TimeMaster: Training Time-Series Multimodal LLMs to Reason via Reinforcement Learning

• URL: http://arxiv.org/abs/2506.13705v1
• Date: Mon, 16 Jun 2025 17:12:26 GMT
• Title: TimeMaster: Training Time-Series Multimodal LLMs to Reason via Reinforcement Learning
• Abstract summary: Time-series reasoning remains a significant challenge in multimodal large language models (MLLMs)<n>We introduce TimeMaster, a reinforcement learning-based method that enables time-series MLLMs to perform structured, interpretable reasoning directly over visualized time-series inputs and task prompts.<n>TimeMaster achieves state-of-the-art performance, outperforming both classical time-series models and few-shot GPT-4o by over 14.6% and 7.3% performance gain, respectively.
• Score: 6.860030590512406
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Time-series reasoning remains a significant challenge in multimodal large language models (MLLMs) due to the dynamic temporal patterns, ambiguous semantics, and lack of temporal priors. In this work, we introduce TimeMaster, a reinforcement learning (RL)-based method that enables time-series MLLMs to perform structured, interpretable reasoning directly over visualized time-series inputs and task prompts. TimeMaster adopts a three-part structured output format, reasoning, classification, and domain-specific extension, and is optimized via a composite reward function that aligns format adherence, prediction accuracy, and open-ended insight quality. The model is trained using a two-stage pipeline: we first apply supervised fine-tuning (SFT) to establish a good initialization, followed by Group Relative Policy Optimization (GRPO) at the token level to enable stable and targeted reward-driven improvement in time-series reasoning. We evaluate TimeMaster on the TimerBed benchmark across six real-world classification tasks based on Qwen2.5-VL-3B-Instruct. TimeMaster achieves state-of-the-art performance, outperforming both classical time-series models and few-shot GPT-4o by over 14.6% and 7.3% performance gain, respectively. Notably, TimeMaster goes beyond time-series classification: it also exhibits expert-like reasoning behavior, generates context-aware explanations, and delivers domain-aligned insights. Our results highlight that reward-driven RL can be a scalable and promising path toward integrating temporal understanding into time-series MLLMs.

