# Explainable time-series forecasting with sampling-free SHAP for Transformers

Source: [Explainable time-series forecasting with sampling-free SHAP for Transformers](<https://www.sota2.com/research/paper/694b739a53fdd5c361739b96>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2025-12-25T05:00:04.588Z

Authors: Matthias Hertel, Sebastian P\\"utz, Ralf Mikut, Veit Hagenmeyer, Benjamin Sch\\"afer

Published: 2025-12-23T17:02:35.000Z

## Abstract

Time-series forecasts are essential for planning and decision-making in many domains. Explainability is key to building user trust and meeting transparency requirements. Shapley Additive Explanations (SHAP) is a popular explainable AI framework, but it lacks efficient implementations for time series and often assumes feature independence when sampling counterfactuals. We introduce SHAPformer, an accurate, fast and sampling-free explainable time-series forecasting model based on the Transformer architecture. It leverages attention manipulation to make predictions based on feature subsets. SHAPformer generates explanations in under one second, several orders of magnitude faster than the SHAP Permutation Explainer. On synthetic data with ground truth explanations, SHAPformer provides explanations that are true to the data. Applied to real-world electrical load data, it achieves competitive predictive performance and delivers meaningful local and global insights, such as identifying the past load as the key predictor and revealing a distinct model behavior during the Christmas period.

## Recorded benchmark results

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

- [Time-series forecasting on TransnetBW 6 months (test)](<https://www.sota2.com/research/sota/time-series-forecasting-on-transnetbw-6-months-test.md>): Time Series Forecasting; TransnetBW 6 months (test); RMSE: 265.9.

- [Time-series forecasting on Synthetic dataset 10,000 examples (test)](<https://www.sota2.com/research/sota/time-series-forecasting-on-synthetic-dataset-10-000-examples-test.md>): Time Series Forecasting; Synthetic dataset 10,000 examples (test); RMSE: 0.06.

## Original sources and implementation

[Original paper](<https://arxiv.org/abs/2512.20514>)

[PDF](<https://arxiv.org/pdf/2512.20514.pdf>)