Various typographical errors are corrected in the above-named article [ibid., vol. 12, no. 6, pp. 1206???1212, Nov. 2013]: for Equations (1), (5), and (6), as well as the text in Section III-C (page 1210); Equation (2); and on page 1208, left column, fourth row from the bottom: Fig. 1(b) is wrongly referred to as (1b). Furthermore, our stochastic simulations had a systematic error that lead to our reporting an error higher than the correct estimate in Figs. 2 and 4 of the original paper. The corrected versions of these figures are Figs. 1 and 2, respectively, in this corrigendum. However, the new plots convey the same physics and overall message: dipole-coupled architectures are unlikely to meet the stringent error requirements of conventional Boolean logic. Nonetheless, such architectures may have niche applications, where lowering energy dissipation is the main consideration and accuracy is not. One example is implanted medical devices that cannot drain the battery too quickly, but need not be too accurate since they usually diagnose a patient???s condition by approximately matching signals. Hence, the key message and our conclusions remain unchanged.