-
PICLe: Eliciting Diverse Behaviors from Large Language Models with Persona In-Context Learning
Authors:
Hyeong Kyu Choi,
Yixuan Li
Abstract:
Large Language Models (LLMs) are trained on massive text corpora, which are encoded with diverse personality traits. This triggers an interesting goal of eliciting a desired personality trait from the LLM, and probing its behavioral preferences. Accordingly, we formalize the persona elicitation task, aiming to customize LLM behaviors to align with a target persona. We present Persona In-Context Le…
▽ More
Large Language Models (LLMs) are trained on massive text corpora, which are encoded with diverse personality traits. This triggers an interesting goal of eliciting a desired personality trait from the LLM, and probing its behavioral preferences. Accordingly, we formalize the persona elicitation task, aiming to customize LLM behaviors to align with a target persona. We present Persona In-Context Learning (PICLe), a novel persona elicitation framework grounded in Bayesian inference. At the core, PICLe introduces a new ICL example selection criterion based on likelihood ratio, which is designed to optimally guide the model in eliciting a specific target persona. We demonstrate the effectiveness of PICLe through extensive comparisons against baseline methods across three contemporary LLMs. Code is available at https://github.com/deeplearning-wisc/picle.
△ Less
Submitted 14 May, 2024; v1 submitted 3 May, 2024;
originally announced May 2024.
-
NuTrea: Neural Tree Search for Context-guided Multi-hop KGQA
Authors:
Hyeong Kyu Choi,
Seunghun Lee,
Jaewon Chu,
Hyunwoo J. Kim
Abstract:
Multi-hop Knowledge Graph Question Answering (KGQA) is a task that involves retrieving nodes from a knowledge graph (KG) to answer natural language questions. Recent GNN-based approaches formulate this task as a KG path searching problem, where messages are sequentially propagated from the seed node towards the answer nodes. However, these messages are past-oriented, and they do not consider the f…
▽ More
Multi-hop Knowledge Graph Question Answering (KGQA) is a task that involves retrieving nodes from a knowledge graph (KG) to answer natural language questions. Recent GNN-based approaches formulate this task as a KG path searching problem, where messages are sequentially propagated from the seed node towards the answer nodes. However, these messages are past-oriented, and they do not consider the full KG context. To make matters worse, KG nodes often represent proper noun entities and are sometimes encrypted, being uninformative in selecting between paths. To address these problems, we propose Neural Tree Search (NuTrea), a tree search-based GNN model that incorporates the broader KG context. Our model adopts a message-passing scheme that probes the unreached subtree regions to boost the past-oriented embeddings. In addition, we introduce the Relation Frequency-Inverse Entity Frequency (RF-IEF) node embedding that considers the global KG context to better characterize ambiguous KG nodes. The general effectiveness of our approach is demonstrated through experiments on three major multi-hop KGQA benchmark datasets, and our extensive analyses further validate its expressiveness and robustness. Overall, NuTrea provides a powerful means to query the KG with complex natural language questions. Code is available at https://github.com/mlvlab/NuTrea.
△ Less
Submitted 23 October, 2023;
originally announced October 2023.
-
The one-message-per-cell-cycle rule: A conserved minimum transcription level for essential genes
Authors:
Teresa W. Lo,
Han Kyou James Choi,
Dean Huang,
Paul A. Wiggins
Abstract:
The inherent stochasticity of cellular processes leads to significant cell-to-cell variation in protein abundance. Although this noise has already been characterized and modeled, its broader implications and significance remain unclear. In this paper, we revisit the noise model and identify the number of messages transcribed per cell cycle as the critical determinant of noise. In yeast, we demonst…
▽ More
The inherent stochasticity of cellular processes leads to significant cell-to-cell variation in protein abundance. Although this noise has already been characterized and modeled, its broader implications and significance remain unclear. In this paper, we revisit the noise model and identify the number of messages transcribed per cell cycle as the critical determinant of noise. In yeast, we demonstrate that this quantity predicts the non-canonical scaling of noise with protein abundance, as well as quantitatively predicting its magnitude. We then hypothesize that growth robustness requires an upper ceiling on noise for the expression of essential genes, corresponding to a lower floor on the transcription level. We show that just such a floor exists: a minimum transcription level of one message per cell cycle is conserved between three model organisms: Escherichia coli, yeast, and human. Furthermore, all three organisms transcribe the same number of messages per gene, per cell cycle. This common transcriptional program reveals that robustness to noise plays a central role in determining the expression level of a large fraction of essential genes, and that this fundamental optimal strategy is conserved from E. coli to human cells.
