Spectral and Spatial Shaping of Spatiotemporal Nonlinearities in Multimode Fibers

Single-mode optical fibers have been used for various applications such as communications, fiber lasers, optical imaging and nonlinear optics. Recently multimode fibers started to attract attention for the aforementioned applications by including the spatial degree of freedom. Specifically, in graded-index multimode fibers (GIMFs), ultrashort pulse propagation with complex modal interactions revealed new nonlinear dynamics for beam shaping, frequency conversion and ultrashort pulse generation [1 , 2] . Here, we present methods to control the spectral and spatial properties of propagating light in multimode fibers and laser cavities. The multimode fiber nonlinearities are learned and shaped with a machine learning approach for single-pass pulse propagation. In spatiotemporal mode-locked fiber lasers, complex multimode laser structures, the generation of high-quality beam shapes are reported by changing temporal dynamics of mode-locked pulses and beam profile improvement with spatiotemporal mode-locking is achieved.