Search

GNNs are powerful models based on node representation learning that perform particularly well in many machine learning problems related to graphs. The major obstacle to the deployment of GNNs is mostly a problem of societal acceptability and trustworthiness, properties which require making explicit the internal functioning of such models. Here, we propose to mine activation rules in the hidden layers to understand how the GNNs perceive the world. The problem is not to discover activation rules that are individually highly discriminating for an output of the model. Instead, the challenge is to provide a small set of rules that cover all input graphs. To this end, we introduce the subjective activation pattern domain. We define an effective and principled algorithm to enumerate activations rules in each hidden layer. The proposed approach for quantifying the interest of these rules is rooted in information theory and is able to account for background knowledge on the input graph data. The activation rules can then be redescribed thanks to pattern languages involving interpretable features. We show that the activation rules provide insights on the characteristics used by the GNN to classify the graphs. Especially, this allows to identify the hidden features built by the GNN through its different layers. Also, these rules can subsequently be used for explaining GNN decisions. Experiments on both synthetic and real-life datasets show highly competitive performance, with up to 200% improvement in fidelity on explaining graph classification over the SOTA methods.

Developers use different means to document the security concerns of their code. Because of all of these opportunities, they may forget where the information is stored, or others may not be aware of it, and leave it unmaintained for so long that it becomes obsolete, if not useless. In this work, we analyzed different sources of code documentation from four large-scale, real-world, open-source projects in an industrial setting to understand where developers report their security concerns. In particular, we manually inspected 2.559 instances taken from source code comments, commit messages, and issue trackers. Overall, we found that developers prefer to document security concerns in source code comments and issue trackers. We also found that the longer the comments stay unfixed, the more likely they remain unfixed. Thus, to create awareness among developers, we implemented a pipeline to remind them about the introduction or removal of comments pointing to a security problem., Comment: Accepted as an extended abstract for XP'24

A great deal of interest is directed nowadays towards the development of innovative technologies in the field of quantum information and quantum computing, with emphasis on obtaining reliable qubits as building blocks. The realization of highly stable, controllable and accessible hole spin qubits is strongly dependent on the quality of the materials hosting them. Ultra-clean germanium/silicon-germanium heterostructures have been predicted and proven to be promising candidates and due to their large scalability potential, they are opening the door towards the development of realistic and reliable solid state, all-electric, silicon-based quantum computers. In order to obtain ultra-clean germanium/silicon-germanium heterostructures in a reverse grading approach, the understanding and control over the growth of Ge virtual substrates and thin films is key. Here, we present a detailed study on the growth kinetics, morphology, and crystal quality of Ge thin films grown via chemical vapor deposition by investigating the effects of growth temperature, partial pressure of the precursor gas and the use of Ar or H2 atmosphere. The presence of carrier gases catalyzes the deposition rate and induces a smoothening on the surfaces of films grown at low temperatures. We investigated the surface roughness and threading dislocation density as a function of deposition temperature, partial pressure and gas mixture. Ge thin films deposited by diluting GeH4 in Ar or H2 were employed as virtual substrates for the growth of full Ge/SiGe QW heterostructures. Their defect density was analyzed and their electric transport properties were characterized via Hall measurements. Similar results were obtained for both carrier gases used.

The requirements for broadband photodetection are becoming exceedingly demanding in hyperspectral imaging. Whilst intrinsic photoconductor arrays based on mercury cadmium telluride represent the most sensitive and suitable technology, their optical spectrum imposes a narrow spectral range with a sharp absorption edge that cuts their operation to < 25 um. Here, we demonstrate a giant ultra-broadband photoconductivity in twisted double bilayer graphene heterostructures spanning a spectral range of 2 - 100 um with internal quantum efficiencies ~ 40 % at speeds of 100 kHz. The giant response originates from unique properties of twist-decoupled heterostructures including pristine, crystal field induced terahertz band gaps, parallel photoactive channels, and strong photoconductivity enhancements caused by interlayer screening of electronic interactions by respective layers acting as sub-atomic spaced proximity screening gates. Our work demonstrates a rare instance of an intrinsic infrared-terahertz photoconductor that is complementary metal-oxide-semiconductor compatible and array integratable, and introduces twist-decoupled graphene heterostructures as a viable route for engineering gapped graphene photodetectors with 3D scalability.