△ Less
Submitted 6 July, 2023;
originally announced July 2023.
-
MELTR: Meta Loss Transformer for Learning to Fine-tune Video Foundation Models
Authors:
Dohwan Ko,
Joonmyung Choi,
Hyeong Kyu Choi,
Kyoung-Woon On,
Byungseok Roh,
Hyunwoo J. Kim
Abstract:
Foundation models have shown outstanding performance and generalization capabilities across domains. Since most studies on foundation models mainly focus on the pretraining phase, a naive strategy to minimize a single task-specific loss is adopted for fine-tuning. However, such fine-tuning methods do not fully leverage other losses that are potentially beneficial for the target task. Therefore, we…
▽ More
Foundation models have shown outstanding performance and generalization capabilities across domains. Since most studies on foundation models mainly focus on the pretraining phase, a naive strategy to minimize a single task-specific loss is adopted for fine-tuning. However, such fine-tuning methods do not fully leverage other losses that are potentially beneficial for the target task. Therefore, we propose MEta Loss TRansformer (MELTR), a plug-in module that automatically and non-linearly combines various loss functions to aid learning the target task via auxiliary learning. We formulate the auxiliary learning as a bi-level optimization problem and present an efficient optimization algorithm based on Approximate Implicit Differentiation (AID). For evaluation, we apply our framework to various video foundation models (UniVL, Violet and All-in-one), and show significant performance gain on all four downstream tasks: text-to-video retrieval, video question answering, video captioning, and multi-modal sentiment analysis. Our qualitative analyses demonstrate that MELTR adequately `transforms' individual loss functions and `melts' them into an effective unified loss. Code is available at https://github.com/mlvlab/MELTR.
△ Less
Submitted 22 March, 2023;
originally announced March 2023.
-
Observation of Electronic Modes in Open Cavity Resonator
Authors:
Hwanchul Jung,
Dongsung T. Park,
Seokyeong Lee,
Uh** Kim,
Chanuk Yang,
Jehyun Kim,
V. Umansky,
Dohun Kim,
H. -S. Sim,
Yunchul Chung,
Hyoungsoon Choi,
Hyung Kook Choi
Abstract:
The resemblance between electrons and optical waves has strongly driven the advancement of mesoscopic physics. However, electron waves have yet to be understood in open cavity structures which have provided contemporary optics with rich insight towards non-Hermitian systems and complex interactions between resonance mode. Here, we report the realization of an open cavity resonator in a two-dimensi…
▽ More
The resemblance between electrons and optical waves has strongly driven the advancement of mesoscopic physics. However, electron waves have yet to be understood in open cavity structures which have provided contemporary optics with rich insight towards non-Hermitian systems and complex interactions between resonance mode. Here, we report the realization of an open cavity resonator in a two-dimensional electronic system. We studied the resonant electron modes within the cavity and resolved the signatures of longitudinal and transverse quantization, showing that the modes are robust despite the openness of the cavity being highly open to the background continuum. The transverse modes were investigated by applying a controlled deformation to the cavity, and their spatial distributions were further analyzed using magnetoconductance measurements and numerical simulation. These results lay the groundwork to exploring electronic wavefunctions in the context of modern optical systems, such as the dielectric microcavity.
