@misc{indiciae4f8c2455f7fa, title = {Linear interpolation gives better gradients than Gaussian smoothing in derivative-free optimization}, author = {Albert S Berahas and Liyuan Cao and Krzysztof Choromanski and Katya Scheinberg}, year = {2019}, url = {https://arxiv.org/abs/1905.13043}, note = {Source identifier: 1905.13043} }