Bias Generator Research Articles

Cryogenic quantum computer control signal generation using high-electron-mobility transistors

Multiplexed local charge storage, close to quantum processors at cryogenic temperatures could generate a multitude of control signals, for electronics or qubits, in an efficient manner. Such cryogenic electronics require generating quasi-static control signals with small area footprint, low noise, high stability, low power dissipation and, ideally, in a multiplexed fashion to reduce the number of input/outputs. In this work, we integrate capacitors with cryogenic high-electron mobility transistor (HEMT) arrays and demonstrate quasi-static bias generation using gate pulses controlled in time and frequency domains. Multi-channel bias generation is also demonstrated. Operation at 4 K exhibits improved bias signal variability and greatly reduced subthreshold swing, reaching values of ~6 mV/decade. Due to the very low threshold voltage of 80 mV at 4 K and the steep subthreshold swing, these circuits can provide an advantage over the silicon-based complementary metal-oxide-semiconductor equivalents by allowing operation at significantly reduced drive bias in the low output voltage regime <1 V. Together with their high-speed operation, this makes HEMTs an attractive platform for future cryogenic signal generation electronics in quantum computers.

DACG: Dual Attention and Context Guidance model for radiology report generation

Medical images are an essential basis for radiologists to write radiology reports and greatly help subsequent clinical treatment. The task of generating automatic radiology reports aims to alleviate the burden of clinical doctors writing reports and has received increasing attention this year, becoming an important research hotspot. However, there are severe issues of visual and textual data bias and long text generation in the medical field. Firstly, Abnormal areas in radiological images only account for a small portion, and most radiological reports only involve descriptions of normal findings. Secondly, there are still significant challenges in generating longer and more accurate descriptive texts for radiology report generation tasks. In this paper, we propose a new Dual Attention and Context Guidance (DACG) model to alleviate visual and textual data bias and promote the generation of long texts. We use a Dual Attention Module, including a Position Attention Block and a Channel Attention Block, to extract finer position and channel features from medical images, enhancing the image feature extraction ability of the encoder. We use the Context Guidance Module to integrate contextual information into the decoder and supervise the generation of long texts. The experimental results show that our proposed model achieves state-of-the-art performance on the most commonly used IU X-ray and MIMIC-CXR datasets. Further analysis also proves that our model can improve reporting through more accurate anomaly detection and more detailed descriptions. The source code is available at https://github.com/LangWY/DACG.

A Modified Current-Mode VCSEL Driver for Short-Range LiDAR Sensor Applications in 180 nm CMOS

This paper presents a modified current-mode vertical-cavity surface-emitting laser (VCSEL) driver as a transmitter for short-range light detection and ranging (LiDAR) sensors, where a stable bias generator is suggested with a regulated cascode current mirror circuit to provide the bias current of 1 mA with a trivial deviation of 5.4%, even at the worst-case process–voltage–temperature (PVT) variations. Also, a modified current-steering logic circuit is exploited with N-type MOSFET (NMOS) switches to deliver the modulation currents of 0.1~10 mApp to the VCSEL diode simultaneously, with no overshoot distortions. Post-layout simulations of the modified current-mode VCSEL driver (m-CMVD), using 180 nm CMOS technology, demonstrate very large and clean output pulses with significantly reduced signal distortions. Hereby, the VCSEL diode is transformed into an equivalent circuit with a 1.6 V DC voltage and a 50 Ω resistor for circuit simulations. The proposed m-CMVD consumes a maximum of 11 mW from a 3.3 V supply voltage and the chip core occupies an area of 0.196 mm2.

