LATIDIA · Investigación

TimeThink: Eliciting Compositional Reasoning in Timeseries Large Language Models

arXiv: 2609.13457v1Tipo de anuncio: nuevo Resumen: Los modelos de lenguaje grandes multimodales (TS-MLLM) de series temporales han comenzado recientemente a aprovechar las capacidades de razonamiento de los modelos de lenguaje grandes (LLM) para la tarea de respuesta a preguntas

La noticia

arXiv:2609.13457v1 Announce Type: new Abstract: Timeseries multimodal large language models (TS-MLLMs) have recently begun leveraging the reasoning capabilities of large language models (LLMs) for question-answering tasks. However, these models often fail to capture dynamic temporal patterns, providing only implicit reasoning that lacks the underlying explanations critical for high-stakes applications like healthcare. While reinforcement learning (RL)-based timeseries language models aim to address this, they often fall short because they are trained on narrow, in-distribution data and struggle with out-of-distribution compositional questions. To address these challenges, we present TimeThink, a synthetic framework for eliciting compositional timeseries reasoning. Core timeseries primitives (e.g., trend, seasonality) are domain-independent and can be deterministically generated. Guided by