{ "title": "EpochCore Benchmark Engine v1 \u2014 measured quantum-vs-classical verdicts", "honesty_framing": "All quantum execution is on the qiskit-aer CPU simulator (no QPU in this loop; IBM Heron attestation is a separate, job-ID-verifiable artifact). The engine reports the axes the quantum path LOSES \u2014 wall-clock on a simulator, volume-law tensor-network cells, classical-friendly datasets \u2014 alongside what it wins. No number here is quotable unless marked claim_eligible, which additionally requires replication in a second independent sweep.", "wall_clock_caveat": "every wall-clock figure on this page measures CPU-simulator overhead, not hardware speed \u2014 classical usually wins that axis and the receipts say so", "replication_status": "PENDING until a second independent full sweep (different master seed) replicates a verdict; no axis is claim-eligible before then", "suite_ts": "2026-06-11T23:56:51.857616+00:00", "sweep_id": "cli-full-20260611T233832Z", "suite_self_hash": "1f8122a9ea850749f2131b871753c025ebc43d4d9fff4e63fc21ad6f15544e4b", "benchmarks": { "sim_crossval": { "verdicts": [ { "axis": "sampled_distribution_agreement", "winner": "agreement_pass", "tvd_cross_method_indep_seeds": { "n": 5, "mean": 0.568115234375, "std": 0.0038947890951034707, "median": 0.568359375, "min": 0.562255859375, "max": 0.57275390625, "ci95_lo": 0.564460128219815, "ci95_hi": 0.570751953125 }, "tvd_self_noise_baseline": { "n": 7, "mean": 0.5242396763392857, "std": 0.10731857088580994, "median": 0.5699462890625, "min": 0.284912109375, "max": 0.5775146484375, "ci95_lo": 0.39444123450786384, "ci95_hi": 0.5712367466517857 }, "tvd_clifford_vs_stabilizer": { "n": 2, "mean": 0.3961181640625, "std": 0.18040199861814996, "median": 0.3961181640625, "min": 0.2685546875, "max": 0.523681640625, "ci95_lo": 0.2685546875, "ci95_hi": 0.523681640625 }, "marginal_max_dev": { "n": 7, "mean": 0.014753069196428572, "std": 0.003250700235910933, "median": 0.0152587890625, "min": 0.00927734375, "max": 0.0196533203125, "ci95_lo": 0.0123291015625, "ci95_hi": 0.016775948660714284 }, "marginal_tolerance_5sigma": 0.032621, "rule": "cross-method TVD (independent seeds) <= max(1.5x self-noise TVD, 0.05) AND per-qubit marginal deviation <= 5-sigma", "claim_eligible": false } ], "ok": true, "receipt_self_hash": "d42c7fc76dc4893a140ae83208f464c1f68e309340d7d00686af87f0f1226275", "limitations": [] }, "grover_search": { "verdicts": [ { "axis": "oracle_queries_to_99pct_success_n3_vs_classical_mean", "winner": "quantum", "higher_is_better": false, "quantum": { "n": 10, "mean": 4.0, "std": 0.0, "median": 4.0, "min": 4.0, "max": 4.0, "ci95_lo": 4.0, "ci95_hi": 4.0 }, "classical": { "n": 10, "mean": 4.5144, "std": 0.08300361437913419, "median": 4.4965, "min": 4.359, "max": 4.633, "ci95_lo": 4.461637172299037, "ci95_hi": 4.5592999999999995 }, "p_corrected": 0.03125, "effect_size_hl": -0.5219999999999998, "claim_eligible": true, "note": "quantum = r* x k* oracle calls to >=99% success (both factors in per_seed); classical comparator here is MEAN probes of a measured seeded scan \u2014 note the confidence targets differ on this axis (99% vs expected-case); see the _vs_classical_p99 axis for the matched-confidence comparison", "replication_sweep_id": "cli-full-20260611T235654Z" }, { "axis": "oracle_queries_to_99pct_success_n3_vs_classical_p99", "winner": "quantum", "higher_is_better": false, "quantum": { "n": 10, "mean": 4.0, "std": 0.0, "median": 4.0, "min": 4.0, "max": 4.0, "ci95_lo": 4.0, "ci95_hi": 4.0 }, "classical": { "n": 10, "mean": 8.0, "std": 0.0, "median": 8.0, "min": 8.0, "max": 8.0, "ci95_lo": 8.0, "ci95_hi": 8.0 }, "p_corrected": 0.03125, "effect_size_hl": -4.0, "claim_eligible": true, "note": "matched 99%-confidence comparison: classical comparator is the MEASURED 99th percentile of probes (queries to have found the item with probability 0.99 under a no-replacement scan)", "replication_sweep_id": "cli-full-20260611T235654Z" }, { "axis": "wall_clock_ms_n3", "winner": "classical", "higher_is_better": false, "quantum": { "n": 10, "mean": 18.593, "std": 0.6470625935719049, "median": 18.65, "min": 17.66, "max": 20.05, "ci95_lo": 18.277, "ci95_hi": 19.062242496124473 }, "classical": { "n": 10, "mean": 1.8914700000000004, "std": 0.07901852173875298, "median": 1.87745, "min": 1.7943, "max": 2.0589, "ci95_lo": 1.854373307661534, "ci95_hi": 1.95055666349971 }, "p_corrected": 0.0, "effect_size_hl": 16.6833, "claim_eligible": true, "note": "CPU-simulator overhead; expected and confirmed classical win", "replication_sweep_id": "cli-full-20260611T235654Z" }, { "axis": "oracle_queries_to_99pct_success_n4_vs_classical_mean", "winner": "quantum", "higher_is_better": false, "quantum": { "n": 10, "mean": 6.0, "std": 0.0, "median": 6.0, "min": 6.0, "max": 6.0, "ci95_lo": 6.0, "ci95_hi": 6.0 }, "classical": { "n": 10, "mean": 8.5391, "std": 0.11264245301937534, "median": 8.530999999999999, "min": 8.354, "max": 8.69, "ci95_lo": 8.4715, "ci95_hi": 8.602412713826729 }, "p_corrected": 0.03125, "effect_size_hl": -2.5489999999999995, "claim_eligible": true, "note": "quantum = r* x k* oracle calls to >=99% success (both factors in per_seed); classical comparator here is MEAN probes of a measured seeded scan \u2014 note the confidence targets differ on this axis (99% vs expected-case); see the _vs_classical_p99 axis for the matched-confidence comparison", "replication_sweep_id": "cli-full-20260611T235654Z" }, { "axis": "oracle_queries_to_99pct_success_n4_vs_classical_p99", "winner": "quantum", "higher_is_better": false, "quantum": { "n": 10, "mean": 6.0, "std": 0.0, "median": 6.0, "min": 6.0, "max": 