Code for the paper "DSpAST: Disentangled Representations for Spatial Audio Reasoning with Large Language Models"