Showing 1–50 of 77 results for author: Teo, C

Exposing Privacy Gaps: Membership Inference Attack on Preference Data for LLM Alignment

Authors: Qizhang Feng, Siva Rajesh Kasa, Hyokun Yun, Choon Hui Teo, Sravan Babu Bodapati

Abstract: Large Language Models (LLMs) have seen widespread adoption due to their remarkable natural language capabilities. However, when deploying them in real-world settings, it is important to align LLMs to generate texts according to acceptable human standards. Methods such as Proximal Policy Optimization (PPO) and Direct Preference Optimization (DPO) have made significant progress in refining LLMs usin… ▽ More Large Language Models (LLMs) have seen widespread adoption due to their remarkable natural language capabilities. However, when deploying them in real-world settings, it is important to align LLMs to generate texts according to acceptable human standards. Methods such as Proximal Policy Optimization (PPO) and Direct Preference Optimization (DPO) have made significant progress in refining LLMs using human preference data. However, the privacy concerns inherent in utilizing such preference data have yet to be adequately studied. In this paper, we investigate the vulnerability of LLMs aligned using human preference datasets to membership inference attacks (MIAs), highlighting the shortcomings of previous MIA approaches with respect to preference data. Our study has two main contributions: first, we introduce a novel reference-based attack framework specifically for analyzing preference data called PREMIA (\uline{Pre}ference data \uline{MIA}); second, we provide empirical evidence that DPO models are more vulnerable to MIA compared to PPO models. Our findings highlight gaps in current privacy-preserving practices for LLM alignment. △ Less

Submitted 8 July, 2024; originally announced July 2024.

PEFA: Parameter-Free Adapters for Large-scale Embedding-based Retrieval Models

Authors: Wei-Cheng Chang, Jyun-Yu Jiang, Jiong Zhang, Mutasem Al-Darabsah, Choon Hui Teo, Cho-Jui Hsieh, Hsiang-Fu Yu, S. V. N. Vishwanathan

Abstract: Embedding-based Retrieval Models (ERMs) have emerged as a promising framework for large-scale text retrieval problems due to powerful large language models. Nevertheless, fine-tuning ERMs to reach state-of-the-art results can be expensive due to the extreme scale of data as well as the complexity of multi-stages pipelines (e.g., pre-training, fine-tuning, distillation). In this work, we propose th… ▽ More Embedding-based Retrieval Models (ERMs) have emerged as a promising framework for large-scale text retrieval problems due to powerful large language models. Nevertheless, fine-tuning ERMs to reach state-of-the-art results can be expensive due to the extreme scale of data as well as the complexity of multi-stages pipelines (e.g., pre-training, fine-tuning, distillation). In this work, we propose the PEFA framework, namely ParamEter-Free Adapters, for fast tuning of ERMs without any backward pass in the optimization. At index building stage, PEFA equips the ERM with a non-parametric k-nearest neighbor (kNN) component. At inference stage, PEFA performs a convex combination of two scoring functions, one from the ERM and the other from the kNN. Based on the neighborhood definition, PEFA framework induces two realizations, namely PEFA-XL (i.e., extra large) using double ANN indices and PEFA-XS (i.e., extra small) using a single ANN index. Empirically, PEFA achieves significant improvement on two retrieval applications. For document retrieval, regarding Recall@100 metric, PEFA improves not only pre-trained ERMs on Trivia-QA by an average of 13.2%, but also fine-tuned ERMs on NQ-320K by an average of 5.5%, respectively. For product search, PEFA improves the Recall@100 of the fine-tuned ERMs by an average of 5.3% and 14.5%, for PEFA-XS and PEFA-XL, respectively. Our code is available at https://github.com/amzn/pecos/tree/mainline/examples/pefa-wsdm24. △ Less

Submitted 5 December, 2023; v1 submitted 4 December, 2023; originally announced December 2023.

Comments: Accept by WSDM 2024

On Measuring Fairness in Generative Models

Authors: Christopher T. H. Teo, Milad Abdollahzadeh, Ngai-Man Cheung

Abstract: Recently, there has been increased interest in fair generative models. In this work, we conduct, for the first time, an in-depth study on fairness measurement, a critical component in gauging progress on fair generative models. We make three contributions. First, we conduct a study that reveals that the existing fairness measurement framework has considerable measurement errors, even when highly a… ▽ More Recently, there has been increased interest in fair generative models. In this work, we conduct, for the first time, an in-depth study on fairness measurement, a critical component in gauging progress on fair generative models. We make three contributions. First, we conduct a study that reveals that the existing fairness measurement framework has considerable measurement errors, even when highly accurate sensitive attribute (SA) classifiers are used. These findings cast doubts on previously reported fairness improvements. Second, to address this issue, we propose CLassifier Error-Aware Measurement (CLEAM), a new framework which uses a statistical model to account for inaccuracies in SA classifiers. Our proposed CLEAM reduces measurement errors significantly, e.g., 4.98% $\rightarrow$ 0.62% for StyleGAN2 w.r.t. Gender. Additionally, CLEAM achieves this with minimal additional overhead. Third, we utilize CLEAM to measure fairness in important text-to-image generator and GANs, revealing considerable biases in these models that raise concerns about their applications. Code and more resources: https://sutd-visual-computing-group.github.io/CLEAM/. △ Less

Submitted 30 October, 2023; originally announced October 2023.

Comments: Accepted in NeurIPS23

A Survey on Generative Modeling with Limited Data, Few Shots, and Zero Shot

Authors: Milad Abdollahzadeh, Touba Malekzadeh, Christopher T. H. Teo, Keshigeyan Chandrasegaran, Guimeng Liu, Ngai-Man Cheung

Abstract: In machine learning, generative modeling aims to learn to generate new data statistically similar to the training data distribution. In this paper, we survey learning generative models under limited data, few shots and zero shot, referred to as Generative Modeling under Data Constraint (GM-DC). This is an important topic when data acquisition is challenging, e.g. healthcare applications. We discus… ▽ More In machine learning, generative modeling aims to learn to generate new data statistically similar to the training data distribution. In this paper, we survey learning generative models under limited data, few shots and zero shot, referred to as Generative Modeling under Data Constraint (GM-DC). This is an important topic when data acquisition is challenging, e.g. healthcare applications. We discuss background, challenges, and propose two taxonomies: one on GM-DC tasks and another on GM-DC approaches. Importantly, we study interactions between different GM-DC tasks and approaches. Furthermore, we highlight research gaps, research trends, and potential avenues for future exploration. Project website: https://gmdc-survey.github.io. △ Less

Submitted 26 July, 2023; originally announced July 2023.

