RESEARCH / Focused target
Binomial and kappa states do not exhaust truncated entropy minimizers
RELEASED FOUNDATIONS / PROPOSED CONNECTION
Research connections
All source anchors (1)
- D5/S3/Quantum/QuantumChannels/TruncatedLossDephasingOptimizerRefutation.resultAbsent from this graph
Problem
The question
Laleh Memarzadeh and Stefano Mancini, Minimum output entropy of a non-Gaussian quantum channel, arXiv:1605.04525v1, Phys. Rev. A 94, 022341 (2016), section IV, Conjecture 1, state:
In a truncated Hilbert space of dimension K+1, the minimal output entropy of the quantum channel (3) is achieved either by binomial states of Eq.(14) or by states |κ_α⟩ of Eq. (11), depending on the values of ε and t.
Issue #11680 specifies the fixed-energy reading: for every K ≥ 1, N ∈ [0,K], ε ∈ [0,1] and t ≥ 0, some binomial or κ candidate minimizes output entropy among positive semidefinite trace-one inputs whose number expectation is N. Binomial candidates satisfy 1 ≤ M ≤ K, μ ∈ [0,1] and Mμ=N; κ candidates allow every real phase α.
Motivation
Our foothold
The frozen declaration
D5/S3/Quantum/QuantumChannels/TruncatedLossDephasingOptimizerRefutation.result
proves the negation of this reading. The witness is a qutrit input of energy
3/2 outside both candidate families; its output entropy is strictly smaller
than that of every candidate at the same channel parameters and energy.
Gap
Missing bridges
Issue #11680 records the source version and the bounded literature search: 15 OpenAlex citing records, arXiv searches and MathDB queries. No settlement was found in that searched scope. Not all 15 citing full texts were read. This is not an exhaustive literature or priority claim.
Route
Proposed approach
- Define the channel using the source's Kraus operators A_j P_k. The amplitude index is finite in the truncation; the phase index remains an infinite natural-number series. Sum the exponential series to derive the dephasing factor exp(−εt(m−n)²).
- Take K=2, N=3/2, ε=6/7, t=(7/2)log(2) and ψ=(|1⟩+|2⟩)/√2. Derive the exact output matrices from those operators. The binomial restrictions force M=2 and μ=3/4. Diagonal unitary conjugation covers all real κ phases.
- Construct strict Rayleigh bounds for the ψ output: an eigenvalue above 1/2 and another below 1/8. Exact positive-definiteness certificates put every candidate eigenvalue strictly between 1/8 and 1/2. All spectra have total mass one. Two convex transfers and strict concavity of −x log(x) give the strict entropy inequalities.
- Prove that the ψ input is admissible and contradict both candidate branches of the proposed minimum.
Falsifier
What would falsify this route
The refutation concerns the Kraus-defined channel, the fixed-energy input constraint and the two families stated in the source. It does not assert that ψ itself is a global minimizer. The source's section-III displayed κ output uses (1−f)^K for its off-diagonal coefficient, whereas its defining Kraus equations give (1−f)^(K/2). At this witness the Kraus-defined coefficient is √3/32768. Changing the channel to the displayed coefficient would change the mathematical problem.
Evidence
Evidence to collect
The canonical source is
D5/S3/Quantum/QuantumChannels/TruncatedLossDephasingOptimizerRefutation.lean.
Its only public theorem is result : ¬ claim; the other public declarations
are the channel, observables, entropy, input constraint and candidate
families needed by that statement. It reuses
D5/S3/Entropy/MaxEntropy.shannonEntropy and
D5/S3/Quantum/PureState/PureStateHandshake.rankOneDensity.
The compiled result uses propext, Classical.choice and Quot.sound;
there is no sorry, native_decide or new axiom.
Triage
Scope assessment
Tier 1 externally named conjecture, preregistered in issue #11680.
Resolution: Refuted. proof_shape: bind-only for result;
escape_witness: none;
admission_basis: open-problem-resolution (#11680; Refuted). Utility:
kind=certified-instance; basis=refutes, with typed claim and result.
What the settlement shows
- Proved in this module: the energy-constrained qutrit instance admits an input whose output entropy is strictly below both candidate families, including every κ phase. The failure is the exhaustiveness of the proposed optimizer families; their allowed parameter ranges do not prevent this instance.
- Proved in this module: a Hermitian trace-one qutrit matrix strictly between (1/8)I and (1/2)I has greater entropy than the explicit ψ output. This is the sufficient spectral criterion used for both families. The Kraus series, exact outputs, strict spectral bounds and admissibility facts are established in the proof rather than assumed numerical data.
- Open: global optimality of ψ, the actual minimizers at these parameters, the parameter region where the two candidate families do minimize entropy, and the analogous conclusion for K=1.
- Open beyond this settlement: global optimization conclusions that use Conjecture 1 require independent justification. Comparisons restricted to the two families, including the source's ε,t crossing curve and t_* comparison, do not establish a minimum over all admissible inputs. The source's pure-state reduction lemmas and its infinite-dimensional low-entropy construction are not formalized or settled by this module.
ASSUMED-UNVERIFIED
Unverified assumptions
The literature absence is bounded to the search in issue #11680; the unread citing full texts remain unverified. The Lean kernel authenticates neither the external publication nor its version history. The spectral entropy uses natural logarithms; the source uses base two, whose common positive scale preserves the ordering. No separate base-conversion or all-parameter complete-positivity theorem is exported.
Information-escape registration is paused under CLAUDE.md §3.9
「信息逃逸登记暂缓」; no Reg source or declared_validated record is part of
this delivery.