Then
</p>
$$
    u_0(t)v'-v_0u'(t)
    =
    -tz+O(t^2).
$$<p>
Since $B$ is quadratic in its first argument,
</p>
$$
\begin{equation}
    \label{eq:B-expansion}
    B(u_0(t)v'-v_0u'(t), Y)
    =
    t^2 B(z, Y)+o(t^2).
\end{equation}
$$<p>On the other hand, setting $u=v=e_0$ in \eqref{eq:B-norm} gives
</p>
$$
    \left\|P(e_0)F(Y)R(e_0)\right\|_{\op}^2
    =
    B(0, Y)=0,
$$<p>
and hence
</p>
$$
    P(e_0)F(Y)R(e_0)=0.
$$<p>
Differentiating the matrix-valued expression $P(e_0)F(Y)R(u(t))$ at $t=0$, we obtain a matrix-valued map
</p>
$$
    \Psi(z, Y)
    :=
    P(e_0)F(Y)
    \left.
    \frac{d}{dt}R(u(t))
    \right|_{t=0}.
$$<p>
The derivative of $R(u)$ depends linearly on the tangent vector $z$, while $F(Y)$ is linear in $Y$. Consequently $\Psi(z, Y)$ is bilinear in $(z, Y)$. Moreover,
</p>
$$
    P(e_0)F(Y)R(u(t))
    =
    t\Psi(z, Y)+o(t).
$$<p>
Dividing \eqref{eq:B-norm} by $t^2$ and letting $t\to0$, using \eqref{eq:B-expansion}, yields
</p>
$$
\begin{equation}
    \label{eq:F-norm}
    B(z, Y)=\|\Psi(z, Y)\|_{\op}^2
    \qquad
    \text{for all }z\in\R^{m-1},\ Y\in\R^{k}.
\end{equation}
$$<p>Write the matrix entries of $\Psi$ as
</p>
$$
    \Psi(z, Y)=\big(\ell_{ij}(z, Y)\big)_{i,j}.
$$<p>
Each $\ell_{ij}$ is a bilinear form. Since the absolute value of every matrix entry is bounded by the operator norm,
</p>
$$
    \ell_{ij}(z, Y)^2
    \leq
    \|\Psi(z, Y)\|_{\op}^2
    =
    B(z, Y).
$$<p>
By \eqref{eq:weak-extremality}, it follows that $\ell_{ij}=0$ for every $i,j$. Hence $\Psi\equiv0$.
Equation \eqref{eq:F-norm} then implies $B\equiv0$,
contradicting the assumption that $B$ is a nonzero extremal nonnegative biquadratic form which is not a sum of squares.</p>
<p>Therefore $H_\Phi$ cannot be spectrahedral.</p>
<p>$\Box$</p>
</div>
</details>
<div class="references">
<h4 id="references">References</h4>
<ol class="references-list">
  <li id="ref-helton2007linear" value="1">J. W. Helton and V. Vinnikov. <a href="https://doi.org/10.1002/cpa.20155"><em>Linear matrix inequality representation of sets</em></a>. Communications on Pure and Applied Mathematics, vol. 60, no. 5, pp. 654–674 (2007)</li>
  <li id="ref-vinnikov2012lmi" value="2">V. Vinnikov. <a href="https://doi.org/10.1007/978-3-0348-0411-0_19"><em>LMI representations of convex semialgebraic sets and determinantal representations of algebraic hypersurfaces: past, present, and future</em></a>. Mathematical Methods in Systems, Optimization, and Control, Operator Theory: Advances and Applications, vol. 222, pp. 325–349 (2012)</li>
  <li id="ref-netzer2012smooth" value="3">T. Netzer and R. Sanyal. <a href="https://doi.org/10.1007/s10107-014-0744-6"><em>Smooth hyperbolicity cones are spectrahedral shadows</em></a>. Mathematical Programming, vol. 153, no. 1, pp. 213–221 (2015)</li>
  <li id="ref-paulsen2002completely" value="4">V. Paulsen. <em>Completely Bounded Maps and Operator Algebras</em>. Cambridge Studies in Advanced Mathematics 78, Cambridge University Press (2002)</li>
  <li id="ref-vinberg1963homogeneous" value="5">È. B. Vinberg. <em>The theory of convex homogeneous cones</em>. Transactions of the Moscow Mathematical Society, vol. 12, pp. 340–403 (1963)</li>
  <li id="ref-choi1975positive" value="6">M.-D. Choi. <a href="https://doi.org/10.1016/0024-3795%2875%2990058-0"><em>Positive semidefinite biquadratic forms</em></a>. Linear Algebra and its Applications, vol. 12, pp. 95–100 (1975)</li>
  <li id="ref-choi1977extremal" value="7">M.-D. Choi and T.-Y. Lam. <a href="https://doi.org/10.1007/BF01360024"><em>Extremal positive semidefinite forms</em></a>. Mathematische Annalen, vol. 231, pp. 1–18 (1977)</li>
  <li id="ref-openai2026lax" value="8">OpenAI. <em>A counterexample to the generalized Lax conjecture</em>. Preprint (2026). Add the canonical URL or DOI once the preprint is public</li>
</ol>
</div>
]]></content:encoded>
    </item>
    <item>
      <title>Surprising Positive Semidefiniteness</title>
      <link>https://www.quasicoherent.blog/posts/psd_completion/</link>
      <pubDate>Tue, 20 Sep 2022 00:05:13 -0400</pubDate>
      <guid>https://www.quasicoherent.blog/posts/psd_completion/</guid>
      <description>&lt;p&gt;They say that in Euclidean geometry, the solutions to all problems start by drawing a line somewhere in the picture. I think that this is characteristic of my favorite kind of proof, where you can&amp;rsquo;t just move forward in a linear fashion to get to an answer, but you actually have to add something new, that you might not have thought was related to the question.&lt;/p&gt;&#xA;&lt;p&gt;Here are two proofs that I only recently thought of as being related to each other (despite the fact that they are both named after Issai Schur). They are both about proving that a certain matrix is positive semidefinite (PSD), meaning that they have nonnegative eigenvalues, and we will prove them by finding that matrix inside a much larger PSD matrix.&lt;/p&gt;</description>