Comments: Technical Survey. Touba Malekzadeh, Christopher T.H. Teo, Keshigeyan Chandrasegaran contribute equally

Partial fillings of the bosonic $E_8$ quantum Hall state

Authors: Pak Kau Lim, Michael Mulligan, Jeffrey C. Y. Teo

Abstract: We study bosonic topological phases constructed from electrons. In addition to a bulk excitation energy gap, these bosonic phases also have a fermion energy gap, below which all local excitations in the bulk and on the edge are even combinations of electrons. We focus on chiral phases, in which all low-energy edge excitations move in the same direction, that arise from the short-range entangled… ▽ More We study bosonic topological phases constructed from electrons. In addition to a bulk excitation energy gap, these bosonic phases also have a fermion energy gap, below which all local excitations in the bulk and on the edge are even combinations of electrons. We focus on chiral phases, in which all low-energy edge excitations move in the same direction, that arise from the short-range entangled $E_8$ quantum Hall state, the bosonic analog of the filled lowest Landau level of electrons. The $E_8$ edge-state theory features an $E_8$ Kac-Moody symmetry that can be decomposed into ${\cal G}_A \times {\cal G}_B$ subalgebras, such as $SU(3) \times E_6$, $SO(M) \times SO(16-M)$, and $G_2 \times F_4$. (Here, $\{SO(M) \}$, $\{SU(N)\}$, and $\{E_8, G_2, F_4 \}$ denote orthogonal, unitary, and exceptional Lie algebras.) Using these symmetry decompositions, we construct exactly solvable coupled-wire model Hamiltonians for families of long-range entangled ${\cal G}_A$ or ${\cal G}_B$ bosonic fractional quantum Hall states that ``partially fill" the $E_8$ state and are pairwise related by a generalized particle-hole symmetry. These long-range entangled states feature either Abelian or non-Abelian topological order. Some support the emergence of non-local Dirac and Majorana fermions, Ising anyons, metaplectic anyons, Fibonacci anyons, as well as deconfined $\mathbb{Z}_2$ gauge fluxes and charges. △ Less

Submitted 24 July, 2023; v1 submitted 30 December, 2022; originally announced December 2022.

Comments: 48 pages, 6 figures

Fair Generative Models via Transfer Learning

Authors: Christopher TH Teo, Milad Abdollahzadeh, Ngai-Man Cheung

Abstract: This work addresses fair generative models. Dataset biases have been a major cause of unfairness in deep generative models. Previous work had proposed to augment large, biased datasets with small, unbiased reference datasets. Under this setup, a weakly-supervised approach has been proposed, which achieves state-of-the-art quality and fairness in generated samples. In our work, based on this setup,… ▽ More This work addresses fair generative models. Dataset biases have been a major cause of unfairness in deep generative models. Previous work had proposed to augment large, biased datasets with small, unbiased reference datasets. Under this setup, a weakly-supervised approach has been proposed, which achieves state-of-the-art quality and fairness in generated samples. In our work, based on this setup, we propose a simple yet effective approach. Specifically, first, we propose fairTL, a transfer learning approach to learn fair generative models. Under fairTL, we pre-train the generative model with the available large, biased datasets and subsequently adapt the model using the small, unbiased reference dataset. We find that our fairTL can learn expressive sample generation during pre-training, thanks to the large (biased) dataset. This knowledge is then transferred to the target model during adaptation, which also learns to capture the underlying fair distribution of the small reference dataset. Second, we propose fairTL++, where we introduce two additional innovations to improve upon fairTL: (i) multiple feedback and (ii) Linear-Probing followed by Fine-Tuning (LP-FT). Taking one step further, we consider an alternative, challenging setup when only a pre-trained (potentially biased) model is available but the dataset that was used to pre-train the model is inaccessible. We demonstrate that our proposed fairTL and fairTL++ remain very effective under this setup. We note that previous work requires access to the large, biased datasets and is incapable of handling this more challenging setup. Extensive experiments show that fairTL and fairTL++ achieve state-of-the-art in both quality and fairness of generated samples. The code and additional resources can be found at bearwithchris.github.io/fairTL/. △ Less

Submitted 1 December, 2022; originally announced December 2022.

Comments: Accepted in AAAI-2023

MICO: Selective Search with Mutual Information Co-training

Authors: Zhanyu Wang, Xiao Zhang, Hyokun Yun, Choon Hui Teo, Trishul Chilimbi

Abstract: In contrast to traditional exhaustive search, selective search first clusters documents into several groups before all the documents are searched exhaustively by a query, to limit the search executed within one group or only a few groups. Selective search is designed to reduce the latency and computation in modern large-scale search systems. In this study, we propose MICO, a Mutual Information CO-… ▽ More In contrast to traditional exhaustive search, selective search first clusters documents into several groups before all the documents are searched exhaustively by a query, to limit the search executed within one group or only a few groups. Selective search is designed to reduce the latency and computation in modern large-scale search systems. In this study, we propose MICO, a Mutual Information CO-training framework for selective search with minimal supervision using the search logs. After training, MICO does not only cluster the documents, but also routes unseen queries to the relevant clusters for efficient retrieval. In our empirical experiments, MICO significantly improves the performance on multiple metrics of selective search and outperforms a number of existing competitive baselines. △ Less

Submitted 9 September, 2022; originally announced September 2022.

Journal ref: Proceedings of the 29th International Conference on Computational Linguistics (COLING). 2022

On the Value of Behavioral Representations for Dense Retrieval

Authors: Nan Jiang, Dhivya Eswaran, Choon Hui Teo, Yexiang Xue, Yesh Dattatreya, Sujay Sanghavi, Vishy Vishwanathan

Abstract: We consider text retrieval within dense representational space in real-world settings such as e-commerce search where (a) document popularity and (b) diversity of queries associated with a document have a skewed distribution. Most of the contemporary dense retrieval literature presents two shortcomings in these settings. (1) They learn an almost equal number of representations per document, agnost… ▽ More We consider text retrieval within dense representational space in real-world settings such as e-commerce search where (a) document popularity and (b) diversity of queries associated with a document have a skewed distribution. Most of the contemporary dense retrieval literature presents two shortcomings in these settings. (1) They learn an almost equal number of representations per document, agnostic to the fact that a few head documents are disproportionately more critical to achieving a good retrieval performance. (ii) They learn purely semantic document representations inferred from intrinsic document characteristics which may not contain adequate information to determine the queries for which the document is relevant--especially when the document is short. We propose to overcome these limitations by augmenting semantic document representations learned by bi-encoders with behavioral document representations learned by our proposed approach MVG. To do so, MVG (1) determines how to divide the total budget for behavioral representations by drawing a connection to the Pitman-Yor process, and (2) simply clusters the queries related to a given document (based on user behavior) within the representational space learned by a base bi-encoder, and treats the cluster centers as its behavioral representations. Our central contribution is the finding such a simple intuitive light-weight approach leads to substantial gains in key first-stage retrieval metrics by incurring only a marginal memory overhead. We establish this via extensive experiments over three large public datasets comparing several single-vector and multi-vector bi-encoders, a proprietary e-commerce search dataset compared to production-quality bi-encoder, and an A/B test. △ Less

Submitted 11 August, 2022; originally announced August 2022.

