{"ID":2858965,"CreatedAt":"2026-06-01T04:54:23.091178241Z","UpdatedAt":"2026-06-01T04:54:23.091178241Z","DeletedAt":null,"paper_url":"https://arxiv.org/abs/2510.07432","arxiv_id":"2510.07432","title":"TS-Agent: Understanding and Reasoning Over Raw Time Series via Iterative Insight Gathering","abstract":"Large language models (LLMs) exhibit strong symbolic and compositional reasoning, yet they struggle with time series question answering as the data is typically transformed into an LLM-compatible modality, e.g., serialized text, plotted images, or compressed time series embeddings. Such conversions impose representation bottlenecks, often require cross-modal alignment or finetuning, and can exacerbate hallucination and knowledge leakage. To address these limitations, we propose TS-Agent, an agentic, tool-grounded framework that uses LLMs strictly for iterative evidence-based reasoning, while delegating statistical and structural extraction to time series analytical tools operating on raw sequences. Our framework solves time series tasks through an evidence-driven agentic process: (1) it alternates between thinking, tool execution, and observation in a ReAct-style loop, (2) records intermediate results in an explicit evidence log and corrects the reasoning trace via a self-refinement critic, and (3) enforces a final answer-verification step to prevent hallucinations and leakage. Across four benchmarks spanning time series understanding and reasoning, TS-Agent matches or exceeds strong text-based, vision-based, and time-series language model baselines, with the largest gains on reasoning tasks where multimodal LLMs are prone to hallucination and knowledge leakage in zero-shot settings.","short_abstract":"Large language models (LLMs) exhibit strong symbolic and compositional reasoning, yet they struggle with time series question answering as the data is typically transformed into an LLM-compatible modality, e.g., serialized text, plotted images, or compressed time series embeddings. Such conversions impose representatio...","url_abs":"https://arxiv.org/abs/2510.07432","url_pdf":"https://arxiv.org/pdf/2510.07432v2","authors":"[\"Penghang Liu\",\"Elizabeth Fons\",\"Annita Vapsi\",\"Mohsen Ghassemi\",\"Svitlana Vyetrenko\",\"Daniel Borrajo\",\"Vamsi K. Potluru\",\"Manuela Veloso\"]","published":"2025-10-08T18:31:53Z","proceeding":"cs.AI","tasks":"[\"cs.AI\"]","methods":"[\"Large Language Model\",\"Language Model\"]","has_code":false}