6.0, "ci95_lo": 6.0, "ci95_hi": 6.0 }, "classical": { "n": 10, "mean": 16.0, "std": 0.0, "median": 16.0, "min": 16.0, "max": 16.0, "ci95_lo": 16.0, "ci95_hi": 16.0 }, "p_corrected": 0.03125, "effect_size_hl": -10.0, "claim_eligible": true, "note": "matched 99%-confidence comparison: classical comparator is the MEASURED 99th percentile of probes (queries to have found the item with probability 0.99 under a no-replacement scan)", "replication_sweep_id": "cli-full-20260611T235654Z" }, { "axis": "wall_clock_ms_n4", "winner": "classical", "higher_is_better": false, "quantum": { "n": 10, "mean": 24.292, "std": 0.9754975938235599, "median": 23.835, "min": 23.34, "max": 26.43, "ci95_lo": 23.84174888074363, "ci95_hi": 25.0299185329009 }, "classical": { "n": 10, "mean": 1.92344, "std": 0.08485247590180658, "median": 1.90225, "min": 1.8319, "max": 2.1128, "ci95_lo": 1.884487180346344, "ci95_hi": 1.9886459958584095 }, "p_corrected": 0.0, "effect_size_hl": 22.1636, "claim_eligible": true, "note": "CPU-simulator overhead; expected and confirmed classical win", "replication_sweep_id": "cli-full-20260611T235654Z" }, { "axis": "oracle_queries_to_99pct_success_n5_vs_classical_mean", "winner": "quantum", "higher_is_better": false, "quantum": { "n": 10, "mean": 4.0, "std": 0.0, "median": 4.0, "min": 4.0, "max": 4.0, "ci95_lo": 4.0, "ci95_hi": 4.0 }, "classical": { "n": 10, "mean": 16.536399999999997, "std": 0.20225407563535253, "median": 16.573, "min": 16.143, "max": 16.894, "ci95_lo": 16.407548461122694, "ci95_hi": 16.646616638182294 }, "p_corrected": 0.03125, "effect_size_hl": -12.559000000000001, "claim_eligible": true, "note": "quantum = r* x k* oracle calls to >=99% success (both factors in per_seed); classical comparator here is MEAN probes of a measured seeded scan \u2014 note the confidence targets differ on this axis (99% vs expected-case); see the _vs_classical_p99 axis for the matched-confidence comparison", "replication_sweep_id": "cli-full-20260611T235654Z" }, { "axis": "oracle_queries_to_99pct_success_n5_vs_classical_p99", "winner": "quantum", "higher_is_better": false, "quantum": { "n": 10, "mean": 4.0, "std": 0.0, "median": 4.0, "min": 4.0, "max": 4.0, "ci95_lo": 4.0, "ci95_hi": 4.0 }, "classical": { "n": 10, "mean": 32.0, "std": 0.0, "median": 32.0, "min": 32.0, "max": 32.0, "ci95_lo": 32.0, "ci95_hi": 32.0 }, "p_corrected": 0.03125, "effect_size_hl": -28.0, "claim_eligible": true, "note": "matched 99%-confidence comparison: classical comparator is the MEASURED 99th percentile of probes (queries to have found the item with probability 0.99 under a no-replacement scan)", "replication_sweep_id": "cli-full-20260611T235654Z" }, { "axis": "wall_clock_ms_n5", "winner": "classical", "higher_is_better": false, "quantum": { "n": 10, "mean": 30.672000000000004, "std": 0.7709849400460285, "median": 30.445, "min": 29.75, "max": 32.24, "ci95_lo": 30.315, "ci95_hi": 31.266 }, "classical": { "n": 10, "mean": 2.1249400000000005, "std": 0.10394135312227226, "median": 2.1033999999999997, "min": 1.9949, "max": 2.3262, "ci95_lo": 2.074613203713927, "ci95_hi": 2.19959 }, "p_corrected": 0.0, "effect_size_hl": 28.4167, "claim_eligible": true, "note": "CPU-simulator overhead; expected and confirmed classical win", "replication_sweep_id": "cli-full-20260611T235654Z" }, { "axis": "oracle_queries_to_99pct_success_n6_vs_classical_mean", "winner": "quantum", "higher_is_better": false, "quantum": { "n": 10, "mean": 6.0, "std": 0.0, "median": 6.0, "min": 6.0, "max": 6.0, "ci95_lo": 6.0, "ci95_hi": 6.0 }, "classical": { "n": 10, "mean": 32.5466, "std": 0.7669876429542026, "median": 32.634, "min": 31.335, "max": 33.698, "ci95_lo": 32.081571222654055, "ci95_hi": 32.9865 }, "p_corrected": 0.03125, "effect_size_hl": -26.5165, "claim_eligible": true, "note": "quantum = r* x k* oracle calls to >=99% success (both factors in per_seed); classical comparator here is MEAN probes of a measured seeded scan \u2014 note the confidence targets differ on this axis (99% vs expected-case); see the _vs_classical_p99 axis for the matched-confidence comparison", "replication_sweep_id": "cli-full-20260611T235654Z" }, { "axis": "oracle_queries_to_99pct_success_n6_vs_classical_p99", "winner": "quantum", "higher_is_better": false, "quantum": { "n": 10, "mean": 6.0, "std": 0.0, "median": 6.0, "min": 6.0, "max": 6.0, "ci95_lo": 6.0, "ci95_hi": 6.0 }, "classical": { "n": 10, "mean": 64.0, "std": 0.0, "median": 64.0, "min": 64.0, "max": 64.0, "ci95_lo": 64.0, "ci95_hi": 64.0 }, "p_corrected": 0.03125, "effect_size_hl": -58.0, "claim_eligible": true, "note": "matched 99%-confidence comparison: classical comparator is the MEASURED 99th percentile of probes (queries to have found the item with probability 0.99 under a no-replacement scan)", "replication_sweep_id": "cli-full-20260611T235654Z" }, { "axis": "wall_clock_ms_n6", "winner": "classical", "higher_is_better": false, "quantum": { "n": 10, "mean": 42.858000000000004, "std": 1.0757715577409748, "median": 42.894999999999996, "min": 41.16, "max": 44.5, "ci95_lo": 42.22803087754312, "ci95_hi": 43.47797263269668 }, "classical": { "n": 10, "mean": 2.32533, "std": 0.0567722946906715, "median": 2.324, "min": 2.2603, "max": 2.4366, "ci95_lo": 2.2967124481653403, "ci95_hi": 2.363987888735777 }, "p_corrected": 0.0, "effect_size_hl": 40.5403, "claim_eligible": true, "note": "CPU-simulator overhead; expected and confirmed classical win", "replication_sweep_id": "cli-full-20260611T235654Z" }, { "axis": "oracle_queries_to_99pct_success_n7_vs_classical_mean", "winner": "quantum", "higher_is_better": false, "quantum": { "n": 10, "mean": 8.0, "std": 0.0, "median": 