We consider a heterogeneous network of quantum computing modules, sparsely connected via Bell states. Operations across these connections constitute a computational bottleneck and they are likely to add more noise to the computation than operations performed within a module. We introduce several techniques for transforming a given quantum circuit into one implementable on a network of the aforementioned type, minimising the number of Bell states required to do so. We extend previous works on circuit distribution over fully connected networks to the case of heterogeneous networks. On the one hand, we extend the hypergraph approach of [Andres-Martinez & Heunen. 2019] to arbitrary network topologies. We additionally make use of Steiner trees to find efficient realisations of the entanglement sharing within the network, reusing already established connections as often as possible. On the other hand, we extend the embedding techniques of [Wu, et al. 2022] to networks with more than two modules. Furthermore, we discuss how these two seemingly incompatible approaches can be made to cooperate. Our proposal is implemented and benchmarked; the results confirming that, when orchestrated, the two approaches complement each other's weaknesses., Comment: 30 pages, 18 figures, comments welcome; v2 - Add link to experiment data; v3 - 31 pages, small updates to discussion and references

Frequency-modulated (FM) combs feature flat intensity spectra with a linear frequency chirp, useful for metrology and sensing applications. Generating FM combs in semiconductor lasers generally requires a fast saturable gain, usually limited by the intrinsic gain medium properties. Here, we show how a spatial modulation of the laser gain medium can enhance the gain saturation dynamics and nonlinearities to generate self-starting FM combs. We demonstrate this with tapered planarized THz quantum cascade lasers (QCLs). While simple ridge THz QCLs typically generate combs which are a mixture of amplitude and frequency modulation, the on-chip field enhancement resulting from extreme spatial confinement leads to an ultrafast saturable gain regime, generating a pure FM comb with a flatter intensity spectrum, a clear linear frequency chirp and very intense beatnotes up to -30 dBm. The observed linear frequency chirp is reproduced using a spatially inhomogeneous mean-field theory model which confirms the crucial role of field enhancement. In addition, the modified spatial temperature distribution within the waveguide results in an improved hightemperature comb operation, up to a heat sink temperature of 115 K, with comb bandwidths of 600 GHz at 90 K. The spatial inhomogeneity also leads to dynamic switching between various harmonic states in the same device., Comment: 9 pages, 6 figures; corrected reference typos

In noisy intermediate-scale quantum computing, the limited scalability of a single quantum processing unit (QPU) can be extended through distributed quantum computing (DQC), in which one can implement global operations over two QPUs by entanglement-assisted local operations and classical communication. To facilitate this type of DQC in experiments, we need an entanglement-efficient protocol. To this end, we extend the protocol in [Eisert et. al., PRA, 62:052317(2000)] implementing each nonlocal controlled-unitary gate locally with one maximally entangled pair to a packing protocol, which can pack multiple nonlocal controlled-unitary gates locally using one maximally entangled pair. In particular, two types of packing processes are introduced as the building blocks, namely the distributing processes and embedding processes. Each distributing process distributes corresponding gates locally with one entangled pair. The efficiency of entanglement is then enhanced by embedding processes, which merge two non-sequential distributing processes and hence save the entanglement cost. We show that the structure of distributability and embeddability of a quantum circuit can be fully represented by the corresponding packing graphs and conflict graphs. Based on these graphs, we derive heuristic algorithms for finding an entanglement-efficient packing of distributing processes for a given quantum circuit to be implemented by two parties. These algorithms can determine the required number of local auxiliary qubits in the DQC. We apply these algorithms for bipartite DQC of unitary coupled-cluster circuits and find a significant reduction of entanglement cost through embeddings. This method can determine a constructive upper bound on the entanglement cost for the DQC of quantum circuits., Comment: Published version: 30+10 pages, 22+2 figures, updated with numerical benchmarking of the packing algorithm

We study the performance of an hot-electron bolometer (HEB) operating at THz frequencies based on superconducting niobium nitride films. We report on the voltage response of the detector over a large optical bandwidth carried out with different THz sources. We show that the impulse response of the fully packaged HEB at 7.5 K has a 3 dB cut-off around 2 GHz. Remarkably, detection capability is still observed above 30 GHz in an heterodyne beating experiment using a THz quantum cascade laser frequency comb. Additionally, the HEB sensitivity has been evaluated and an optical noise equivalent power NEP of 0.8 pW/sqrt(Hz) has been measured at 1 MHz.