△ Less
Submitted 15 March, 2023;
originally announced March 2023.
-
Relation-Aware Language-Graph Transformer for Question Answering
Authors:
**young Park,
Hyeong Kyu Choi,
Juyeon Ko,
Hyeon** Park,
Ji-Hoon Kim,
Jisu Jeong,
Kyungmin Kim,
Hyunwoo J. Kim
Abstract:
Question Answering (QA) is a task that entails reasoning over natural language contexts, and many relevant works augment language models (LMs) with graph neural networks (GNNs) to encode the Knowledge Graph (KG) information. However, most existing GNN-based modules for QA do not take advantage of rich relational information of KGs and depend on limited information interaction between the LM and th…
▽ More
Question Answering (QA) is a task that entails reasoning over natural language contexts, and many relevant works augment language models (LMs) with graph neural networks (GNNs) to encode the Knowledge Graph (KG) information. However, most existing GNN-based modules for QA do not take advantage of rich relational information of KGs and depend on limited information interaction between the LM and the KG. To address these issues, we propose Question Answering Transformer (QAT), which is designed to jointly reason over language and graphs with respect to entity relations in a unified manner. Specifically, QAT constructs Meta-Path tokens, which learn relation-centric embeddings based on diverse structural and semantic relations. Then, our Relation-Aware Self-Attention module comprehensively integrates different modalities via the Cross-Modal Relative Position Bias, which guides information exchange between relevant entites of different modalities. We validate the effectiveness of QAT on commonsense question answering datasets like CommonsenseQA and OpenBookQA, and on a medical question answering dataset, MedQA-USMLE. On all the datasets, our method achieves state-of-the-art performance. Our code is available at http://github.com/mlvlab/QAT.
△ Less
Submitted 25 April, 2023; v1 submitted 2 December, 2022;
originally announced December 2022.
-
TokenMixup: Efficient Attention-guided Token-level Data Augmentation for Transformers
Authors:
Hyeong Kyu Choi,
Joonmyung Choi,
Hyunwoo J. Kim
Abstract:
Mixup is a commonly adopted data augmentation technique for image classification. Recent advances in mixup methods primarily focus on mixing based on saliency. However, many saliency detectors require intense computation and are especially burdensome for parameter-heavy transformer models. To this end, we propose TokenMixup, an efficient attention-guided token-level data augmentation method that a…
▽ More
Mixup is a commonly adopted data augmentation technique for image classification. Recent advances in mixup methods primarily focus on mixing based on saliency. However, many saliency detectors require intense computation and are especially burdensome for parameter-heavy transformer models. To this end, we propose TokenMixup, an efficient attention-guided token-level data augmentation method that aims to maximize the saliency of a mixed set of tokens. TokenMixup provides x15 faster saliency-aware data augmentation compared to gradient-based methods. Moreover, we introduce a variant of TokenMixup which mixes tokens within a single instance, thereby enabling multi-scale feature augmentation. Experiments show that our methods significantly improve the baseline models' performance on CIFAR and ImageNet-1K, while being more efficient than previous methods. We also reach state-of-the-art performance on CIFAR-100 among from-scratch transformer models. Code is available at https://github.com/mlvlab/TokenMixup.
△ Less
Submitted 14 October, 2022;
originally announced October 2022.