Grad's Distribution Function for 13 Moments based Moment Gas Kinetic Solver for Steady and Unsteady Rarefied flows: Discrete and Explicit Forms

Authors: W. Liu, Z. J. Liu, Z. L. Zhang, C. J. Teo, C. Shu

Abstract: Efficient modeling of rarefied flow has drawn widespread interest for practical engineering applications. In the present work, we proposed the Grad's distribution function for 13 moments-based moment gas kinetic solver (G13-MGKS) and the macroscopic governing equations are derived based on the moment integral of discrete Boltzmann equation in the finite volume framework. Numerical fluxes at the ce… ▽ More Efficient modeling of rarefied flow has drawn widespread interest for practical engineering applications. In the present work, we proposed the Grad's distribution function for 13 moments-based moment gas kinetic solver (G13-MGKS) and the macroscopic governing equations are derived based on the moment integral of discrete Boltzmann equation in the finite volume framework. Numerical fluxes at the cell interface related to the macroscopic variables, stress and heat flux can be reconstructed from the Boltzmann integration equation at surrounding points of the cell interface directly, so the complicated partial differential equations with tedious implementation of boundary conditions in the moment method can be avoided. Meanwhile, the explicit expression of numerical fluxes is proposed, which could release the present solver the from the discretization and numerical summation in molecular velocity space. To evaluate the Grad's distribution function for 13 moments in the present framework, the G13-MGKS with the discrete and explicit form of numerical fluxes are examined by several test cases covering the steady and unsteady rarefied flows. Numerical results indicate that the G13-MGKS could simulate continuum flows accurately and present reasonable prediction for rarefied flows at moderate Knudsen number. Moreover, the tests of computations and memory costs demonstrate that the present framework could preserve the highly efficient feature. △ Less

Submitted 19 July, 2022; originally announced July 2022.

Embracing Structure in Data for Billion-Scale Semantic Product Search

Authors: Vihan Lakshman, Choon Hui Teo, Xiaowen Chu, Priyanka Nigam, Abhinandan Patni, Pooja Maknikar, SVN Vishwanathan

Abstract: We present principled approaches to train and deploy dyadic neural embedding models at the billion scale, focusing our investigation on the application of semantic product search. When training a dyadic model, one seeks to embed two different types of entities (e.g., queries and documents or users and movies) in a common vector space such that pairs with high relevance are positioned nearby. Durin… ▽ More We present principled approaches to train and deploy dyadic neural embedding models at the billion scale, focusing our investigation on the application of semantic product search. When training a dyadic model, one seeks to embed two different types of entities (e.g., queries and documents or users and movies) in a common vector space such that pairs with high relevance are positioned nearby. During inference, given an embedding of one type (e.g., a query or a user), one seeks to retrieve the entities of the other type (e.g., documents or movies, respectively) that are highly relevant. In this work, we show that exploiting the natural structure of real-world datasets helps address both challenges efficiently. Specifically, we model dyadic data as a bipartite graph with edges between pairs with positive associations. We then propose to partition this network into semantically coherent clusters and thus reduce our search space by focusing on a small subset of these partitions for a given input. During training, this technique enables us to efficiently mine hard negative examples while, at inference, we can quickly find the nearest neighbors for a given embedding. We provide offline experimental results that demonstrate the efficacy of our techniques for both training and inference on a billion-scale Amazon.com product search dataset. △ Less

Submitted 12 October, 2021; originally announced October 2021.

Comments: 10 pages

Towards Generalized and Incremental Few-Shot Object Detection

Authors: Yiting Li, Haiyue Zhu, Jun Ma, Chek Sing Teo, Cheng Xiang, Prahlad Vadakkepat, Tong Heng Lee

Abstract: Real-world object detection is highly desired to be equipped with the learning expandability that can enlarge its detection classes incrementally. Moreover, such learning from only few annotated training samples further adds the flexibility for the object detector, which is highly expected in many applications such as autonomous driving, robotics, etc. However, such sequential learning scenario wi… ▽ More Real-world object detection is highly desired to be equipped with the learning expandability that can enlarge its detection classes incrementally. Moreover, such learning from only few annotated training samples further adds the flexibility for the object detector, which is highly expected in many applications such as autonomous driving, robotics, etc. However, such sequential learning scenario with few-shot training samples generally causes catastrophic forgetting and dramatic overfitting. In this paper, to address the above incremental few-shot learning issues, a novel Incremental Few-Shot Object Detection (iFSOD) method is proposed to enable the effective continual learning from few-shot samples. Specifically, a Double-Branch Framework (DBF) is proposed to decouple the feature representation of base and novel (few-shot) class, which facilitates both the old-knowledge retention and new-class adaption simultaneously. Furthermore, a progressive model updating rule is carried out to preserve the long-term memory on old classes effectively when adapt to sequential new classes. Moreover, an inter-task class separation loss is proposed to extend the decision region of new-coming classes for better feature discrimination. We conduct experiments on both Pascal VOC and MS-COCO, which demonstrate that our method can effectively solve the problem of incremental few-shot detection and significantly improve the detection accuracy on both base and novel classes. △ Less

Submitted 23 September, 2021; originally announced September 2021.

Comments: 12 pages, 4 figures

Measuring Fairness in Generative Models

Authors: Christopher T. H Teo, Ngai-Man Cheung

Abstract: Deep generative models have made much progress in improving training stability and quality of generated data. Recently there has been increased interest in the fairness of deep-generated data. Fairness is important in many applications, e.g. law enforcement, as biases will affect efficacy. Central to fair data generation are the fairness metrics for the assessment and evaluation of different gener… ▽ More Deep generative models have made much progress in improving training stability and quality of generated data. Recently there has been increased interest in the fairness of deep-generated data. Fairness is important in many applications, e.g. law enforcement, as biases will affect efficacy. Central to fair data generation are the fairness metrics for the assessment and evaluation of different generative models. In this paper, we first review fairness metrics proposed in previous works and highlight potential weaknesses. We then discuss a performance benchmark framework along with the assessment of alternative metrics. △ Less

Submitted 16 July, 2021; originally announced July 2021.