8.0, "min": 8.0, "max": 8.0, "ci95_lo": 8.0, "ci95_hi": 8.0 }, "classical": { "n": 10, "mean": 64.57390000000001, "std": 1.1834284891318492, "median": 64.667, "min": 62.25, "max": 66.974, "ci95_lo": 63.87199422944273, "ci95_hi": 65.28387799885368 }, "p_corrected": 0.03125, "effect_size_hl": -56.592000000000006, "claim_eligible": true, "note": "quantum = r* x k* oracle calls to >=99% success (both factors in per_seed); classical comparator here is MEAN probes of a measured seeded scan \u2014 note the confidence targets differ on this axis (99% vs expected-case); see the _vs_classical_p99 axis for the matched-confidence comparison", "replication_sweep_id": "cli-full-20260611T235654Z" }, { "axis": "oracle_queries_to_99pct_success_n7_vs_classical_p99", "winner": "quantum", "higher_is_better": false, "quantum": { "n": 10, "mean": 8.0, "std": 0.0, "median": 8.0, "min": 8.0, "max": 8.0, "ci95_lo": 8.0, "ci95_hi": 8.0 }, "classical": { "n": 10, "mean": 127.5, "std": 0.5270462766947299, "median": 127.5, "min": 127.0, "max": 128.0, "ci95_lo": 127.2, "ci95_hi": 127.8 }, "p_corrected": 0.03125, "effect_size_hl": -119.5, "claim_eligible": true, "note": "matched 99%-confidence comparison: classical comparator is the MEASURED 99th percentile of probes (queries to have found the item with probability 0.99 under a no-replacement scan)", "replication_sweep_id": "cli-full-20260611T235654Z" }, { "axis": "wall_clock_ms_n7", "winner": "classical", "higher_is_better": false, "quantum": { "n": 10, "mean": 58.602, "std": 1.6241907386620444, "median": 58.35, "min": 56.21, "max": 61.02, "ci95_lo": 57.69016886676067, "ci95_hi": 59.561 }, "classical": { "n": 10, "mean": 3.0337899999999998, "std": 0.2425003479951684, "median": 2.9306, "min": 2.8624, "max": 3.5533, "ci95_lo": 2.922900552009126, "ci95_hi": 3.2274549414550187 }, "p_corrected": 0.0, "effect_size_hl": 55.508300000000006, "claim_eligible": true, "note": "CPU-simulator overhead; expected and confirmed classical win", "replication_sweep_id": "cli-full-20260611T235654Z" }, { "axis": "oracle_queries_to_99pct_success_n8_vs_classical_mean", "winner": "quantum", "higher_is_better": false, "quantum": { "n": 10, "mean": 12.0, "std": 0.0, "median": 12.0, "min": 12.0, "max": 12.0, "ci95_lo": 12.0, "ci95_hi": 12.0 }, "classical": { "n": 10, "mean": 128.22629999999998, "std": 2.744112730515596, "median": 127.69200000000001, "min": 122.97, "max": 132.678, "ci95_lo": 126.54443007853797, "ci95_hi": 129.78289999999998 }, "p_corrected": 0.03125, "effect_size_hl": -115.824, "claim_eligible": true, "note": "quantum = r* x k* oracle calls to >=99% success (both factors in per_seed); classical comparator here is MEAN probes of a measured seeded scan \u2014 note the confidence targets differ on this axis (99% vs expected-case); see the _vs_classical_p99 axis for the matched-confidence comparison", "replication_sweep_id": "cli-full-20260611T235654Z" }, { "axis": "oracle_queries_to_99pct_success_n8_vs_classical_p99", "winner": "quantum", "higher_is_better": false, "quantum": { "n": 10, "mean": 12.0, "std": 0.0, "median": 12.0, "min": 12.0, "max": 12.0, "ci95_lo": 12.0, "ci95_hi": 12.0 }, "classical": { "n": 10, "mean": 253.8, "std": 0.4216370213557839, "median": 254.0, "min": 253.0, "max": 254.0, "ci95_lo": 253.5, "ci95_hi": 254.0 }, "p_corrected": 0.03125, "effect_size_hl": -242.0, "claim_eligible": true, "note": "matched 99%-confidence comparison: classical comparator is the MEASURED 99th percentile of probes (queries to have found the item with probability 0.99 under a no-replacement scan)", "replication_sweep_id": "cli-full-20260611T235654Z" }, { "axis": "wall_clock_ms_n8", "winner": "classical", "higher_is_better": false, "quantum": { "n": 10, "mean": 90.218, "std": 2.7651554105410465, "median": 89.99000000000001, "min": 86.37, "max": 94.07, "ci95_lo": 88.59700000000001, "ci95_hi": 91.81629508243125 }, "classical": { "n": 10, "mean": 4.27691, "std": 0.33702181086689337, "median": 4.1278, "min": 4.0346, "max": 5.0755, "ci95_lo": 4.128100184852501, "ci95_hi": 4.549783886966027 }, "p_corrected": 0.0, "effect_size_hl": 85.94300000000001, "claim_eligible": true, "note": "CPU-simulator overhead; expected and confirmed classical win", "replication_sweep_id": "cli-full-20260611T235654Z" }, { "axis": "oracle_queries_to_99pct_success_n9_vs_classical_mean", "winner": "quantum", "higher_is_better": false, "quantum": { "n": 10, "mean": 17.0, "std": 0.0, "median": 17.0, "min": 17.0, "max": 17.0, "ci95_lo": 17.0, "ci95_hi": 17.0 }, "classical": { "n": 10, "mean": 255.3457, "std": 3.8066720096284747, "median": 254.6655, "min": 250.378, "max": 262.313, "ci95_lo": 253.3743815266437, "ci95_hi": 257.9382487258613 }, "p_corrected": 0.03125, "effect_size_hl": -238.086, "claim_eligible": true, "note": "quantum = r* x k* oracle calls to >=99% success (both factors in per_seed); classical comparator here is MEAN probes of a measured seeded scan \u2014 note the confidence targets differ on this axis (99% vs expected-case); see the _vs_classical_p99 axis for the matched-confidence comparison", "replication_sweep_id": "cli-full-20260611T235654Z" }, { "axis": "oracle_queries_to_99pct_success_n9_vs_classical_p99", "winner": "quantum", "higher_is_better": false, "quantum": { "n": 10, "mean": 17.0, "std": 0.0, "median": 17.0, "min": 17.0, "max": 17.0, "ci95_lo": 17.0, "ci95_hi": 17.0 }, "classical": { "n": 10, "mean": 506.804, "std": 1.873091324818711, "median": 507.01, "min": 503.01, "max": 509.01, "ci95_lo": 505.505, "ci95_hi": 507.802 }, "p_corrected": 0.03125, "effect_size_hl": -490.0, "claim_eligible": true, "note": "matched 99%-confidence