      <content:encoded><![CDATA[<p>They say that in Euclidean geometry, the solutions to all problems start by drawing a line somewhere in the picture. I think that this is characteristic of my favorite kind of proof, where you can&rsquo;t just move forward in a linear fashion to get to an answer, but you actually have to add something new, that you might not have thought was related to the question.</p>
<p>Here are two proofs that I only recently thought of as being related to each other (despite the fact that they are both named after Issai Schur). They are both about proving that a certain matrix is positive semidefinite (PSD), meaning that they have nonnegative eigenvalues, and we will prove them by finding that matrix inside a much larger PSD matrix.</p>
<p>It is actually not too hard to show that if you take a principal submatrix of a PSD matrix, that will also be PSD. But we can make use of a slightly stronger theorem to strengthen our results.</p>
<h4 id="theorem-cauchy-interlacing-theorem">Theorem (Cauchy Interlacing Theorem)</h4>
<blockquote>
<p>Let $X$ be an $n \times n$ symmetric matrix, and let $Y$ be a $k\times k$ principal submatrix of $X$ (that is, we obtain $Y$ by deleting $n-k$ rows of $X$ and their corresponding columns). Let $\lambda_i(\cdot)$ denote the $i^{th}$ smallest eigenvalue function. Then, for each $i = 1,\dots,k$,
\begin{equation*}
\lambda_{i}(X) \le \lambda_i(Y) \le \lambda_{i+n-k}(X).
\end{equation*}</p>
</blockquote>
<p>Disclaimer: There are almost certainly better proofs of both of these results, but I think highlighting their similarities is interesting enough for us. It will also allow us to prove what I think is a previously unknown fact about Schur complements.</p>
<h1 id="the-schur-product-theorem">The Schur Product Theorem</h1>
<p>We all know how to multiply square matrices. If $X$ and $Y$ are two matrices, then we just multiply them entriwise, so that each entry of $X \cdot Y$ is the product of the corresponding entries of $X$ and $Y$.</p>
<p>This style of multiplication, called the Hadamard product, is sort of weird, but what is weirder is that it actually interacts well with linear algebra, which usually use a different kind of multiplication.</p>
<h4 id="theorem-schur-product-theorem">Theorem (Schur Product Theorem)</h4>
<blockquote>
<p>Let $X$ and $Y$ be PSD matrices, then the Hadamard product, $X \cdot Y$ is also PSD.</p>
</blockquote>
<p>How can we go about proving this? The difficulty is that the Hadamard product seems to depend a lot on the basis of the underlying vector space, whereas the eigenvalues of a matrix are naturally basis independent. What we might try to do is to relate the basis dependent $X \cdot Y$ with a naturally basis independent thing.</p>
<p>To do this, we will need the notion of a <a href="https://en.wikipedia.org/wiki/Tensor_product">tensor product</a>. There are actually two related notions here: the tensor product of vector spaces, and of matrices. The tensor product of vector spaces $V$ and $W$ is the vector space spanned by simple tensors of the form $v \otimes w$, where $v \in V$ and $w \in W$, together with some natural relations that make the tensor product behave like the usual product.</p>
<p>The tensor product of matrices $X$ and $Y$, $X \otimes Y$, is given as follows: it is an $n^2 \times n^2$ matrix where, for a tensor product vector $v \otimes w$, $(X \otimes Y)(v\otimes w) = (Xv)\otimes(Yw)$.</p>
<p>If $v$ is a vector satisfying $Xv = \lambda v$, and $w$ is a vector satisfying $Yw = \mu w$, then $v \otimes w$ satisfies $(X\otimes Y)(v \otimes w) = \lambda\mu v\otimes w$. So, we can see that the eigenvalues of $X \otimes Y$ are all of the form $\lambda \mu$, where $\lambda$ and $\mu$ are eigenvalues of $X$ and $Y$ respectively.</p>
<p>In particular, it is easy to see that if $X$ and $Y$ are PSD, then $X \otimes Y$ is also PSD. This proof was relatively straightforward, since the object that we are looking at $X \otimes Y$ is basis invariant.</p>
<p>Now we relate $X \otimes Y$ to $X \cdot Y$ by looking at the entries of $X \otimes Y$ in the natural basis consisting of vectors of the form $e_i \otimes e_j$ for basis vectors $e_i$ and $e_j$. We can find the entries of this matrix using the natural inner product of $V \otimes W$, given by $\langle v_1 \otimes w_1, v_2 \otimes w_2\rangle = \langle v_1, w_1 \rangle \langle v_2, w_2\rangle$, in which case,
\begin{equation*}
(e_i \otimes e_j)^{\intercal} (X \otimes Y) (e_{\ell}\otimes e_k) = X_{i,\ell}Y_{j,k},
\end{equation*}
That is, the entries of $X \otimes Y$ in this basis are really just pairwise products of entries of entries of $X$ and $Y$, which sounds suspiciously like the definition of the Hadamard product. Indeed, let take the principal submatrix of $X \otimes Y$ that you get from just looking at basis elements of the form $e_i \otimes e_i$, in which case, you get that
\begin{equation*}
(e_i \otimes e_i)^{\intercal} (X \otimes Y) (e_j\otimes e_j) = X_{i,j}Y_{i,j} = (X \cdot Y)_{i,j}.