Comments: Accepted in ICML 2021 Workshop - Machine Learning for Data: Automated Creation, Privacy, Bias

Dihedral twist liquid models from emergent Majorana fermions

Authors: Jeffrey C. Y. Teo, Yichen Hu

Abstract: We present a family of electron-based coupled-wire models of bosonic orbifold topological phases, referred to as twist liquids, in two spatial dimensions. All local fermion degrees of freedom are gapped and removed from the topological order by many-body interactions. Bosonic chiral spin liquids and anyonic superconductors are constructed on an array of interacting wires, each supports emergent ma… ▽ More We present a family of electron-based coupled-wire models of bosonic orbifold topological phases, referred to as twist liquids, in two spatial dimensions. All local fermion degrees of freedom are gapped and removed from the topological order by many-body interactions. Bosonic chiral spin liquids and anyonic superconductors are constructed on an array of interacting wires, each supports emergent massless Majorana fermions that are non-local (fractional) and constitute the $SO(N)$ Kac-Moody Wess-Zumino-Witten algebra at level 1. We focus on the dihedral $D_k$ symmetry of $SO(2n)_1$, and its promotion to a gauge symmetry by manipulating the locality of fermion pairs. Gauging the symmetry (sub)group generates the $\mathcal{C}/G$ twist liquids, where $G=\mathbb{Z}_2$ for $\mathcal{C}=U(1)_l$, $SU(n)_1$, and $G=\mathbb{Z}_2$, $\mathbb{Z}_k$, $D_k$ for $\mathcal{C}=SO(2n)_1$. We construct exactly solvable models for all of these topological states. We prove the presence of a bulk excitation energy gap and demonstrate the appearance of edge orbifold conformal field theories corresponding to the twist liquid topological orders. We analyze the statistical properties of the anyon excitations, including the non-Abelian metaplectic anyons and a new class of quasiparticles referred to as Ising-fluxons. We show an eight-fold periodic gauging pattern in $SO(2n)_1/G$ by identifying the non-chiral components of the twist liquids with discrete gauge theories. △ Less

Submitted 27 March, 2023; v1 submitted 15 July, 2021; originally announced July 2021.

Comments: 75 pages, 4 figures

Journal ref: Quantum 7, 967 (2023)

Extreme Multi-label Learning for Semantic Matching in Product Search

Authors: Wei-Cheng Chang, Daniel Jiang, Hsiang-Fu Yu, Choon-Hui Teo, Jiong Zhang, Kai Zhong, Kedarnath Kolluri, Qie Hu, Nikhil Shandilya, Vyacheslav Ievgrafov, Japinder Singh, Inderjit S. Dhillon

Abstract: We consider the problem of semantic matching in product search: given a customer query, retrieve all semantically related products from a huge catalog of size 100 million, or more. Because of large catalog spaces and real-time latency constraints, semantic matching algorithms not only desire high recall but also need to have low latency. Conventional lexical matching approaches (e.g., Okapi-BM25)… ▽ More We consider the problem of semantic matching in product search: given a customer query, retrieve all semantically related products from a huge catalog of size 100 million, or more. Because of large catalog spaces and real-time latency constraints, semantic matching algorithms not only desire high recall but also need to have low latency. Conventional lexical matching approaches (e.g., Okapi-BM25) exploit inverted indices to achieve fast inference time, but fail to capture behavioral signals between queries and products. In contrast, embedding-based models learn semantic representations from customer behavior data, but the performance is often limited by shallow neural encoders due to latency constraints. Semantic product search can be viewed as an eXtreme Multi-label Classification (XMC) problem, where customer queries are input instances and products are output labels. In this paper, we aim to improve semantic product search by using tree-based XMC models where inference time complexity is logarithmic in the number of products. We consider hierarchical linear models with n-gram features for fast real-time inference. Quantitatively, our method maintains a low latency of 1.25 milliseconds per query and achieves a 65% improvement of Recall@100 (60.9% v.s. 36.8%) over a competing embedding-based DSSM model. Our model is robust to weight pruning with varying thresholds, which can flexibly meet different system requirements for online deployments. Qualitatively, our method can retrieve products that are complementary to existing product search system and add diversity to the match set. △ Less

Submitted 23 June, 2021; originally announced June 2021.

Comments: Accepted in KDD 2021 Applied Data Science Track

Disentangling (2+1)d Topological States of Matter with the Entanglement Negativity

Authors: Pak Kau Lim, Hamed Asasi, Jeffrey C. Y. Teo, Michael Mulligan

Abstract: We use the entanglement negativity, a bipartite measure of entanglement in mixed quantum states, to study how multipartite entanglement constrains the real-space structure of the ground state wavefunctions of $(2+1)$-dimensional topological phases. We focus on the (Abelian) Laughlin and (non-Abelian) Moore-Read states at filling fraction $ν=1/m$. We show that a combination of entanglement negativi… ▽ More We use the entanglement negativity, a bipartite measure of entanglement in mixed quantum states, to study how multipartite entanglement constrains the real-space structure of the ground state wavefunctions of $(2+1)$-dimensional topological phases. We focus on the (Abelian) Laughlin and (non-Abelian) Moore-Read states at filling fraction $ν=1/m$. We show that a combination of entanglement negativities, calculated with respect to specific cylinder and torus geometries, determines a necessary condition for when a topological state can be disentangled, i.e., factorized into a tensor product of states defined on cylinder subregions. This condition, which requires the ground state to lie in a definite topological sector, is sufficient for the Laughlin state. On the other hand, we find that a general Moore-Read ground state cannot be disentangled even when the disentangling condition holds. △ Less

Submitted 28 September, 2021; v1 submitted 14 June, 2021; originally announced June 2021.

Comments: 35 pages, 5 figures

Journal ref: Phys. Rev. B 104, 115155 (2021)

Limousine Service Management: Capacity Planning with Predictive Analytics and Optimization

Authors: Peng Liu, Ying Chen, Chung-Piaw Teo

Abstract: The limousine service in luxury hotels is an integral component of the whole customer journey in the hospitality industry. One of the largest hotels in Singapore manages a fleet of both in-house and outsourced vehicles around the clock, serving 9000 trips per month on average. The need for vehicles may scale up rapidly, especially during special events and festive periods in the country. The exces… ▽ More The limousine service in luxury hotels is an integral component of the whole customer journey in the hospitality industry. One of the largest hotels in Singapore manages a fleet of both in-house and outsourced vehicles around the clock, serving 9000 trips per month on average. The need for vehicles may scale up rapidly, especially during special events and festive periods in the country. The excess demand is met by having additional outsourced vehicles on standby, incurring millions of dollars of additional expenses per year for the hotel. Determining the required number of limousines by hour of the day is a challenging service capacity planning problem. In this paper, a recent transformational journey to manage this problem in the hotel is introduced, driving up to S\$3.2 million of savings per year with improved service level. The approach builds on widely available open-source statistical and spreadsheet optimization tools, along with robotic process automation, to optimize the schedule of its fleet of limousines and drivers, and to support decision-making for planners/controllers to drive sustained business value. △ Less

Submitted 11 September, 2020; originally announced September 2020.

