{"paper":{"title":"SMT-AD: a scalable quantum-inspired anomaly detection approach","license":"http://creativecommons.org/licenses/by/4.0/","headline":"A quantum-inspired tensor network detects anomalies competitively with linear parameter growth.","cross_cats":["cond-mat.stat-mech","quant-ph"],"primary_cat":"cs.LG","authors_text":"Apimuk Sornsaeng, Dario Poletti, Jonathan Pan, Joshua Lim, Si Min Chan, Swee Liang Wong, Wenxuan Zhang","submitted_at":"2026-04-07T02:37:45Z","abstract_excerpt":"Quantum-inspired tensor networks algorithms have shown to be effective and efficient models for machine learning tasks, including anomaly detection. Here, we propose a highly parallelizable quantum-inspired approach which we call SMT-AD from Superposition of Multiresolution Tensors for Anomaly Detection. It is based upon the superposition of bond-dimension-1 matrix product operators to transform the input data with Fourier-assisted feature embedding, where the number of learnable parameters grows linearly with feature size, embedding resolutions, and the number of additional components in the "},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"even with minimal configurations, it achieves competitive performance against established anomaly detection baselines","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"That the superposition of bond-dimension-1 matrix product operators combined with Fourier-assisted multiresolution embedding can reliably separate anomalous from normal patterns in real data.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"SMT-AD applies superposition of bond-dimension-1 matrix product operators with multiresolution Fourier embedding to achieve competitive anomaly detection on standard datasets with linear parameter growth.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"A quantum-inspired tensor network detects anomalies competitively with linear parameter growth.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"d3b14bd0daf93ef710a2d729d2d74abba8ae313416371f3e36bd93b1801e49a9"},"source":{"id":"2604.06265","kind":"arxiv","version":2},"verdict":{"id":"ce3d4fbe-3582-4fd4-80af-49809c700ad6","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-10T19:46:52.102278Z","strongest_claim":"even with minimal configurations, it achieves competitive performance against established anomaly detection baselines","one_line_summary":"SMT-AD applies superposition of bond-dimension-1 matrix product operators with multiresolution Fourier embedding to achieve competitive anomaly detection on standard datasets with linear parameter growth.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"That the superposition of bond-dimension-1 matrix product operators combined with Fourier-assisted multiresolution embedding can reliably separate anomalous from normal patterns in real data.","pith_extraction_headline":"A quantum-inspired tensor network detects anomalies competitively with linear parameter growth."},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2604.06265/integrity.json","findings":[],"available":true,"detectors_run":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938"},"references":{"count":0,"sample":[],"resolved_work":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","internal_anchors":0},"formal_canon":{"evidence_count":1,"snapshot_sha256":"7cfbf706b6ed019892ca32fddf10ea5182a668f41896aca6afc85fac12396f0a"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"}