\end{equation*}
That is, if you delete all of the rows and columns of $X\otimes Y$ that are indexed by basis elements of the form $e_i \otimes e_j$ for $i \neq j$, then you get $X \cdot Y$. So, $X \cdot Y$ is actually a principal submatrix of $X \otimes Y$!</p>
<p>In particular, since $X \otimes Y$ is &lsquo;obviously PSD&rsquo;, that means that $X \cdot Y$ is PSD! We can even get bounds on the eigenvalues of $X \cdot Y$ in terms of the eigenvalues of $X$ and $Y$. Let $v(X,Y)$ denote the $n^2$ long vector that you get from taking all of the pairwise products of eigenvalues of $X$ and $Y$, and sorting them from smallest to largest. Then, by Cauchy interlacing,</p>
<p>\begin{equation*}
v(X,Y)_i \le \lambda_i(X \cdot Y) \le v(X,Y)_{i + n^2 - n}.
\end{equation*}</p>
<h1 id="schur-complements">Schur Complements</h1>
<p>I have heard it said that the difference between a pure mathematician and an applied mathematician is that an applied mathematician knows what a Schur complement is (and likes it). When I was an undergrad studying mathematics, I had vaguely heard of a Schur complement, and dismissed it as an ugly trick. Now, I suppose I have turned over a new leaf.</p>
<p>Let $X$ be an $n\times n $matrix, and let $T \subseteq [n]$ index a subset of the columns of $X$. Also,
assume that $X_{T,T}$ is invertible. Then
\begin{equation*}
X \setminus T = X_{T^c,T^c} - X_{T^c, T} X_{T,T}^{-1} X_{T,T^c}.
\end{equation*}
Here, $X_{S,T}$ is the submatrix of $X$ whose columns are indexed by $S$ and whose rows are indexed by $T$.</p>
<p>One nice thing about Schur complements is that they interact nicely with other Schur complements, and also the operations of taking submatrices.</p>
<p>Precisely, let $S$ and $U$ be disjoint from $T$, then
\begin{equation*}
(X \setminus S)\setminus T = X\setminus (S \cup T),
\end{equation*}
and
\begin{equation*}
(X_{S\cup T,U\cup T})\setminus T = (X\setminus T)_{S,U}.
\end{equation*}
(Yes, the notation is quite terrible, but again, this is applied math, so that is perhaps to be expected.)</p>
<p>We also have the determinant formula for Schur complements, which we will use without proof.</p>
<h4 id="theorem-determinant-formula">Theorem (Determinant Formula)</h4>
<blockquote>
<p>$\det(X) = \det(X_{T,T})\det(X \setminus T)$.</p>
</blockquote>
<p>Using the above properties, it can be seen that the entries of the Schur complement satisfy</p>
<p>\begin{equation*}
(X \setminus T)_{i,j} = \frac{1}{\det(X_{T,T})}\det(X_{T+i, T+j}).
\end{equation*}</p>
<p>We should expect that the following theorem hold then:</p>
<h4 id="theorem-schur-complement-theorem">Theorem (Schur Complement Theorem)</h4>
<blockquote>
<p>Let $X$ be a PSD matrix, then $X \setminus T$ is PSD for any $T \subseteq [n]$.</p>
</blockquote>
<p>This is often stated as a consequence of a more powerful fact, known as Haynesworth inertia additivity, named after <a href="https://en.wikipedia.org/wiki/Emilie_Virginia_Haynsworth">Emilie Virginia Haynsworth</a>, who I think should be a better known mathematician.
Let&rsquo;s give a proof more in line with our previous proof.</p>
<p>To do this, we will make use an object called the <a href="https://en.wikipedia.org/wiki/Exterior_algebra">exterior power</a> of a vector space, which is related to the tensor product, but slightly more complicated.</p>
<p>Given a vector space $V$, $\bigwedge^{(k)}V$ is a vector space spanned by elements of the form
\begin{equation*}
v_1 \wedge v_2 \dots \wedge v_k,
\end{equation*}
where $v_i \in V$ for each $i$.
