{"paper":{"title":"DistributedEstimator: Distributed Training of Quantum Neural Networks via Circuit Cutting","license":"http://creativecommons.org/licenses/by/4.0/","headline":"A staged distributed pipeline for circuit-cut quantum neural network training preserves test accuracy and robustness on standard benchmarks.","cross_cats":["cs.LG","quant-ph"],"primary_cat":"cs.DC","authors_text":"Adel N. Toosi, Prabhjot Singh, Rajkumar Buyya","submitted_at":"2026-02-18T07:17:29Z","abstract_excerpt":"Circuit cutting decomposes a large quantum circuit into smaller subcircuits executed independently; expectation values are recovered by classically combining subcircuit outcomes. Prior work characterises cutting overhead via subcircuit counts and sampling complexity, but its end-to-end impact on iterative, estimator-driven training pipelines remains under-measured from a systems perspective. We propose DistributedEstimator, a cut-aware estimator execution pipeline that treats circuit cutting as a staged distributed workload, instrumenting each query across four phases: partitioning, subexperim"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"Despite these overheads, test accuracy is fully preserved on Iris and maintained without systematic degradation on MNIST across all cut configurations. Robustness under Gaussian noise and FGSM perturbations is similarly preserved, with several cut configurations exhibiting comparable or improved robustness relative to the uncut baseline.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"The workloads (Iris and MNIST binary classification) and cut configurations tested are representative of broader QNN training scenarios, and the logged runtime traces accurately capture end-to-end overheads without unaccounted hardware variability.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"DistributedEstimator demonstrates that circuit cutting preserves test accuracy and robustness in QNN training on Iris and MNIST while revealing that classical reconstruction dominates runtime and exponential subcircuit growth limits scaling.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"A staged distributed pipeline for circuit-cut quantum neural network training preserves test accuracy and robustness on standard benchmarks.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"b3d961110f92575a5808e40d4ffc81b6d34cf65bbcdc19d5f98141a7db87c0ad"},"source":{"id":"2602.16233","kind":"arxiv","version":3},"verdict":{"id":"d2e2a134-c1b3-4e56-8ebf-d1e8478b291f","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-15T21:40:44.322906Z","strongest_claim":"Despite these overheads, test accuracy is fully preserved on Iris and maintained without systematic degradation on MNIST across all cut configurations. Robustness under Gaussian noise and FGSM perturbations is similarly preserved, with several cut configurations exhibiting comparable or improved robustness relative to the uncut baseline.","one_line_summary":"DistributedEstimator demonstrates that circuit cutting preserves test accuracy and robustness in QNN training on Iris and MNIST while revealing that classical reconstruction dominates runtime and exponential subcircuit growth limits scaling.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"The workloads (Iris and MNIST binary classification) and cut configurations tested are representative of broader QNN training scenarios, and the logged runtime traces accurately capture end-to-end overheads without unaccounted hardware variability.","pith_extraction_headline":"A staged distributed pipeline for circuit-cut quantum neural network training preserves test accuracy and robustness on standard benchmarks."},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2602.16233/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":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"}