发表机构
University of California, San Diego; West Virginia University（加利福尼亚大学圣迭戈分校; 西弗吉尼亚大学）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
T-SMART通过分离语言推理、确定性计算和感知，揭示工具增强时间序列问答的主要收益来自可靠数值执行，而非额外语言推理，准确率提升31.7个百分点。
AI 中文摘要
大型语言模型（LLM）在处理时间序列问答（TS-QA）时可能遇到困难，尤其是当数值信号以文本形式序列化并需要显式计算时。工具增强方法提升了性能，但现有系统常常将语言推理、计算和感知交织在一起，使得难以确定哪些组件推动了性能提升。我们提出了T-SMART，一个神经符号框架，将这些角色分离：一个冻结的LLM解释问题并选择操作，确定性工具执行数值计算，结构化感知仅在需要时被调用。受控配对消融实验表明，确定性计算提供了主要收益，在序列化时间序列上相比直接LLM推理将准确率提高了31.7个百分点，而语言理解和感知提供了较小的互补性增益。这些结果表明，工具增强的TS-QA主要受益于可靠的数值执行，而非额外的语言模型推理，并为分析神经符号时间序列系统中组件贡献提供了一个受控框架。
英文摘要
Large language models (LLMs) can struggle with time-series question answering (TS-QA), especially when numerical signals are serialized as text and require explicit computation. Tool-augmented approaches improve performance, but existing systems often intertwine language reasoning, computation, and perception, making it difficult to determine which components drive the gains. We present T-SMART, a neurosymbolic framework that separates these roles: a frozen LLM interprets questions and selects operations, deterministic tools perform numerical computation, and structured perception is invoked only when needed. Controlled paired ablations show that deterministic computation provides the dominant benefit, improving accuracy by 31.7 percentage points over direct LLM reasoning on serialized time series, while language understanding and perception offer smaller complementary gains. These results indicate that tool-augmented TS-QA benefits primarily from reliable numerical execution rather than additional language-model reasoning and provide a controlled framework for analyzing component contributions in neurosymbolic time-series systems.
CommentsAccepted by the 38th IEEE International Conference on Tools with Artificial Intelligence (ICTAI'26)
