{"id":"d28a23db-b8a3-4d8a-942a-1e9e9be97154","arxiv_id":"2508.11308","paper_version":2,"verdict":"UNVERDICTED","confidence":"LOW","novelty_score":5.0,"correctness_risk":"unknown","formal_verification":"none","parameter_count":0,"one_line_summary":"Spectral extrema of normalized entanglement witnesses are characterized, with a proven divergence between decomposable and nondecomposable witnesses and a universality theorem for nondecomposable detection of NPT states.","lead":"This paper characterizes the possible spectra of unit-trace entanglement witnesses, proving that the smallest eigenvalue's infimum is reached by decomposable but never by nondecomposable witnesses. It also shows that any entangled state with non-positive partial transpose, beyond qubit-qubit and qubit-qutrit systems, is detected by some nondecomposable witness.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"No readable proof is supplied: the central DEW/NDEW spectral dichotomy and NPT-detection claim are currently unverifiable, so the verdict remains UNVERDICTED.","rationale":"I read the abstract as a set of plausible, checkable spectral-optimization claims over unit-trace entanglement witnesses. The central results would require careful compactness and duality arguments, and they are not self-evidently true or false from the abstract alone. However, the provided manuscript text is unreadable and appears to belong to a different arXiv identifier, so no derivation can be examined. The reader's weakest assumption about closure and compactness of the feasible set is reasonable, but the immediate blocker is not a hypothesized mathematical gap: it is the total absence of readable proof. Therefore I cannot certify the claims, and I also cannot demonstrate a specific false step. This is precisely the situation in which the verdict should remain UNVERDICTED rather than being moved to ACCEPT or REJECT. I partially agree with the reader because we reach the same unverified conclusion, though through different emphasis: the reader focuses on a possible compactness failure, while I flag the unreadability of the evidence itself. A clean retrieval of the source and targeted re-derivation of the smallest-eigenvalue theorem would resolve the concern.","tokens_in":47571,"tokens_out":5124,"duration_ms":58791,"concrete_test":"Retrieve the clean source of arXiv:2508.11308 and re-examine the proof of the smallest-eigenvalue theorem. Specifically: (i) identify the explicit DEW attaining the infimum; (ii) for the NDEW case, exhibit the extremizing sequence and verify that its limit is not an NDEW, checking whether the non-attainment genuinely follows from non-closedness of the NDEW set rather than from a hidden assumption; (iii) for the NPT-detection claim, construct the claimed NDEW for a representative NPT state in dimension 3x3 and verify that it is non-decomposable and detects the state, while confirming the 2x2 and 2x3 exclusions are exactly the Peres–Horodecki cases. If any of these steps cannot be reproduced from stated assumptions, the corresponding theorem should be marked unproved.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The supplied full text is unreadable mojibake and even embeds the header 'arXiv:2508.11306v1 [math.AG]' rather than the target quant-ph identifier 2508.11308. No derivation, lemma, or reference list can be inspected, so none of the central claims can be checked. The most load-bearing claim—that the infimum of the smallest eigenvalue is attained by DEWs but strictly unattainable by NDEWs—depends on subtle closure/compactness facts about the unit-trace witness set. Attainment for DEWs requires either a compactness argument or an explicit optimizer; strict non-attainment for NDEWs requires showing that every extremizing sequence either leaves the NDEW set or converges to a witness that is not nondecomposable under the relevant topology. None of that argument is visible. The additional NPT-detection claim also relies on the Peres–Horodecki low-dimension exclusion and on a dual-cone correspondence between witnesses and states; these are plausible but unsupported by the available text. This is an evidential unverifiability concern, not a demonstrated mathematical flaw.