comparison: classical comparator is the MEASURED 99th percentile of probes (queries to have found the item with probability 0.99 under a no-replacement scan)", "replication_sweep_id": "cli-full-20260611T235654Z" }, { "axis": "wall_clock_ms_n9", "winner": "classical", "higher_is_better": false, "quantum": { "n": 10, "mean": 143.414, "std": 4.631064672405255, "median": 143.245, "min": 136.91, "max": 152.73, "ci95_lo": 141.02399999999997, "ci95_hi": 146.42811853162186 }, "classical": { "n": 10, "mean": 6.588390000000001, "std": 0.19216609973897283, "median": 6.54875, "min": 6.4296, "max": 7.0863, "ci95_lo": 6.512229339081973, "ci95_hi": 6.76947206006912 }, "p_corrected": 0.0, "effect_size_hl": 136.44164999999998, "claim_eligible": true, "note": "CPU-simulator overhead; expected and confirmed classical win", "replication_sweep_id": "cli-full-20260611T235654Z" }, { "axis": "oracle_queries_to_99pct_success_n10_vs_classical_mean", "winner": "quantum", "higher_is_better": false, "quantum": { "n": 10, "mean": 24.9, "std": 0.3162277660168379, "median": 25.0, "min": 24.0, "max": 25.0, "ci95_lo": 24.5, "ci95_hi": 25.0 }, "classical": { "n": 10, "mean": 508.6540999999999, "std": 6.592268290286078, "median": 508.3705, "min": 497.462, "max": 518.796, "ci95_lo": 504.6412769428197, "ci95_hi": 512.5326084849071 }, "p_corrected": 0.03125, "effect_size_hl": -483.74800000000005, "claim_eligible": true, "note": "quantum = r* x k* oracle calls to >=99% success (both factors in per_seed); classical comparator here is MEAN probes of a measured seeded scan \u2014 note the confidence targets differ on this axis (99% vs expected-case); see the _vs_classical_p99 axis for the matched-confidence comparison", "replication_sweep_id": "cli-full-20260611T235654Z" }, { "axis": "oracle_queries_to_99pct_success_n10_vs_classical_p99", "winner": "quantum", "higher_is_better": false, "quantum": { "n": 10, "mean": 24.9, "std": 0.3162277660168379, "median": 25.0, "min": 24.0, "max": 25.0, "ci95_lo": 24.5, "ci95_hi": 25.0 }, "classical": { "n": 10, "mean": 1012.1120000000001, "std": 5.3376769811928115, "median": 1013.02, "min": 999.02, "max": 1018.0, "ci95_lo": 1007.5150000000001, "ci95_hi": 1014.408 }, "p_corrected": 0.03125, "effect_size_hl": -988.015, "claim_eligible": true, "note": "matched 99%-confidence comparison: classical comparator is the MEASURED 99th percentile of probes (queries to have found the item with probability 0.99 under a no-replacement scan)", "replication_sweep_id": "cli-full-20260611T235654Z" }, { "axis": "wall_clock_ms_n10", "winner": "classical", "higher_is_better": false, "quantum": { "n": 10, "mean": 252.38400000000001, "std": 12.952696158629584, "median": 251.27, "min": 234.99, "max": 270.83, "ci95_lo": 245.19924961284732, "ci95_hi": 260.3615268585233 }, "classical": { "n": 10, "mean": 12.58017, "std": 3.0533745659697753, "median": 11.46855, "min": 11.2558, "max": 21.1991, "ci95_lo": 11.53580241332227, "ci95_hi": 16.35062549973749 }, "p_corrected": 0.0, "effect_size_hl": 239.5534, "claim_eligible": true, "note": "CPU-simulator overhead; expected and confirmed classical win", "replication_sweep_id": "cli-full-20260611T235654Z" } ], "ok": true, "receipt_self_hash": "9d4a418739b56085bcbc6a8a5cf82e2941714789f00cac599e96e2ecb3f9c35e", "limitations": [] }, "mlae_vs_mc": { "verdicts": [ { "axis": "queries_to_eps_0.01_truth_free_a0.05", "winner": "classical", "higher_is_better": false, "quantum": { "n": 20, "mean": 4096.0, "std": 0.0, "median": 4096.0, "min": 4096.0, "max": 4096.0, "ci95_lo": 4096.0, "ci95_hi": 4096.0 }, "classical": { "n": 20, "mean": 1850.0, "std": 127.73327473170102, "median": 1800.0, "min": 1600.0, "max": 2100.0, "ci95_lo": 1795.0, "ci95_hi": 1905.0 }, "p_corrected": 1.1444091796875e-05, "effect_size_hl": 2246.0, "claim_eligible": true, "note": "quantum: oracle applications (sum shots_k*(2k+1)) at Fisher-CI crossing; classical: Bernoulli samples at Wilson-CI crossing", "replication_sweep_id": "cli-full-20260611T235654Z" }, { "axis": "queries_to_eps_0.01_truth_free_a0.1", "winner": "classical", "higher_is_better": false, "quantum": { "n": 20, "mean": 4096.0, "std": 0.0, "median": 4096.0, "min": 4096.0, "max": 4096.0, "ci95_lo": 4096.0, "ci95_hi": 4096.0 }, "classical": { "n": 20, "mean": 3535.0, "std": 166.30662866176473, "median": 3550.0, "min": 3100.0, "max": 3900.0, "ci95_lo": 3460.0, "ci95_hi": 3605.0 }, "p_corrected": 1.1444091796875e-05, "effect_size_hl": 546.0, "claim_eligible": true, "note": "quantum: oracle applications (sum shots_k*(2k+1)) at Fisher-CI crossing; classical: Bernoulli samples at Wilson-CI crossing", "replication_sweep_id": "cli-full-20260611T235654Z" }, { "axis": "queries_to_eps_0.01_truth_free_a0.25", "winner": "quantum", "higher_is_better": false, "quantum": { "n": 20, "mean": 4096.0, "std": 0.0, "median": 4096.0, "min": 4096.0, "max": 4096.0, "ci95_lo": 4096.0, "ci95_hi": 4096.0 }, "classical": { "n": 20, "mean": 7240.0, "std": 118.7655806953123, "median": 7300.0, "min": 7000.0, "max": 7400.0, "ci95_lo": 7185.0, "ci95_hi": 7285.0 }, "p_corrected": 1.1444091796875e-05, "effect_size_hl": -3154.0, "claim_eligible": true, "note": "quantum: oracle applications (sum shots_k*(2k+1)) at Fisher-CI crossing; classical: Bernoulli samples at Wilson-CI crossing", "replication_sweep_id": "cli-full-20260611T235654Z" }, { "axis": "queries_to_eps_0.01_truth_free_a0.5", "winner": "quantum", "higher_is_better": false, "quantum": { "n": 20, "mean": 4096.0, "std": 0.0, "median": 4096.0, "min": 4096.0, "max": 4096.0, "ci95_lo": 4096.0, "ci95_hi": 4096.0 }, "classical": { "n": 20, "mean": 9600.0, "std": 0.0, "median": 9600.0, "min": 