人工代理在交流压力下自发形成数字符号，但无法泛化到未见数量。
The emergence of numerical representations in communicating artificial agents
- 通过对话游戏让神经网络代理自主发展数字符号与图像表达
- 在训练范围内准确传递数量，但对未见数量表现差
- 适合研究语言演化与符号系统形成的机器学习研究者
人类语言能高效表达数量，但仅靠交流压力是否足以促使人工代理自发产生数字符号，且这些符号是否类似人类数字仍不清楚。我们研究了两种基于神经网络的代理，在参照游戏中需通过离散标记或连续草图传递数量信息，从而探索符号化与图像化表征。在无预设数字符号的情况下，代理在两种通信方式中均达到高精度的分布内通信准确率，并收敛于高精度的符号-意义映射。然而，生成的代码是非组合性的：代理无法推导出未训练数量的系统性表达，通常重复使用最高训练数量的符号（离散）或把外推值压缩为单一草图（连续）。结论是，仅靠交流压力可实现已学数量的精确传递，但要获得组合性代码与泛化能力，还需额外压力机制。
原文摘要 · Abstract (English)
Human languages provide efficient systems for expressing numerosities, but whether the sheer pressure to communicate is enough for numerical representations to arise in artificial agents, and whether the emergent codes resemble human numerals at all, remains an open question. We study two neural network-based agents that must communicate numerosities in a referential game using either discrete tokens or continuous sketches, thus exploring both symbolic and iconic representations. Without any pre-defined numeric concepts, the agents achieve high in-distribution communication accuracy in both communication channels and converge on high-precision symbol-meaning mappings. However, the emergent code is non-compositional: the agents fail to derive systematic messages for unseen numerosities, typically reusing the symbol of the highest trained numerosity (discrete), or collapsing extrapolated values onto a single sketch (continuous). We conclude that the communication pressure alone suffices for precise transmission of learned numerosities, but additional pressures are needed to yield compositional codes and generalisation abilities.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。