Hierarchical Learning-Enhanced Chaotic Crayfish Optimization Algorithm: Improving Extreme Learning Machine Diagnostics in Breast Cancer

Extreme learning machines (ELMs), single hidden-layer feedforward neural networks, are renowned for their speed and efficiency in classification and regression tasks. However, their generalization ability is often undermined by the random generation of hidden layer weights and biases. To address this issue, this paper introduces a Hierarchical Learning-based Chaotic Crayfish Optimization Algorithm (HLCCOA) aimed at enhancing the generalization ability of ELMs. Initially, to resolve the problems of slow search speed and premature convergence typical of traditional crayfish optimization algorithms (COAs), the HLCCOA utilizes chaotic sequences for population position initialization. The ergodicity of chaos is leveraged to boost population diversity, laying the groundwork for effective global search efforts. Additionally, a hierarchical learning mechanism encourages under-performing individuals to engage in extensive cross-layer learning for enhanced global exploration, while top performers directly learn from elite individuals at the highest layer to improve their local exploitation abilities. Rigorous testing with CEC2019 and CEC2022 suites shows the HLCCOA’s superiority over both the original COA and nine renowned heuristic algorithms. Ultimately, the HLCCOA-optimized extreme learning machine model, the HLCCOA-ELM, exhibits superior performance over reported benchmark models in terms of accuracy, sensitivity, and specificity for UCI breast cancer diagnosis, underscoring the HLCCOA’s practicality and robustness, as well as the HLCCOA-ELM’s commendable generalization performance.

Impact of generative artificial intelligence models on the performance of citizen data scientists in retail firms

Generative Artificial Intelligence (AI) models serve as powerful tools for organizations aiming to integrate advanced data analysis and automation into their applications and services. Citizen data scientists—individuals without formal training but skilled in data analysis—combine domain expertise with analytical skills, making them invaluable assets in the retail sector. Generative AI models can further enhance their performance, offering a cost-effective alternative to hiring professional data scientists. However, it is unclear how AI models can effectively contribute to this development and what challenges may arise. This study explores the impact of generative AI models on citizen data scientists in retail firms. We investigate the strengths, weaknesses, opportunities, and threats of these models. Survey data from 268 retail companies is used to develop and validate a new model. Findings highlight that misinformation, lack of explainability, biased content generation, and data security and privacy concerns in generative AI models are major factors affecting citizen data scientists’ performance. Practical implications suggest that generative AI can empower retail firms by enabling advanced data science techniques and real-time decision-making. However, firms must address drawbacks and threats in generative AI models through robust policies and collaboration between domain experts and AI developers.

Strain-Induced Robust Exchange Bias Effect in Epitaxial La0.7Sr0.3MnO3/LaFeO3 Bilayers.

The ground state of correlated electrons in complex oxide films can be controlled by applying epitaxial strain, offering the potential to produce unexpected phenomena applicable to modern spintronic devices. In this study, we demonstrate that substrate-induced strain strongly affects the coupling mode of interfacial magnetic moments in a ferromagnetic (FM)/antiferromagnetic (AFM) system. In an epitaxial bilayer comprising AFM LaFeO3 (LFO) and FM La0.7Sr0.3MnO3 (LSMO), samples grown on a LaAlO3 (LAO) substrate exhibit a larger exchange bias field than those grown on a SrTiO3 substrate. Our results indicate a transition in the alignment of magnetic moments from perpendicular to collinear due to the large compressive strain exerted by the LAO substrate. Collinear magnetic moments at the LSMO/LFO interface generate strong exchange coupling, leading to a considerable exchange bias effect. Thus, our findings provide a method for tailoring and manipulating the orientations of magnetic moments at the FM/AFM heterogeneous interface using strain engineering, thereby augmenting methods for exchange bias generation.

A simultaneous area, delay and leakage current reduction method for Linear Feedback Shift Register (LFSR) using pulsed latches and DTMOS transistors

In digital circuits, the shift registers are the significant elements having the ability of storing and transmitting the data in sequential manner.The linear feedback shift register (LFSR) uses large number of flips-flops that consumes maximum amount of power. To overcome this issue, this research work concentrates on the design of LFSR for the area efficient, power reduction, minimization of delay and leakage current. By replacing the d-flip-flop with the bi-enabled pulsed latch to reduce the power consumption and delay of the proposed LFSR design. For the minimization of area, leakage current and delay, the hybrid body bias generator with dynamic threshold MOS (BBG-DTMOS) is used. Moreover, the pulsed clock signal is generated in the digital circuit with the aid of hybrid BBG-DTMOS technique. The proposed LFSR design is implemented in the tanner EDA tool with the available CMOS model libraries. Based on the implementation result, the proposed LFSR area is 1824 μm2and its power consumption is 172μW at a 100 MHz clock frequency withVDD=5V. Furthermore, having the better performance in different corners of the method, reduces the propagation delay and leakage current in the digital circuit. The proposed LFSR design is compared with some of the shift registers such as unidirectional and bidirectional shift registers. The existing clock pulse generator such as conventional, delayed clock pulse, TIMBER logic, PTL-AND logic and multiplexer based clock pulse generator (MCPG) are compared with the proposed BBG-DTMOS technique.

