A weaker bound on the stabilizer-inradius conjecture
by Rafael Wagner · accepted 18 Nov 2026 · partially AI-assisted
Result
Using Claude Fable 5.2 and a few prompts, it was possible to prove the following weaker statement:
Proposition: Let ρ∈D(n timesC2⊗⋯⊗C2) be an
arbitrary n-qubit state, with n≥2. Denote, furthermore, dn=2n and the function c(⋅):N→Q given by
cn=7dn+22dn+7>72.
Then,
Tr[ρ2]≤dn−cn1⟹ρ∈STABn.
In particular, for all n≥2 it holds that
Tr[ρ2]≤dn−721⟹ρ∈STABn.
Note that this is only a partial result because the conjecture claims that the implication holds with cn=21 for arbitrary n.
Background
The key technical idea is to use the fact that one can investigate the easier properties associated with the convex cones generated by the set of states and the stabilizer polytopes, as opposed to the sets themselves. In this case, one replaces rotations and the more complicated geometry of the inradius by properties of cone-preservation and related ideas.
Isotropic Pauli subgroups
Definition (isotropic Pauli subgroup). Let Pn be the set of 4n Pauli strings, i.e. tensor products of Pauli matricesI,X,Y,Z on n qubits without phases. A subset T⊆Pn is an isotropic Pauli subgroup if
(1) it contains the identity, i.e., if I∈T;
(2) it is Abelian, i.e. any two elements of T commute;
(3) it is closed under multiplication up to a sign, i.e., ∀R,R′∈T, one has that either RR′∈T or −RR′∈T.
Such a subgroup has ∣T∣=2k elements for some 0≤k≤n. A sign pattern τ assigns an eigenvalue τ(R)=±1 to each R∈T consistently with multiplication (so, for example, τ(I)=1 and τ(R)τ(R′)=±τ(R′′) if RR′=±R′′, with the same sign on both sides). For any fixed T there are finitely many sign patterns, namely 2k of them. The operator
Πτ=∣T∣1R∈T∑τ(R)R
is the projector onto the corresponding joint eigenspace, which has dimension 2n−k. The centraliser C(T) is the set of Pauli strings commuting with every element of T.
Stabilizer codes
These constructions are very common in quantum error correction (QEC). An important family of quantum codes is that of stabilizer codes. Each pair (T,τ) defines a stabilizer group, and Πτ projects onto its stabilizer code.
Definition (stabilizer code space). Given an isotropic Pauli subgroup T with ∣T∣=2k and a sign pattern τ, the stabilizer code space is the joint eigenspace given by
CT,τ={∣ψ⟩:R∣ψ⟩=τ(R)∣ψ⟩ for all R∈T}=range[Πτ].
Mathematical preliminaries
Lemma: Let T be an isotropic Pauli subgroup with ∣T∣=2k, and Πτ the projector onto its joint eigenspace with sign pattern τ. IfA is non-negative on all n-qubit stabilizer states, i.e., A is a Hermitian operator on n qubits with
⟨s∣A∣s⟩≥0
for every n-qubit stabilizer state ∣s⟩, then there is a Clifford unitaryU and an operator B on n−k qubits such that
(a) UΠτAΠτU†=∣0k⟩⟨0k∣⊗B;
(b) ⟨t∣B∣t⟩≥0 for every (n−k)-qubit stabilizer state ∣t⟩.
In other words, the block of A on the code space is, up to a Clifford change of basis, an (n−k)-qubit operator B that satisfies the same kind of hypothesis as A. We can also gain some intuition for why such a result might be useful. From (a), we have that
Tr[B]=Tr[ΠτA]
and
Tr[B2]=Tr[(ΠτAΠτ)2].
From (b), any inequality Tr[B2]≤M(Tr[B])2 known for (n−k)-qubit operators satisfying the positivity constraint holds for B. Together they imply that
Tr[(ΠτAΠτ)2]≤M(Tr[ΠτA])2.
This is the way the lemma enters the final theorem.
Proof. We want to show that there is a Clifford unitary U and an operator B on n−k qubits such that
UΠτAΠτU†=∣0k⟩⟨0k∣⊗B,
where ⟨t∣B∣t⟩≥0 for every (n−k)-qubit stabilizer state ∣t⟩. The argument has four steps. First, we move the code space to a standard position. Pick independent generators R1,…,Rk of T and set gi=τ(Ri)Ri.
These are k independent, commuting, Hermitian Pauli operators, and they allow us to rewrite Πτ as
Πτ=i=1∏k21(I+gi).
It is a standard fact about the Clifford group that any such list can be mapped to Z1,…,Zk: there is a Clifford unitary U with
UgiU†=Zi
for all i. (In more mathematical terms, Cliffords realise every symplectic transformation of the Pauli labels together with every choice of signs, and any k independent commuting Paulis can be completed to a symplectic basis.) Consequently
UΠτU†=i=1∏k21(I+Zi)=∣0k⟩⟨0k∣⊗In−k.
