mlsanityData to 14 Oct 2026

ReasonCast: Agentic Demand Forecasting with Selective Semantic Reasoning

arXiv 2608.15291 ↗First posted 15 Aug 2026cs.AIcited by 0

On the map: LLM agents and tool use → Financial trading and forecasting agents

Abstract (arXiv)

Demand forecasting increasingly requires combining two complementary sources of information: historical sales reveal recurring numerical dynamics, while future promotions, holidays, price changes, and platform interventions provide forward-looking knowledge. Existing text-enhanced forecasting methods often encode such context into generic representations and fuse it uniformly with time-series features, without explicitly distinguishing which semantic effects are forecast-relevant or how they should modify future dynamics. We introduce ReasonCast, a structured semantic intervention framework that translates event knowledge into forecast-specific operations. An agent examines the event context, the no-text forecast, and its uncertainty to determine whether textual reasoning is needed. Rather than injecting free-form text, ReasonCast represents event knowledge through structured fields describing event relevance, demand direction, temporal shape, amplitude, and peak intensity. These fields interact selectively with temporal components of a time-series foundation model. An additive path corrects local trends and temporal shapes, while a multiplicative path captures event-driven level shifts. ReasonCast introduces a forecast-grounded post-training curriculum. Schema SFT establishes semantic fields; semantic-field RL calibrates direction, shape, amplitude, and peak judgments; and forecast-utility RL evaluates semantic interventions through a frozen forecaster, aligning reasoning outputs with marginal forecast improvement. ReasonCast lowers WMAPE by 3.29, 1.25, and 0.47 percentage points on holiday-sensitive categories, mega-sale-sensitive categories, and M5 event windows, respectively. On stable-sales periods, indiscriminate semantic intervention increases WMAPE by 1.68 percentage points, whereas suppressing unnecessary intervention preserves the numerical backbone.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Placed in Financial trading and forecasting agents, one of 16 directions in LLM agents and tool use. Builds on 21 of its 32 references that are in the map. First posted on arXiv 15 Aug 2026.

Builds on

21 of 32 references.

Closest ancestors · 1

Same direction, newest first

  1. 2024From News to Forecast: Integrating Event Analysis in LLM-Based Time Series Forecasting with Reflection26 Sep

Other references · 20

  1. 2023Toolformer: Language Models Can Teach Themselves to Use Toolscited by 4,182
  2. 2021Scaling Up Visual and Vision-Language Representation Learning With Noisy Text Supervisioncited by 3,864
  3. 2022A Time Series is Worth 64 Words: Long-term Forecasting with Transformerscited by 2,039
  4. 2017DeepAR: Probabilistic Forecasting with Autoregressive Recurrent Networkscited by 1,013
  5. 2023iTransformer: Inverted Transformers Are Effective for Time Series Forecastingcited by 931
  6. 2019Temporal Fusion Transformers for Interpretable Multi-horizon Time Series Forecastingcited by 761
  7. 2024Chronos: Learning the Language of Time Seriescited by 759
  8. 2023One Fits All:Power General Time Series Analysis by Pretrained LMcited by 632
  9. 2024Unified Training of Universal Time Series Forecasting Transformerscited by 614
  10. 2023Large Language Models Are Zero-Shot Time Series Forecasterscited by 500
  11. 2024MOMENT: A Family of Open Time-series Foundation Modelscited by 482
  12. 2023TEST: Text Prototype Aligned Embedding to Activate LLM's Ability for Time Seriescited by 169
8 more
  1. 2024CALF: Aligning LLMs for Time Series Forecasting via Cross-modal Fine-Tuningcited by 108
  2. 2024ChatTime: A Unified Multimodal Time Series Foundation Model Bridging Numerical and Textual Datacited by 100
  3. 2024Timer-XL: Long-Context Transformers for Unified Time Series Forecastingcited by 87
  4. 2024Context is Key: A Benchmark for Forecasting with Essential Textual Informationcited by 87
  5. 2025Language in the Flow of Time: Time-Series-Paired Texts Weaved into a Unified Temporal Narrativecited by 40
  6. 2025Aurora: Towards Universal Generative Multimodal Time Series Forecastingcited by 29
  7. 2025BALM-TSF: Balanced Multimodal Alignment for LLM-Based Time Series Forecastingcited by 6
  8. 2026Unlocking the Value of Text: Event-Driven Reasoning and Multi-Level Alignment for Time Series Forecastingcited by 5

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2026ReasonCast: Towards Explainable Time Series Forecasting with Reasoning
  2. 2026EventCast: Hybrid Demand Forecasting in E-Commerce with LLM-Based Event Knowledge
  3. 2025AlphaCast: A Human Wisdom-LLM Intelligence Co-Reasoning Framework for Interactive Time Series Forecasting
  4. 2026CastFSR: A Fast--Slow--Reflect Agentic Reasoning Framework for Context-Aware Time Series Forecasting
  5. 2026From Narrative to Auditable Forecasts: A Structured Scaffold for Agentic Forecasting
  6. 2026TemporalBench: A Benchmark for Evaluating LLM-Based Agents on Contextual and Event-Informed Time Series Tasks
  7. 2026Nexus : An Agentic Framework for Time Series Forecasting
  8. 2026KairosAgent: Agentic Time Series Forecasting with Fused Semantic Reasoning