Luca Pedrelli wins Best Paper Award at the international workshop “Deep Learning Meets Neuromorphic Hardware”

Luca Pedrelli wins Best Paper Award at the international workshop "Deep Learning Meets Neuromorphic Hardware"

Luca Pedrelli, a research fellow at the TeCIP Institute (Telecommunications, Computer Engineering and Photonics) of the Sant'Anna School of Advanced Studies in Pisa, has received the Best Paper Award at the 4th International Workshop on Deep Learning Meets Neuromorphic Hardware, held in Naples on 7 September 2026 as part of ECML-PKDD 2026, one of the leading international conferences in the fields of machine learning and knowledge discovery in data.
The award, sponsored by APL Machine Learning, an AIP Publishing journal, was presented by an award committee chaired by Daniel Brunner (CNRS, Institut FEMTO-ST, Besançon) to Pedrelli in his capacity as the paper's presenting author.

Pedrelli was recognised for the study entitled "Random Attention Memory for Online Generative Reservoir Computing: a Preliminary Study", co-authored with Stefano Dettori and Valentina Colla.
The paper introduces a generative artificial intelligence model for time-series processing based on a random attention mechanism. Unlike the most widely used deep learning models, in which the entire attention component needs to be trained, the proposed approach keeps the attention mechanism random and untrained, while learning only a small portion of the overall model.
This approach makes it possible to drastically reduce energy consumption and computational requirements, while maintaining performance competitive with leading state-of-the-art models.