-
Consistency Learning via Decoding Path Augmentation for Transformers in Human Object Interaction Detection
Authors:
Jihwan Park,
SeungJun Lee,
Hwan Heo,
Hyeong Kyu Choi,
Hyunwoo J. Kim
Abstract:
Human-Object Interaction detection is a holistic visual recognition task that entails object detection as well as interaction classification. Previous works of HOI detection has been addressed by the various compositions of subset predictions, e.g., Image -> HO -> I, Image -> HI -> O. Recently, transformer based architecture for HOI has emerged, which directly predicts the HOI triplets in an end-t…
▽ More
Human-Object Interaction detection is a holistic visual recognition task that entails object detection as well as interaction classification. Previous works of HOI detection has been addressed by the various compositions of subset predictions, e.g., Image -> HO -> I, Image -> HI -> O. Recently, transformer based architecture for HOI has emerged, which directly predicts the HOI triplets in an end-to-end fashion (Image -> HOI). Motivated by various inference paths for HOI detection, we propose cross-path consistency learning (CPC), which is a novel end-to-end learning strategy to improve HOI detection for transformers by leveraging augmented decoding paths. CPC learning enforces all the possible predictions from permuted inference sequences to be consistent. This simple scheme makes the model learn consistent representations, thereby improving generalization without increasing model capacity. Our experiments demonstrate the effectiveness of our method, and we achieved significant improvement on V-COCO and HICO-DET compared to the baseline models. Our code is available at https://github.com/mlvlab/CPChoi.
△ Less
Submitted 10 April, 2022;
originally announced April 2022.
-
Numerical Reconstruction of 2D Magnetic Focusing Experiments
Authors:
Dongsung T. Park,
Seokyeong Lee,
Uh** Kim,
Yunchul Chung,
Hyoungsoon Choi,
Hyung Kook Choi
Abstract:
Spatial aspects in quantum mechanics are often difficult to model in geometrically intricate settings that are typical of mesoscopic physics. In such cases, predicting the device behaviors is a vital but difficult challenge. Transverse magnetic focusing (TMF) is a prime example where a classically simple effect becomes difficult to approach in the quantum regime. Here, we have simulated a realisti…
▽ More
Spatial aspects in quantum mechanics are often difficult to model in geometrically intricate settings that are typical of mesoscopic physics. In such cases, predicting the device behaviors is a vital but difficult challenge. Transverse magnetic focusing (TMF) is a prime example where a classically simple effect becomes difficult to approach in the quantum regime. Here, we have simulated a realistic TMF device and compared the results to those from experiments performed on GaAs/AlGaAs two-dimensional electron gas systems. Unlike previous studies, device features such as quantum point contacts and disorder were realized within the simulation. The simulated and experimental focusing spectra showed good agreement, and the analysis was extended to multichannel and energy-modulated scenarios. By revisiting the energy-modulated simulation with a quantum dot (QD) emitter, we confirmed that the unique geometry of a QD does not affect the focusing spectra, thereby validating the feasibility of such experiments in the study of monoenergetic excitations.
△ Less
Submitted 28 May, 2020;
originally announced May 2020.
-
Ultrafast energy relaxation of quantum dot-generated 2D hot electrons
Authors:
Dongsung T. Park,
Dongkun Kim,
Uh** Kim,
Hwanchul Jung,
Juho Choi,
Cheolhee Han,
Yunchul Chung,
H. -S. Sim,
V. Umansky,
Hyoungsoon Choi,
Hyung Kook Choi
Abstract:
Through a series of transverse magnetic focusing experiments, we show that hot electrons in a two-dimensional electron gas system undergo an ultrafast relaxation when generated by a quantum dot (QD) instead of a quantum point contact (QPC). We find here that QPC hot electrons were well described by the non-interacting Fermi gas model for excitations up to 1.5 meV above the Fermi level of 7.44 meV,…
▽ More
Through a series of transverse magnetic focusing experiments, we show that hot electrons in a two-dimensional electron gas system undergo an ultrafast relaxation when generated by a quantum dot (QD) instead of a quantum point contact (QPC). We find here that QPC hot electrons were well described by the non-interacting Fermi gas model for excitations up to 1.5 meV above the Fermi level of 7.44 meV, whereas QD hot electrons exhibited an energy loss quadratic to the excitation. The energy relaxation was a sizeable fraction of the tested excitations, up to about 55%. With the proposal that the hot electrons are relaxed by the QD immediately after emission, we present a toy model in which a capacitive coupling between the QD and its leads results in a finite, ultrafast energy relaxation.
△ Less
Submitted 23 April, 2020;
originally announced April 2020.