This exterior power also satisfies two relations:</p>
<ul>
<li>
<p><strong>Multilinearity</strong>
\begin{equation*}
(v_1+\lambda w) \wedge v_2 \dots \wedge v_k =
v_1 \wedge v_2 \dots \wedge v_k + \lambda (w \wedge v_2 \dots \wedge v_k)
\end{equation*}</p>
</li>
<li>
<p><strong>Alternativity</strong> For any $i \in [k]$,
\begin{equation*}
v_1 \wedge v_2 \dots \wedge v_i \wedge v_{i+1} \dots \wedge v_k = - v_1 \wedge v_2 \dots \wedge v_{i+1} \wedge v_{i} \dots \wedge v_k
\end{equation*}</p>
</li>
</ul>
<p>These two properties make the exterior power related to the determinant of a matrix. To see this, we can define the wedge power of a matrix in a similar way to the tensor power:
\begin{equation*}
(\bigwedge^{(k)}X)(v_1 \wedge v_2 \dots \wedge v_k) =
Xv_1 \wedge Xv_2 \dots \wedge Xv_k.
\end{equation*}</p>
<p>In this case, it can be seent that $\bigwedge^{(n)} V$ is a 1-dimensional vector space spanned by $e_1\wedge e_2\dots \wedge e_n$, and that
\begin{equation*}
(\bigwedge^{(n)}X)(e_1 \wedge \dots \wedge e_n) = \det(X) e_1 \wedge \dots \wedge e_n.
\end{equation*}</p>
<p>Generally, let $S \subseteq [n]$ be of size $k$, and let
\begin{equation*}
e_S = e_{s_1} \wedge e_{s_2} \wedge \dots e_{s_k}.
\end{equation*}
Then
$\bigwedge^{(k)} V$ is an $\binom{n}{k}$ dimensional vector space with $\{e_S\}$ as a basis.</p>
<p>There is also a natural inner product on wedge powers given by
\begin{equation*}
\langle v_1 \wedge v_2 \dots \wedge v_k, w_1 \wedge w_2 \dots \wedge w_k \rangle =
\det \begin{bmatrix} \langle v_1, w_1 \rangle &amp; \langle v_1, w_2 \rangle &amp; \dots &amp; \langle v_1, w_k\rangle\\
\langle v_2, w_1 \rangle &amp; \langle v_2, w_2 \rangle &amp; \dots &amp; \langle v_2, w_k\rangle\\
&amp;&amp;\dots&amp;\\
\langle v_k, w_1 \rangle &amp; \langle v_k, w_2 \rangle &amp; \dots &amp; \langle v_k, w_k\rangle
\end{bmatrix}.
\end{equation*}
The fact that this is an inner product is implied by the <a href="https://en.wikipedia.org/wiki/Cauchy%E2%80%93Binet_formula">Cauchy-Binet Theorem</a>.</p>
<p>Now, we can perform the same trick as before: let&rsquo;s look at the eigenvalues of this matrix, and also look at its entries.</p>
<p>If $v_1, \dots, v_k$ are eigenvectors of $X$, with eigenvalues $\lambda_1, \dots, \lambda_k$, then
\begin{equation*}
(\bigwedge^{(k)}X)(v_1 \wedge v_2 \dots \wedge v_k) =
Xv_1 \wedge Xv_2 \dots \wedge Xv_k = \prod_{i=1}^k \lambda_i (v_1 \wedge \dots \wedge v_k).
\end{equation*}
So, it is clear that the eigenvalues of $\bigwedge^{(k)}X$ are precisely $\prod_{i=1}^k \lambda_i$, where $\lambda_i$ are distinct eigenvalues of $X$. In particular, if $X$ is PSD, then so is $\bigwedge^{(k)}X$.</p>
<p>On the other hand, we can look at the entries of $\bigwedge^{(k)}X$ as follows: let $S = \{i_1, \dots, i_k\}$, and let $U = \{j_1, \dots, j_k\}$, then
\begin{equation*}
(e_{i_1} \wedge e_{i_2} \dots \wedge e_{i_k})^{\intercal} (X \otimes Y) (e_{j_1} \wedge e_{j_2} \dots \wedge e_{j_k}) = \det(X_{S,U}).
\end{equation*}</p>
<p>Now, we can connect this back to Schur complements: let $k = |T| + 1$, consider the submatrix of $\wedge^{(k)}X$ indexed by basis vectors of the form $e_{t_1} \wedge e_{t_2} \wedge e_{t_{k-1}} \wedge e_{j}$, for some $j \not \in T$, then then from our earlier properties,</p>
<p>\begin{equation*}
(X \setminus T)_{j,k} = \frac{1}{\det(X_{T,T})}\det(X_{T+j,T+k}) = \frac{1}{\det(X_{T,T})} (\bigwedge^{(k)} X)_{T+i, T+k}.
\end{equation*}
Thus, we have discovered $(X \setminus T)$ as a principal submatrix of $\frac{1}{\det(X_{T,T})}\bigwedge^{(k)} X$, and so is PSD.</p>
<p>In fact, we have a slightly stronger fact, from the Cauchy interlacing theorem.