Quality-Diversity Generative Sampling for Learning with Synthetic Data

Generative models can serve as surrogates for some real data sources by creating synthetic training datasets, but in doing so they may transfer biases to downstream tasks. We focus on protecting quality and diversity when generating synthetic training datasets. We propose quality-diversity generative sampling (QDGS), a framework for sampling data uniformly across a user-defined measure space, despite the data coming from a biased generator. QDGS is a model-agnostic framework that uses prompt guidance to optimize a quality objective across measures of diversity for synthetically generated data, without fine-tuning the generative model. Using balanced synthetic datasets generated by QDGS, we first debias classifiers trained on color-biased shape datasets as a proof-of-concept. By applying QDGS to facial data synthesis, we prompt for desired semantic concepts, such as skin tone and age, to create an intersectional dataset with a combined blend of visual features. Leveraging this balanced data for training classifiers improves fairness while maintaining accuracy on facial recognition benchmarks. Code available at: https://github.com/Cylumn/qd-generative-sampling.

Bias-Conflict Sample Synthesis and Adversarial Removal Debias Strategy for Temporal Sentence Grounding in Video

Temporal Sentence Grounding in Video (TSGV) is troubled by dataset bias issue, which is caused by the uneven temporal distribution of the target moments for samples with similar semantic components in input videos or query texts. Existing methods resort to utilizing prior knowledge about bias to artificially break this uneven distribution, which only removes a limited amount of significant language biases. In this work, we propose the bias-conflict sample synthesis and adversarial removal debias strategy (BSSARD), which dynamically generates bias-conflict samples by explicitly leveraging potentially spurious correlations between single-modality features and the temporal position of the target moments. Through adversarial training, its bias generators continuously introduce biases and generate bias-conflict samples to deceive its grounding model. Meanwhile, the grounding model continuously eliminates the introduced biases, which requires it to model multi-modality alignment information. BSSARD will cover most kinds of coupling relationships and disrupt language and visual biases simultaneously. Extensive experiments on Charades-CD and ActivityNet-CD demonstrate the promising debiasing capability of BSSARD. Source codes are available at https://github.com/qzhb/BSSARD.

Chiroptical properties of membrane glycerophospholipids and their chiral backbones.

Glycerophospholipid membranes are one of the key cellular components. Still, their species-dependent composition and homochirality remain an elusive subject. In the context of the astrophysical circularly polarized light scenario likely involved in the generation of a chiral bias in meteoritic amino and sugar acids in space, and consequently in the origin of life's homochirality on Earth, this study reports the first measurements of circular dichroism and anisotropy spectra of a selection of glycerophospholipids, their chiral backbones and their analogs. The rather low asymmetry in the interaction of UV/VUV circularly polarized light with sn-glycerol-1/3-phosphate indicates that chiral photons would have been unlikely to directly induce symmetry breaking to membrane lipids. In contrast, the anisotropy spectra of d-3-phosphoglyceric acid and d-glyceraldehyde-3-phosphate unveil up to 20 and 100 times higher maximum anisotropy factor values, respectively. This first experimental report, targeted on investigating the origins of phospholipid symmetry breaking, opens up new avenues of research to explore alternative mechanisms leading to membrane lipid homochirality, while providing important clues for the search for chiral biosignatures of extant and/or extinct life in space, in particular for the ExoMars 2028 mission.

