# Improving Interpretability of Deep Neural Networks with Semantic Information

arXiv: https://arxiv.org/abs/1703.04096 · HTML: https://mlsanity.com/paper/1703.04096/

## Key facts

- Cited by 15 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 12 Mar 2017.

## Built on by (most cited first, 15 of 15)

- [From Anecdotal Evidence to Quantitative Evaluation Methods: A Systematic Review on Evaluating Explainable AI](https://mlsanity.com/paper/2201.08164/) (2022)
- [Interpretable Deep Learning: Interpretation, Interpretability, Trustworthiness, and Beyond](https://mlsanity.com/paper/2103.10689/) (2021)
- [LLMs for Explainable AI: A Comprehensive Survey](https://mlsanity.com/paper/2504.00125/) (2025)
- [Multimodal Explainable Artificial Intelligence: A Comprehensive Review of Methodological Advances and Future Research Directions](https://mlsanity.com/paper/2306.05731/) (2023)
- [Video Captioning: a comparative review of where we are and which could be the route](https://mlsanity.com/paper/2204.05976/) (2022)
- [Human-Understandable Decision Making for Visual Recognition](https://mlsanity.com/paper/2103.03429/) (2021)
- [NeuroView-RNN: It's About Time](https://mlsanity.com/paper/2202.11811/) (2022)
- [Joint framework with deep feature distillation and adaptive focal loss for weakly supervised audio tagging and acoustic event detection](https://mlsanity.com/paper/2103.12388/) (2021)
- [Explainability-aided Domain Generalization for Image Classification](https://mlsanity.com/paper/2104.01742/) (2021)
- [Leveraging Causal Inference for Explainable Automatic Program Repair](https://mlsanity.com/paper/2205.13342/) (2022)
- [Beyond One-Hot-Encoding: Injecting Semantics to Drive Image Classifiers](https://mlsanity.com/paper/2308.00607/) (2023)
- [On Generating Monolithic and Model Reconciling Explanations in Probabilistic Scenarios](https://mlsanity.com/paper/2405.19229/) (2024)
- [Two Approaches to Supervised Image Segmentation](https://mlsanity.com/paper/2307.10123/) (2023)
- [Learning Decomposable and Debiased Representations via Attribute-Centric Information Bottlenecks](https://mlsanity.com/paper/2403.14140/) (2024)
- [Rethinking the Harmonic Loss via Non-Euclidean Distance Layers](https://mlsanity.com/paper/2603.10225/) (2026)

## Abstract (arXiv)

Interpretability of deep neural networks (DNNs) is essential since it enables users to understand the overall strengths and weaknesses of the models, conveys an understanding of how the models will behave in the future, and how to diagnose and correct potential problems. However, it is challenging to reason about what a DNN actually does due to its opaque or black-box nature. To address this issue, we propose a novel technique to improve the interpretability of DNNs by leveraging the rich semantic information embedded in human descriptions. By concentrating on the video captioning task, we first extract a set of semantically meaningful topics from the human descriptions that cover a wide range of visual concepts, and integrate them into the model with an interpretive loss. We then propose a prediction difference maximization algorithm to interpret the learned features of each neuron. Experimental results demonstrate its effectiveness in video captioning using the interpretable features, which can also be transferred to video action recognition. By clearly understanding the learned features, users can easily revise false predictions via a human-in-the-loop procedure.