Since the two qubits are encoded in identical bosons, the full two-particle wavefunction must be symmetric under particle exchange, e.g.,, where the two atoms are labelled 1 and 2. (In the merged trap, the subscripts L and R are replaced by e and g, , and an independent 1D lattice along the vertical direction. By controlling the laser polarization, the unit cell of the 2D lattice can be continuously changed between single-well (λ-lattice) or double-well (λ/2-lattice)configurations (see Fig. 1a), where λ=816 nm. We start with a magnetically trapped We can prepare every pair of atoms in any non-entangled two-qubit state by selectively addressing the atoms in the L and R sites. We exploit the spin-dependence of the potential, which can be manipulated through the same polarization control used to adjust the lattice topology 2,12 . We first induce a state-dependence in the optical potential that produces an effective magnetic field gradient between the two adjacent sites of the double well. This introduces a differential shift The > 10 ms decay of the swap oscillations in Fig. 3 , but there the underlying noise arises from the inherent fluctuating background of nuclear spins. In contrast, here the inhomogeneous broadening arises