Now, let A′=UAU†. For any n-qubit stabilizer state ∣s⟩,
⟨s∣A′∣s⟩=⟨U†s∣A∣U†s⟩≥0.
Here the assumption on the positivity of A is used: Cliffords map stabilizer states to stabilizer states, so U†∣s⟩ is an n-qubit stabilizer state, and A is non-negative on it by assumption. So A′ satisfies the same positivity requirements as A. Let B=(⟨0k∣⊗In−k)A′(∣0k⟩⊗In−k), an operator on the last n−k qubits. Then, using our findings above,
To conclude, let ∣t⟩ be a stabilizer state of n−k qubits, with stabilizer generators h1,…,hn−k. Then ∣0k⟩⊗∣t⟩ is stabilized by the n independent commuting Paulis Z1,…,Zk,I⊗h1,…,I⊗hn−k, so it is an n-qubit stabilizer state. Hence, by the above (that is, by the positivity requirements in the assumption of our lemma applied to this particular n-qubit stabilizer state),
⟨t∣B∣t⟩=(⟨0k∣⊗⟨t∣)A′(∣0k⟩⊗∣t⟩)≥0.
This concludes the proof.
Proof of the main result
Let us write some definitions in order to state the main result differently. From now on we write d:=dn=2n. Write A=d1∑PaPP over the n-qubit Paulis, with aI:=Tr[A]. Let Λn be the set of Hermitian operators A with Tr[A]=1 and ⟨s∣A∣s⟩≥0 for every stabilizer state s. This is known as the Λ-polytope. Furthermore, define the number
Mn=A∈ΛnmaxTr[A2].
We know that ρ∈STABn iff Tr[Aρ]≥0 for all A∈Λn (this is standard convex duality; the polytope Λn was introduced in the work of Zurel, Okay, and Raussendorf). If we write ρ=I/d+Y with Tr[Y]=0, we can take any A∈Λn and then, since Tr[A]=1 and Tr[Y]=0, we find
Above we have used the Cauchy–Schwarz inequality and ∥A−I/d∥22=Tr[A2]−d1≤Mn−d1. The right-hand side is non-negative as soon as ∥Y∥22≤1/(d(dMn−1)). Because Tr[ρ2]=d1+∥Y∥22, this condition reads Tr[ρ2]≤Mn/(dMn−1)=1/(d−1/Mn).
Tr[Aρ]≥d1−Mn−d1∥Y∥2.
Hence
Tr[ρ2]≤d−1/Mn1⟹ρ∈STABn,
and the conjecture is exactly that Mn=2 for all n. With the Lemma, we can now prove the result through a re-statement in terms of Mn, namely the bound Mn≤1/cn.
Theorem 2. For all n≥2,
Mn≤2d+77d+2<27.
Proof. Take A∈Λn and an isotropic T with ∣T∣=2n−2=d/4, so that its centraliser has ∣C(T)∣=4d. By the Lemma, each block ΠτAΠτ corresponds to a two-qubit operator B that is non-negative on all two-qubit stabilizer states, with trace qτ=Tr[ΠτA].
The two-qubit case of the conjecture, M2=2, is known from the exhaustive enumeration of the facets of the two-qubit stabilizer polytope (see Reichardt); it gives Tr[B2]≤2Tr[B]2 (if Tr[B]=0 then B=0, and otherwise B/Tr[B]∈Λ2). Hence Tr[(ΠτAΠτ)2]≤2qτ2. Summing over τ:
Left side: ∑τTr[(ΠτAΠτ)2]=Tr[(∣T∣1∑R∈TRAR)2]=d1∑Q∈C(T)aQ2, because averaging over T keeps exactly the Paulis commuting with T.
Right side: qτ=∣T∣1∑R∈Tτ(R)aR, so ∑τqτ2=∣T∣1∑R∈TaR2.
Therefore
Q∈C(T)∑aQ2≤8R∈T∑aR2.
Now average this inequality over all such T. The symplectic group is transitive on non-identity Paulis, so each Q=I lies in C(T) for a fraction (4d−1)/(d2−1) of the T's and in T for a fraction (d/4−1)/(d2−1). With Σ=∑Q=IaQ2 and aI=1:
1+d2−14d−1Σ≤8+d2−12d−8Σ⟹Σ≤2d+77(d2−1).
Finally Tr[A2]=(1+Σ)/d≤(7d+2)/(2d+7). Since this gives 1/Mn≥(2d+7)/(7d+2)=cn>2/7, the implication above yields the Proposition.
Other results in the literature
The most notable recent result which is better than the above is the result shown recently by Liu and Liu in ther arXiv preprint:2609.03944v3 titled "On the geometry, typicality, and detectability of quantum magic".
Their theorem 1 shows that one can get much better than the bound cn above. Their value is ≈0.458.