To see this, note that for $1\le i \le n-|T|$, the $i^{th}$ smallest eigenvalue of $\bigwedge^{(|T|+1)}X$ is
\begin{equation*}
\lambda_{i}(\bigwedge^{(|T|+1)}X) = (\prod_{j=1}^{|T|}\lambda_j(X)) \lambda_{i+|T|+1}(X),
\end{equation*}
\begin{equation*}
\frac{\prod_{j=1}^{|T|}\lambda_j(X)}{\det(X_{T,T})}\lambda_{i+|T|+1}(X) \le \lambda_i(X\setminus T).
\end{equation*}
Note that Cauchy interlacing in this case also implies that $\frac{\prod_{j=1}^{|T|}\lambda_j(X)}{\det(X_{T,T})} \le 1$, so this lower bound is somewhat smaller than $\lambda_{i+|T|+1}(X)$.</p>
<p>The upper bound from the Cauchy interlacing inequality is not very useful here, since it is obvious from the definition that
\begin{equation*}
\lambda_i(X\setminus T) \le \lambda_{i+|T|+1}(X),
\end{equation*}
which is already smaller than the upper bound given by the Cauchy interlacing inequality.</p>
]]></content:encoded>
    </item>
    <item>
      <title>A Short Proof of Brickman&#39;s Theorem</title>
      <link>https://www.quasicoherent.blog/posts/brickman_theorem/</link>
      <pubDate>Thu, 09 Sep 2021 00:05:13 -0400</pubDate>
      <guid>https://www.quasicoherent.blog/posts/brickman_theorem/</guid>
      <description>&lt;p&gt;This was based on a conversation with Alex Wang and Mark Gillespie. &lt;a href=&#34;https://www.cs.cmu.edu/~alw1/blog/dines.html&#34;&gt;Alex has a blog post&lt;/a&gt; about Dine&amp;rsquo;s theorem, which is very related to this theorem.&lt;/p&gt;&#xA;&lt;p&gt;There are a lot of places in algebraic geometry where convexity appears in surprising places.&#xA;One well known result coming from quadratic programming is a theorem of Brickman.&lt;/p&gt;&#xA;&lt;h4 id=&#34;theorem-brickman&#34;&gt;Theorem (Brickman)&lt;/h4&gt;&#xA;&lt;blockquote&gt;&#xA;&lt;p&gt;Let $n \ge 2$.&#xA;Let $Q_1, Q_2$ be homogeneous quadratic maps on $\mathbb{R}^{n+1}$, and let $Q : \mathbb{R}^{n+1} \rightarrow \mathbb{R}^2$ be given by $Q(x) = (Q_1(x),Q_2(x))$.&#xA;Let $S^{n} = \\{x \in \mathbb{R}^{n+1} : \|x\|_2 = 1\\}$.&#xA;Then $Q(S^{n}) \subseteq \mathbb{R}^2$ is convex.&lt;/p&gt;&#xA;&lt;/blockquote&gt;</description>
      <content:encoded><![CDATA[<p>This was based on a conversation with Alex Wang and Mark Gillespie. <a href="https://www.cs.cmu.edu/~alw1/blog/dines.html">Alex has a blog post</a> about Dine&rsquo;s theorem, which is very related to this theorem.</p>
<p>There are a lot of places in algebraic geometry where convexity appears in surprising places.
One well known result coming from quadratic programming is a theorem of Brickman.</p>
<h4 id="theorem-brickman">Theorem (Brickman)</h4>
<blockquote>
<p>Let $n \ge 2$.
Let $Q_1, Q_2$ be homogeneous quadratic maps on $\mathbb{R}^{n+1}$, and let $Q : \mathbb{R}^{n+1} \rightarrow \mathbb{R}^2$ be given by $Q(x) = (Q_1(x),Q_2(x))$.
Let $S^{n} = \\{x \in \mathbb{R}^{n+1} : \|x\|_2 = 1\\}$.
Then $Q(S^{n}) \subseteq \mathbb{R}^2$ is convex.</p>
</blockquote>
<p>There are a few well known proofs of this theorem (see for example, Barvinok&rsquo;s <a href="https://bookstore.ams.org/gsm-54/">A Course on Convexity</a>), which tend to involve passing to some associated convex set of positive semidefinite matrices, and then showing that the image of that convex set under a linear projection agrees with the image of the sphere under these quadratic maps.</p>
<p>Our goal is to give a short, geometric proof of this fact, which uses only some elementary observations.</p>
<h4 id="proof">Proof</h4>
<p>We base our proof on two facts:</p>
<ul>
<li>
<p>The image of the unit circle $S^1 = \\{(x,y) : x^2+y^2 = 1\\}$ under a homogeneous quadratic map $Q : \mathbb{R}^2 \rightarrow \mathbb{R}^n$ is also the image of the unit circle under an affine map.
We will call the image of $S^1$ under an affine map an <strong>ellipse</strong>.
We prove this below the fold.
<details class="proof-toggle">
<summary style="color:rgb(157, 157, 230)">Click to expand</summary>
<div class="proof-body">

<pre><code>    This surprising geometric fact is the main reason the result holds.