9600.0, "max": 9600.0, "ci95_lo": 9600.0, "ci95_hi": 9600.0 }, "p_corrected": 1.1444091796875e-05, "effect_size_hl": -5504.0, "claim_eligible": true, "note": "quantum: oracle applications (sum shots_k*(2k+1)) at Fisher-CI crossing; classical: Bernoulli samples at Wilson-CI crossing", "replication_sweep_id": "cli-full-20260611T235654Z" }, { "axis": "queries_to_eps_0.01_truth_free_a0.75", "winner": "quantum", "higher_is_better": false, "quantum": { "n": 20, "mean": 4096.0, "std": 0.0, "median": 4096.0, "min": 4096.0, "max": 4096.0, "ci95_lo": 4096.0, "ci95_hi": 4096.0 }, "classical": { "n": 20, "mean": 7215.0, "std": 93.33020044867295, "median": 7200.0, "min": 7000.0, "max": 7400.0, "ci95_lo": 7175.0, "ci95_hi": 7255.0 }, "p_corrected": 1.1444091796875e-05, "effect_size_hl": -3104.0, "claim_eligible": true, "note": "quantum: oracle applications (sum shots_k*(2k+1)) at Fisher-CI crossing; classical: Bernoulli samples at Wilson-CI crossing", "replication_sweep_id": "cli-full-20260611T235654Z" }, { "axis": "queries_to_eps_0.01_truth_free_a0.9", "winner": "classical", "higher_is_better": false, "quantum": { "n": 20, "mean": 4096.0, "std": 0.0, "median": 4096.0, "min": 4096.0, "max": 4096.0, "ci95_lo": 4096.0, "ci95_hi": 4096.0 }, "classical": { "n": 20, "mean": 3530.0, "std": 130.18205875255097, "median": 3500.0, "min": 3300.0, "max": 3800.0, "ci95_lo": 3475.0, "ci95_hi": 3585.0 }, "p_corrected": 1.1444091796875e-05, "effect_size_hl": 546.0, "claim_eligible": true, "note": "quantum: oracle applications (sum shots_k*(2k+1)) at Fisher-CI crossing; classical: Bernoulli samples at Wilson-CI crossing", "replication_sweep_id": "cli-full-20260611T235654Z" }, { "axis": "wall_clock_ms_a0.05", "winner": "classical", "higher_is_better": false, "quantum": { "n": 20, "mean": 4.614035000486183, "std": 0.2437952910439195, "median": 4.577700001391349, "min": 4.283300004317425, "max": 5.218900005274918, "ci95_lo": 4.519610157401761, "ci95_hi": 4.733815332704407 }, "classical": { "n": 20, "mean": 0.465349998921738, "std": 0.019334902136621264, "median": 0.45930000123917125, "min": 0.43429999641375616, "max": 0.5201000021770597, "ci95_lo": 0.45845499847463983, "ci95_hi": 0.47559431813759023 }, "p_corrected": 0.0, "effect_size_hl": 4.128725002374267, "claim_eligible": true, "note": "CPU-simulator overhead; expected classical win, reported anyway", "replication_sweep_id": "cli-full-20260611T235654Z" }, { "axis": "wall_clock_ms_a0.1", "winner": "classical", "higher_is_better": false, "quantum": { "n": 20, "mean": 4.611220000515459, "std": 0.35098072359683735, "median": 4.52210000003106, "min": 4.227000004902948, "max": 5.563700004131533, "ci95_lo": 4.490244661231304, "ci95_hi": 4.80676555655707 }, "classical": { "n": 20, "mean": 0.9123499992711004, "std": 0.0922095405453134, "median": 0.8787999977357686, "min": 0.8432999966316856, "max": 1.2473999959183857, "ci95_lo": 0.8852163199026067, "ci95_hi": 0.974924725143805 }, "p_corrected": 0.0, "effect_size_hl": 3.6397249987203395, "claim_eligible": true, "note": "CPU-simulator overhead; expected classical win, reported anyway", "replication_sweep_id": "cli-full-20260611T235654Z" }, { "axis": "wall_clock_ms_a0.25", "winner": "classical", "higher_is_better": false, "quantum": { "n": 20, "mean": 4.718494999906397, "std": 0.333020725928932, "median": 4.766399997606641, "min": 4.196900001261383, "max": 5.461000000650529, "ci95_lo": 4.5839162773233335, "ci95_hi": 4.87118431190152 }, "classical": { "n": 20, "mean": 1.89710999984527, "std": 0.046997557114686686, "median": 1.875450001534773, "min": 1.8416000020806678, "max": 1.9921999992220663, "ci95_lo": 1.8793337916178827, "ci95_hi": 1.9199882469991492 }, "p_corrected": 0.0, "effect_size_hl": 2.810299998600385, "claim_eligible": true, "note": "CPU-simulator overhead; expected classical win, reported anyway", "replication_sweep_id": "cli-full-20260611T235654Z" }, { "axis": "wall_clock_ms_a0.5", "winner": "classical", "higher_is_better": false, "quantum": { "n": 20, "mean": 4.683170000498649, "std": 0.2687089081364418, "median": 4.6446500018646475, "min": 4.321000000345521, "max": 5.31520000367891, "ci95_lo": 4.578332501242958, "ci95_hi": 4.814407389532252 }, "classical": { "n": 20, "mean": 2.5008149994391715, "std": 0.1302185454738703, "median": 2.4621499978820793, "min": 2.3603000008733943, "max": 2.792899998894427, "ci95_lo": 2.4514918518473006, "ci95_hi": 2.565301824313264 }, "p_corrected": 0.0, "effect_size_hl": 2.1565750012086937, "claim_eligible": true, "note": "CPU-simulator overhead; expected classical win, reported anyway", "replication_sweep_id": "cli-full-20260611T235654Z" }, { "axis": "wall_clock_ms_a0.75", "winner": "classical", "higher_is_better": false, "quantum": { "n": 20, "mean": 4.858479999893461, "std": 0.2953346220330347, "median": 4.82315000044764, "min": 4.4572000042535365, "max": 5.410200006735977, "ci95_lo": 4.734048687529844, "ci95_hi": 4.989747510950267 }, "classical": { "n": 20, "mean": 1.9450249990768498, "std": 0.1718134209618123, "median": 1.8702000015764497, "min": 1.8194999938714318, "max": 2.4230999988503754, "ci95_lo": 1.889169062982287, "ci95_hi": 2.0504553490407624 }, "p_corrected": 0.0, "effect_size_hl": 2.9099250059516635, "claim_eligible": true, "note": "CPU-simulator overhead; expected classical win, reported anyway", "replication_sweep_id": "cli-full-20260611T235654Z" }, { "axis": "wall_clock_ms_a0.9", "winner": "classical", "higher_is_better": false, "quantum": { "n": 20, "mean": 4.756090000228141, "std": 0.5493511203431295, "median": 4.660499998863088, "min": 4.1482000015093945, "max": 6.337599996186327, "ci95_lo": 4.57101788083286, "ci95_hi": 