Quantum Cascade Lasers (QCL) constitute an intriguing opportunity for the production of on-chip optical Dissipative Kerr Solitons (DKS): self-organized optical waves which can travel while preserving their shape thanks to the interplay between Kerr effect and dispersion. Originally demonstrated in passive microresonators, DKS were recently observed in mid-IR ring QCL paving the way for their achievement even at longer wavelengths. To this end we realized defect-less THz ring QCLs featuring anomalous dispersion leveraging on a technological platform based on waveguide planarization. A concentric coupled-waveguide approach is implemented for dispersion compensation whilst a passive broadband bullseye antenna improves the device power extraction and far field. In these devices, comb spectra featuring sech$^2$ envelopes are presented for free-running operation. This first hint of the presence of solitons is further supported by the observation of highly hysteretic behaviour and by phase-sensitive measurements which show the presence of self-starting 12 ps-long pulses in the reconstructed time profile of the emission intensity. These observations are in very good agreement with our numeric simulations based on a Complex Ginzburg-Landau equation time-domain solver. Such devices constitute a new experimental platform for the study of soliton phenomena in the THz range, allowing as well on-chip, passive ultrashort THz pulse generation appealing for a variety of applications.

We developed a process to fabricate nanoscale metallic gate electrodes on scanning probe cantilevers, including on the irregular surface of protruding cantilever tips. The process includes a floating-layer technique to coat the cantilevers in electron-beam resist. We demonstrate gate definition through a lift-off process, as well as through an etching process. The cantilevers maintain a high force sensitivity after undergoing the patterning process. Our method allows the patterning of nanoscale devices on fragile scanning probes, extending their functionality as sensors.

Recently, there has been a growing interest in integrated THz photonics for various applications in communications, spectroscopy and sensing. We present a new integrated photonic platform based on active and passive elements integrated in a double-metal, high confinement waveguide layout planarized with a low-loss polymer. An extended top metallization results in low waveguide losses and improved dispersion, thermal and RF properties, as it enables to decouple the design of THz and microwave cavities. Free-running on-chip quantum cascade laser combs spanning 800 GHz, harmonic states over 1.1 THz and RF-injected broadband incoherent states spanning over nearly 1.6 THz are observed. With a strong external RF drive, actively mode-locked pulses as short as 4.4 ps can be produced, as measured by SWIFTS. We demonstrate as well passive waveguides with low insertion loss, enabling the tuning of the laser cavity boundary conditions and the co-integration of active and passive components on the same THz photonic chip., Comment: 27 pages, 5 figures

We show the possibility of tailoring the molecular arrangement, as well as the chemical and structural modifications, of porphyrins at the monolayer saturation coverage on TiO2(110) by Synchrotron photoemission, electron diffraction, STM topography and DFT calculations. Free-base tetra-phenyl-porphyrins (2H-TPP) adsorb on the oxygen rows, where they can spontaneously capture two additional hydrogen atoms at their iminic nitrogens (4H-TPP). Both 2H-TPP and 4H-TPP molecules aggregate into a commensurate phase at the saturation coverage of one monolayer. Upon sample heating, a self-metalation reaction sets in at 100C, yielding full metalation of the saturated monolayer at ~200C. The Ti atoms are extracted from the substrate and, by simultaneous dehydrogenation of the pyrrolic nitrogen atoms, incorporated into the porphyrin macrocycle, where they remain coordinated to two oxygen atoms underneath. Neither the adsorption geometry (on-bridge, atop the oxygen rows) nor the molecular arrangement change across the self-metalation transition up to 300C. On one side, the robustness of this saturation phase makes it a promising system for its implementation into applications for photocatalysis and photovoltaic devices. On the other side, the possible manifestation of metal exchange with the very reactive Ti atoms must be taken into account when designing porphyrin-sensitized solar cells since the critical temperature for the onset of self-metalation is very close to the normal operating temperature of photovoltaic devices., Comment: Final Title and Abstract; Pre-final manuscript

Fast (sub-second) spectroscopy with high spectral resolution is of vital importance for revealing quantum chemistry kinetics of complex chemical and biological reactions. Fourier transform (FT) spectrometers can achieve high spectral resolution and operate at hundreds of ms time scales in rapid-scan mode. However, the linear translation of a scanning mirror imposes stringent time-resolution limitations to these systems, which makes simultaneous high spectral and temporal resolution impossible. Here, we demonstrate an FT spectrometer whose operational principle is based on continuous rotational, rather than linear, motion of the scanning mirror, decoupling the spectral resolution from the temporal one. This enables 0.5 cm${}^{-1}$ resolution on sub-ms time scales. Furthermore, we show that such rotational FT spectrometers can perform dual-comb spectroscopy with a single comb source, since the Doppler-shifted version of the comb serves as the second comb. In this way, we combine the advantages of dual-comb and FT spectroscopy using a single quantum cascade laser frequency comb as a light source. Our technique does not require any diffractive or dispersive optical elements and hence preserve the Jacquinot's-, Fellgett's-, and Connes'-advantages of FT spectrometers. The system supports a large optical bandwidth from visible to THz frequencies. The combination of a rotational delay line with collimated coherent or non-coherent light sources pave the way for FT spectrometers in applications where high speed, large optical bandwidth, and high spectral resolution are desired.

