TY - RPRT TI - Evaluating the Ability of Large Language Models to Reason about Cardinal Directions, Revisited AU - Anthony G Cohn AU - Robert E Blackwell PY - 2025 UR - https://arxiv.org/abs/2507.12059 ID - 2507.12059 ER -