Energetic primary recoil atoms from ion implantation or fast neutron irradiation produce isolated point defects and clusters of both vacancies and interstitials. The migration energies and mechanisms for these defects are crucial to successful multiscale modeling of microstructural evolution during ion-implantation, thermal annealing, or under irradiation over long periods of time. The dimer method is employed to search for possible transition states of interstitials and small interstitial clusters in SiC and α-Fe. The method uses only the first derivatives of the potential energy to find saddle points without knowledge of the final state of the transition. In SiC, the possible migration pathway for the C interstitial is found to consist of the first neighbor jump via a Si site or second neighbor jump, but the relative probability for the second neighbor jump is very low. In α-Fe, the possible transition states are studied as a function of interstitial cluster size, and the lowest energy barriers correspond to defect migration along 〈1 1 1〉 directions. However, this paper addresses whether migrating interstitial clusters can thermally change their direction, and the activation energies and corresponding mechanisms for changing the direction of these clusters are determined.