{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,2,12]],"date-time":"2026-02-12T06:15:42Z","timestamp":1770876942317,"version":"3.50.1"},"reference-count":21,"publisher":"MDPI AG","issue":"2","license":[{"start":{"date-parts":[[2026,1,31]],"date-time":"2026-01-31T00:00:00Z","timestamp":1769817600000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["Entropy"],"abstract":"<jats:p>We introduce Lehmer Activation Units (LAUs), a class of aggregation-based neural activations derived from the Lehmer transform that unify feature weighting and nonlinearity within a single differentiable operator. Unlike conventional pointwise activations, LAUs operate on collections of features and adapt their aggregation behavior through learnable parameters, yielding intrinsically interpretable representations. We develop both real-valued and complex-valued formulations, with the complex extension enabling phase-sensitive interactions and enhanced expressive capacity. We establish a universal approximation theorem for LAU-based networks, providing formal guarantees of expressive completeness. Empirically, we show that LAUs enable highly compact architectures to achieve strong predictive performance under tightly controlled experimental settings, demonstrating that expressive power can be concentrated within individual neurons rather than architectural depth. These results position LAUs as a principled, interpretable, and efficient alternative to conventional activation functions.<\/jats:p>","DOI":"10.3390\/e28020157","type":"journal-article","created":{"date-parts":[[2026,2,2]],"date-time":"2026-02-02T09:48:08Z","timestamp":1770025688000},"page":"157","update-policy":"https:\/\/doi.org\/10.3390\/mdpi_crossmark_policy","source":"Crossref","is-referenced-by-count":0,"title":["Compact and Interpretable Neural Networks Using Lehmer Activation Units"],"prefix":"10.3390","volume":"28","author":[{"ORCID":"https:\/\/orcid.org\/0000-0002-1114-1760","authenticated-orcid":false,"given":"Masoud","family":"Ataei","sequence":"first","affiliation":[{"name":"Department of Mathematical and Computational Sciences, University of Toronto, Mississauga, ON L5L 1C6, Canada"}]},{"given":"Sepideh","family":"Forouzi","sequence":"additional","affiliation":[{"name":"Department of Computer Science, University of Colorado Boulder, Boulder, CO 80309, USA"}]},{"given":"Xiaogang","family":"Wang","sequence":"additional","affiliation":[{"name":"Department of Mathematics and Statistics, York University, Toronto, ON M3J 1P3, Canada"}]}],"member":"1968","published-online":{"date-parts":[[2026,1,31]]},"reference":[{"key":"ref_1","doi-asserted-by":"crossref","first-page":"3663","DOI":"10.1038\/s41598-022-07413-y","article-title":"Theory of Lehmer transform and its applications in identifying the electroencephalographic signature of major depressive disorder","volume":"12","author":"Ataei","year":"2022","journal-title":"Sci. Rep."},{"key":"ref_2","doi-asserted-by":"crossref","first-page":"741","DOI":"10.1109\/TRPMS.2021.3066428","article-title":"On interpretability of artificial neural networks: A survey","volume":"5","author":"Fan","year":"2021","journal-title":"IEEE Trans. Radiat. Plasma Med. Sci."},{"key":"ref_3","doi-asserted-by":"crossref","first-page":"206","DOI":"10.1038\/s42256-019-0048-x","article-title":"Stop explaining black box machine learning models for high stakes decisions and use interpretable models instead","volume":"1","author":"Rudin","year":"2019","journal-title":"Nat. Mach. Intell."},{"key":"ref_4","unstructured":"Doshi-Velez, F., and Kim, B. (2017). Towards a rigorous science of interpretable machine learning. arXiv."},{"key":"ref_5","unstructured":"Simonyan, K., and Zisserman, A. (2014). Very deep convolutional networks for large-scale image recognition. arXiv."},{"key":"ref_6","doi-asserted-by":"crossref","first-page":"211","DOI":"10.1016\/j.patcog.2016.11.008","article-title":"Explaining nonlinear classification decisions with deep Taylor decomposition","volume":"65","author":"Montavon","year":"2017","journal-title":"Pattern Recognit."},{"key":"ref_7","doi-asserted-by":"crossref","unstructured":"Ribeiro, M.T., Singh, S., and Guestrin, C. (2016, January 13\u201317). Why should I trust you? Explaining the predictions of any classifier. Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, San Francisco, CA, USA.","DOI":"10.1145\/2939672.2939778"},{"key":"ref_8","unstructured":"Bahdanau, D., Cho, K., and Bengio, Y. (2015, January 7\u20139). Neural machine translation by jointly learning to align and translate. Proceedings of the International Conference on Learning Representations (ICLR), San Diego, CA, USA."},{"key":"ref_9","unstructured":"Vaswani, A., Shazeer, N., Parmar, N., Uszkoreit, J., Jones, L., Gomez, A.N., Kaiser, L., and Polosukhin, I. (2017). Attention is all you need. arXiv."},{"key":"ref_10","unstructured":"Alvarez-Melis, D., and Jaakkola, T. (2018). Towards interpretable reinforcement learning models with self-explaining neural networks. Advances in Neural Information Processing Systems (NeurIPS), Curran Associates, Inc."},{"key":"ref_11","unstructured":"Kim, B., Wattenberg, M., Gilmer, J., Cai, C., Wexler, J., Viegas, F., and Sayres, R. (2018, January 10\u201315). Interpretability beyond feature attribution: Quantitative testing with concept activation vectors (TCAV). Proceedings of the International Conference on Machine Learning (ICML), Stockholm, Sweden."},{"key":"ref_12","unstructured":"Kipf, T.N., and Welling, M. (2017, January 24\u201326). Semi-supervised classification with graph convolutional networks. Proceedings of the International Conference on Learning Representations (ICLR), Toulon, France."},{"key":"ref_13","doi-asserted-by":"crossref","unstructured":"Hirose, A., and Yoshida, S. (2012). Complex-Valued Neural Networks: Advances and Applications, John Wiley & Sons.","DOI":"10.1007\/978-3-642-27632-3"},{"key":"ref_14","unstructured":"Trabelsi, C., Bilaniuk, O., Zhang, Y., Serdyuk, D., Subramanian, S., Santos, J.F., Mehri, S., Rostamzadeh, N., Bengio, Y., and Pal, C.J. (May, January 30). Deep complex networks. Proceedings of the International Conference on Learning Representations (ICLR), Vancouver, BC, Canada."},{"key":"ref_15","first-page":"1013","article-title":"The complex gradient operator and the Wirtinger calculus","volume":"58","year":"2009","journal-title":"IEEE Trans. Signal Process."},{"key":"ref_16","first-page":"147","article-title":"Learning complex-valued neural networks with adaptive activation functions","volume":"423","author":"Zhang","year":"2021","journal-title":"Neurocomputing"},{"key":"ref_17","first-page":"4237","article-title":"Complex-valued neural networks for PolSAR image classification","volume":"58","author":"Chen","year":"2020","journal-title":"IEEE Trans. Geosci. Remote Sens."},{"key":"ref_18","doi-asserted-by":"crossref","first-page":"104862","DOI":"10.1016\/j.bspc.2023.104862","article-title":"A hybrid complex-valued neural network framework with applications to electroencephalogram (EEG)","volume":"85","author":"Du","year":"2023","journal-title":"Biomed. Signal Process. Control"},{"key":"ref_19","first-page":"3567","article-title":"Complex-valued neural networks for signal processing","volume":"69","author":"Amin","year":"2021","journal-title":"IEEE Trans. Signal Process."},{"key":"ref_20","unstructured":"Bullen, P.S. (2013). Handbook of Means and Their Inequalities, Springer Science & Business Media."},{"key":"ref_21","doi-asserted-by":"crossref","first-page":"5510","DOI":"10.22436\/jnsa.009.10.02","article-title":"Schur-convexity for Lehmer mean of n variables","volume":"9","author":"Fub","year":"2016","journal-title":"J. Nonlinear Sci. Appl."}],"container-title":["Entropy"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/www.mdpi.com\/1099-4300\/28\/2\/157\/pdf","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,2,12]],"date-time":"2026-02-12T05:22:12Z","timestamp":1770873732000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.mdpi.com\/1099-4300\/28\/2\/157"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,1,31]]},"references-count":21,"journal-issue":{"issue":"2","published-online":{"date-parts":[[2026,2]]}},"alternative-id":["e28020157"],"URL":"https:\/\/doi.org\/10.3390\/e28020157","relation":{},"ISSN":["1099-4300"],"issn-type":[{"value":"1099-4300","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,1,31]]}}}