Background: Many animals use celestial cues to determine their heading. In the insect brain, the central complex is crucial for spatial orientation. In desert locusts and fruit flies, compass neurons in the central complex encode the animal’s heading relative to the sun [2, 5]. However, the neural computations underlying the transformations of visual inputs to a compass signal remain elusive. In fruit flies, it is hypothesised that visual information is relayed from tangential neu- rons to compass neurons through likely inhibitory all-to-all connections with plastic synapses [1]. We aim to explore how a similar system functions in the desert locust. Model and Results: We constructed a dynamical firing rate model of TL neurons supplying visual inputs to CL1a (compass) neurons (homologous to ER- and E-PG neurons in fruit flies). The circuit’s firing behaviour was modelled after physiological data [2], and we assumed an all-to-all connectivity from TL- to CL1a-neurons [7]. Using a machine learning algorithm, we determined the synaptic weights needed for a functioning compass. We present two different possible connectivities: the first one is derived by constraining all synapses from TL- onto CL1a neurons to be inhibitory (in line with [3, 6]) and requires an excitatory overall luminance input to the CL1a neurons. The second TL-CL1a connectivity is unconstrained. In this case, the overall luminance input is not necessary. In both cases, the network maintains its activity state when the visual input is ”switched off” and updates its activity to match available visual cues about the sun position. Conclusion: Our work offers testable hypotheses for future studies on functional connectiv- ity in the central complex of the desert locust. We plan to integrate this model with our previous one [4], allowing the simulation of locomotor behaviour and the computation of heading using both self-motion integration and external-cue tethering.