JigShape：通过拼图任务评估视觉语言模型中的视觉-几何推理能力

摘要

拼图求解需要同时推理视觉内容与几何约束，然而现有基准采用矩形切割，在纹理重复区域会产生模糊的真实标签。我们提出了 \ours{}，这是一个采用凸块与凹槽互锁拼块的基准，其几何约束提供了强局部兼容性要求，与视觉内容结合后能够产生无歧义的真实标签。在四种网格密度（4×4 到 16×16）的 95K 个实例中，我们发现零样本视觉语言模型（VLM）在很大程度上缺乏几何推理能力：五个前沿模型中仅有一个（GPT-5.5）在 4×4 拼图上超过随机基线，其余所有模型均处于随机水平。虽然监督微调在 4×4 上能达到 97% 以上的准确率，但所有模型在更大网格上均出现性能崩溃：GPT-5.5 在 8×8 上从 70% 跌至接近随机水平，甚至微调后的模型在 12×12 上也低于 5%。这种“规模悬崖”表明，随着拼块数量增加，当前架构无法维持一致的约束满足。\ours{} 将可扩展的几何推理确立为视觉语言模型面临的一项开放挑战。

English

Jigsaw puzzle solving requires jointly reasoning about visual content and geometric constraints, yet existing benchmarks use rectangular cuts that create ambiguous ground truth in texture-repeated regions. We introduce \ours{}, a benchmark with tab-and-blank interlocking pieces where geometric constraints provide strong local compatibility requirements that, combined with visual content, yield unambiguous ground truth. Across 95K instances at four grid densities (4times4 to 16times16), we find that zero-shot VLMs largely lack geometric reasoning: only one of five frontier models (GPT-5.5) exceeds random baseline on 4times4 puzzles, while all others perform at chance level. While supervised fine-tuning achieves >97\% on 4times4, all models collapse on larger grids: GPT-5.5 drops from 70\% to near-random on 8times8, and even fine-tuned models fall below 5\% on 12times12. This ``scaling cliff'' suggests current architectures cannot maintain consistent constraint satisfaction as the number of pieces increases. establishes scalable geometric reasoning as an open challenge for vision-language models.