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript claims a systematic spectral characterization of unit-trace entanglement witnesses (EWs), analyzing the infimum and supremum of the largest eigenvalue, the smallest eigenvalue, the negativity, and the squared Frobenius norm, together with attainment conditions. The main claimed results are a fundamental divergence between decomposable witnesses (DEWs) and nondecomposable witnesses (NDEWs): the infimum of the smallest eigenvalue is attained by DEWs but strictly unattainable by NDEWs; necessary conditions for an EW to possess a mirrored EW; and a detection statement that every NPT state beyond the two-qubit and qubit-qutrit systems can be detected by an NDEW. The abstract is readable, but the supplied full text is corrupted mojibake and embeds a different arXiv header, so no proofs, lemmas, or references can be inspected.","tokens_in":47758,"tokens_out":3270,"duration_ms":35931,"significance":"If the stated theorems are correct, the paper would provide a clean spectral characterization of DEWs versus NDEWs and a strong, general detection guarantee for NDEWs, which would be of genuine interest to the entanglement-witness community. The work appears to be purely mathematical: the abstract shows no fitted parameters, no empirically derived predictions, and no self-referential circularity. However, the significance is conditional: because the full text is unreadable in the supplied form, no proof can be checked and no machine-checked or reproducible derivation is available. The potential value is high, but the current submission does not permit verification.","major_comments":[{"comment":"The supplied full text is corrupted mojibake and embeds the header “arXiv:2508.11306v1 [math.AG]” rather than the target quant-ph identifier 2508.11308. No derivation, lemma, or reference list can be inspected. This prevents verification of every central claim, including the DEW/NDEW attainment dichotomy and the NPT-detection theorem. This is a load-bearing presentation defect, not a mere typo.","section":"Full Text (entirety)"},{"comment":"The claim that the infimum of the smallest eigenvalue is attained by DEWs yet strictly unattainable by NDEWs requires a precise closure/compactness argument for the unit-trace witness set, or explicit extremizing sequences and an optimizer. Neither the topology of the feasible set nor a limiting argument is visible in the provided text. Without such an argument, the attainment/non-attainment distinction is unsupported.","section":"Abstract, final paragraph"},{"comment":"The claim that every NPT state beyond 2×2 and 2×3 is detected by an NDEW rests on the Peres–Horodecki low-dimensional equivalence and on a dual-cone correspondence between witnesses and states. The supplied text shows neither the proof of this correspondence nor a dimension-dependent construction of the detecting NDEW. As stated, this is an assertion rather than a demonstrated theorem.","section":"Abstract, NPT-detection claim"}],"minor_comments":[{"comment":"The embedded arXiv header “2508.11306 [math.AG]” is inconsistent with the stated quant-ph identifier 2508.11308. Please confirm that the correct source file was submitted.","section":"Full Text, header"},{"comment":"No references are visible in the provided text; the discussion of Peres–Horodecki results and dual-cone arguments requires proper citations.","section":"General"},{"comment":"The abstract would benefit from explicitly stating the Hilbert-space dimension conventions and the precise definition of negativity beyond “absolute value of the sum of negative eigenvalues”.","section":"Abstract, first paragraph"}],"recommendation":"uncertain","confidential_remarks":"I could not review the mathematics because the supplied full text is corrupted mojibake and contains an arXiv header for a different paper (2508.11306 [math.AG]). This is not a reflection on the authors' work; the editor may wish to request a clean source file and confirm the manuscript identifier before assigning a new review. The abstract-level claims are interesting but are unverifiable in the current form."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Here's the bottom line: this looks like a serious mathematical paper on entanglement witnesses, but from what we actually received, no one can check it. The abstract claims a complete spectral envelope for unit-trace witnesses, a sharp attainment difference between DEWs and NDEWs for the smallest eigenvalue infimum, and a universality result that every NPT state beyond 2x2 and 2x3 is NDEW-detectable. If those theorems are right, that is real structural progress and worth the attention of the entanglement-detection community.\n\nWhat the paper does well, as far as the abstract shows, is to come with a clean, well-scoped question and a pair of non-obvious headline claims. The low-dimensional exceptions line up with the known Peres-Horodecki equivalence, and the abstract doesn't look like it is fitting parameters or massaging data. The claims are theorem-shaped, not curve-fit shaped.