    To see why it is the case, notice that each point on the circle can be written as $(\cos(t), \sin(t))$, and that any homogeneous quadratic function $a\cos(t)^2 + b \cos(t) \sin(t) + c\sin(t)^2$ can be expressed as a linear combination of the functions $\cos(2t)$, $\sin(2t)$ and 1.

    This implies that if we apply a quadratic function to the unit circle, the image can be parameterized as $v_1 \cos(2t) + v_2 \sin(2t) + v_3$ for some $v_1, v_2, v_3 \in \mathbb{R}^n$, and it is therefore the image of $S^1$ under the affine map $(x,y) \mapsto v_1 x + v_2 y + v_3$.
</code></pre>


$\Box$

</div>
</details>
</p>
</li>
<li>
<p>If $X$ is a simply connected topological space (for example $S^{n}$ for $n \ge 2$), $\gamma$ is a closed curve in $X$, and $f : X \rightarrow Y$ is a continuous map, then the image of  $\gamma$ is contractible in the image of $f$.</p>
</li>
</ul>
<p>With these facts, we can show the result.</p>
<p>Let $a = Q(x)$, and $b = Q(y)$, where $x, y\in S^n$.
We want to show that the line segment from $a$ to $b$ is in the image of $S^n$ under $Q$.</p>
<p>Consider the great circle from $x$ to $y$ in $S^n$, which is the image of $S^1$ under the linear map $(u,v) \mapsto u \hat{x} + v \hat{y}$, where $\\{\hat{x}, \hat{y}\\}$ is an orthonormal basis for the linear span of $x$ and $y$.
Now, if we compose $Q$ with this linear map from $S^1$, we obtain a quadratic map from $S^1$ to $\mathbb{R}^2$, whose image we have seen is an ellipse, which contains both $a$ and $b$.</p>
<p>The ellipse divides $\mathbb{R}^2$ into two pieces: the compact inside part, and the unbounded outside part. The compact inside part is convex, and therefore, it contains the line segment from $a$ to $b$.
Because the $S^{n}$ is simply connected, it must be the case that every point in the inside part of the ellipse must be contained in the image.
(To see this, say that some subset of the inside part of the ellipse is missing from the image, then the ellipse would not be contractible in the image set.)</p>
<p>Therefore, the line segment from $a$ to $b$ is contained in the image, and we are done.</p>
<p><figure><img src="/img/brickman/great_circle.jpg"
			alt="A Great Circle on the Sphere."><figcaption>
			<p>A Great Circle on the Sphere.</p>
		</figcaption>
</figure>

<figure><img src="/img/brickman/image3.jpeg"
			alt="The image of a sphere with a special great circle highlighted."><figcaption>
			<p>The image of a sphere with a special great circle highlighted.</p>
		</figcaption>
</figure>
</p>
]]></content:encoded>
    </item>
    <item>
      <title>Inequalities between Binomial Coefficients in a Needlessly Fancy Way</title>
      <link>https://www.quasicoherent.blog/posts/representations_sl2/</link>
      <pubDate>Wed, 20 Jan 2021 00:44:12 -0500</pubDate>
      <guid>https://www.quasicoherent.blog/posts/representations_sl2/</guid>
      <description>&lt;p&gt;Most high school students already know far more about the binomial coefficients than they need to, and one of the things that they might know is the fact that if $k &lt; \frac{n}{2}$, then&#xA;&lt;/p&gt;&#xA;$$&#xA;    \binom{n}{k} \le \binom{n}{k+1}.&#xA;    $$&lt;p&gt;We&amp;rsquo;re going to talk about a needlessly complicated proof of this fact that uses some pretty cool math.&lt;/p&gt;&#xA;&lt;h1 id=&#34;categorification-and-inequalities&#34;&gt;Categorification and Inequalities&lt;/h1&gt;&#xA;&lt;p&gt;When we see an inequality like&#xA;&lt;/p&gt;&#xA;$$&#xA;    \binom{n}{k} \le \binom{n}{k+1},&#xA;    $$&lt;p&gt;&#xA;we might want to write down an equation for the binomial coefficients like $\binom{n}{k} = \frac{n!}{(n-k)!k!}$, and then do some algebra. This might be fast, but it&amp;rsquo;s hardly satisfying.&lt;/p&gt;</description>
      <content:encoded><![CDATA[<p>Most high school students already know far more about the binomial coefficients than they need to, and one of the things that they might know is the fact that if $k < \frac{n}{2}$, then
</p>
$$
    \binom{n}{k} \le \binom{n}{k+1}.
    $$<p>We&rsquo;re going to talk about a needlessly complicated proof of this fact that uses some pretty cool math.</p>
<h1 id="categorification-and-inequalities">Categorification and Inequalities</h1>
<p>When we see an inequality like
</p>
$$
    \binom{n}{k} \le \binom{n}{k+1},
    $$<p>
we might want to write down an equation for the binomial coefficients like $\binom{n}{k} = \frac{n!}{(n-k)!k!}$, and then do some algebra. This might be fast, but it&rsquo;s hardly satisfying.</p>
<p>Maybe more satisfying would be to first think about what $\binom{n}{k}$ means; if $S_k$ is the number of sets of size $k$ in a set of $n$ elements, then $\binom{n}{k}$ is the size of $S_k$.