Simple contrastive learning in a self-supervised manner for robust visual question answering

Recent observations have revealed that Visual Question Answering models are susceptible to learning the spurious correlations formed by dataset biases, i.e., the language priors, instead of the intended solution. For instance, given a question and a relative image, some VQA systems are prone to provide the frequently occurring answer in the dataset while disregarding the image content. Such a preferred tendency has caused them to be brittle in real-world settings, harming the robustness of VQA models. We experimentally found that conventional VQA methods often confuse negative samples that with identical questions but different images, which results in the generation of linguistic bias. In this paper, we propose a simple contrastive learning scheme, namely SCLSM, to mitigate the above issues in a self-supervised manner. We construct several special negative samples and introduce a debiasing-aware contrastive learning approach to help the model learn more discriminative multimodal features, thus improving the ability of debiasing. The SCLSM is compatible with numerous VQA baselines. Experimental results on the widely-used public datasets VQA-CP v2 and VQA v2 validate the effectiveness of our proposed model.

Atomic Layer Etching of GaN and AlGaN Using CH4/H2, H2 and HCl Chemistry

Modern microelectronic components for high-power and high-frequency applications require reliability and reproducibility of the production processes. Atomic layer etching (ALE) represents a key technology for achieving these requirements.An ALE sequence is characterized by a chemical modification step that affects only the top atomic layer of the surface (adhesion) and an ion assisted non-reactive inert gas etching step that removes only this chemically modified area due to creation of volatile byproducts. These two steps are separated by inert gas purge. This grants the etching of single atomic layers. The etching step can be triggered by thermal energy or kinetic energy by impinging particles (e.g. ion bombardment by a plasma). The low energy required for ALE guarantees low-damage etching.In literature ALE processes using Cl2 are often reported. In this work, the results of plasma-enhanced ALE of GaN and AlGaN using different chemistries of CH4/H2, H2 and HCl as etch gases are presented. The experiments were performed with a highly customized etch facility equipped with a capacitively coupled plasma source (CCP, 60 MHz) and a substrate bias generator (RF, 2 MHz). The investigations were carried out on epitaxially deposited GaN on sapphire with masks of protective resist and SiO2 as well as GaN-AlGaN heterostructures with patterned SiO2 maskings. For process analysis optical emission spectroscopy (OES) was carried out. The etching rate and roughness were determined by ellipsometry and atomic force microscopy (AFM) as well as scanning electron microscopy (SEM).Observations by OES indicate a successful excitation and dissociation of the etching molecules. With rising plasma powers physical sputtering effects were revealed. At the beginning of the process development continuous etching processes were examined. With CH4/H2 chemistry using GaN samples it was found that the etching behavior is strongly influenced by the type of masking material. Protective resist resulted in a low etch rate of 7.5 nm/min and significant damage of the masking material. The cause of this behavior is assumed to be redeposition of material by interaction of methane, resist, and plasma. With the change to a hard mask of SiO2 etch rates of 45 to 72 nm/min were achieved, increasing with RF-power and CH4/H2 ratio. Adding methane led to detrimental side effects such as dendrite growth and wall formation on the transient region of unmasked to former masked sample areas. In the further development of the ALE process an etch chemistry using only hydrogen as the chemical etchant was used resulting in an etch per cycle (EPC) of about 3 Å for GaN samples. Further investigations of ALE with hydrogen etch chemistry on GaN-AlGaN samples revealed that AlGaN, in contrast to pure GaN, could not be etched successfully with this chemistry.Substituting CH4/H2 or H2 to HCl enables the etch of AlGaN. The investigations carried out indicate that this etching process is also a self-limiting ALE. AFM measurements revealed an EPC of 1.2 Å for lower RF-powers (50 W). Increasing the RF power to 150 W only slightly enhance the EPC to 2.0 Å. The surface roughness tends to decrease with rising RF-power from 50 to 300 W.