A laser exhibits both controllable gain and loss and, under proper design conditions, is an ideal non-Hermitian system allowing the direct observation and engineering of spectral singularities such as exceptional points (EPs). A dual section distributed feedback (DFB) quantum cascade laser (QCL) is a prototype of such a system, allowing the controlled coupling of a ladder of cavity Fabry-Perot (FP) modes to a quarter wave shifted DFB mode. Tuning the coupling strength and the gain difference between these two set of modes enables probing the regimes from weak coupling to strong coupling and the robust observation of exceptional point singularities. At these exceptional points, the laser exhibits a sequence of lasing and coherent prefect absorption dynamics1,2 when pumped above transparency. Additionally, the pumping scheme allows the deliberate lifting of the exceptional point degeneracy. These results show that dual section QCL is a perfect platform to study exceptional points because the coupling parameter and system loss can be tuned in a single device.

Crystal phase engineering gives access to new types of superlattices where, rather than different materials, different crystal phases of the same material are juxtaposed. Here, by means of atomistic nonequilibrium molecular dynamics calculations, we study to what extent these periodic systems can be used to alter phonon transport, similarly to what has been predicted and observed in conventional superlattices based on heterointerfaces. We focus on twin superlattices in GaAs and InAs and highlight the existence of two different transport regimes: in one each interface behaves like an independent scatterer; in the other, a segment with a sufficiently large number of closely-spaced interfaces, is seen by propagating phonons as a metamaterial with its own thermal properties.

Harmonic comb state has proven to be emerging in quantum cascade lasers and promoted by an interplay between parametric gain and spatial hole burning. We report here on robust, pure, self-starting harmonic mode locking in Copper-based double-metal THz quantum cascade lasers. Different harmonic orders can be excited in the same laser cavity depending on the pumping condition and stable harmonic combs spanning more than 600 GHz bandwidth at 80 K are reported. Such devices can be RF injected and the free running coherence is assessed by means of self-mixing technique performed at 50 GHz. A theoretical model based on Maxwell-Bloch equations including an asymmetry in the gain profile is used to interpret the data., Comment: Submitted to APL, 6 pages, 5 figures

We present an antipodal Vivaldi antenna for broadband double metal waveguide terahertz quantum cascade lasers and frequency combs. Its exponentially curved flare profile results in an adiabatic in-plane mode expansion, producing an improved far-field with a single-lobed beam of (23{\deg}x19{\deg}) full width half maximum with an octave-spanning bandwidth. The antenna also acts as a wave retarder, rotating the polarization from vertical toward horizontal polarization by a frequency-dependent angle. The laser's emission spectrum and current-voltage characteristics are not affected, as well as frequency comb operation. Measurements agree well with numerical simulations, and the proposed antenna covers a broad spectral range (1.5-4.5 THz).

For power grid operations, a large body of research focuses on using generation redispatching, load shedding or demand side management flexibilities. However, a less costly and potentially more flexible option would be grid topology reconfiguration, as already partially exploited by Coreso (European RSC) and RTE (French TSO) operations. Beyond previous work on branch switching, bus reconfigurations are a broader class of action and could provide some substantial benefits to route electricity and optimize the grid capacity to keep it within safety margins. Because of its non-linear and combinatorial nature, no existing optimal power flow solver can yet tackle this problem. We here propose a new framework to learn topology controllers through imitation and reinforcement learning. We present the design and the results of the first "Learning to Run a Power Network" challenge released with this framework. We finally develop a method providing performance upper-bounds (oracle), which highlights remaining unsolved challenges and suggests future directions of improvement.

We present homogeneous quantum cascade lasers (QCLs) emitting around 3 THz which display bandwidths up to 950 GHz with a single stable beatnote. Devices are spontaneously operating in a harmonic comb state, and when in a dense mode regime they can be injection locked at the cavity roundtrip frequency with very small RF powers down to -55 dBm. When operated in the electrically unstable region of negative differential resistance, the device displays ultra-broadband operation exceeding 1.83 THz ($\Delta f/f=50\%$) with high phase noise, exhibiting self-sustained, periodic voltage oscillations. The low CW threshold (115 A$\cdot$ cm$^{-2}$) and broadband comb operation ($\Delta f/f=25\%$) make these sources extremely appealing for on-chip frequency comb applications.