# Challenges of Explainability in Continual Learning for Time Series Forecasting

arXiv: https://arxiv.org/abs/2607.19382 · HTML: https://mlsanity.com/paper/2607.19382/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 11 of its 14 references that are in the map.
- First posted on arXiv 1 Jul 2026.

## Builds on (most cited first, 11 of 11)

- [A Time Series is Worth 64 Words: Long-term Forecasting with Transformers](https://mlsanity.com/paper/2211.14730/) (2022)
- [Are Transformers Effective for Time Series Forecasting?](https://mlsanity.com/paper/2205.13504/) (2022)
- [Explainable Artificial Intelligence (XAI) on TimeSeries Data: A Survey](https://mlsanity.com/paper/2104.00950/) (2021)
- [Meta-attention for ViT-backed Continual Learning](https://mlsanity.com/paper/2203.11684/) (2022)
- [ICICLE: Interpretable Class Incremental Continual Learning](https://mlsanity.com/paper/2303.07811/) (2023)
- [Adaptive Memory Replay for Continual Learning](https://mlsanity.com/paper/2404.12526/) (2024)
- [PatchMixer: A Patch-Mixing Architecture for Long-Term Time Series Forecasting](https://mlsanity.com/paper/2310.00655/) (2023)
- [Continual Learning for Multivariate Time Series Tasks with Variable Input Dimensions](https://mlsanity.com/paper/2203.06852/) (2022)
- [Task-Attentive Transformer Architecture for Continual Learning of Vision-and-Language Tasks Using Knowledge Distillation](https://mlsanity.com/paper/2303.14423/) (2023)
- [Self-Attention Meta-Learner for Continual Learning](https://mlsanity.com/paper/2101.12136/) (2021)
- [Prediction error-driven memory consolidation for continual learning. On the case of adaptive greenhouse models](https://mlsanity.com/paper/2006.12616/) (2020)

## Abstract (arXiv)

Deep learning models have shown strong potential for time series forecasting, yet their deployment in real-world environmental monitoring remains challenging due to non-stationary dynamics and limited explainability. In this work, we investigate explainability as a central tool for understanding continual learning in adaptive time series forecasting, with Experience Replay strategies. We study neural forecasting architectures such as PatchMixer, PatchTST and DLinear, augmented with attention-based sampling mechanisms to support model adaptation over time. Explainability is leveraged through attention rollout and gradient-based attribution methods (Grad-CAM) to analyze both predictive behavior and sampling strategies within a continual learning framework. Experiments conducted on real-world piezometric time series exhibiting heterogeneous patterns and regime shifts show that analyzing model and sampling behaviors provides valuable insights into the dynamics of the continual learning framework. Beyond predictive performance, our results highlight the challenges and opportunities of using explainability to understand continual learning behaviors, revealing how attribution patterns evolve over time and how they can inform data selection and adaptation strategies in non-stationary forecasting scenarios.