Showing 1–7 of 7 results for author: Severino, S

Vortex flow anisotropy in nematic superconductors

Authors: F. Castillo Menegotto, R. S. Severino, P. D. Mininni, E. Fradkin, V. Bekeris, G. Pasquini, G. S. Lozano

Abstract: We investigate the vortex flow anisotropy in the mixed state of nematic superconductors, focusing on the effects of nematic-superconducting coupling on vortex dynamics. Using numerical simulations within a time-dependent Ginzburg-Landau (TDGL) approach, we analyze vortex viscosity in a model featuring an s-wave superconducting order parameter coupled to an Ising-like nematic order parameter, suita… ▽ More We investigate the vortex flow anisotropy in the mixed state of nematic superconductors, focusing on the effects of nematic-superconducting coupling on vortex dynamics. Using numerical simulations within a time-dependent Ginzburg-Landau (TDGL) approach, we analyze vortex viscosity in a model featuring an s-wave superconducting order parameter coupled to an Ising-like nematic order parameter, suitable for systems with $C_4$ symmetry. Our results indicate that nematicity induces a significant anisotropy in the flux-flow resistivity, which depends on both vortex core shape anisotropy and normal-phase conductivity anisotropy. These two effects can either compete or cooperate with each other. We discuss the implications of these findings for identifying nematic superconductivity in the superconducting phase. Our work provides new insights into the interplay between nematic and superconducting order parameters, leading to new possibilities for experimental and theoretical exploration of anisotropic transport properties in unconventional superconductors. △ Less

Submitted 30 April, 2025; originally announced April 2025.

Enhanced optical conductivity and many-body effects in strongly-driven photo-excited semi-metallic graphite

Authors: T. P. H. Sidiropoulos, N. Di Palo, D. E. Rivas, A. Summers, S. Severino, M. Reduzzi, J. Biegert

Abstract: The excitation of quasi-particles near the extrema of the electronic band structure is a gateway to electronic phase transitions in condensed matter. In a many-body system, quasi-particle dynamics are strongly influenced by the electronic single-particle structure and have been extensively studied in the weak optical excitation regime. Yet, under strong optical excitation, where light fields coher… ▽ More The excitation of quasi-particles near the extrema of the electronic band structure is a gateway to electronic phase transitions in condensed matter. In a many-body system, quasi-particle dynamics are strongly influenced by the electronic single-particle structure and have been extensively studied in the weak optical excitation regime. Yet, under strong optical excitation, where light fields coherently drive carriers, the dynamics of many-body interactions that can lead to new quantum phases remain largely unresolved. Here, we induce such a highly non-equilibrium many-body state through strong optical excitation of charge carriers near the van Hove singularity in graphite. We investigate the system's evolution into a strongly-driven photo-excited state with attosecond soft X-ray core-level spectroscopy. Surprisingly, we find an enhancement of the optical conductivity of nearly ten times the quantum conductivity and pinpoint it to carrier excitations in flat bands. This interaction regime is robust against carrier-carrier interaction with coherent optical phonons acting as an attractive force reminiscent of superconductivity. The strongly-driven non-equilibrium state is markedly different from the single-particle structure and macroscopic conductivity and is a consequence of the non-adiabatic many-body state. △ Less

Submitted 11 August, 2023; originally announced August 2023.

Accelerating Transformer Inference for Translation via Parallel Decoding

Authors: Andrea Santilli, Silvio Severino, Emilian Postolache, Valentino Maiorca, Michele Mancusi, Riccardo Marin, Emanuele Rodolà