-
Magnetic proximity and nonreciprocal current switching in a monolayer WTe2 helical edge
Authors:
Wen** Zhao,
Zaiyao Fei,
Tiancheng Song,
Han Kyou Choi,
Tauno Palomaki,
Bosong Sun,
Paul Malinowski,
Michael A. McGuire,
Jiun-Haw Chu,
Xiaodong Xu,
David H. Cobden
Abstract:
The integration of diverse electronic phenomena, such as magnetism and nontrivial topology, into a single system is normally studied either by seeking materials that contain both ingredients, or by layered growth of contrasting materials. The ability to simply stack very different two dimensional (2D) van der Waals materials in intimate contact permits a different approach. Here we use this approa…
▽ More
The integration of diverse electronic phenomena, such as magnetism and nontrivial topology, into a single system is normally studied either by seeking materials that contain both ingredients, or by layered growth of contrasting materials. The ability to simply stack very different two dimensional (2D) van der Waals materials in intimate contact permits a different approach. Here we use this approach to couple the helical edges states in a 2D topological insulator, monolayer WTe2, to a 2D layered antiferromagnet, CrI3. We find that the edge conductance is sensitive to the magnetization state of the CrI3, and the coupling can be understood in terms of an exchange field from the nearest and next-nearest CrI3 layers that produces a gap in the helical edge. We also find that the nonlinear edge conductance depends on the magnetization of the nearest CrI3 layer relative to the current direction. At low temperatures this produces an extraordinarily large nonreciprocal current that is switched by changing the antiferromagnetic state of the CrI3.
△ Less
Submitted 2 January, 2020;
originally announced January 2020.
-
Stock Price Correlation Coefficient Prediction with ARIMA-LSTM Hybrid Model
Authors:
Hyeong Kyu Choi
Abstract:
Predicting the price correlation of two assets for future time periods is important in portfolio optimization. We apply LSTM recurrent neural networks (RNN) in predicting the stock price correlation coefficient of two individual stocks. RNNs are competent in understanding temporal dependencies. The use of LSTM cells further enhances its long term predictive properties. To encompass both linearity…
▽ More
Predicting the price correlation of two assets for future time periods is important in portfolio optimization. We apply LSTM recurrent neural networks (RNN) in predicting the stock price correlation coefficient of two individual stocks. RNNs are competent in understanding temporal dependencies. The use of LSTM cells further enhances its long term predictive properties. To encompass both linearity and nonlinearity in the model, we adopt the ARIMA model as well. The ARIMA model filters linear tendencies in the data and passes on the residual value to the LSTM model. The ARIMA LSTM hybrid model is tested against other traditional predictive financial models such as the full historical model, constant correlation model, single index model and the multi group model. In our empirical study, the predictive ability of the ARIMA-LSTM model turned out superior to all other financial models by a significant scale. Our work implies that it is worth considering the ARIMA LSTM model to forecast correlation coefficient for portfolio optimization.
△ Less
Submitted 1 October, 2018; v1 submitted 5 August, 2018;
originally announced August 2018.
-
Interaction-induced interference in the integer quantum Hall effect
Authors:
I. Sivan,
R. Bhattacharyya,
H. K. Choi,
M. Heiblum,
D. E. Feldman,
D. Mahalu,
V. Umansky
Abstract:
In recent interference experiments with an electronic Fabry-Perot interferometer (FPI), implemented in the integer quantum Hall effect regime, a flux periodicity of $h/2e$ was observed at bulk fillings $ν_B>2.5$. The halved periodicity was accompanied by an interfering charge $e^*=2e$, determined by shot noise measurements. Here, we present measurements, demonstrating that, counterintuitively, the…
▽ More
In recent interference experiments with an electronic Fabry-Perot interferometer (FPI), implemented in the integer quantum Hall effect regime, a flux periodicity of $h/2e$ was observed at bulk fillings $ν_B>2.5$. The halved periodicity was accompanied by an interfering charge $e^*=2e$, determined by shot noise measurements. Here, we present measurements, demonstrating that, counterintuitively, the coherence and the interference periodicity of the interfering chiral edge channel are solely determined by the coherence and the enclosed flux of the adjacent edge channel. Our results elucidate the important role of the latter and suggest that a neutral chiral edge mode plays a crucial role in the pairing phenomenon. Our findings reveal that the observed pairing of electrons is not a curious isolated phenomenon, but one of many manifestations of unexpected edge physics in the quantum Hall effect regime.