Regularised Text Logistic Regression: Key Word Detection and Sentiment Classification for Online Reviews

Authors: Ying Chen, Peng Liu, Chung Piaw Teo

Abstract: Online customer reviews have become important for managers and executives in the hospitality and catering industry who wish to obtain a comprehensive understanding of their customers' demands and expectations. We propose a Regularized Text Logistic (RTL) regression model to perform text analytics and sentiment classification on unstructured text data, which automatically identifies a set of statis… ▽ More Online customer reviews have become important for managers and executives in the hospitality and catering industry who wish to obtain a comprehensive understanding of their customers' demands and expectations. We propose a Regularized Text Logistic (RTL) regression model to perform text analytics and sentiment classification on unstructured text data, which automatically identifies a set of statistically significant and operationally insightful word features, and achieves satisfactory predictive classification accuracy. We apply the RTL model to two online review datasets, Restaurant and Hotel, from TripAdvisor. Our results demonstrate satisfactory classification performance compared with alternative classifiers with a highest true positive rate of 94.9%. Moreover, RTL identifies a small set of word features, corresponding to 3% for Restaurant and 20% for Hotel, which boosts working efficiency by allowing managers to drill down into a much smaller set of important customer reviews. We also develop the consistency, sparsity and oracle property of the estimator. △ Less

Submitted 9 September, 2020; originally announced September 2020.

Gras** Detection Network with Uncertainty Estimation for Confidence-Driven Semi-Supervised Domain Adaptation

Authors: Haiyue Zhu, Yiting Li, Fengjun Bai, Wenjie Chen, Xiaocong Li, Jun Ma, Chek Sing Teo, Pey Yuen Tao, Wei Lin

Abstract: Data-efficient domain adaptation with only a few labelled data is desired for many robotic applications, e.g., in gras** detection, the inference skill learned from a gras** dataset is not universal enough to directly apply on various other daily/industrial applications. This paper presents an approach enabling the easy domain adaptation through a novel gras** detection network with confiden… ▽ More Data-efficient domain adaptation with only a few labelled data is desired for many robotic applications, e.g., in gras** detection, the inference skill learned from a gras** dataset is not universal enough to directly apply on various other daily/industrial applications. This paper presents an approach enabling the easy domain adaptation through a novel gras** detection network with confidence-driven semi-supervised learning, where these two components deeply interact with each other. The proposed gras** detection network specially provides a prediction uncertainty estimation mechanism by leveraging on Feature Pyramid Network (FPN), and the mean-teacher semi-supervised learning utilizes such uncertainty information to emphasizing the consistency loss only for those unlabelled data with high confidence, which we referred it as the confidence-driven mean teacher. This approach largely prevents the student model to learn the incorrect/harmful information from the consistency loss, which speeds up the learning progress and improves the model accuracy. Our results show that the proposed network can achieve high success rate on the Cornell gras** dataset, and for domain adaptation with very limited data, the confidence-driven mean teacher outperforms the original mean teacher and direct training by more than 10% in evaluation loss especially for avoiding the overfitting and model diverging. △ Less

Submitted 20 August, 2020; originally announced August 2020.

Comments: 6 pages, 7 figures, accepted in IROS 2020

An encryption-decryption framework for validating single-particle imaging

Authors: Zhou Shen, Colin Zhi Wei Teo, Kartik Ayyer, N. Duane Loh

Abstract: We propose an encryption-decryption framework for validating diffraction intensity volumes reconstructed using single-particle imaging (SPI) with x-ray free-electron lasers (XFELs) when the ground truth volume is absent. This framework exploits each reconstructed volumes' ability to decipher latent variables (e.g. orientations) of unseen sentinel diffraction patterns. Using this framework, we quan… ▽ More We propose an encryption-decryption framework for validating diffraction intensity volumes reconstructed using single-particle imaging (SPI) with x-ray free-electron lasers (XFELs) when the ground truth volume is absent. This framework exploits each reconstructed volumes' ability to decipher latent variables (e.g. orientations) of unseen sentinel diffraction patterns. Using this framework, we quantify novel measures of orientation disconcurrence, inconsistency, and disagreement between the decryptions by two independently reconstructed volumes. We also study how these measures can be used to define data sufficiency and its relation to spatial resolution, and the practical consequences of focusing XFEL pulses to smaller foci. This framework overcomes critical ambiguities in using Fourier Shell Correlation (FSC) as a validation measure for SPI. Finally, we show how this encryption-decryption framework naturally leads to an information-theoretic reformulation of the resolving power of XFEL-SPI, which we hope will lead to principled frameworks for experiment and instrument design. △ Less

Submitted 31 July, 2020; v1 submitted 6 July, 2020; originally announced July 2020.

Data-Driven Multi-Objective Controller Optimization for a Magnetically-Levitated Nanopositioning System

Authors: Xiaocong Li, Haiyue Zhu, Jun Ma, Tat Joo Teo, Chek Sing Teo, Masayoshi Tomizuka, Tong Heng Lee

Abstract: The performance achieved with traditional model-based control system design approaches typically relies heavily upon accurate modeling of the motion dynamics. However, modeling the true dynamics of present-day increasingly complex systems can be an extremely challenging task; and the usually necessary practical approximations often render the automation system to operate in a non-optimal condition… ▽ More The performance achieved with traditional model-based control system design approaches typically relies heavily upon accurate modeling of the motion dynamics. However, modeling the true dynamics of present-day increasingly complex systems can be an extremely challenging task; and the usually necessary practical approximations often render the automation system to operate in a non-optimal condition. This problem can be greatly aggravated in the case of a multi-axis magnetically-levitated nanopositioning system where the fully floating behavior and multi-axis coupling make extremely accurate identification of the motion dynamics largely impossible. On the other hand, in many related industrial automation applications, e.g., the scanning process with the maglev system, repetitive motions are involved which could generate a large amount of motion data under non-optimal conditions. These motion data essentially contain rich information; therefore, the possibility exists to develop an intelligent automation system to learn from these motion data and to drive the system to operate towards optimality in a data-driven manner. Along this line then, this paper proposes a data-driven controller optimization approach that learns from the past non-optimal motion data to iteratively improve the motion control performance. Specifically, a novel data-driven multi-objective optimization approach is proposed that is able to automatically estimate the gradient and Hessian purely based on the measured motion data; the multi-objective cost function is suitably designed to take into account both smooth and accurate trajectory tracking. Experiments are then conducted on the maglev nanopositioning system to demonstrate the effectiveness of the proposed method, and the results show rather clearly the practical appeal of our methodology for related complex robotic systems with no accurate model available. △ Less

Submitted 6 July, 2020; originally announced July 2020.

