Natural microswimmers use various swimming gaits to propel under low Reynolds number conditions in their fluid surrounding for various reasons: search for nutrition, escape from predators or search for prey. A very common strategy for propulsion is the non-reciprocal deformation of the shape of the swimmer in an effort to realize its motion through the medium. In recent times efforts have been made to design artificial swimmers that can successfully mimic their natural counterparts and are thus able to perform specific tasks, such as targeted drug delivery in the case of nanomedical applications. In this work, we train a two-dimensional, triangular swimmer to move in a desired direction and to detect in an efficient manner nutrition sources. This is achieved – similar as in previous work on a one-dimensional linear swimmer [1] – with adaptive neural networks (which connect the degrees of freedom and the forces acting on the swimmer), involving thereby the NEAT algorithm [2] which optimizes the internal architecture of these networks. In preceding work the one-dimensional, linear three-bead swimmer was successfully trained to swim in a chemical landscape [1]. Here we proceed to the considerably more challenging case of the triangular swimmer (see Fig. 1(I)): now rotation (see Fig. 1(A)-(C)), translation and the coupling of these two kinds of motion have to be taken into account. We demonstrate that also in this setup the swimmer can be trained to propagate in a desired direction and to develop swimming gaits that allows to find nutrient in a chemical landscape. Again we are able to extract valuable information about the swimmer’s strategies by analyzing the internal structure of the emerging networks. [1] B. Hartl, M. Hϋbl, G. Kahl, and A. Zӧttl, Proc. Natl. Acad. Sci. USA., 118 (2021), e2019683118. [2] K. O. Stanley, R. Miikkulainen, Evol. Comput., 10 (2002), 99.