5.063958042911702 }, "classical": { "n": 20, "mean": 1.1282650000794092, "std": 1.126089567398329, "median": 0.8608000025560614, "min": 0.8328999974764884, "max": 5.905500001972541, "ci95_lo": 0.8679694520442058, "ci95_hi": 2.1433796223985455 }, "p_corrected": 0.0, "effect_size_hl": 3.708399999595713, "claim_eligible": true, "note": "CPU-simulator overhead; expected classical win, reported anyway", "replication_sweep_id": "cli-full-20260611T235654Z" }, { "axis": "scaling_exponent_queries_vs_inv_eps", "winner": "informational", "quantum_exponent": { "mean": 1.0121218670057879, "ci95_lo": 1.0021539563920148, "ci95_hi": 1.0214555784960437, "n": 120 }, "classical_exponent": { "mean": 1.9438067876659313, "ci95_lo": 1.9312692849818762, "ci95_hi": 1.9548798351399703, "n": 120 }, "quantum_fit_r2": { "n": 120, "mean": 0.9303456383797915, "std": 0.02267953332420668, "median": 0.9362968308688856, "min": 0.8868507948392621, "max": 0.9500357119639445, "ci95_lo": 0.9259641675081458, "ci95_hi": 0.9340409850427276 }, "classical_fit_r2": { "n": 120, "mean": 0.9987172655696492, "std": 0.0020197745521664443, "median": 0.9992746482277441, "min": 0.9890232349405974, "max": 0.9999772057413556, "ci95_lo": 0.9982773253720566, "ci95_hi": 0.9990100817080924 }, "fit_caveat": "exponents are fitted on few points with prefix-granular quantum query counts; R2 aggregates above quantify fit quality per rep \u2014 treat exponents as descriptive, not claim-bearing", "positive_control_gate": "classical exponent must lie in [1.7, 2.3]", "positive_control_pass": true, "claim_eligible": false, "note": "theoretical: MLAE ~1/eps (finite-schedule estimates run above 1), classical MC ~1/eps^2" } ], "ok": true, "receipt_self_hash": "6bb2bda85f3f3d1e6693669b714301ea9dff247ee701f1c6ecf316fe17ceafe4", "limitations": [] }, "mps_memory": { "verdicts": [ { "axis": "peak_memory_brickwork_n16", "winner": "below_measurement_floor", "comparison": "mps_vs_dense_statevector (method-vs-method, both classical simulators)", "ratio_dense_over_mps": null, "ratio_lower_bound": null, "ratio_provenance": "below_floor_no_verdict", "fidelity_verified": true, "fidelity_tvd_independent_seeds": 0.824219, "fidelity_selfnoise_tvd": 0.833008, "fidelity_marginal_max_dev": 0.061523, "fidelity_pass": true, "claim_eligible": false }, { "axis": "peak_memory_brickwork_n20", "winner": "below_measurement_floor", "comparison": "mps_vs_dense_statevector (method-vs-method, both classical simulators)", "ratio_dense_over_mps": null, "ratio_lower_bound": null, "ratio_provenance": "below_floor_no_verdict", "fidelity_verified": true, "fidelity_tvd_independent_seeds": 0.93457, "fidelity_selfnoise_tvd": 0.945312, "fidelity_marginal_max_dev": 0.054688, "fidelity_pass": true, "claim_eligible": false }, { "axis": "peak_memory_brickwork_n24", "winner": "mps", "comparison": "mps_vs_dense_statevector (method-vs-method, both classical simulators)", "ratio_dense_over_mps": null, "ratio_lower_bound": 8.0, "ratio_provenance": "measured_lower_bound_mps_below_floor", "fidelity_verified": true, "fidelity_tvd_independent_seeds": 0.995117, "fidelity_selfnoise_tvd": 0.990234, "fidelity_marginal_max_dev": 0.038086, "fidelity_pass": true, "claim_eligible": true, "replication_sweep_id": "cli-full-20260611T235654Z" }, { "axis": "peak_memory_brickwork_n26", "winner": "mps", "comparison": "mps_vs_dense_statevector (method-vs-method, both classical simulators)", "ratio_dense_over_mps": null, "ratio_lower_bound": 32.0, "ratio_provenance": "measured_lower_bound_mps_below_floor", "fidelity_verified": true, "fidelity_tvd_independent_seeds": 0.99707, "fidelity_selfnoise_tvd": 0.995117, "fidelity_marginal_max_dev": 0.03418, "fidelity_pass": true, "claim_eligible": true, "replication_sweep_id": "cli-full-20260611T235654Z" }, { "axis": "peak_memory_brickwork_n28", "winner": "mps", "comparison": "mps_vs_dense_statevector (method-vs-method, both classical simulators)", "ratio_dense_over_mps": null, "ratio_lower_bound": 127.9, "ratio_provenance": "measured_lower_bound_mps_below_floor", "fidelity_verified": true, "fidelity_tvd_independent_seeds": 0.998047, "fidelity_selfnoise_tvd": 0.999023, "fidelity_marginal_max_dev": 0.039062, "fidelity_pass": true, "claim_eligible": true, "replication_sweep_id": "cli-full-20260611T235654Z" }, { "axis": "peak_memory_alltoall_n16", "winner": "below_measurement_floor", "comparison": "mps_vs_dense_statevector (method-vs-method, both classical simulators)", "ratio_dense_over_mps": null, "ratio_lower_bound": null, "ratio_provenance": "below_floor_no_verdict", "fidelity_verified": true, "fidelity_tvd_independent_seeds": 0.873047, "fidelity_selfnoise_tvd": 0.848633, "fidelity_marginal_max_dev": 0.041016, "fidelity_pass": true, "claim_eligible": false }, { "axis": "peak_memory_alltoall_n20", "winner": "dense", "comparison": "mps_vs_dense_statevector (method-vs-method, both classical simulators)", "ratio_dense_over_mps": 0.102, "ratio_lower_bound": null, "ratio_provenance": "measured", "fidelity_verified": true, "fidelity_tvd_independent_seeds": 0.986328, "fidelity_selfnoise_tvd": 0.993164, "fidelity_marginal_max_dev": 0.043945, "fidelity_pass": true, "claim_eligible": false }, { "axis": "peak_memory_trotter_zz_n16", "winner": "below_measurement_floor", "comparison": "mps_vs_dense_statevector (method-vs-method, both classical simulators)", "ratio_dense_over_mps": null, "ratio_lower_bound": null, "ratio_provenance": "below_floor_no_verdict", "fidelity_verified": true, "fidelity_tvd_independent_seeds": 0.802734, "fidelity_selfnoise_tvd": 0.783203, "fidelity_marginal_max_dev": 0.057617, "fidelity_pass": true, "claim_eligible": false }, { "axis": "peak_memory_trotter_zz_n20", "winner": "below_measurement_floor", "comparison": "mps_vs_dense_statevector (method-vs-method, both classical simulators)", "ratio_dense_over_mps": null, "ratio_lower_bound": null, "ratio_provenance": "below_floor_no_verdict", "fidelity_verified": true, "fidelity_tvd_independent_seeds": 0.922852, "fidelity_selfnoise_tvd": 0.933594, "fidelity_marginal_max_dev": 0.052734, "fidelity_pass": true, "claim_eligible": false }, { "axis": "peak_memory_trotter_zz_n24", "winner": "mps", "comparison": "mps_vs_dense_statevector (method-vs-method, both classical simulators)", "ratio_dense_over_mps": null, "ratio_lower_bound": 8.0, "ratio_provenance": "measured_lower_bound_mps_below_floor", "fidelity_verified": true, "fidelity_tvd_independent_seeds": 0.992188, "fidelity_selfnoise_tvd": 0.984375, "fidelity_marginal_max_dev": 0.045898, "fidelity_pass": true, "claim_eligible": true, "replication_sweep_id": "cli-full-20260611T235654Z" }, { "axis": "peak_memory_trotter_zz_n26", "winner": "mps", "comparison": "mps_vs_dense_statevector (method-vs-method, both classical simulators)", "ratio_dense_over_mps": null, "ratio_lower_bound": 32.0, "ratio_provenance": "measured_lower_bound_mps_below_floor", "fidelity_verified": true, "fidelity_tvd_independent_seeds": 0.996094, "fidelity_selfnoise_tvd": 0.996094, "fidelity_marginal_max_dev": 0.048828, "fidelity_pass": true, "claim_eligible": true, "replication_sweep_id": "cli-full-20260611T235654Z" }, { "axis": "peak_memory_trotter_zz_n28", "winner": "mps", "comparison": "mps_vs_dense_statevector (method-vs-method, both classical simulators)", "ratio_dense_over_mps": null, "ratio_lower_bound": 128.0, "ratio_provenance": "measured_lower_bound_mps_below_floor", "fidelity_verified": true, "fidelity_tvd_independent_seeds": 0.99707, "fidelity_selfnoise_tvd": 0.99707, "fidelity_marginal_max_dev": 0.041992, "fidelity_pass": true, "claim_eligible": true, "replication_sweep_id": "cli-full-20260611T235654Z" }, { "axis": "dense_allocation_limit", "winner": "informational", "result": { "n": 32, "result": "probe_error_not_oom", "error": "subprocess failed rc=1: y\", line 385, in _get_experiment\n raise QiskitError(result_status, \", \", exp_status)\nqiskit.exceptions.QiskitError: 'ERROR: [Experiment 0] Insufficient memory to run circuit circuit-46 using the statevector simulator. Required memory: 65536M, max memory: 64881M , ERROR: " }, "claim_eligible": false } ], "ok": true, "receipt_self_hash": "24581a46d230ad7a699f928fb7d7901e1147b174caba90a600c8e711f7df3950", "limitations": [ "alltoall/n24: skipped \u2014 exact (untruncated) MPS on this volume-law family exceeds tractable bond growth beyond n=20; analytically MPS memory n*chi^2*16 with chi->2^(n/2) meets or exceeds dense 2^n*16 here, i.e. MPS has no memory advantage on this family", "alltoall/n26: skipped \u2014 exact (untruncated) MPS on this volume-law family exceeds tractable bond growth beyond n=20; analytically MPS memory n*chi^2*16 with chi->2^(n/2) meets or exceeds dense 2^n*16 here, i.e. MPS has no memory advantage on this family", "alltoall/n28: skipped \u2014 exact (untruncated) MPS on this volume-law family exceeds tractable bond growth beyond n=20; analytically MPS memory n*chi^2*16 with chi->2^(n/2) meets or exceeds dense 2^n*16 here, i.e. MPS has no memory advantage on this family" ] }, "qsvm_vs_classical": { "verdicts": [ { "axis": "test_accuracy_gauss_mixture", "winner": "classical", "higher_is_better": true, "quantum": { "n": 10, "mean": 0.596875, "std": 0.1288049370387469, "median": 0.59375, "min": 0.375, "max": 0.8125, "ci95_lo": 0.515625, "ci95_hi": 0.665625 }, "classical": { "n": 10, "mean": 0.81875, "std": 0.06718548123582126, "median": 0.8125, "min": 0.71875, "max": 0.9375, "ci95_lo": 0.78125, "ci95_hi": 0.8625 }, "p_corrected": 0.005859375, "effect_size_hl": -0.21875, "claim_eligible": true, "note": "quantum = exact ZZ fidelity kernel + SVC(precomputed); classical = best of SVC-RBF and HistGradientBoosting per seed", "replication_sweep_id": "cli-full-20260611T235654Z" }, { "axis": "test_accuracy_make_classification", "winner": "classical", "higher_is_better": true, "quantum": { "n": 10, "mean": 0.7375, "std": 0.09340213238108289, "median": 0.765625, "min": 0.59375, "max": 0.84375, "ci95_lo": 0.675, "ci95_hi": 0.784375 }, "classical": { "n": 10, "mean": 0.878125, "std": 0.0579039254570304, "median": 0.875, "min": 0.78125, "max": 0.96875, "ci95_lo": 0.846875, "ci95_hi": 0.915625 }, "p_corrected": 0.005859375, "effect_size_hl": -0.140625, "claim_eligible": true, "note": "quantum = exact ZZ fidelity kernel + SVC(precomputed); classical = best of SVC-RBF and HistGradientBoosting per seed", "replication_sweep_id": "cli-full-20260611T235654Z" }, { "axis": "test_accuracy_linear_control", "winner": "classical", "higher_is_better": true, "quantum": { "n": 10, "mean": 0.53125, "std": 0.09991315673568166, "median": 0.5, "min": 0.4375, "max": 0.75, "ci95_lo": 0.48749999999999993, "ci95_hi": 0.609375 }, "classical": { "n": 10, "mean": 0.88125, "std": 0.08690272339422588, "median": 0.90625, "min": 0.6875, "max": 0.96875, "ci95_lo": 0.809375, "ci95_hi": 0.9197844188985186 }, "p_corrected": 0.005859375, "effect_size_hl": -0.34375, "claim_eligible": true, "note": "quantum = exact ZZ fidelity kernel + SVC(precomputed); classical = best of SVC-RBF and HistGradientBoosting per seed", "replication_sweep_id": "cli-full-20260611T235654Z" }, { "axis": "wall_clock_ms_gauss_mixture", "winner": "informational_simulator_artifact", "higher_is_better": false, "quantum": { "n": 10, "mean": 75.30503000016324, "std": 