arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2512.06205cs.AIcs.CLcs.LG
收藏

关于测量 grounding 并推广 grounding 问题

A framework for auditing grounding claims

Center for Possible Minds(可能心智中心);Indiana University(印第安纳大学)

Daniel Quigley, Eric Maynard

AI总结:

本文提出了一种框架,用于测量和推广 grounding 问题,通过定义不同需求和评估标准,分析了符号、参照、向量和关系四种模式的 grounding 现状,并探讨了其在不同任务中的表现。

摘要:

符号 grounding 问题询问如何使像 cat 这样的符号能指代猫,而不是仅仅在算术中被操纵。我们将 grounding 从二元判断重新表述为一种跨需求的审计,每个都由一个评估元组(上下文、意义类型、威胁模型、参考分布)索引:真实性(机制位于智能体内部,并且对于强主张,是通过学习或进化获得的);保存(原子意义保持完整);忠实性,包括相关性(实现的意义与预期意义相符)和因果性(内部机制因果地促成成功);鲁棒性(在声明的扰动下优雅退化);组合性(整体是系统地由部分构建的)。我们将此框架应用于四种 grounding 模式(符号;参照;向量;关系)和三个案例研究:模型理论语义学实现了精确组合但缺乏因果性证明;大型语言模型在语言任务中显示出相关性契合和局部鲁棒性,但缺乏无 grounded 交互的世界任务中的选择性成功;人类语言在强真实性下满足了这些需求,通过进化和发展获取。通过将哲学 inquiry 转化为操作性问题,我们为科学家、计算机科学家、语言学家和数学家提供了一种共同的语言和技术框架,以系统地研究 grounding 和意义。

英文摘要:

The symbol grounding problem asks how a token such as cat can be about cats. We propose a framework for auditing grounding claims against a declared semantic standard. The audit reports measurements and evidence, with overall verdicts conditional on explicit acceptance criteria. Its profiles assess accuracy, robustness, and composition alongside evidence about how the system acquired its mechanisms, how they contribute to performance, and why they were retained. In a toy gridworld, an agent interprets individual symbols accurately but fails a withheld combination. Composing its interpretations by the declared rule would succeed. This comparison identifies a departure from the composition rule within the observed failure. Both this audit and a pilot on pretrained word vectors provide evidence that a designated mechanism contributes to present performance. Whether that contribution explains its retention remains uncertified. The framework evaluates the evidence for grounding claims; candidate accounts remain responsible for explaining how meaning emerges.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