Explore the vast range of titles available.

Detectors of microwave photons find applications in different fields ranging from security to cosmology. Due to the intrinsic difficulties related to the detection of vanishingly small energy quanta ℏ ω , significant portions of the microwave electromagnetic spectrum are still uncovered by suitable techniques. No prevailing technology has clearly emerged yet, although different solutions have been tested in different contexts. Here, we focus on semiconductor quantum dots, which feature wide tunability by external gate voltages and scalability for large architectures. We discuss possible pathways for the development of microwave photon detectors based on photon-assisted tunneling in semiconducting double quantum dot circuits. In particular, we consider implementations based on either broadband transmission lines or resonant cavities, and we discuss how developments in charge sensing techniques and hybrid architectures may be beneficial for the development of efficient photon detectors in the microwave range.

With downscaling of electronic circuits, components based on semiconductor quantum dots are assuming increasing relevance for future technologies. Their response under external stimuli intrinsically depend on their quantum properties. Here we investigate single-electron tunneling in hard-wall InAs/InP nanowires in the presence of an off-resonant microwave drive. Our heterostructured nanowires include InAs quantum dots (QDs) and exhibit different tunnel-current regimes. In particular, for source-drain bias up to few mV Coulomb diamonds spread with increasing contrast as a function of microwave power and present multiple current polarity reversals. This behavior can be modelled in terms of voltage fluctuations induced by the microwave field and presents features that depend on the interplay of the discrete energy levels that contribute to the tunneling process.

In this work, we show the design of a silicon photonic-based polarization converting device based on the integration of semiconduction InP nanowires on the silicon photonic platform. We present a comprehensive numerical analysis showing that full polarization conversion (from quasi-TE modes to quasi-TM modes, and vice versa) can be achieved in devices exhibiting small footprints (total device lengths below 20 µm) with minimal power loss (<2 dB). The approach described in this work can pave the way to the realization of complex and re-configurable photonic processors based on the manipulation of the state of polarization of guided light beams.

We numerically investigated the use of graphene nanoribbons placed on top of silicon-on-insulator (SOI) strip waveguides for light polarization control in silicon photonic-integrated waveguides. We found that two factors mainly affected the polarization control: the graphene chemical potential and the geometrical parameters of the waveguide, such as the waveguide and nanoribbon widths and distance. We show that the graphene chemical potential influences both TE and TM polarizations almost in the same way, while the waveguide width tapering enables both TE-pass and TM-pass polarizing functionalities. Overall, by increasing the oxide spacer thickness between the silicon waveguide and the top graphene layer, the device insertion losses can be reduced, while preserving a high polarization extinction ratio.

Since the 1980s, researchers have taken giant steps in understanding how to use quantum mechanics for solving real problems—for example, making a computer that works according to the laws of quantum mechanics. In recent decades, researchers have tried to develop a platform for quantum information and computation that can be integrated into digital and telecom technologies without the need of a cryogenic environment. The current status of research in the field of quantum integrated photonics will be reviewed. A review of the most common integrated photonic platforms will be given, together with the main achievements and results in the last decade.

Single-photon detectors have gained significant attention recently, driven by advancements in quantum information technology. Applications such as quantum key distribution, quantum cryptography, and quantum computation demand the ability to detect individual quanta of light and distinguish between single-photon states and multi-photon states, particularly when operating within waveguide systems. Although single-photon detector fabrication has been established for some time, integrating detectors with waveguides using new materials with suitable structural and electronic properties, especially at telecommunication wavelengths, creates more compact source-line-detector systems. This review explores the state of the art of single-photon detector research and examines the potential breakthroughs offered by novel low-dimensional materials in this field.

Integrated photonic platforms have rapidly emerged as highly promising and extensively investigated systems for advancing classical and quantum information technologies, since their ability to seamlessly integrate photonic components within the telecommunication band with existing silicon-based industrial processes offers significant advantages. However, despite this integration facilitating the development of novel devices, fostering fast and reliable communication protocols and the manipulation of quantum information, traditional integrated silicon photonics faces inherent physical limitations that necessitate a challenging trade-off between device efficiency and spatial footprint. To address this issue, researchers are focusing on the integration of nanoscale materials into photonic platforms, offering a novel approach to enhance device performance while reducing spatial requirements. These developments are of paramount importance in both classical and quantum information technologies, potentially revolutionizing the industry. In this review, we explore the latest endeavors in hybrid photonic platforms leveraging the combination of integrated silicon photonic platforms and nanoscale materials, allowing for the unlocking of increased device efficiency and compact form factors. Finally, we provide insights into future developments and the evolving landscape of hybrid integrated photonic nanomaterial platforms.

With downscaling of electronic circuits, components based on semiconductor quantum dots are assuming increasing relevance for future technologies. Their response under external stimuli intrinsically depend on their quantum properties. Here we investigate single-electron tunneling in hard-wall InAs/InP nanowires in the presence of an off-resonant microwave drive. Our heterostructured nanowires include InAs quantum dots (QDs) and exhibit different tunnel-current regimes. In particular, for source-drain bias up to few mV Coulomb diamonds spread with increasing contrast as a function of microwave power and present multiple current polarity reversals. This behavior can be modelled in terms of voltage fluctuations induced by the microwave field and presents features that depend on the interplay of the discrete energy levels that contribute to the tunneling process.

Machine Unlearning is an emerging paradigm for selectively removing the impact of training datapoints from a network. Unlike existing methods that target a limited subset or a single class, our framework unlearns all classes in a single round. We achieve this by modulating the network's components using memory matrices, enabling the network to demonstrate selective unlearning behavior for any class after training. By discovering weights that are specific to each class, our approach also recovers a representation of the classes which is explainable by design. We test the proposed framework on small- and medium-scale image classification datasets, with both convolution- and Transformer-based backbones, showcasing the potential for explainable solutions through unlearning.

Large Language Models (LLMs) are increasingly evaluated on multiple-choice question answering (MCQA) tasks using *first-token probability* (FTP), which selects the answer option whose initial token has the highest likelihood. While efficient, FTP can be fragile: models may assign high probability to unrelated tokens (*misalignment*) or use a valid token merely as part of a generic preamble rather than as a clear answer choice (*misinterpretation*), undermining the reliability of symbolic evaluation. We propose a simple solution: the *prefilling attack*, a structured natural-language prefix (e.g., \"*The correct option is:*\") prepended to the model output. Originally explored in AI safety, we repurpose prefilling to steer the model to respond with a clean, valid option, without modifying its parameters. Empirically, the FTP with prefilling strategy substantially improves accuracy, calibration, and output consistency across a broad set of LLMs and MCQA benchmarks. It outperforms standard FTP and often matches the performance of open-ended generation approaches that require full decoding and external classifiers, while being significantly more efficient. Our findings suggest that prefilling is a simple, robust, and low-cost method to enhance the reliability of FTP-based evaluation in multiple-choice settings.