△ Less
Submitted 8 March, 2018; v1 submitted 21 July, 2017;
originally announced July 2017.
-
Charge frustration in a triangular triple quantum dot
Authors:
M. Seo,
H. K. Choi,
S. -Y. Lee,
N. Kim,
Y. Chung,
H. -S. Sim,
V. Umansky,
D. Mahalu
Abstract:
We experimentally investigate the charge (isospin) frustration induced by a geometrical symmetry in a triangular triple quantum dot. We observe the ground-state charge configurations of six-fold degeneracy, the manifestation of the frustration. The frustration results in omnidirectional charge transport, and it is accompanied by nearby nontrivial triple degenerate states in the charge stability di…
▽ More
We experimentally investigate the charge (isospin) frustration induced by a geometrical symmetry in a triangular triple quantum dot. We observe the ground-state charge configurations of six-fold degeneracy, the manifestation of the frustration. The frustration results in omnidirectional charge transport, and it is accompanied by nearby nontrivial triple degenerate states in the charge stability diagram. The findings agree with a capacitive interaction model. We also observe unusual transport by the frustration, which might be related to elastic cotunneling and the interference of trajectories through the dot. This work demonstrates a unique way of studying geometrical frustration in a controllable way.
△ Less
Submitted 26 February, 2014;
originally announced February 2014.
-
Multi-Valued Logic Gates based on Ballistic Transport in Quantum Point Contacts
Authors:
M. Seo,
C. Hong,
S. -Y. Lee,
H. K. Choi,
N. Kim,
Y. Chung,
V. Umansky,
D. Mahalu
Abstract:
Multi-valued logic gates, which can handle quaternary numbers as inputs, are developed by exploiting the ballistic transport properties of quantum point contacts in series. The principle of a logic gate that finds the minimum of two quaternary number inputs is demonstrated. The device is scalable to allow multiple inputs, which makes it possible to find the minimum of multiple inputs in a single g…
▽ More
Multi-valued logic gates, which can handle quaternary numbers as inputs, are developed by exploiting the ballistic transport properties of quantum point contacts in series. The principle of a logic gate that finds the minimum of two quaternary number inputs is demonstrated. The device is scalable to allow multiple inputs, which makes it possible to find the minimum of multiple inputs in a single gate operation. Also, the principle of a half-adder for quaternary number inputs is demonstrated. First, an adder that adds up two quaternary numbers and outputs the sum of inputs is demonstrated. Second, a device to express the sum of the adder into two quaternary digits [Carry (first digit) and Sum (second digit)] is demonstrated. All the logic gates presented in this paper can in principle be extended to allow decimal number inputs with high quality QPCs.
△ Less
Submitted 25 February, 2014;
originally announced February 2014.
