Realistic Leviathan Movement

Implement physics-based, procedural animation or neural network learning for large fauna? Currently, when massive creatures turn, they often look like they are simply pivoting around a central origin point using standard spherical linear interpolation (slerp). Transitioning to procedural locomotion—similar to the systems seen in ARC Raiders—where creatures physically propel themselves based on their environment. For example, the Collector Leviathan could actively use its tentacles to bank, drag, and turn itself through the water column rather than just rotating on an axis.