Lapras：面向时间序列语言模型的潜在推理

中文摘要 AI 辅助

时间序列语言模型（TSLMs）通过对时间信号进行推理并生成自然语言答案与解释，为时间序列理解提供了一条有前景的路径。一种常见方法是思维链（CoT），它生成将相关信号模式与最终答案关联起来的分步推理依据。尽管这些模型在后续训练中从参考CoT轨迹中学习，但在推理时生成输入时间序列的忠实描述仍然具有挑战性。将高维、连续的时间表示形式表达为离散语言标记可能会导致模型忽略任务相关模式或对其描述不准确。由于后续推理步骤基于这些描述构建，早期错误会传播，导致产生与输入信号不一致的看似合理的解释的错误答案。我们提出Lapras（Latent Post-trained Reasoning Across Series，跨序列潜在后训练推理），这是一种为TSLMs配备潜在推理的后训练框架。经Lapras训练的模型在联合时间序列-语言空间中通过一系列连续思维进行推理，仅为最终答案生成文本。它通过教师-学生自蒸馏学习这一点，其中在CoT参考轨迹上训练的教师通过文本显式推理。学生在答案阶段将其隐藏状态与教师的隐藏状态对齐，将教师的推理能力转移到其潜在计算中。我们在五个时间序列问答基准上的四个TSLM骨干上评估Lapras。Lapras比显式CoT平均F1提升高达10.79%，同时生成的标记少23.9倍。Lapras的连续思维还可通过标准语言解码解码为可读的推理轨迹，保留文本解释。这些结果共同表明，Lapras是一种用于高效、有效且可解释的TSLM推理的有前景的后训练范式。

英文摘要

Time Series Language Models (TSLMs) offer a promising path toward time series understanding by reasoning over temporal signals and producing natural language answers and explanations. A common approach is Chain-of-Thought (CoT), which generates step-by-step rationales linking relevant signal patterns to final answers. Although these models learn from reference CoT traces during post-training, generating faithful descriptions of input time series at inference remains challenging. Expressing high-dimensional, continuous temporal representations in discrete language tokens may cause the model to neglect task-relevant patterns or describe them inaccurately. Because later reasoning steps build on these descriptions, early errors propagate, leading to incorrect answers with plausible explanations that are inconsistent with the input signal. We propose Lapras (Latent Post-trained Reasoning Across Series), a post-training framework that equips TSLMs with latent reasoning. A model trained with Lapras reasons through a sequence of continuous thoughts in the joint time series-language space, producing text only for the final answer. It learns this through teacher-student self-distillation, where a teacher trained on CoT reference traces reasons explicitly through text. The student aligns its hidden states with the teacher's at the answer stage, transferring the teacher's reasoning ability into its latent computation. We evaluate Lapras across four TSLM backbones on five time series question answering benchmarks. Lapras improves average F1 by up to 10.79% over explicit CoT while generating 23.9x fewer tokens. Lapras's continuous thoughts can also be decoded into readable reasoning traces via standard language decoding, preserving textual explanations. Together, these results highlight Lapras as a promising post-training paradigm for efficient, effective, and interpretable TSLM reasoning.

发表机构

• Dartmouth College（达特茅斯学院）
• Aionic Labs（艾奥尼克实验室）
• ETH Zurich（苏黎世联邦理工学院）
• Stanford University（斯坦福大学）
• National University of Singapore（新加坡国立大学）

机构由 AI 辅助整理，请以论文原文为准。