Abstract: Autoregressive decoding limits the efficiency of transformers for Machine Translation (MT). The community proposed specific network architectures and learning-based methods to solve this issue, which are expensive and require changes to the MT model, trading inference speed at the cost of the translation quality. In this paper, we propose to address the problem from the point of view of decoding a… ▽ More Autoregressive decoding limits the efficiency of transformers for Machine Translation (MT). The community proposed specific network architectures and learning-based methods to solve this issue, which are expensive and require changes to the MT model, trading inference speed at the cost of the translation quality. In this paper, we propose to address the problem from the point of view of decoding algorithms, as a less explored but rather compelling direction. We propose to reframe the standard greedy autoregressive decoding of MT with a parallel formulation leveraging Jacobi and Gauss-Seidel fixed-point iteration methods for fast inference. This formulation allows to speed up existing models without training or modifications while retaining translation quality. We present three parallel decoding algorithms and test them on different languages and models showing how the parallelization introduces a speedup up to 38% w.r.t. the standard autoregressive decoding and nearly 2x when scaling the method on parallel resources. Finally, we introduce a decoding dependency graph visualizer (DDGviz) that let us see how the model has learned the conditional dependence between tokens and inspect the decoding procedure. △ Less

Submitted 17 May, 2023; originally announced May 2023.

Comments: Accepted at ACL 2023 main conference

Sparse Vicious Attacks on Graph Neural Networks

Authors: Giovanni Trappolini, Valentino Maiorca, Silvio Severino, Emanuele Rodolà, Fabrizio Silvestri, Gabriele Tolomei

Abstract: Graph Neural Networks (GNNs) have proven to be successful in several predictive modeling tasks for graph-structured data. Amongst those tasks, link prediction is one of the fundamental problems for many real-world applications, such as recommender systems. However, GNNs are not immune to adversarial attacks, i.e., carefully crafted malicious examples that are designed to fool the predictive mo… ▽ More Graph Neural Networks (GNNs) have proven to be successful in several predictive modeling tasks for graph-structured data. Amongst those tasks, link prediction is one of the fundamental problems for many real-world applications, such as recommender systems. However, GNNs are not immune to adversarial attacks, i.e., carefully crafted malicious examples that are designed to fool the predictive model. In this work, we focus on a specific, white-box attack to GNN-based link prediction models, where a malicious node aims to appear in the list of recommended nodes for a given target victim. To achieve this goal, the attacker node may also count on the cooperation of other existing peers that it directly controls, namely on the ability to inject a number of ``vicious'' nodes in the network. Specifically, all these malicious nodes can add new edges or remove existing ones, thereby perturbing the original graph. Thus, we propose SAVAGE, a novel framework and a method to mount this type of link prediction attacks. SAVAGE formulates the adversary's goal as an optimization task, striking the balance between the effectiveness of the attack and the sparsity of malicious resources required. Extensive experiments conducted on real-world and synthetic datasets demonstrate that adversarial attacks implemented through SAVAGE indeed achieve high attack success rate yet using a small amount of vicious nodes. Finally, despite those attacks require full knowledge of the target model, we show that they are successfully transferable to other black-box methods for link prediction. △ Less

Submitted 20 September, 2022; originally announced September 2022.

Non-Adiabatic Electronic and Vibrational Ring-Opening Dynamics resolved with Attosecond Core-Level Spectroscopy

Authors: S. Severino, K. M. Ziems, M. Reduzzi, A. Summers, H. -W. Sun, Y. -H. Chien, S. Gräfe, J. Biegert

Abstract: Non-adiabatic dynamics and conical intersections play a central role in the chemistry of most polyatomic molecules, ranging from isomerization to heterocyclic ring opening and avoided photo-damage of DNA. Studying the underpinning correlated dynamics of electronic and nuclear wave packets is a major challenge in real-time and, many times involves optically dark transient states. We show that attos… ▽ More Non-adiabatic dynamics and conical intersections play a central role in the chemistry of most polyatomic molecules, ranging from isomerization to heterocyclic ring opening and avoided photo-damage of DNA. Studying the underpinning correlated dynamics of electronic and nuclear wave packets is a major challenge in real-time and, many times involves optically dark transient states. We show that attosecond core-level spectroscopy reveals the pathway dynamics of neutral furan across its conical intersections and dark states. Our method measures electronic-nuclear correlations to detect the dephasing of electronic coherence due to nuclear motion and identifies the ring-opened isomer as the dominant product. These results demonstrate the efficacy of attosecond core level spectroscopy as a potent method to investigate the real-time dynamics of photochemical reaction pathways in complex molecular systems. △ Less

Submitted 7 October, 2022; v1 submitted 9 September, 2022; originally announced September 2022.