\n\nThe soft spot is not the math, it's the evidence. The provided 'full text' is mojibake, and the embedded header even refers to a different arXiv ID. I'm assuming that's an extraction failure rather than something wrong with the paper, but it means there are no readable lemmas, derivations, or references. The stress-test concern about closure and compactness of the unit-trace witness set is exactly the sort of fine topological point that the attainment dichotomy would hinge on, but that is a genuine question for the proofs, not a demonstrated flaw. Similarly, the NPT-detection claim leans on the dual-cone correspondence and the low-dimension exclusions; those are standard tools, but the application has to be right. I see no internal contradiction in the abstract; I just cannot see the arguments.\n\nWho is this for? People working on entanglement witnesses and their geometry. If the results hold, they are citable. But I wouldn't cite them yet, and I wouldn't put them in front of a reading group unless we can lay hands on the actual PDF.\n\nMy recommendation: a good editor should send this to peer review, not desk-reject it, provided the manuscript itself is legible. The claims are plausible and significant enough to justify referee time. The reviewer should check the attainment proof in particular, and the NDEW-universality proof's use of the dual cone. Treat the current artifact as 'unverified,' not 'wrong.'","headline":"Plausible and potentially significant spectral results for entanglement witnesses, but the supplied full text is unreadable, so the paper is unverifiable as-is; it deserves a referee once a clean copy is available.","tokens_in":48300,"tokens_out":2682,"would_cite":false,"duration_ms":29305,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["81P40","15A42"],"pacs":["03.67.Mn"],"model":"deepseek-v4-flash","headline":"Spectral characterization of entanglement witnesses: unit-trace witnesses have known eigenvalue extrema, and the smallest-eigenvalue infimum is attainable by decomposable but never by nondecomposable witnesses.","keywords":["entanglement witnesses","decomposable witnesses","nondecomposable witnesses","spectral bounds","eigenvalue extremization","negative partial transpose","quantum entanglement detection","unit-trace normalization"],"falsifier":"Fix a finite dimension such as $3 \\times 3$ and numerically minimize the smallest eigenvalue over unit-trace nondecomposable witnesses; if a sequence of NDEWs is found whose smallest eigenvalue actually reaches the claimed infimum rather than merely approaching it, the strict-unattainability claim fails. For the detection claim, take a known NPT state in $3 \\times 3$ and attempt to construct an NDEW that detects it; an explicit NPT state that no nondecomposable witness detects would refute the universality result.","tokens_in":47392,"feed_emoji":"⚛️","tokens_out":4343,"duration_ms":50871,"temperature":0.7,"pith_summary":"The paper aims to settle what spectral shapes an entanglement witness can have once its trace is fixed to one. It identifies the attainable extremes of four spectral quantities—largest eigenvalue, smallest eigenvalue, negativity, and squared Frobenius norm—separately for decomposable and nondecomposable witnesses, and says exactly when those extremes are reached. The headline discovery is that the two witness classes are spectrally almost interchangeable except at one point: the smallest possible smallest eigenvalue can be attained by a decomposable witness but is only approached, never reached, by any nondecomposable one. A second result makes nondecomposable witnesses look more useful: beyond the smallest dimensions, every non-positive-transpose state is detected by one. If accepted, these results give a complete spectral map of normalized witnesses and a sharp new distinction between decomposable and nondecomposable detection.","feed_headline":"Smallest eigenvalue exposes a hard split among entanglement witnesses","feed_subtitle":"Unit-trace witnesses: decomposable ones hit the spectral floor, nondecomposable ones approach but never reach it.","key_machinery":"The central object is the unit-trace entanglement witness: a Hermitian operator on a bipartite Hilbert space with nonnegative expectation on every separable state and trace one. Witnesses are split into decomposable witnesses, written as $W = P + Q^{T_B}$ with $P,Q \\ge 0$, and nondecomposable witnesses, which require an additional term that cannot be written this way. The proof machinery translates witness inequalities plus the trace normalization into eigenvalue constraints, then optimizes the four spectral functionals separately over the two convex witness sets; the operative identity is the attainment dichotomy: the smallest-eigenvalue infimum lies in the closure for both classes, but the","core_discovery":"For unit-trace entanglement witnesses, the extreme values of the largest eigenvalue, smallest