So, another way of saying the inequality is that $S_k$ has fewer elements than $S_{k+1}$.</p>
<p>One way to show that one set has more elements than another is to find an injective map $f$ from one set to the other, and such maps are usually a lot more satisfying than a computational proof.</p>
<p>It turns out that this is kind of tricky in this case. For instance, we might naturally want to choose, for each $k$ element set $S$, an element $t(S) \in [n] \setminus S$, and then set $f(S) = S \cup \{t(S)\}$. We might hope that if $t(S)$ is well chosen, then $S$ can be recovered from  $f(S)$ uniquely. There are in fact ways to do this, (see <a href="https://math.stackexchange.com/a/2894340/428575">this</a>), but it&rsquo;s actually a little tricky to find such an injection. We&rsquo;ll see in the sequal that this is because there is too much symmetry here; to find such a $t(S)$, we need some clever way of choosing an element of $[n] \setminus S$, and all elements of $[n] \setminus S$ basically look the same to me!</p>
<p>As a first towards getting this bijective proof we want, we can apply an an idea that shows up a lot in combinatorics (that we&rsquo;ll also do in a fancier way in the sequel): instead of assigning to each $S \subseteq [n]$ with $|S| = k$, we can form a <strong>relation</strong>, which we might think of as a bipartite graph. This will let us get around the symmetry issue.</p>
<p>The relation here is just the subset relation. For each element $S$ of $S_k$, there are $(n-k)$ elements of $[n] \setminus S$ that we can add to $S_k$ to get an element of $S_{k+1}$. For each element of $S_{k+!}$, there are $k+1$ elements in $S_{k+1}$ that we can remove to get a set of size $k$. So, by a double counting argument, we know that $\frac{|S_{k}|}{|S_{k+1}|} = \frac{k+1}{n-k} < 1$. This doesn&rsquo;t use the equation for binomial coefficients, but requires an unacceptably large amount of computation! Let&rsquo;s see if we can do better.</p>
<h1 id="symmetry-and-vector-spaces">Symmetry and vector spaces</h1>
<p>Usually, we think of symmetry as being our friend, but in the above paragraph, we saw that symmetry was actually kind of an obstruction to getting a nice map. A slight modification of the above ideas though will let us get symmetry on our side again.</p>
<p>Instead of thinking of $\binom{n}{k}$ as the number of elements in some set, let&rsquo;s think of it as the <strong>dimension</strong> of the vector space spanned by elements of $S_k$, say $V_k$. This is the set of formal linear combinations of elements of $S_k$. Once again, if we have an injective map from $V_k$ to $V_{k+1}$, then linear algebra tells us that $\binom{n}{k} \le \binom{n}{k+1}$.</p>
<p>The value of doing this is that we can now turn the relation that we had above into an honest map of vector spaces; for each $S \in S_k$, let
</p>
$$
f_k(S) = \sum_{S \subseteq T \in S_{k+1}}T.
$$<p>
That $f_k(S)$ is the sum of all sets of sets of size $k+1$ that contain $S$, and is sort of the &lsquo;vectorization&rsquo; of the above relation.  If we can show that this map is injective, then we are done, and we will have satisfied our desire for a synthetic proof of this fact.</p>
<p>In fact, this is true; the map is injective and it can be shown for example by looking at the matrix associated with this map, computing its Gram matrix and showing that is positive definite (see for example <a href="https://www.jstor.org/stable/2958520?seq=1">this</a>). Once again though, this is an unacceptable amount of computation that you can&rsquo;t just handwave away. We must go deeper!</p>
<h1 id="playing-around-with-commutators">Playing around with commutators</h1>
<p>There are a few steps we need to give a bijective proof `without computations&rsquo;.</p>
<p>Firstly, the map we defined above has a transpose, specifically, this is a map $g_{k+1} : V_{k+1} \rightarrow V_k$ given by
</p>
$$
    g_{k+1}(T) = \sum_{T \supseteq S \in S_k} S
    $$<p>Secondly, we need to assemble all of these $V_k$ into a single (graded) vector space, namely
</p>
$$
V = V_1 \oplus V_2 \dots \oplus V_n
$$<p>We can define $f$ and $g$ as follows: if $v \in V_i$, then $f(v) = f_i(v)$, and similarly, $g(v) = g_i(v)$.</p>
<p>Now, we can do something a little strange; we can try computing the comutator between this map $g$ and $f$ for some $S \in V_k$:
</p>
$$
    [g, f] = g(f(S)) - f(g(S)) = (n-2k)S
    $$<p>
The way you get this is by noticing that if $S' \in S_k$, and $S' \neq S$, then the coefficient in front of $S'$ in $g_{k+1}(f_k(S))$ is the number of elements in $S_{k+1}$ that contain both $S$ and $S'$, but there&rsquo;s only one of these, namely $S \cup S'$. A similar idea for $f_{k-1}(g_k(S))$ gives us that the coefficient in front of $S'$ will also be 1, so these cancel, and the only surviving term is proportional to $X$, and again, it&rsquo;s not that hard to figure out what this coefficient should be.</p>
<p>This is really nice! It says that the commutator of $f$ and $g$ acts diagonally on the $V$. Let&rsquo;s call this commutator $h = gf - fg$.</p>
<p>We also have other commutation relations, which can also be verified `by hand waving&rsquo; as above: $[h, f] = -2f$ and $[h,g] = 2g$.</p>
<p>These relations are important because they define a <strong>representation</strong>, which we&rsquo;ll look at next:</p>
<h1 id="lie-algebras-and">Lie algebras and $sl_2$</h1>
<p>There are in fact $2 \times 2$ matrices which satisfy these same commutation relations, namely
</p>
$$
    h' = \left(\begin{matrix} 1 \& 0 \\\\ 0 \& -1\end{matrix}\right)
$$<p>
</p>
$$
    f' = \left(\begin{matrix} 0 \& 1 \\\\ 0 \& 0\end{matrix}\right)
    $$<p>
</p>
$$
    g' = \left(\begin{matrix} 0 \& 0 \\\\ 1 \& 0\end{matrix}\right)
    $$<p>
These form what is called a <strong>Lie algebra</strong>, which is a vector space with an antisymmetric, bilinear map called the Lie bracket, denoted $[-,-]$. There are some additional properties that you need, but let&rsquo;s not say what they are here.