Ingroup/Outgroup Bias and Emotional Perception

With the globalization of the world, it has become a main tendency to establish cooperative relations among countries. However, in the course of historical development, countries have formed their own unique culture and established profound national consciousness and identity. These gaps between countries cause prejudice and stereotypes which hinder cooperation. To get to the root of the problem, it is crucial to find out what factors would influence the formation of group bias. People always associate group identity with intergroup bias, but few have thought about the relationship between group bias and emotional perception. However, as a prerequisite for understanding others' behavior, emotional perception will greatly affect people's behavioral responses to others. Recent studies give more concern about the relationship between group bias and emotional perception, and they confirmed that they do influence each other. This review discussed how group bias and emotional perception affect each other and also factors that work in this relationship. It was found that people tend to judge more positively towards ingroup members’ emotions. In addition, competition plays an important role in the generation of emotional perception bias. With these conclusions, it might be easier to reduce group bias and promote cooperations in society.

Causal Disentanglement for Semantic-Aware Intent Learning in Recommendation

Traditional recommendation models trained on observational interaction data have generated large impacts in a wide range of applications, it faces bias problems that cover users true intent and thus deteriorate the recommendation effectiveness. Existing methods tracks this problem as eliminating bias for the robust recommendation, e.g., by re-weighting training samples or learning disentangled representation. The disentangled representation methods as the state-of-the-art eliminate bias through revealing cause-effect of the bias generation. However, how to design the semantics-aware and unbiased representation for users true intents is largely unexplored. To bridge the gap, we are the first to propose an unbiased and semantics-aware disentanglement learning called CaDSI (Causal Disentanglement for Semantics-Aware Intent Learning) from a causal perspective. Particularly, CaDSI explicitly models the causal relations underlying recommendation task, and thus produces semantics-aware representations via disentangling users true intents aware of specific item context. Moreover, the causal intervention mechanism is designed to eliminate confounding bias stemmed from context information, which further to align the semantics-aware representation with users true intent. Extensive experiments and case studies both validate the robustness and interpretability of our proposed model.

Low-power Relaxation Oscillator with Temperature-compensated Thyristor Decision Elements

This paper presents a low-power 140 kHz relaxation oscillator (ROSC) for low-frequency clock generators and timers. In voltage-mode ROSCs, unavoidable shunt current consumption results from voltage slewing at the integration capacitor. The proposed circuit employs CMOS thyristor-based decision elements which effectively reduce shunt currents by exploiting internal positive feedback. A complementary-to-absolute temperature (CTAT) current reference compensates for the frequency’s temperature sensitivity. In order to achieve high negative temperature coefficient with small area and power overhead, the circuit reuses parts of the positive-to-absolute temperature (PTAT) bias generation block. Moreover, a modified start-up circuit with 3 times faster oscillator power-on is presented. The 0.09 mm{}^2 oscillator consumes 6.5 nW/kHz at 1.5 V to 2.5 V supply if only the CTAT source is considered, resulting in a power consumption of 907.4 nW at 140 kHz. The measured temperature coefficient of -514.7 ppm/K in the range of −40 °C to 85 °C shows an improvement of 5.5 times compared to the uncompensated case. A supply sensitivity of 2.62 %/V, a frequency resolution of 2.67 kHz/step, and an average clock jitter of 6.02 ns are achieved. The oscillator is embedded in an ultra-low power system-on-chip for autonomous environmental sensing.

Political Bias on Instagram: Generation Z and the 2022 Midterm Elections

Generation Z, today's 10 to 25-year-olds, is emerging into the political sphere. In fact, the outcome of the 2022 United States Midterm Elections was heavily influenced by the power of progressive young voters. With the increased use of social media in recent years, many young activists have adopted online platforms as spaces for political conversation. This study investigates the influence social media activism, specifically from Generation Z creators, had on the outcome of the 2022 United States midterm elections. The study adopts a content analysis approach to analyze a selected amount of Generation Z Instagram posts from the midterms. The goal of the study is to determine the political bias Generation Z expressed online at this time and the effect social media activism had on the outcome of the election. The analysis includes 177 carefully selected posts, and each post was evaluated for its popularity, public response, and political bias. We found a slight progressive political bias within the post content and an unexpected amount of popularity and public response for conservative-leaning posts. The findings from this study suggest a developing progressive political bias among Generation Z. Further research on the connection between social media activism and election results is suggested.