-
An Electronic Quantum Eraser
Authors:
E. Weisz,
H. K. Choi,
I. Sivan,
M. Heiblum,
Y. Gefen,
D. Mahalu,
V. Umansky
Abstract:
Complementarity, the incomplete nature of a quantum measurement - a core concept in quantum mechanics - stems from the choice of the measurement apparatus. The notion of complementarity is closely related to Heisenberg's uncertainty principle, but the exact relation between the two remains a source of debate. For example, knowledge of a particle's position in a double slit interference experiment…
▽ More
Complementarity, the incomplete nature of a quantum measurement - a core concept in quantum mechanics - stems from the choice of the measurement apparatus. The notion of complementarity is closely related to Heisenberg's uncertainty principle, but the exact relation between the two remains a source of debate. For example, knowledge of a particle's position in a double slit interference experiment will quench its wave-like nature and, vice versa, observing the wave property via interference implies lack of knowledge of the particle's path. A canonical system for exploring complementarity is the quantum eraser (QE), predominantly studied thus far in photonic systems. A QE is an interference experiment consisting of two stages. First, one of the interfering paths is coupled to a 'which path' (WP) detector - demonstrating loss of interference due to acquisition of WP information. Second, the WP information is being 'erased' by projecting the detector's wavefunction on a particular basis; this renders the WP information inaccessible, thus allowing reconstruction of the interference pattern. In this work, we present a first implementation of a QE in an electronic system. Our system consists of two identical electronic Mach-Zehnder interferometers (MZIs) entangled via Coulomb interactions. Such novel setup has already attracted a considerable theoretical attention. With one MZI serving as a path detector and the other as the system interferometer, the visibility of the Aharonov-Bohm oscillation in the System can be controlled by the Detector. We demonstrate how a continuous change of the measurement basis, followed by post selection (via cross correlation of current fluctuations), allows a smooth transition between kee** and erasing the WP information.
△ Less
Submitted 14 September, 2013; v1 submitted 8 September, 2013;
originally announced September 2013.
-
Controlled dephasing of an electron interferometer with a path detector at equilibrium
Authors:
E. Weisz,
H. K. Choi,
M. Heiblum,
Yuval Gefen,
V. Umansky,
D. Mahalu
Abstract:
Controlled dephasing of electrons, via 'which path' detection, involves, in general, coupling a coherent system to a current driven noise source. However, here, we present a case in which a nearly isolated electron puddle at thermal equilibrium strongly affects the coherence of a nearby electronic interferometer. Moreover, for certain average electron occupations of the puddle, the interferometer…
▽ More
Controlled dephasing of electrons, via 'which path' detection, involves, in general, coupling a coherent system to a current driven noise source. However, here, we present a case in which a nearly isolated electron puddle at thermal equilibrium strongly affects the coherence of a nearby electronic interferometer. Moreover, for certain average electron occupations of the puddle, the interferometer exhibits complete dephasing. This robust phenomenon stems from the Friedel Sum Rule, which relates a system's occupation with its scattering phases. The interferometer opens a pee** window into physics of the isolated electron puddle, which cannot be accessed otherwise.
△ Less
Submitted 24 December, 2012;
originally announced December 2012.
-
Switching teraherz waves with gate-controlled active graphene metamaterials
Authors:
Seung Hoon Lee,
Muhan Choi,
Teun-Teun Kim,
Seungwoo Lee,
Ming Liu,
Xiaobo Yin,
Hong Kyw Choi,
Seung S. Lee,
Choon-Gi Choi,
Sung-Yool Choi,
Xiang Zhang,
Bumki Min
Abstract:
The extraordinary electronic properties of graphene, such as its continuously gate-variable ambipolar field effect and the resulting steep change in resistivity, provided the main thrusts for the rapid advance of graphene electronics. The gate-controllable electronic properties of graphene provide a route to efficiently manipulate the interaction of low-energy photons with massless Dirac fermions,…
▽ More
The extraordinary electronic properties of graphene, such as its continuously gate-variable ambipolar field effect and the resulting steep change in resistivity, provided the main thrusts for the rapid advance of graphene electronics. The gate-controllable electronic properties of graphene provide a route to efficiently manipulate the interaction of low-energy photons with massless Dirac fermions, which has recently sparked keen interest in graphene plasmonics. However, the electro-optic tuning capability of unpatterned graphene alone is still not strong enough for practical optoelectronic applications due to its nonresonant Drude-like behaviour. Here, we experimentally demonstrate that substantial gate-induced persistent switching and linear modulation of terahertz waves can be achieved in a two-dimensional artificial material, referred to as a metamaterial, into which an atomically thin, gated two-dimensional graphene layer is integrated. The gate-controllable light-matter interaction in the graphene layer can be greatly enhanced by the strong resonances and the corresponding field enhancement in the metamaterial. Although the thickness of the embedded single-layer graphene is more than 'six' orders of magnitude smaller than the wavelength (< λ/1,000,000), the one-atom-thick layer, in conjunction with the metamaterial, can modulate both the amplitude of the transmitted wave by up to 90 per cent and its phase by more than 40 degrees at room temperature. More interestingly, the gate-controlled active graphene metamaterials show hysteretic behaviour in the transmission of terahertz waves, especially when fabricated with multilayer graphene, which is indicative of persistent photonic memory effects.