Comments: 11 pages, 11 figures, to appear in IEEE/ASME Transactions on Mechatronics

Incremental Few-Shot Object Detection for Robotics

Authors: Yiting Li, Haiyue Zhu, Sichao Tian, Fan Feng, Jun Ma, Chek Sing Teo, Cheng Xiang, Prahlad Vadakkepat, Tong Heng Lee

Abstract: Incremental few-shot learning is highly expected for practical robotics applications. On one hand, robot is desired to learn new tasks quickly and flexibly using only few annotated training samples; on the other hand, such new additional tasks should be learned in a continuous and incremental manner without forgetting the previous learned knowledge dramatically. In this work, we propose a novel Cl… ▽ More Incremental few-shot learning is highly expected for practical robotics applications. On one hand, robot is desired to learn new tasks quickly and flexibly using only few annotated training samples; on the other hand, such new additional tasks should be learned in a continuous and incremental manner without forgetting the previous learned knowledge dramatically. In this work, we propose a novel Class-Incremental Few-Shot Object Detection (CI-FSOD) framework that enables deep object detection network to perform effective continual learning from just few-shot samples without re-accessing the previous training data. We achieve this by equip** the widely-used Faster-RCNN detector with three elegant components. Firstly, to best preserve performance on the pre-trained base classes, we propose a novel Dual-Embedding-Space (DES) architecture which decouples the representation learning of base and novel categories into different spaces. Secondly, to mitigate the catastrophic forgetting on the accumulated novel classes, we propose a Sequential Model Fusion (SMF) method, which is able to achieve long-term memory without additional storage cost. Thirdly, to promote inter-task class separation in feature space, we propose a novel regularization technique that extends the classification boundary further away from the previous classes to avoid misclassification. Overall, our framework is simple yet effective and outperforms the previous SOTA with a significant margin of 2.4 points in AP performance. △ Less

Submitted 23 March, 2022; v1 submitted 6 May, 2020; originally announced May 2020.

Comments: ICRA 2022

Entanglement Entropy of Generalized Moore-Read Fractional Quantum Hall State Interfaces

Authors: Ramanjit Sohal, Bo Han, Luiz H. Santos, Jeffrey C. Y. Teo

Abstract: Topologically ordered phases of matter can be characterized by the presence of a universal, constant contribution to the entanglement entropy known as the topological entanglement entropy (TEE). The TEE can been calculated for Abelian phases via a "cut-and-glue" approach by treating the entanglement cut as a physical cut, coupling the resulting gapless edges with explicit tunneling terms, and comp… ▽ More Topologically ordered phases of matter can be characterized by the presence of a universal, constant contribution to the entanglement entropy known as the topological entanglement entropy (TEE). The TEE can been calculated for Abelian phases via a "cut-and-glue" approach by treating the entanglement cut as a physical cut, coupling the resulting gapless edges with explicit tunneling terms, and computing the entanglement between the two edges. We provide a first step towards extending this methodology to non-Abelian topological phases, focusing on the generalized Moore-Read (MR) fractional quantum Hall states at filling fractions $ν=1/n$. We consider interfaces between different MR states, write down explicit gap** interactions, which we motivate using an anyon condensation picture, and compute the entanglement entropy for an entanglement cut lying along the interface. Our work provides new insight towards understanding the connections between anyon condensation, gapped interfaces of non-Abelian phases, and TEE. △ Less

Submitted 7 July, 2020; v1 submitted 3 April, 2020; originally announced April 2020.

Comments: 34 pages; 2 figures. v2: updated references. v3: minor edits

Journal ref: Phys. Rev. B 102, 045102 (2020)

A Study of Context Dependencies in Multi-page Product Search

Authors: Ke** Bi, Choon Hui Teo, Yesh Dattatreya, Vijai Mohan, W. Bruce Croft

Abstract: In product search, users tend to browse results on multiple search result pages (SERPs) (e.g., for queries on clothing and shoes) before deciding which item to purchase. Users' clicks can be considered as implicit feedback which indicates their preferences and used to re-rank subsequent SERPs. Relevance feedback (RF) techniques are usually involved to deal with such scenarios. However, these metho… ▽ More In product search, users tend to browse results on multiple search result pages (SERPs) (e.g., for queries on clothing and shoes) before deciding which item to purchase. Users' clicks can be considered as implicit feedback which indicates their preferences and used to re-rank subsequent SERPs. Relevance feedback (RF) techniques are usually involved to deal with such scenarios. However, these methods are designed for document retrieval, where relevance is the most important criterion. In contrast, product search engines need to retrieve items that are not only relevant but also satisfactory in terms of customers' preferences. Personalization based on users' purchase history has been shown to be effective in product search. However, this method captures users' long-term interest, which does not always align with their short-term interest, and does not benefit customers with little or no purchase history. In this paper, we study RF techniques based on both long-term and short-term context dependencies in multi-page product search. We also propose an end-to-end context-aware embedding model which can capture both types of context. Our experimental results show that short-term context leads to much better performance compared with long-term and no context. Moreover, our proposed model is more effective than state-of-art word-based RF models. △ Less

Submitted 9 January, 2020; v1 submitted 9 September, 2019; originally announced September 2019.

Comments: Accepted by CIKM 2019. arXiv admin note: substantial text overlap with arXiv:1909.02065

Leverage Implicit Feedback for Context-aware Product Search

Authors: Ke** Bi, Choon Hui Teo, Yesh Dattatreya, Vijai Mohan, W. Bruce Croft

Abstract: Product search serves as an important entry point for online shop**. In contrast to web search, the retrieved results in product search not only need to be relevant but also should satisfy customers' preferences in order to elicit purchases. Previous work has shown the efficacy of purchase history in personalized product search. However, customers with little or no purchase history do not benefi… ▽ More Product search serves as an important entry point for online shop**. In contrast to web search, the retrieved results in product search not only need to be relevant but also should satisfy customers' preferences in order to elicit purchases. Previous work has shown the efficacy of purchase history in personalized product search. However, customers with little or no purchase history do not benefit from personalized product search. Furthermore, preferences extracted from a customer's purchase history are usually long-term and may not always align with her short-term interests. Hence, in this paper, we leverage clicks within a query session, as implicit feedback, to represent users' hidden intents, which further act as the basis for re-ranking subsequent result pages for the query. It has been studied extensively to model user preference with implicit feedback in recommendation tasks. However, there has been little research on modeling users' short-term interest in product search. We study whether short-term context could help promote users' ideal item in the following result pages for a query. Furthermore, we propose an end-to-end context-aware embedding model which can capture long-term and short-term context dependencies. Our experimental results on the datasets collected from the search log of a commercial product search engine show that short-term context leads to much better performance compared with long-term and no context. Our results also show that our proposed model is more effective than word-based context-aware models. △ Less

Submitted 9 January, 2020; v1 submitted 4 September, 2019; originally announced September 2019.