16.97205523810595, "median": 70.36460000017541, "min": 63.62729999818839, "max": 122.06139999761945, "ci95_lo": 69.08991837872534, "ci95_hi": 95.87754979313704 }, "classical": { "n": 10, "mean": 240.13150999890058, "std": 9.104651701775268, "median": 239.22979999770178, "min": 229.77030000038212, "max": 260.67609999881824, "ci95_lo": 235.8121475753009, "ci95_hi": 247.15072438014295 }, "p_corrected": null, "effect_size_hl": -166.91625000021304, "claim_eligible": false, "note": "INFORMATIONAL ONLY \u2014 each arm timed including its own hyperparameter search; classical side is BOTH baselines combined. At this tiny scale (n_train<=48, d=4) the exact-statevector kernel is cheap; that does NOT generalize (exact kernel cost scales O(n^2 * 2^d) and shot-based kernels are far slower), so no winner is declared on this axis" }, { "axis": "geometric_difference_gauss_mixture", "winner": "informational", "g_aggregate": { "n": 10, "mean": 2.908870653186189, "std": 0.25990753303450614, "median": 2.992187186821765, "min": 2.3944423183691295, "max": 3.21633603910428, "ci95_lo": 2.716578997540391, "ci95_hi": 3.035073049914563 }, "interpretation": "Huang et al. (Power of Data): small g => the classical RBF kernel can match the quantum kernel on this data \u2014 no prospective quantum-kernel advantage exists here, as expected for d=4 classically simulable features", "label_shuffle_null": { "quantum": 0.59375, "svc_rbf": 0.34375, "hist_gradient_boosting": 0.59375 }, "label_shuffle_null_pass": true, "claim_eligible": false }, { "axis": "wall_clock_ms_make_classification", "winner": "informational_simulator_artifact", "higher_is_better": false, "quantum": { "n": 10, "mean": 70.30625000043074, "std": 3.443583765816283, "median": 69.22839999970165, "min": 66.66149999364279, "max": 75.29790000262437, "ci95_lo": 68.44170531272572, "ci95_hi": 72.52805869562597 }, "classical": { "n": 10, "mean": 238.15288999976474, "std": 6.109790792724284, "median": 239.11109999971814, "min": 229.68009999749484, "max": 247.78040000092005, "ci95_lo": 234.43624790128615, "ci95_hi": 241.69145547876536 }, "p_corrected": null, "effect_size_hl": -167.09755000192672, "claim_eligible": false, "note": "INFORMATIONAL ONLY \u2014 each arm timed including its own hyperparameter search; classical side is BOTH baselines combined. At this tiny scale (n_train<=48, d=4) the exact-statevector kernel is cheap; that does NOT generalize (exact kernel cost scales O(n^2 * 2^d) and shot-based kernels are far slower), so no winner is declared on this axis" }, { "axis": "geometric_difference_make_classification", "winner": "informational", "g_aggregate": { "n": 10, "mean": 3.1069864187424514, "std": 0.4665358694130003, "median": 3.070462661153443, "min": 2.4567751907067263, "max": 4.035899144722595, "ci95_lo": 2.851966078034019, "ci95_hi": 3.4073413404301385 }, "interpretation": "Huang et al. (Power of Data): small g => the classical RBF kernel can match the quantum kernel on this data \u2014 no prospective quantum-kernel advantage exists here, as expected for d=4 classically simulable features", "label_shuffle_null": { "quantum": 0.5, "svc_rbf": 0.4375, "hist_gradient_boosting": 0.40625 }, "label_shuffle_null_pass": true, "claim_eligible": false }, { "axis": "wall_clock_ms_linear_control", "winner": "informational_simulator_artifact", "higher_is_better": false, "quantum": { "n": 10, "mean": 69.62100999953691, "std": 2.8830131072730443, "median": 68.49194999813335, "min": 65.89420000091195, "max": 75.76699999481207, "ci95_lo": 68.23526016058437, "ci95_hi": 71.72177999891574 }, "classical": { "n": 10, "mean": 248.94729999941774, "std": 11.369893921360449, "median": 246.04919999546837, "min": 234.25529999803985, "max": 271.8611000018427, "ci95_lo": 243.42270758160058, "ci95_hi": 257.27363721029974 }, "p_corrected": null, "effect_size_hl": -177.00199999671895, "claim_eligible": false, "note": "INFORMATIONAL ONLY \u2014 each arm timed including its own hyperparameter search; classical side is BOTH baselines combined. At this tiny scale (n_train<=48, d=4) the exact-statevector kernel is cheap; that does NOT generalize (exact kernel cost scales O(n^2 * 2^d) and shot-based kernels are far slower), so no winner is declared on this axis" }, { "axis": "geometric_difference_linear_control", "winner": "informational", "g_aggregate": { "n": 10, "mean": 3.0476436224092667, "std": 0.21148220195714043, "median": 3.049603542366593, "min": 2.655206118891542, "max": 3.509561743183376, "ci95_lo": 2.932973729819645, "ci95_hi": 3.1854787871585466 }, "interpretation": "Huang et al. (Power of Data): small g => the classical RBF kernel can match the quantum kernel on this data \u2014 no prospective quantum-kernel advantage exists here, as expected for d=4 classically simulable features", "label_shuffle_null": { "quantum": 0.4375, "svc_rbf": 0.34375, "hist_gradient_boosting": 0.5625 }, "label_shuffle_null_pass": true, "claim_eligible": false } ], "ok": true, "receipt_self_hash": "03f91137ff4ce7d585035d9498344b1eae7342ec3b18fb3a9518d7c1772fb68a", "limitations": [] } }, "supersession": "These receipts supersede and retire the earlier T2/T3 hash-proxy and formula-derived 'advantage' receipts, which are non-citable.", "replication": { "verdict_self_hash": "d5f02212685c42a0b71982814bbe39b57aaa4c29e7e44be93518457cb928f946", "sweep_b": "cli-full-20260611T235654Z", "claim_eligible_axes": 45, "note": "the alltoall_n20 memory cell did not replicate: the replication sweep drew a different random volume-law instance whose exact-MPS simulation exceeded the per-cell timeout (failure recorded in its receipt); that axis is therefore NOT claim-eligible" } }