Journal ref: Nat. Photon. 18, 731-737 (2024)

Vortices in a Ginzburg Landau Theory of Superconductors with Nematic Order

Authors: Ramiro Sebastián Severino, Pablo Daniel Mininni, Eduardo Fradkin, Victoria Bekeris, Gabriela Pasquini, Gustavo Sergio Lozano

Abstract: In this work we explore the interplay between superconductivity and nematicity in the framework of a Ginzburg Landau theory with a nematic order parameter coupled to the superconductor order parameter, often used in the description of superconductivity of Fe based materials. In particular, we focus on the study of the vortex-vortex interaction in order to determine the way nematicity affects its a… ▽ More In this work we explore the interplay between superconductivity and nematicity in the framework of a Ginzburg Landau theory with a nematic order parameter coupled to the superconductor order parameter, often used in the description of superconductivity of Fe based materials. In particular, we focus on the study of the vortex-vortex interaction in order to determine the way nematicity affects its attractive or repulsive character. To do so, we use a dynamical method based on the solutions of the Time Dependent Ginzburg Landau equations in a bulk superconductor. An important contribution of our work is the implementation of a pseudo-spectral method to solve the dynamics, known to be highly efficient and of very high order in comparison to the usual finite differences/elements methods. The coupling between the superconductor and the (real) nematic order parameters is represented by two terms in the free energy: a biquadratic term and a coupling of the nematic order parameter to the covariant derivatives of the superconductor order parameter. Our results show that there is a competing effect: while the former independently of its competitive or cooperative character generates an attractive vortex-vortex interaction, the latter always generates a repulsive interaction. △ Less

Submitted 25 April, 2022; v1 submitted 14 April, 2022; originally announced April 2022.

Comments: 37 pages; 12 figures; corrected typos; added references

Probing the energy conversion pathways between light, carriers and lattice in real time with attosecond core-level spectroscopy

Authors: T. P. H. Sidiropoulos, N. Di Palo, D. E. Rivas, S. Severino, M. Reduzzi, B. Nandy, B. Bauerhenne, S. Krylow, T. Vasileiadis, T. Danz, P. Elliott, S. Sharma, K. Dewhurst, C. Ropers, Y. Joly, K. M. E. Garcia, M. Wolf, R. Ernstorfer, J. Biegert

Abstract: Detection of the energy conversion pathways, between photons, charge carriers, and the lattice is of fundamental importance to understand fundamental physics and to advance materials and devices. Yet, such insight remains incomplete due to experimental challenges in disentangling the various signatures on overlapping time scales. Here, we show that attosecond core-level X-ray spectroscopy can iden… ▽ More Detection of the energy conversion pathways, between photons, charge carriers, and the lattice is of fundamental importance to understand fundamental physics and to advance materials and devices. Yet, such insight remains incomplete due to experimental challenges in disentangling the various signatures on overlapping time scales. Here, we show that attosecond core-level X-ray spectroscopy can identify these interactions with attosecond precision and across a picosecond range. We demonstrate this methodology on graphite since its investigation is complicated by a variety of mechanisms occurring across a wide range of temporal scales. Our methodology reveals, through the simultaneous real-time detection of electrons and holes, the different dephasing mechanisms for each carrier type dependent on excitation with few-cycle-duration light fields. These results demonstrate the general ability of our methodology to detect and distinguish the various dynamic contributions to the flow of energy inside materials on their native time scales. △ Less

Submitted 13 October, 2021; originally announced October 2021.