Comments: Presented at 2019 SIGIR Workshop on eCommerce (ECOM'19)

Magnetic Effects on Topological Chiral Channels in Bilayer Graphene

Authors: Aaron Winn, John Maier, Andrew Jiao, Jeffrey C. Y. Teo

Abstract: We study the effect of a magnetic field on topological chiral channels of bilayer graphene at electric domain walls. The persistence of chiral edge states is attributed to the difference in valley Chern number in the regions of opposite electric field. We explore the regime of large electric and magnetic fields perpendicular to the lattice. The magnetic field shifts the channel away from our elect… ▽ More We study the effect of a magnetic field on topological chiral channels of bilayer graphene at electric domain walls. The persistence of chiral edge states is attributed to the difference in valley Chern number in the regions of opposite electric field. We explore the regime of large electric and magnetic fields perpendicular to the lattice. The magnetic field shifts the channel away from our electric interface in a way that is inconsistent with the semiclassical expectation from the Lorentz force. Moreover, the magnetic field causes an imbalanced layer occupation preference to the chiral channels. These behaviors admit analytic solutions in the limits that either the electric or the magnetic field dominates. We numerically show in the general case that the system can be well-approximated as a weighted sum of the two limits. △ Less

Submitted 15 July, 2019; originally announced July 2019.

Comments: 12 pages, 23 figures

Semantic Product Search

Authors: Priyanka Nigam, Yiwei Song, Vijai Mohan, Vihan Lakshman, Weitian, Ding, Ankit Shingavi, Choon Hui Teo, Hao Gu, Bing Yin

Abstract: We study the problem of semantic matching in product search, that is, given a customer query, retrieve all semantically related products from the catalog. Pure lexical matching via an inverted index falls short in this respect due to several factors: a) lack of understanding of hypernyms, synonyms, and antonyms, b) fragility to morphological variants (e.g. "woman" vs. "women"), and c) sensitivity… ▽ More We study the problem of semantic matching in product search, that is, given a customer query, retrieve all semantically related products from the catalog. Pure lexical matching via an inverted index falls short in this respect due to several factors: a) lack of understanding of hypernyms, synonyms, and antonyms, b) fragility to morphological variants (e.g. "woman" vs. "women"), and c) sensitivity to spelling errors. To address these issues, we train a deep learning model for semantic matching using customer behavior data. Much of the recent work on large-scale semantic search using deep learning focuses on ranking for web search. In contrast, semantic matching for product search presents several novel challenges, which we elucidate in this paper. We address these challenges by a) develo** a new loss function that has an inbuilt threshold to differentiate between random negative examples, impressed but not purchased examples, and positive examples (purchased items), b) using average pooling in conjunction with n-grams to capture short-range linguistic patterns, c) using hashing to handle out of vocabulary tokens, and d) using a model parallel training architecture to scale across 8 GPUs. We present compelling offline results that demonstrate at least 4.7% improvement in Recall@100 and 14.5% improvement in mean average precision (MAP) over baseline state-of-the-art semantic search methods using the same tokenization method. Moreover, we present results and discuss learnings from online A/B tests which demonstrate the efficacy of our method. △ Less

Submitted 1 July, 2019; originally announced July 2019.

Comments: 10 pages, 7 figures, KDD 2019 (Applied Data Science Track)

A non-Abelian twist to integer quantum Hall states

Authors: Pedro L. S. Lopes, V. L. Quito, Bo Han, Jeffrey C. Y. Teo

Abstract: Through a theoretical coupled wire model, we construct strongly correlated electronic \emph{integer} quantum Hall states. As a distinguishing feature, these states support electric and thermal Hall transport violating the Wiedemann-Franz law as $\left(κ_{xy}/σ_{xy}\right)/\left[\left(π^{2}k_{B}^{2}T\right)/3e^{2}\right]<1$.We propose a new Abelian incompressible fluid at filling $ν=16$ that suppor… ▽ More Through a theoretical coupled wire model, we construct strongly correlated electronic \emph{integer} quantum Hall states. As a distinguishing feature, these states support electric and thermal Hall transport violating the Wiedemann-Franz law as $\left(κ_{xy}/σ_{xy}\right)/\left[\left(π^{2}k_{B}^{2}T\right)/3e^{2}\right]<1$.We propose a new Abelian incompressible fluid at filling $ν=16$ that supports a bosonic chiral $(E_{8})_{1}$ conformal field theory at the edge and is intimately related to topological paramagnets in (3+1)D. We further show that this topological phase can be partitioned into two non-Abelian quantum Hall states at filling $ν=8$, each carrying bosonic chiral $(G_{2})_{1}$ or $(F_{4})_{1}$ edge theories, and hosting Fibonacci anyonic excitations in the bulk. Finally, we discover a new notion of particle-hole conjugation based on the $E_{8}$ state that relates the $G_{2}$ and $F_{4}$ Fibonacci states. △ Less

Submitted 13 August, 2019; v1 submitted 25 January, 2019; originally announced January 2019.

Comments: 12 pages. 3 figures

Journal ref: Phys. Rev. B 100, 085116 (2019)

Paired parton quantum Hall states: a coupled wire construction

Authors: Alexander Sirota, Sharmistha Sahoo, Gil Young Cho, Jeffrey C. Y. Teo

Abstract: The Pfaffian fractional quantum Hall (FQH) states are incompressible non-Abelian topological fluids present in a half-filled electron Landau level, where there is a balanced population of electrons and holes. They give rise to half-integral quantum Hall plateaus that divide critical transitions between integer quantum Hall (IQH) states. On the other hand, there are Abelian FQH states, such as the… ▽ More The Pfaffian fractional quantum Hall (FQH) states are incompressible non-Abelian topological fluids present in a half-filled electron Landau level, where there is a balanced population of electrons and holes. They give rise to half-integral quantum Hall plateaus that divide critical transitions between integer quantum Hall (IQH) states. On the other hand, there are Abelian FQH states, such as the Laughlin state, that can be understood using partons, which are fermionic divisions of the electron. In this paper, we propose a new family of incompressible paired parton FQH states at filling $ν=1/6$ (modulo 1) that emerge from critical transitions between IQH states and Abelian FQH states at filling $ν=1/3$ (modulo 1). These paired parton states are originated from a half-filled parton Landau level, where there is an equal amount of partons and holes. They generically support Ising-like anyonic quasiparticle excitations and carry non-Abelian Pfaffian topological orders (TO) for partons. We prove the principle existence of these paired parton states using exactly solvable interacting arrays of electronic wires under a magnetic field. Moreover, we establish a new notion of particle-hole (PH) symmetry for partons and relate the PH symmetric parton Pfaffian TO with the gapped symmetric surface TO of a fractional topological insulator in three dimension. △ Less

Submitted 4 December, 2018; originally announced December 2018.

