-
Hebbian Learning from First Principles
Authors:
Linda Albanese,
Adriano Barra,
Pierluigi Bianco,
Fabrizio Durante,
Diego Pallara
Abstract:
Recently, the original storage prescription for the Hopfield model of neural networks -- as well as for its dense generalizations -- has been turned into a genuine Hebbian learning rule by postulating the expression of its Hamiltonian for both the supervised and unsupervised protocols. In these notes, first, we obtain these explicit expressions by relying upon maximum entropy extremization à la Ja…
▽ More
Recently, the original storage prescription for the Hopfield model of neural networks -- as well as for its dense generalizations -- has been turned into a genuine Hebbian learning rule by postulating the expression of its Hamiltonian for both the supervised and unsupervised protocols. In these notes, first, we obtain these explicit expressions by relying upon maximum entropy extremization à la Jaynes. Beyond providing a formal derivation of these recipes for Hebbian learning, this construction also highlights how Lagrangian constraints within entropy extremization force network's outcomes on neural correlations: these try to mimic the empirical counterparts hidden in the datasets provided to the network for its training and, the denser the network, the longer the correlations that it is able to capture. Next, we prove that, in the big data limit, whatever the presence of a teacher (or its lacking), not only these Hebbian learning rules converge to the original storage prescription of the Hopfield model but also their related free energies (and, thus, the statistical mechanical picture provided by Amit, Gutfreund and Sompolinsky is fully recovered). As a sideline, we show mathematical equivalence among standard Cost functions (Hamiltonian), preferred in Statistical Mechanical jargon, and quadratic Loss Functions, preferred in Machine Learning terminology. Remarks on the exponential Hopfield model (as the limit of dense networks with diverging density) and semi-supervised protocols are also provided.
△ Less
Submitted 13 January, 2024;
originally announced January 2024.
-
LSA-T: The first continuous Argentinian Sign Language dataset for Sign Language Translation
Authors:
Pedro Dal Bianco,
Gastón Ríos,
Franco Ronchetti,
Facundo Quiroga,
Oscar Stanchi,
Waldo Hasperué,
Alejandro Rosete
Abstract:
Sign language translation (SLT) is an active field of study that encompasses human-computer interaction, computer vision, natural language processing and machine learning. Progress on this field could lead to higher levels of integration of deaf people. This paper presents, to the best of our knowledge, the first continuous Argentinian Sign Language (LSA) dataset. It contains 14,880 sentence level…
▽ More
Sign language translation (SLT) is an active field of study that encompasses human-computer interaction, computer vision, natural language processing and machine learning. Progress on this field could lead to higher levels of integration of deaf people. This paper presents, to the best of our knowledge, the first continuous Argentinian Sign Language (LSA) dataset. It contains 14,880 sentence level videos of LSA extracted from the CN Sordos YouTube channel with labels and keypoints annotations for each signer. We also present a method for inferring the active signer, a detailed analysis of the characteristics of the dataset, a visualization tool to explore the dataset and a neural SLT model to serve as baseline for future experiments.
△ Less
Submitted 14 November, 2022;
originally announced November 2022.
-
Quantum Technologies in Space
Authors:
Rainer Kaltenbaek,
Antonio Acin,
Laszlo Bacsardi,
Paolo Bianco,
Philippe Bouyer,
Eleni Diamanti,
Christoph Marquardt,
Yasser Omar,
Valerio Pruneri,
Ernst Rasel,
Bernhard Sang,
Stephan Seidel,
Hendrik Ulbricht,
Rupert Ursin,
Paolo Villoresi,
Mathias van den Bossche,
Wolf von Klitzing,
Hugo Zbinden,
Mauro Paternostro,
Angelo Bassi
Abstract:
Recently, the European Commission supported by many European countries has announced large investments towards the commercialization of quantum technology (QT) to address and mitigate some of the biggest challenges facing today's digital era - e.g. secure communication and computing power. For more than two decades the QT community has been working on the development of QTs, which promise landmark…
▽ More
Recently, the European Commission supported by many European countries has announced large investments towards the commercialization of quantum technology (QT) to address and mitigate some of the biggest challenges facing today's digital era - e.g. secure communication and computing power. For more than two decades the QT community has been working on the development of QTs, which promise landmark breakthroughs leading to commercialization in various areas. The ambitious goals of the QT community and expectations of EU authorities cannot be met solely by individual initiatives of single countries, and therefore, require a combined European effort of large and unprecedented dimensions comparable only to the Galileo or Copernicus programs. Strong international competition calls for a coordinated European effort towards the development of QT in and for space, including research and development of technology in the areas of communication and sensing. Here, we aim at summarizing the state of the art in the development of quantum technologies which have an impact in the field of space applications. Our goal is to outline a complete framework for the design, development, implementation, and exploitation of quantum technology in space.
△ Less
Submitted 3 July, 2021;
originally announced July 2021.
-
Kinetics of double stranded DNA overstretching revealed by 0.5-2 pN force steps
Authors:
Pasquale Bianco,
Lorenzo Bongini,
Luca Melli,
Mario Dolfi,
Vincenzo Lombardi
Abstract:
A detailed description of the conformational plasticity of double stranded DNA (ds) is a necessary framework for understanding protein-DNA interactions. Until now, however structure and kinetics of the transition from the basic conformation of ds-DNA (B state) to the 1.7 times longer and partially unwound conformation (S state) have not been defined. The force-extension relation of the ds-DNA of l…
▽ More
A detailed description of the conformational plasticity of double stranded DNA (ds) is a necessary framework for understanding protein-DNA interactions. Until now, however structure and kinetics of the transition from the basic conformation of ds-DNA (B state) to the 1.7 times longer and partially unwound conformation (S state) have not been defined. The force-extension relation of the ds-DNA of lambda-phage is measured here with unprecedented resolution using a dual laser optical tweezers that can impose millisecond force steps of 0.5-2 pN (25 C). This approach reveals the kinetics of the transition between intermediate states of ds-DNA and uncovers the load-dependence of the rate constant of the unitary reaction step. DNA overstretching transition results essentially a two-state reaction composed of 5.85 nm steps, indicating cooperativity of ~25 base pairs. This mechanism increases the free energy for the unitary reaction to ~94 kBT, accounting for the stability of the basic conformation of DNA, and explains the absence of hysteresis in the force-extension relation at equilibrium. The novel description of the kinetics and energetics of the B-S transition of ds-DNA improves our understanding the biological role of the S state in the interplay between mechanics and enzymology of the DNA-protein machinery.
△ Less
Submitted 3 March, 2011;
originally announced March 2011.
