# Towards a Unified Evaluation of Explanation Methods without Ground Truth

arXiv: https://arxiv.org/abs/1911.09017 · HTML: https://mlsanity.com/paper/1911.09017/

## Key facts

- Cited by 6 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 20 Nov 2019.

## Built on by (most cited first, 6 of 6)

- [From Anecdotal Evidence to Quantitative Evaluation Methods: A Systematic Review on Evaluating Explainable AI](https://mlsanity.com/paper/2201.08164/) (2022)
- [The Solvability of Interpretability Evaluation Metrics](https://mlsanity.com/paper/2205.08696/) (2022)
- [Visualizing Automatic Speech Recognition -- Means for a Better Understanding?](https://mlsanity.com/paper/2202.00673/) (2022)
- [A Survey on Interpretability in Visual Recognition](https://mlsanity.com/paper/2507.11099/) (2025)
- [Interpretable Multimodal Learning for Tumor Protein-Metal Binding: Progress, Challenges, and Perspectives](https://mlsanity.com/paper/2504.03847/) (2025)
- [Evaluating Webcam-based Gaze Data as an Alternative for Human Rationale Annotations](https://mlsanity.com/paper/2402.19133/) (2024)

## Abstract (arXiv)

This paper proposes a set of criteria to evaluate the objectiveness of explanation methods of neural networks, which is crucial for the development of explainable AI, but it also presents significant challenges. The core challenge is that people usually cannot obtain ground-truth explanations of the neural network. To this end, we design four metrics to evaluate explanation results without ground-truth explanations. Our metrics can be broadly applied to nine benchmark methods of interpreting neural networks, which provides new insights of explanation methods.