△ Less
Submitted 4 March, 2012;
originally announced March 2012.
-
Interplay between carrier and impurity concentrations in annealed Ga$_{1-x}$Mn$_{x}$As intrinsic anomalous Hall Effect
Authors:
S. H. Chun,
Y. S. Kim,
H. K. Choi,
I. T. Jeong,
W. O. Lee,
K. S. Suh,
Y. S. OH,
K. H. Kim,
Z. G. Khim,
J. C. Woo,
Y. D. Park
Abstract:
Investigating the scaling behavior of annealed Ga$_{1-x}$Mn$_{x}$As anomalous Hall coefficients, we note a universal crossover regime where the scaling behavior changes from quadratic to linear, attributed to the anomalous Hall Effect intrinsic and extrinsic origins, respectively. Furthermore, measured anomalous Hall conductivities when properly scaled by carrier concentration remain constant, e…
▽ More
Investigating the scaling behavior of annealed Ga$_{1-x}$Mn$_{x}$As anomalous Hall coefficients, we note a universal crossover regime where the scaling behavior changes from quadratic to linear, attributed to the anomalous Hall Effect intrinsic and extrinsic origins, respectively. Furthermore, measured anomalous Hall conductivities when properly scaled by carrier concentration remain constant, equal to theoretically predicated values, spanning nearly a decade in conductivity as well as over 100 K in T$_{C}$. Both the qualitative and quantitative agreement confirms the validity of new equations of motion including the Berry phase contributions as well as tunablility of the intrinsic anomalous Hall Effect.
△ Less
Submitted 30 March, 2006;
originally announced March 2006.
-
Evidence of metallic clustering in annealed Ga1-xMnxAs from atypical scaling behavior of the anomalous Hall coefficient
Authors:
H. K. Choi,
W. O. Lee,
Y. S. OH,
K. H. Kim,
Y. D. Park,
S. S. A. Seo,
T. W. Noh,
Y. S. Kim,
Z. G. Khim,
I. T. Jeong,
J. C. Woo,
S. H. Chun
Abstract:
We report on the anomalous Hall coefficient and longitudinal resistivity scaling relationships on a series of annealed Ga1-xMnxAs epilayers (x~0.055). As-grown samples exhibit scaling parameter n of ~ 1. Near the optimal annealing temperature, we find n ~ 2 to be consistent with recent theories on the intrinsic origins of anomalous Hall Effect in Ga1-xMnxAs. For annealing temperatures far above…
▽ More
We report on the anomalous Hall coefficient and longitudinal resistivity scaling relationships on a series of annealed Ga1-xMnxAs epilayers (x~0.055). As-grown samples exhibit scaling parameter n of ~ 1. Near the optimal annealing temperature, we find n ~ 2 to be consistent with recent theories on the intrinsic origins of anomalous Hall Effect in Ga1-xMnxAs. For annealing temperatures far above the optimum, we note n > 3, similar behavior to certain inhomogeneous systems. This observation of atypical behavior agrees well with characteristic features attributable to spherical resonance from metallic inclusions from optical spectroscopy measurements.
△ Less
Submitted 17 March, 2006;
originally announced March 2006.
