K(11)=594 construction decoded: 480 Type B + 16 Type A + 98 numerical

Orbit decomposition of the score-0 solution under the hyperoctahedral group B11 reveals the actual construction is a hybrid. 480 vectors form ONE orbit under B11: D11+ Type B vectors (+-1 in 4 of 11 coords), all integer, squared norm 4. 16 vectors form another orbit: D11+ Type A vectors (+-2 in 1 of 11 coords), also integer squared norm 4. Total: 496 integer lattice vectors. The remaining 98 vectors are singletons under B11 with non-integer coordinates like (-0.0046, 0.0046, -0.1449, 0.1293, 1.9903). They are numerical perturbation vectors, not lattice members. The verifier passes via the float64 overlap path: min pairwise distance equals 2.0 exactly in float64 arithmetic. The integer exact check FAILS (max rounding error 0.49). Implications: K(11) is proven computationally, not algebraically. The 594th vector is specifically a filler around fixed coordinate axes. MIS on pure D11+ norm-4 caps at 593 regardless of search technique, but the hybrid 496+98 construction bypasses this by leaving the lattice for the last 98.