DeepACTIF: Efficient Feature Attribution via Activation Traces in Neural Sequence Models

• URL: http://arxiv.org/abs/2509.19362v1
• Date: Thu, 18 Sep 2025 15:47:05 GMT
• Title: DeepACTIF: Efficient Feature Attribution via Activation Traces in Neural Sequence Models
• Authors: Benedikt W. Hosp,
• Abstract summary: Feature attribution is essential for interpreting deep learning models in time-series domains such as healthcare, biometrics, and human-AI interaction.<n>Standard attribution methods, such as Integrated Gradients or SHAP, are computationally intensive and not well-suited for real-time applications.<n>We present DeepACTIF, a lightweight and architecture-aware feature attribution method that leverages internal activations of sequence models to estimate feature importance efficiently.
• Score: 0.0
• License: http://creativecommons.org/licenses/by-nc-sa/4.0/
• Abstract: Feature attribution is essential for interpreting deep learning models, particularly in time-series domains such as healthcare, biometrics, and human-AI interaction. However, standard attribution methods, such as Integrated Gradients or SHAP, are computationally intensive and not well-suited for real-time applications. We present DeepACTIF, a lightweight and architecture-aware feature attribution method that leverages internal activations of sequence models to estimate feature importance efficiently. Focusing on LSTM-based networks, we introduce an inverse-weighted aggregation scheme that emphasises stability and magnitude of activations across time steps. Our evaluation across three biometric gaze datasets shows that DeepACTIF not only preserves predictive performance under severe feature reduction (top 10% of features) but also significantly outperforms established methods, including SHAP, IG, and DeepLIFT, in terms of both accuracy and statistical robustness. Using Wilcoxon signed-rank tests and effect size analysis, we demonstrate that DeepACTIF yields more informative feature rankings with significantly lower error across all top-k conditions (10 - 40%). Our experiments demonstrate that DeepACTIF not only reduces computation time and memory usage by orders of magnitude but also preserves model accuracy when using only top-ranked features. That makes DeepACTIF a viable solution for real-time interpretability on edge devices such as mobile XR headsets or embedded health monitors.

