mlsanityData to 7 Oct 2026

KairosVL: Orchestrating Time Series and Semantics for Unified Reasoning

arXiv 2602.20494 ↗First posted 24 Feb 2026cs.AIcited by 0

Not on the map. Why?

Abstract (arXiv)

Driven by the increasingly complex and decision-oriented demands of time series analysis, we introduce the Semantic-Conditional Time Series Reasoning task, which extends conventional time series analysis beyond purely numerical modeling to incorporate contextual and semantic understanding. To further enhance the mode's reasoning capabilities on complex time series problems, we propose a two-round reinforcement learning framework: the first round strengthens the mode's perception of fundamental temporal primitives, while the second focuses on semantic-conditioned reasoning. The resulting model, KairosVL, achieves competitive performance across both synthetic and real-world tasks. Extensive experiments and ablation studies demonstrate that our framework not only boosts performance but also preserves intrinsic reasoning ability and significantly improves generalization to unseen scenarios. To summarize, our work highlights the potential of combining semantic reasoning with temporal modeling and provides a practical framework for real-world time series intelligence, which is in urgent demand.

Key facts. No citing papers in the map yet (data through 7 Oct 2026). Builds on 33 of its 57 references that are in the map. First posted on arXiv 24 Feb 2026.

Builds on

33 of 57 references.

References · 33

  1. 2024DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Modelscited by 8,851
  2. 2025Qwen2.5-VL Technical Reportcited by 5,418
  3. 2025Gemini 2.5: Pushing the Frontier with Advanced Reasoning, Multimodality, Long Context, and Next Generation Agentic Capabilitiescited by 3,655
  4. 2022A Time Series is Worth 64 Words: Long-term Forecasting with Transformerscited by 1,981
  5. 2024HybridFlow: A Flexible and Efficient RLHF Frameworkcited by 1,980
  6. 2025InternVL3: Exploring Advanced Training and Test-Time Recipes for Open-Source Multimodal Modelscited by 1,769
  7. 2024A Survey on LLM-as-a-Judgecited by 1,384
  8. 2025Understanding R1-Zero-Like Training: A Critical Perspectivecited by 1,276
  9. 2024ChatGLM: A Family of Large Language Models from GLM-130B to GLM-4 All Toolscited by 1,096
  10. 2018Deep learning for time series classification: a reviewcited by 575
  11. 2021Anomaly Transformer: Time Series Anomaly Detection with Association Discrepancycited by 387
  12. 2024Time-MoE: Billion-Scale Time Series Foundation Models with Mixture of Expertscited by 280
21 more
  1. 2025Logic-RL: Unleashing LLM Reasoning with Rule-Based Reinforcement Learningcited by 260
  2. 2018Unsupervised Anomaly Detection via Variational Auto-Encoder for Seasonal KPIs in Web Applicationscited by 223
  3. 2025SFT or RL? An Early Investigation into Training R1-Like Reasoning Large Vision-Language Modelscited by 209
  4. 2024ChatTS: Aligning Time Series with LLMs via Synthetic Data for Enhanced Understanding and Reasoningcited by 95
  5. 2024ChatTime: A Unified Multimodal Time Series Foundation Model Bridging Numerical and Textual Datacited by 92
  6. 2024Language Models Still Struggle to Zero-shot Reason about Time Seriescited by 81
  7. 2025Time-MQA: Time Series Multi-Task Question Answering with Context Enhancementcited by 74
  8. 2023ECG-QA: A Comprehensive Question Answering Dataset Combined With Electrocardiogramcited by 52
  9. 2024TimeSeriesExam: A time series understanding examcited by 47
  10. 2023Multi-Modal Self-Supervised Learning for Recommendationcited by 44
  11. 2024Can LLMs Understand Time Series Anomalies?cited by 41
  12. 2024Evaluating Large Language Models on Time Series Feature Understanding: A Comprehensive Taxonomy and Benchmarkcited by 38
  13. 2025ITFormer: Bridging Time Series and Natural Language for Multi-Modal QA with Large-Scale Multitask Datasetcited by 35
  14. 2024A Picture is Worth A Thousand Numbers: Enabling LLMs Reason about Time Series via Visualizationcited by 31
  15. 2021REST: Relational Event-driven Stock Trend Forecastingcited by 30
  16. 2024TimeSeriesBench: An Industrial-Grade Benchmark for Time Series Anomaly Detection Modelscited by 22
  17. 2025When LLM Meets Time Series: Can LLMs Perform Multi-Step Time Series Reasoning and Inferencecited by 12
  18. 2021Truth-Conditional Captioning of Time Series Datacited by 11
  19. 2025BEDTime: A Unified Benchmark for Automatically Describing Time Seriescited by 11
  20. 2025CMoS: Rethinking Time Series Prediction Through the Lens of Chunk-wise Spatial Correlationscited by 8
  21. 2023A Survey of Time Series Anomaly Detection Methods in the AIOps Domaincited by 2

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2026KairosAgent: Agentic Time Series Forecasting with Fused Semantic Reasoning
  2. 2025Time Series Forecasting via Reasoning: A Slow-Thinking Approach with Reinforcement Fine-Tuned LLMs
  3. 2025Achieving Time Series Reasoning Requires Rethinking Model Design, Tasks Formulation, and Evaluation
  4. 2025Eliciting Chain-of-Thought Reasoning for Time Series Analysis using Reinforcement Learning
  5. 2026SenTSR-Bench: Thinking with Injected Knowledge for Time-Series Reasoning
  6. 2026Time Series Reasoning via Process-Verifiable Thinking Data Synthesis and Scheduling for Tailored LLM Reasoning
  7. 2026TSRBench: A Comprehensive Multi-task Multi-modal Time Series Reasoning Benchmark for Generalist Models
  8. 2026From Recognition to Understanding: Unlocking Cognitive Time Series Reasoning with LLMs