The name of this Lie algebra is $sl_2$.</p>
<p>Between two Lie algebras, both of which have an associated Lie bracket, a morphism is just a linear map $\phi : L_1 \rightarrow L_2$ so that $[\phi(x), \phi(y)] = \phi([x,y])$. A <strong>representation</strong> of a Lie algebra $L$ is a map from $L$% to $gl_n$, the Lie algebra of matrices with the normal matrix commutator.</p>
<p>The interesting thing about representations about $sl_2$ is that if you look at $\phi(h')$, we can say some important things about the eigenvectors of $h$, namely that if $v$ is an eigenvector of $\phi(h')$ with eigenvalue $\lambda$, then $\phi(f')v$ and $\phi(g')v$ are also eigenvectors of $h$ with eigenvalues $\lambda+2$ and $\lambda-2$ respectively. In particular, because $\phi(h')$ has only finitely many eigenvalues, $\phi(h')$ and $\phi(g')$ are nilpotent, and so the eigenvalues of $h$ must all be <strong>integers</strong>.</p>
<p>If $gl_n$ acts on an $n$ dimensional vector space $V$, and $\phi$ is a representation of $sl_2$ on $gl_n$, then we have a grading of $V$ according to the eigenvalue of $\phi(h)$, i.e.
</p>
$$
V = \dots V_{-1} \oplus V_{0} \oplus V_{1} \oplus \dots.
$$<p>
where $V_i$ is an eigenspace of $h$. Notice that this is exactly the position we found ourselves in the previous section.</p>
<p>The important thing is that people have studied representations of $sl_2$ for a really long time. Specifically, they know that the <strong>irreducible representations</strong> have the following property: the dimension of $V_i$ when $i > 0$ is always at most the dimension of $V_{i+1}$! Because this is true for all irreducible representations of $sl_2$, and all representations are direct sums of irreducible ones, <strong>any representation of $sl_2$</strong> will have this property! In addition, this theory will give us the fact that $f$ and $g$ above are injective, just because they satisfy these commutation relations, so we get our combinatorial proof, and all it took us was a few years worth of representation theory.</p>
<h1 id="applications-to-betti-numbers-and-hodge-theory">Applications to Betti Numbers and Hodge Theory</h1>
<p>The representation theory of $sl_2$ actually appears in a lot of different fields. In particular, it shows up in the cohomology theory of complex projective varieties. I don&rsquo;t understand this very well, but here&rsquo;s the gist of it from my point of view.</p>
<p>Every complex projective variety has something called a <strong>Kahler structure</strong>, which, if the variety is smooth, we can think of as a differential 2-form that lives on the manifold. This differential 2-form $\omega$ (called the symplectic form)  is closed, and therefore defines an element of the De Rham cohomology for the manifold.</p>
<p>The cohomology rings of a manifold have an attached ring structure, whose multiplciation is given by the cup product, and multipliciation by $\omega$ in this ring structure gives us a linear map from the cohomology ring to itself. Let&rsquo;s call this map $f$, suggestively. There is also an inner product on the cohomology ring given by the Riemannian metric structure of the manifold, and so $f$ has an adjoint, which we might call $g$, again suggestively. As you might expect, $h = [f, g]$ turns out to act diagonally on the cohomology ring, its eigenspaces are the graded pieces of the cohomology ring, and the triple $f, g, h$ defines a representation of $sl_2$ on the cohomology ring of the variety.</p>
<p>The dimension of the degree $k$ graded piece of the cohomology ring is called the $k^{th}$ betti number $\beta_k$.  So, what we get in total is that for $k > \frac{n}{2}$, $\beta_k \le \beta_{k+1}$.</p>
<p>This theory was used by Richard Stanley in the case of toric varieties (whose paper on unimodal sequences forms the substance of this post) to show that the number of facets of simplicial polytopes satisfy the <strong>least upper bound conjecture</strong>, which I don&rsquo;t really get, but am assured is very cool.</p>
]]></content:encoded>
    </item>
  </channel>
</rss>