Related papers

• Interpolation-Driven Machine Learning Approaches for Plume Shine Dose Estimation: A Comparison of XGBoost, Random Forest, and TabNet [0.0]
  An assisted machine learning framework was developed for plume shine dose estimation.<n>The framework was developed using discrete dose datasets generated with the pyEIADOS suite for 17 gamma-emitting radionuclides.<n>Interpretability analysis using permutation importance and attention-based feature attribution revealed that performance differences stem from how the models utilize input features.
  arXiv  Detail & Related papers  (2026-02-23T08:12:49Z)
• Self-Supervised Learning via Flow-Guided Neural Operator on Time-Series Data [57.85958428020496]
  Flow-Guided Neural Operator (FGNO) is a novel framework combining operator learning with flow matching for SSL training.<n>FGNO learns mappings in functional spaces by using Short-Time Fourier Transform to unify different time resolutions.<n>Unlike prior generative SSL methods that use noisy inputs during inference, we propose using clean inputs for representation extraction while learning representations with noise.
  arXiv  Detail & Related papers  (2026-02-12T18:54:57Z)
• FAIM: Frequency-Aware Interactive Mamba for Time Series Classification [87.84511960413715]
  Time series classification (TSC) is crucial in numerous real-world applications, such as environmental monitoring, medical diagnosis, and posture recognition.<n>We propose FAIM, a lightweight Frequency-Aware Interactive Mamba model.<n>We show that FAIM consistently outperforms existing state-of-the-art (SOTA) methods, achieving a superior trade-off between accuracy and efficiency.
  arXiv  Detail & Related papers  (2025-11-26T08:36:33Z)
• Evaluating Ensemble and Deep Learning Models for Static Malware Detection with Dimensionality Reduction Using the EMBER Dataset [0.0]
  This study investigates the effectiveness of several machine learning algorithms for static malware detection using the EMBER dataset.<n>We evaluate eight classification models: LightGBM, XGBoost, CatBoost, Random Forest, Extra Trees, HistGradientBoosting, k-Nearest Neighbors (KNN), and TabNet.<n>The models are assessed on accuracy, precision, recall, F1 score, and AUC to examine both predictive performance and robustness.
  arXiv  Detail & Related papers  (2025-07-22T18:45:10Z)
• CAST: Cross-Attentive Spatio-Temporal feature fusion for Deepfake detection [0.0]
  CNNs are effective at capturing spatial artifacts, and Transformers excel at modeling temporal inconsistencies.<n>We propose a unified CAST model that leverages cross-attention to effectively fuse spatial and temporal features.<n>We evaluate the performance of our model using the FaceForensics++, Celeb-DF, and DeepfakeDetection datasets.
  arXiv  Detail & Related papers  (2025-06-26T18:51:17Z)
• MAWIFlow Benchmark: Realistic Flow-Based Evaluation for Network Intrusion Detection [47.86433139298671]
  This paper introduces MAWIFlow, a flow-based benchmark derived from the MAWILAB v1.1 dataset.<n>The resulting datasets comprise temporally distinct samples from January 2011, 2016, and 2021, drawn from trans-Pacific backbone traffic.<n>Traditional machine learning methods, including Decision Trees, Random Forests, XGBoost, and Logistic Regression, are compared to a deep learning model based on a CNN-BiLSTM architecture.
  arXiv  Detail & Related papers  (2025-06-20T14:51:35Z)
• Dynamic Graph Structure Estimation for Learning Multivariate Point Process using Spiking Neural Networks [14.77536193242342]
  Spiking Dynamic Graph Network is a novel framework that leverages the temporal processing capabilities of spiking neural networks (SNNs) and spike-dependent plasticity (STD-P)<n>It adapts to any dataset by learning dynamic-temporal dependencies directly from event data, enhancing generalizability and modeling.<n>Our evaluations conducted on both synthetic and real-world datasets including NYC Taxi, 911 Reddit, and Stack Overflow, demonstrate superior accuracy while maintaining computational efficiency.
  arXiv  Detail & Related papers  (2025-04-01T23:23:10Z)
• Task-Oriented Real-time Visual Inference for IoVT Systems: A Co-design Framework of Neural Networks and Edge Deployment [61.20689382879937]
  Task-oriented edge computing addresses this by shifting data analysis to the edge. Existing methods struggle to balance high model performance with low resource consumption. We propose a novel co-design framework to optimize neural network architecture.
  arXiv  Detail & Related papers  (2024-10-29T19:02:54Z)
• Uncertainty-Aware Deep Attention Recurrent Neural Network for Heterogeneous Time Series Imputation [0.25112747242081457]
  Missingness is ubiquitous in multivariate time series and poses an obstacle to reliable downstream analysis. We propose DEep Attention Recurrent Imputation (Imputation), which jointly estimates missing values and their associated uncertainty. Experiments show that I surpasses the SOTA in diverse imputation tasks using real-world datasets.
  arXiv  Detail & Related papers  (2024-01-04T13:21:11Z)
• Physics Inspired Hybrid Attention for SAR Target Recognition [61.01086031364307]
  We propose a physics inspired hybrid attention (PIHA) mechanism and the once-for-all (OFA) evaluation protocol to address the issues. PIHA leverages the high-level semantics of physical information to activate and guide the feature group aware of local semantics of target. Our method outperforms other state-of-the-art approaches in 12 test scenarios with same ASC parameters.
  arXiv  Detail & Related papers  (2023-09-27T14:39:41Z)
• Robust Learning with Progressive Data Expansion Against Spurious Correlation [65.83104529677234]
  We study the learning process of a two-layer nonlinear convolutional neural network in the presence of spurious features. Our analysis suggests that imbalanced data groups and easily learnable spurious features can lead to the dominance of spurious features during the learning process. We propose a new training algorithm called PDE that efficiently enhances the model's robustness for a better worst-group performance.
  arXiv  Detail & Related papers  (2023-06-08T05:44:06Z)
• On Efficient Uncertainty Estimation for Resource-Constrained Mobile Applications [0.0]
  Predictive uncertainty supplements model predictions and enables improved functionality of downstream tasks. We tackle this problem by building upon Monte Carlo Dropout (MCDO) models using the Axolotl framework. We conduct experiments on (1) a multi-class classification task using the CIFAR10 dataset, and (2) a more complex human body segmentation task.
  arXiv  Detail & Related papers  (2021-11-11T22:24:15Z)
• AQD: Towards Accurate Fully-Quantized Object Detection [94.06347866374927]
  We propose an Accurate Quantized object Detection solution, termed AQD, to get rid of floating-point computation. Our AQD achieves comparable or even better performance compared with the full-precision counterpart under extremely low-bit schemes.
  arXiv  Detail & Related papers  (2020-07-14T09:07:29Z)
• Influence Functions in Deep Learning Are Fragile [52.31375893260445]
  influence functions approximate the effect of samples in test-time predictions. influence estimates are fairly accurate for shallow networks. Hessian regularization is important to get highquality influence estimates.
  arXiv  Detail & Related papers  (2020-06-25T18:25:59Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.