K301a |
Quantinuum Nexus Helios-1E-lite |
5-step Krump interference ladder (1,280 emulator shots; 0 physical QPU shots) |
P_cons = 0.9961, P_dest = 1.0000, R² = 0.9942 |
JSON d43f26d9 |
K301b |
Quantinuum Nexus Helios-1E-lite |
5-layer Frobenius entangled duet Bell uncompute (1,024 emulator shots) |
1,018 / 1,024 in |01⟩ (99.41% fidelity) |
JSON a411dac7 |
K302 |
DisCoPy v1.2.1 (FHilb) |
Snake identity, swap involution, bifunctorial interchange & CHSH violation |
Exact tensor norm diff < 10⁻¹⁵; S_CHSH = 2.8284 |
JSON Verified |
K303 |
Modal Cloud Quantum Scientist |
4–10 qubit barren plateaus, entanglement entropy & 12q Quantum Fisher Info |
5.9× grad variance retention; I_Q = 40.0 (3.33× SQL) |
JSON Receipt |
K304 |
Modal Cloud + Hugging Face Hub |
QKMLM 4-qubit tri-head VQC (2,000 train / 500 test synthetic benchmark) |
98.60% foundation acc; 89.40% phenotype acc |
JSON Receipt |
K305 |
Apple MPS + Modal Cloud PyTorch |
Motion2MixamoRig (2,953 frames @ 29.99 fps) + 1.01M-param KrumpRobot causal GRU |
8.765×10⁻¹¹ FK; 2.32° in-seq error; 2.70 ms latency |
JSON Receipt |
K306 |
OpenJij Simulated Quantum Annealing |
Act III Bell-pair stage routing QUBO + Haiqu circuit compression |
Ground energy E₀ = −7.3569 (0 vertex collisions); 85.7% depth cut |
Monograph §7 |
K307 |
TypeSafe Jev 1.14 Claim Checklist |
Automated schema-constrained anti-hype & classical-floor prose audit |
0.33 µs 1q / 1.74 µs 4q classical floors disclosed |
JSON Audit |
K308 |
PyTorch + scikit-learn Ablation Suite |
QKMLM classical bottleneck ablations & KrumpRobot 80%/20% temporal block split |
Classical MLP 99.80% (0.04 ms); temporal split 1.42° / 18.58° |
JSON Ablation |
K309 |
Universal Knowledge Graph (JSON-LD) |
10-domain ontological schema connecting pedagogy, DisCoPy, Helios, KrumpRobot & MuJoCo WBC |
92 semantic nodes · 156 directed relations · 0 dangling edges (v2.0.0) |
JSON-LD v2.0 |
K310 |
DisCoPy 1.2 + SymPy 1.14 + pytket 2.18 |
Functorial Krump theorem suite: 1,296 interchange phrases, non-Clifford magic & dual-channel demux |
1,296/1,296 confluent; M₂ = 0.8301 b; max ‖Δ‖₂ = 2.1×10⁻⁸; 0.63 µs CPU |
JSON Proof |
K311 |
AlphaGenome 2 (alphagenome 2.0.0) |
Live fast/slow gastrocnemius myocyte conditioning & 20-SNV empirical AVI Kuramoto pacing |
MYH1 2,307× vs motor neuron; MYHAS 5.94× fast/slow; ΔΦ = 3.1376 rad |
JSON Empirical |
K312 |
MuJoCo 4.0.1 + mujoco_menagerie G1 |
Closed-loop 1,000 Hz floating-base Unitree G1 WBC: 3-controller ablation, 60s endurance & 2-G1 session |
5/5 survived (vs. 0/5 open-loop); 5.40° RMSE; +58.7% jerk; 60s PASS; 0.52° neural |
JSON 33388a82 |