Low-Power Process and Temperature-Invariant Constant Slope-and-Swing Ramp-Based Phase Interpolator

This article presents a process-and temperature-invariant high-resolution and highly linear low-power phase interpolator (PI) as an enabler for discrete-time spatial signal processors (SSPs) and for various mixed-mode RF transceiver architectures. Using current integration techniques, the proposed PI generates an adaptable constant slope-and-swing ramp signal to achieve significantly lower power suited for multiple antenna elements. Switched-capacitor-based bias generation enables tracking the ramp generator over process, voltage, and temperature enhancing the PI linearity. The 7-bit PI realized in the 65-nm CMOS technology can generate full delay range with a resolution of 4.88 ps at 1.6-GHz input frequency. The PI consumes a power of 503 <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$\mu$</tex-math> </inline-formula> W and occupies an active area of 0.021 mm <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$^2$</tex-math> </inline-formula> with a jitter of 0.410 ps for a 1.6-GHz operation and measured DNL and INL of 0.52 and 0.52 LSB, respectively.

Investigation and Modeling of DC Bias Impact on Core Losses at High Frequency

This paper aims to study the core losses behavior at high frequency (MHz range) with superimposed DC bias in ferrite. Inductive compensation method is employed in the measurement circuit to reduce sensitivity to phase errors. An addition to the circuit is proposed in this work to make easier DC bias generation. Furthermore, a detailed measurement accuracy consideration is also discussed. The tested MnZn ferrite has a nominal relative permeability of 1500 and 800, which was tested at frequency from 500 kHz to 3 MHz. The measurement results are explained thoroughly with three controlling parameters: excitation frequency, peak AC flux density, and DC bias. There are several important findings. First, a higher DC bias creates higher losses at the same frequency and AC flux density. Secondly, at the same DC bias and AC flux density, higher frequency generates a lower relative losses increase. This second point has not been seen in previous literature and is elaborated more in Section III.B. Thirdly, the measurement result shows how DC bias increases the hysteresis loop area and coercivity. The first and third findings confirm the the existing literature findings. As a final step, an improved Steinmetz Premagnetization Graph and Artificial Neural Network are used to create core loss prediction model. The measurement data and the built model can be accessed online for use by other magnetics designer.

Multilayer-Cavity Tandem Catalyst for Profiling Sequentially Coupling of Intermediate CO in Electrocatalytic Reduction Reaction of CO2 to Multi-Carbon Products.

Electrochemical CO2 reduction reaction (CO2 RR) is an effective approach to address CO2 emission, promote recycling, and synthesize high-value multi-carbon (C2+ ) chemicals for storing renewable electricity in the long-term. The construction of multilayer-bound nanoreactors to achieve management of intermediate CO is a promising strategy for tandem to C2+ products. In this study, a series of Ag@Cu2 O nanoreactors consisting of an Ag-yolk and a multilayer confined Cu shell is designed to profile electrocatalytic CO2 RR reactions. The optimized Ag@Cu2 O-2nanoreactor exhibits a 74% Faradaic efficiency during the C2+ pathway and remains stable for over 10h at a bias current density of 100mAcm-2 . Using the finite element method, this model determines that the certain volume of cavity in the Ag@Cu2 O nanoreactors facilitates on-site CO retention and that multilayers of Cu species favor CO capture. Density functional theory calculations illustrate that the biased generation of ethanol products may arise from the (100)/(111) interface of the Cu layer. In-depth explorations in multilayer-bound nanoreactors provide structural and interfacial guidance for sequential coupling of CO2 RR intermediates for efficient C2+ generation.

Controlled generation and detection of a thermal bias in Corbino devices under the quantum Hall regime

We present an experimental technique to generate and measure a temperature bias in the quantum Hall effect of GaAs/AlGaAs Corbino samples. The bias is generated by injecting an electric current at a central resistive heater, and the resulting radial temperature drop is determined by conductance measurements at internal and external concentric rings. The experimental results agree with the predictions of numerical simulations of the heat flow through the substrate. We also compare these results with previous predictions based on the thermoelectric response of these devices.