Related papers

• TEMPO: Temporally-grounded Multi-task Post-training for Large Audio-Language Models [58.80225070159013]
  Large audio-language models (LALMs) describe audio at the clip level but cannot assign timestamps to the events, speakers, or sounds they identify.<n>We present TEMPO, the first unified model to handle audio, speech, and music timestamping tasks.
  arXiv  Detail & Related papers  (2026-08-30T19:49:21Z)
• Representing Time Series as Structured Programs for LLM Reasoning [24.83778109852643]
  We introduce Time-Series-to-Structured-Program representation (T2SP), a deterministic, training-free method that represents a time series as a structured symbolic program.<n>By shifting temporal-structure extraction from the model to the representation itself, T2SP enables off-the-shelf LLMs to leverage their existing reasoning capabilities for time-series understanding.
  arXiv  Detail & Related papers  (2026-06-10T07:42:35Z)
• Enhancing Temporal Awareness in LLMs for Temporal Point Processes [53.596733432865626]
  Temporal point processes (TPPs) are crucial for analyzing events over time.<n> TPP-TAL is a novel plug-and-play framework designed to enhance temporal reasoning within large language models.<n> TPP-TAL delivers substantial improvements in temporal likelihood estimation and event prediction accuracy.
  arXiv  Detail & Related papers  (2025-12-29T03:01:24Z)
• TimeLens: Rethinking Video Temporal Grounding with Multimodal LLMs [81.78017865436816]
  We present TimeLens, a systematic investigation into building MLLMs with strong video temporal grounding ability.<n>We first expose critical quality issues in existing VTG benchmarks and introduce TimeLens-Bench, comprising meticulously re-annotated versions of three popular benchmarks.<n>We also address noisy training data through an automated re-annotation pipeline, yielding TimeLens-100K, a large-scale, high-quality training dataset.
  arXiv  Detail & Related papers  (2025-12-16T18:59:58Z)
• TS-Reasoner: Aligning Time Series Foundation Models with LLM Reasoning [17.69077698133793]
  Time series reasoning is crucial to decision-making in diverse domains, including finance, energy usage, traffic, weather, and scientific discovery.<n>LLMs often struggle with the numerical understanding of time series data.<n>We develop a training recipe that applies instruction finetuning after the alignment pretraining.
  arXiv  Detail & Related papers  (2025-10-03T21:20:54Z)
• TimeOmni-1: Incentivizing Complex Reasoning with Time Series in Large Language Models [105.47481207029047]
  We introduce Time Series Reasoning Suite (TSR-Suite), which formalizes four atomic tasks that span three fundamental capabilities for reasoning with time series.<n>We also introduce Time Omni-1, the first unified reasoning model designed to address diverse real-world problems demanding time series reasoning.
  arXiv  Detail & Related papers  (2025-09-29T13:54:34Z)
• Forecasting Time Series with LLMs via Patch-Based Prompting and Decomposition [48.50019311384125]
  We explore simple and flexible prompt-based strategies that enable LLMs to perform time series forecasting without extensive retraining.<n>We propose our own method, PatchInstruct, which enables LLMs to make precise and effective predictions.
  arXiv  Detail & Related papers  (2025-06-15T19:42:58Z)
• LLM-PS: Empowering Large Language Models for Time Series Forecasting with Temporal Patterns and Semantics [56.99021951927683]
  Time Series Forecasting (TSF) is critical in many real-world domains like financial planning and health monitoring.<n>Existing Large Language Models (LLMs) usually perform suboptimally because they neglect the inherent characteristics of time series data.<n>We propose LLM-PS to empower the LLM for TSF by learning the fundamental textitPatterns and meaningful textitSemantics from time series data.
  arXiv  Detail & Related papers  (2025-03-12T11:45:11Z)
• LangTime: A Language-Guided Unified Model for Time Series Forecasting with Proximal Policy Optimization [3.1819993716919472]
  LangTime is a language-guided unified model for time series forecasting.<n>TimePPO is a reinforcement learning-based fine-tuning algorithm.
  arXiv  Detail & Related papers  (2025-03-11T10:40:39Z)
• Small but Mighty: Enhancing Time Series Forecasting with Lightweight LLMs [11.1036247482657]
  We present SMETimes, the first systematic investigation of sub-3B parameter SLMs for efficient and accurate time series forecasting.<n>Our approach centers on three key innovations: A statistically-enhanced prompting mechanism that bridges numerical time series with textual semantics through statistical features; A adaptive fusion embedding architecture that aligns temporal patterns with language model token spaces through learnable parameters.
  arXiv  Detail & Related papers  (2025-03-05T15:27:36Z)
• General Time-series Model for Universal Knowledge Representation of Multivariate Time-Series data [61.163542597764796]
  We show that time series with different time granularities (or corresponding frequency resolutions) exhibit distinct joint distributions in the frequency domain.<n>A novel Fourier knowledge attention mechanism is proposed to enable learning time-aware representations from both the temporal and frequency domains.<n>An autoregressive blank infilling pre-training framework is incorporated to time series analysis for the first time, leading to a generative tasks agnostic pre-training strategy.
  arXiv  Detail & Related papers  (2025-02-05T15:20:04Z)
• Towards Time Series Reasoning with LLMs [0.4369058206183195]
  We propose a novel multi-modal time-series LLM approach that learns generalizable information across various domains with powerful zero-shot performance.<n>We show that our model learns a latent representation that reflects specific time-series features, as well as outperforming GPT-4o on a set of zero-shot reasoning tasks.
  arXiv  Detail & Related papers  (2024-09-17T17:23:44Z)
• Multi-Patch Prediction: Adapting LLMs for Time Series Representation Learning [22.28251586213348]
  aLLM4TS is an innovative framework that adapts Large Language Models (LLMs) for time-series representation learning. A distinctive element of our framework is the patch-wise decoding layer, which departs from previous methods reliant on sequence-level decoding.
  arXiv  Detail & Related papers  (2024-02-07T13:51:26Z)
• TimeSiam: A Pre-Training Framework for Siamese Time-Series Modeling [67.02157180089573]
  Time series pre-training has recently garnered wide attention for its potential to reduce labeling expenses and benefit various downstream tasks. This paper proposes TimeSiam as a simple but effective self-supervised pre-training framework for Time series based on Siamese networks.
  arXiv  Detail & Related papers  (2024-02-04T13:10:51Z)
• AutoTimes: Autoregressive Time Series Forecasters via Large Language Models [67.83502953961505]
  AutoTimes projects time series into the embedding space of language tokens and autoregressively generates future predictions with arbitrary lengths. We formulate time series as prompts, extending the context for prediction beyond the lookback window. AutoTimes achieves state-of-the-art with 0.1% trainable parameters and over $5times$ training/inference speedup.
  arXiv  Detail & Related papers  (2024-02-04T06:59:21Z)
• Time-LLM: Time Series Forecasting by Reprogramming Large Language Models [110.20279343734548]
  Time series forecasting holds significant importance in many real-world dynamic systems. We present Time-LLM, a reprogramming framework to repurpose large language models for time series forecasting. Time-LLM is a powerful time series learner that outperforms state-of-the-art, specialized forecasting models.
  arXiv  Detail & Related papers  (2023-10-03T01:31:25Z)
• LLM4TS: Aligning Pre-Trained LLMs as Data-Efficient Time-Series Forecasters [11.796765525301051]
  We propose a framework for time-series forecasting with pre-trained Large Language Models (LLMs)<n>LLM4TS consists of a two-stage fine-tuning strategy to align LLMs with the nuances of time-series data, and the forecasting fine-tuning stage for downstream time-series forecasting tasks.<n>Our framework features a novel two-level aggregation method that integrates multi-scale temporal data within pre-trained LLMs, enhancing their ability to interpret time-specific information.
  arXiv  Detail & Related papers  (2023-08-16T16:19:50Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.