eigenvalue, negativity, and squared Frobenius norm are characterized separately for decomposable and nondecomposable witnesses. The central divergence is that the infimum of the smallest eigenvalue is attained by a decomposable witness, while no nondecomposable witness can attain it—such witnesses can only approach the value in the limit. As a consequence, reaching the spectral floor is itself a certificate that the witness is decomposable. The paper also shows that nondecomposable witnesses are stronger detectors: every non-positive-transpose state in dimensions beyond two-qubit and qubit-qutrit is","pith_inferences":["If the smallest-eigenvalue infimum is approached but never reached by NDEWs, then the boundary of the normalized NDEW set is not witnessed from inside; numerical or algorithmic searches for extremal NDEWs would need to work with limiting sequences rather than attained optima.","The same trace-normalized spectral optimization could be scaled to witnesses with arbitrary trace and applied to noise tolerance: the Frobenius-norm and negativity extrema are natural robustness parameters for detection in noisy settings.","The universal NPT-detection result suggests a constructive recipe: any NPT state in those dimensions should have a nearby nondecomposable witness, which may turn an existence proof into an explicit detection protocol if the extremal construction can be made algorithmic.","The DEW/NDEW attainment gap resembles complementary-slackness behavior in convex conic optimization, hinting that other trace-normalized operator cones in quantum information could exhibit similar 'attained by one cone, not its dual counterpart' dichotomies."],"forward_implications":["In any fixed finite bipartite dimension, the spectral range of unit-trace entanglement witnesses is fixed for four quantities, with known attainment cases for decomposable and nondecomposable witnesses.","A witness that actually reaches the smallest-eigenvalue infimum cannot be nondecomposable, so reaching the spectral floor becomes a spectral certificate of decomposability.","The mirrored-witness conditions give a spectral necessary test: witnesses whose eigenvalues fall outside the characterized range cannot have a mirror partner.","Every NPT state beyond the two-qubit and qubit-qutrit cases is detected by some nondecomposable witness, giving nondecomposable witnesses a universal detection role in those dimensions.","The extremal values can be used to calibrate normalized witness constructions in optimization tasks that convert spectra into quantitative entanglement bounds."],"supporting_citations":[],"fun_headline_variants":["Entanglement witness spectral floor: only decomposable ones can hit it","For entanglement witnesses, the lowest eigenvalue is a decomposable-only privilege","Nondecomposable witnesses never reach spectral minimum; decomposable do","New spectral rule: reaching the minimum eigenvalue proves decomposability"],"cache_read_input_tokens":2816,"weakest_assumption_plain":"The whole extremal analysis assumes that the set of unit-trace entanglement witnesses is closed and compact enough for infima and suprema to behave like ordinary extrema, and that the witness-state duality is exact; if a sequence of normalized witnesses can leave the witness set in the limit, the claimed DEW/NDEW attainment split could fail.","fun_headline_variants_meta":{"raw":{"variants":["Entanglement witness spectral floor: only decomposable ones can hit it","For entanglement witnesses, the lowest eigenvalue is a decomposable-only privilege","Nondecomposable witnesses never reach spectral minimum; decomposable do","New spectral rule: reaching the minimum eigenvalue proves decomposability"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000192,"raw_usage":{"total_tokens":1160,"prompt_tokens":696,"completion_tokens":464,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":440,"completion_tokens_details":{"reasoning_tokens":390}},"tokens_in":440,"tokens_out":464,"duration_ms":4994,"temperature":1.0,"reasoning_tokens":390,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-05T19:59:52.559738+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Fix a finite dimension such as $3 \\times 3$ and numerically minimize the smallest eigenvalue over unit-trace nondecomposable witnesses; if a sequence of NDEWs is found whose smallest eigenvalue actually reaches the claimed infimum rather than merely approaching it, the strict-unattainability claim fails. For the detection claim, take a known NPT state in $3 \\times 3$ and attempt to construct an NDEW that detects it; an explicit NPT state that no nondecomposable witness detects would refute the universality result.","supporting_citations":[],"review_version":1}