TY - RPRT TI - Efficient Modular Learning through Naive LoRA Summation: Leveraging Orthogonality in High-Dimensional Models AU - Zhanhao Cao AU - Clement Truong AU - Andrew Lizarraga PY - 2025 UR - https://arxiv.org/abs/2508.11985 ID - 2508.11985 ER -