# Interpreting time-series and weather models

Area: [Mechanistic interpretability](https://mlsanity.com/mechinterp/index.md). HTML: https://mlsanity.com/mechinterp/interpreting-time-series-and-weather-models/

## Key facts

- Interpreting time-series and weather models: 14 papers on arXiv since 2024 (data through 14 Oct 2026).
- 2 of its 14 mapped papers (14%) cite “Interpretability for Time Series Transformers using A Concept Bottleneck Framework” (2024, arXiv 2410.06070).
- 2 of its 14 mapped papers (14%) cite “Mechanistic Interpretability for Transformer-based Time Series Classification” (2025, arXiv 2511.21514).
- 2 new papers 24 Sep–14 Oct.

## Most cited here

- [Interpretability for Time Series Transformers using A Concept Bottleneck Framework](https://mlsanity.com/paper/2410.06070/) (2024), cited by 2 papers here
- [Mechanistic Interpretability for Transformer-based Time Series Classification](https://mlsanity.com/paper/2511.21514/) (2025), cited by 2 papers here
- [Dissecting Chronos: Sparse Autoencoders Reveal Causal Feature Hierarchies in Time Series Foundation Models](https://mlsanity.com/paper/2603.10071/) (2026), cited by 1 papers here

## New (24 Sep–14 Oct)

- [Mechanistic Interpretability of Atmospheric Rivers in GraphCast](https://mlsanity.com/paper/2610.07583/) (2026-10-06)
- [TSAE: Structured Sparse Autoencoders for Interpreting Time-Series Forecasting Models](https://mlsanity.com/paper/2610.04925/) (2026-10-04)