Covariance / symmetry reduction of quantum channels — THE WALL: the one recipe every partial PPT² result reuses, and nobody has escaped to the general case

Statement

Covariance reduction is the standard technique for making an otherwise-intractable question about quantum channels solvable by *restricting attention to channels that commute with a chosen symmetry group*. A channel $\Phi:M_n\to M_n$ is $G$-covariant for a group $G$ with unitary representations $U_g, V_g$ if $$ \Phi(U_g\,\rho\,U_g^\dagger)=V_g\,\Phi(\rho)\,V_g^\dagger\qquad\text{for all }g\in G. $$ By Schur's lemma / the double-commutant structure, imposing covariance forces the channel's Choi matrix (see Choi–Jamiołkowski isomorphism — state–channel duality: a map is CP iff its Choi matrix is PSD) to lie in the commutant of the representation — a low-dimensional, often block-diagonal algebra. The infinite-dimensional cone of all channels collapses to a handful of scalar parameters, and cone-membership questions (positivity, PPT, separability, entanglement-breaking) become finite linear-algebra / SDP problems that can be settled in closed form.

Applied to PPT²

every 2020–2026 partial result toward the PPT-squared conjecture (PPT² conjecture (Christandl 2012) — composition of two PPT channels is entanglement-breaking; open for dimension ≥ 4) picks a symmetry group $G$, characterizes the $G$-invariant PPT cone, and shows that within that cone the composition of two PPT channels collapses to entanglement-breaking — i.e. it proves PPT² *only for $G$-covariant channels*.

Facts — the recipe, and the roster that follows it

The uniform 3-step recipe (this repo's ppt2-scan.md §3, §4): 1. Pick a symmetry/covariance group $G$. 2. Characterize the $G$-invariant Choi/PPT cone via representation theory (commutant is block-diagonal → few parameters). 3. Show the invariant cone collapses to entanglement-breaking under composition.

The papers that instantiate it — *the same recipe, a different group each time*:

- Singh–Nechita 2020, "The PPT² Conjecture Holds for All Choi-Type Maps," arXiv:2011.03809 (Ann. Henri Poincaré 23, 2022). $G=$ diagonal unitary group. Introduces the "factor-width" generalization on Choi matrices; covers Choi-type, depolarizing, dephasing, amplitude-damping maps and mixtures. This is the paper that *established* covariance-group reduction as the PPT² workhorse. (Source: ppt2-scan.md §3.6.) - Nechita–Park 2024, "Random covariant quantum channels," arXiv:2403.03667 (Ann. Henri Poincaré 27, 2026). $G=$ diagonal-orthogonal-covariant (DOC); shows composition of two *random* DOC channels is *generically* EB — probabilistic, still within a fixed covariance class. (Source: ppt2-scan.md §2.) - Prudhoe 2025, "Cartan-covariant Quantum Channels and the PPT² conjecture," arXiv:2501.03959. $G=$ Cartan (a 2-parameter covariant generalization of depolarizing channels); abstract (fetched): "we prove that the PPT² conjecture holds for these channels in any dimension." Any $n$ — but only for that class. (Source: ppt2-scan.md §2.) - Park 2026, "k-Positivity and high-dimensional bound entanglement under symplectic group symmetries," arXiv:2602.09860. $G=$ symplectic / conjugate-symplectic. (Source: ppt2-scan.md §2, §3.11.) - García-Velo–Ibort 2026, "Schwarz maps with symmetry," arXiv:2601.02282. $G\in\{U(n),\ \mathrm{DU}(2),\ \mathrm{DU}(3),\ U(2)\otimes U(2),\ U(2)\otimes U(3)\}$ — small dimensions / specific symmetry families. (Source: ppt2-scan.md §2.)

Why this is THE WALL: - Intrinsic ceiling. The method *only ever* proves PPT² for channels *with imposed symmetry*. Each new paper swaps in a different group ($\text{diagonal-unitary}\to\text{Cartan}\to\text{symplectic}\to\ldots$); none touches the general, non-covariant channel. The general $n\ge4$ case has been open since Christandl posed it in 2012 and remains open (July 2026). (Source: ppt2-scan.md §3, §6.) - The obstruction is exactly bound entanglement. Covariance forces the invariant PPT cone to be *provably separable*, sidestepping Bound entanglement — entangled states that are PPT, hence undistillable; the rare structured regime where a PPT² counterexample must live. A general counterexample would be a bound-entangled Choi state from a composed PPT map — and that lives in the *non-symmetric*, structured regime the recipe is designed to avoid. Singh–Nechita's own limitation (quoted in scan §4): outside the covariant class "the validity of the conjecture still remains ambiguous." - No fundamentally new strategy 2020–2026. The roster is thin (Nechita/Singh/Christandl/Müller-Hermes lineage plus two entrants, Prudhoe and Park, both working the same recipe). Nobody has broken out. (Source: ppt2-scan.md §3, §6.)

Related

- PPT² conjecture (Christandl 2012) — composition of two PPT channels is entanglement-breaking; open for dimension ≥ 4 — the conjecture this wall surrounds; every known partial proof is a covariance-reduction instance, and the general case is exactly what covariance reduction *cannot* reach. - Bound entanglement — entangled states that are PPT, hence undistillable; the rare structured regime where a PPT² counterexample must live — the phenomenon the wall exists to avoid: covariance guarantees the invariant cone is separable, so the method structurally never confronts a genuine bound-entangled counterexample. - Choi–Jamiołkowski isomorphism — state–channel duality: a map is CP iff its Choi matrix is PSD — the isomorphism that turns "$G$-covariant channel" into "commutant-block-diagonal Choi matrix," which is *what makes* the reduction computable. - Peres–Horodecki (PPT) criterion — partial transpose positivity is necessary for separability, and sufficient only in 2×2 and 2×3 — the PPT test being restricted to the invariant cone; within symmetric classes it becomes decidable, which is the whole leverage of the method. - Entanglement-breaking channel — measure-and-prepare maps: separable Choi matrix ⇔ destroys all entanglement — the collapse target: the recipe's step 3 is always "show the invariant composed cone is EB." - PPT map — a channel that is CP and completely co-positive: Choi matrix PSD with positive partial transpose — the object class (CP and co-CP maps) whose $G$-covariant sub-cone each paper analyzes. - Entanglement witness — a Hermitian W with Tr(Wσ)≥0 on all separable states but Tr(Wρ)<0 on some entangled ρ — dual certificates; escaping the wall plausibly needs non-decomposable witnesses over the *general* (non-covariant) cone, not the symmetry-simplified one. - Bisociation (Koestler 1964) — creative insight as the collision of two habitually unrelated frames of thought — the thesis edge. Fourteen years of same-recipe results say the wall will not fall to another symmetry group drawn from inside quantum information's own toolbox. Escaping it likely requires a technique bisociated from a distant field — the scan's own under-exploited candidates (§6) are non-covariant representation theory, second-order free probability to close the Collins–Yin–Zhong identical-channel gap, and structured (not blind) ML/SDP counterexample search seeded on bound-entangled states — each a transplant from a matrix domain far from the covariance recipe. Bisociation names exactly this move: the general case is likely unreachable by recombination *within* the field and needs a cross-domain graft.