Comments: 34 pages, 12 figures

Journal ref: Phys. Rev. B 99, 245117 (2019)

Experimental 3D Coherent Diffractive Imaging from photon-sparse random projections

Authors: K. Giewekemeyer, A. Aquila, N. D. Loh, Y. Chushkin, K. S. Shanks, J. Weiss, M. W. Tate, H. T. Philipp, S. Stern, P. Vagovic, M. Mehrjoo, C. Teo, M. Barthelmess, F. Zontone, C. Chang, R. C. Tiberio, A. Sakdinawat, G. J. Williams, S. M. Gruner, A. P. Mancuso

Abstract: The routine atomic-resolution structure determination of single particles is expected to have profound implications for probing the structure-function relationship in systems ranging from energy materials to biological molecules. Extremely-bright, ultrashort-pulse X-ray sources---X-ray Free Electron Lasers (XFELs)---provide X-rays that can be used to probe ensembles of nearly identical nano-scale… ▽ More The routine atomic-resolution structure determination of single particles is expected to have profound implications for probing the structure-function relationship in systems ranging from energy materials to biological molecules. Extremely-bright, ultrashort-pulse X-ray sources---X-ray Free Electron Lasers (XFELs)---provide X-rays that can be used to probe ensembles of nearly identical nano-scale particles. When combined with coherent diffractive imaging, these objects can be imaged; however, as the resolution of the images approaches the atomic scale, the measured data are increasingly difficult to obtain and, during an X-ray pulse, the number of photons incident on the two-dimensional detector is much smaller than the number of pixels. This latter concern, the signal "sparsity," materially impedes the application of the method. We demonstrate an experimental analog using a synchrotron X-ray source that yields signal levels comparable to those expected from single biomolecules illuminated by focused XFEL pulses. The analog experiment provides an invaluable cross-check on the fidelity of the reconstructed data that is not available during XFEL experiments. We establish---using this experimental data---that a sparsity of order $1.3\times10^{-3}$ photons per pixel per frame can be overcome, lending vital insight to the solution of the atomic-resolution XFEL single particle imaging problem by experimentally demonstrating 3D coherent diffractive imaging from photon-sparse random projections. △ Less

Submitted 14 November, 2018; originally announced November 2018.

Comments: 17 pages, 12 figures

On Policies for Single-leg Revenue Management with Limited Demand Information

Authors: Will Ma, David Simchi-Levi, Chung-Piaw Teo

Abstract: In this paper we study the single-item revenue management problem, with no information given about the demand trajectory over time. When the item is sold through accepting/rejecting different fare classes, Ball and Queyranne (2009) have established the tight competitive ratio for this problem using booking limit policies, which raise the acceptance threshold as the remaining inventory dwindles. Ho… ▽ More In this paper we study the single-item revenue management problem, with no information given about the demand trajectory over time. When the item is sold through accepting/rejecting different fare classes, Ball and Queyranne (2009) have established the tight competitive ratio for this problem using booking limit policies, which raise the acceptance threshold as the remaining inventory dwindles. However, when the item is sold through dynamic pricing instead, there is the additional challenge that offering a low price may entice high-paying customers to substitute down. We show that despite this challenge, the same competitive ratio can still be achieved using a randomized dynamic pricing policy. Our policy incorporates the price-skimming technique from Eren and Maglaras (2010), but importantly we show how the randomized price distribution should be stochastically-increased as the remaining inventory dwindles. A key technical ingredient in our policy is a new "valuation tracking" subroutine, which tracks the possible values for the optimum, and follows the most "inventory-conservative" control which maintains the desired competitive ratio. Finally, we demonstrate the empirical effectiveness of our policy in simulations, where its average-case performance surpasses all naive modifications of the existing policies. △ Less

Submitted 17 January, 2020; v1 submitted 25 October, 2018; originally announced October 2018.

Adaptive, Personalized Diversity for Visual Discovery

Authors: Choon Hui Teo, Houssam Nassif, Daniel Hill, Sriram Srinavasan, Mitchell Goodman, Vijai Mohan, SVN Vishwanathan

Abstract: Search queries are appropriate when users have explicit intent, but they perform poorly when the intent is difficult to express or if the user is simply looking to be inspired. Visual browsing systems allow e-commerce platforms to address these scenarios while offering the user an engaging shop** experience. Here we explore extensions in the direction of adaptive personalization and item diversi… ▽ More Search queries are appropriate when users have explicit intent, but they perform poorly when the intent is difficult to express or if the user is simply looking to be inspired. Visual browsing systems allow e-commerce platforms to address these scenarios while offering the user an engaging shop** experience. Here we explore extensions in the direction of adaptive personalization and item diversification within Stream, a new form of visual browsing and discovery by Amazon. Our system presents the user with a diverse set of interesting items while adapting to user interactions. Our solution consists of three components (1) a Bayesian regression model for scoring the relevance of items while leveraging uncertainty, (2) a submodular diversification framework that re-ranks the top scoring items based on category, and (3) personalized category preferences learned from the user's behavior. When tested on live traffic, our algorithms show a strong lift in click-through-rate and session duration. △ Less

Submitted 2 October, 2018; originally announced October 2018.

Comments: Best Paper Award

Journal ref: Adaptive, Personalized Diversity for Visual Discovery. Teo CH, Nassif H, Hill D, Srinavasan S, Goodman M, Mohan V, and Vishwanathan SVN. ACM Conference on Recommender Systems (RecSys'16), Boston, pp. 35-38, 2016

A coupled wire description of surface $ADE$ topological orders

Authors: Bo Han, Jeffrey C. Y. Teo

Abstract: Symmetry-protected and symmetry-enriched topological (SPT/SET) phases in three dimensions are quantum systems that support non-trivial two-dimensional surface states. These surface states develop finite excitation energy gaps when the relevant symmetries are broken. On the other hand, one-dimensional gapless modes can populate along interfaces that separate adjacent gapped surface domains with dis… ▽ More Symmetry-protected and symmetry-enriched topological (SPT/SET) phases in three dimensions are quantum systems that support non-trivial two-dimensional surface states. These surface states develop finite excitation energy gaps when the relevant symmetries are broken. On the other hand, one-dimensional gapless modes can populate along interfaces that separate adjacent gapped surface domains with distinct symmetry-breaking orders. A surface strip pattern in general reduces the low-energy SPT/SET surface degrees of freedom onto a 2D array of gapless 1D channels. These channels can be coupled to one another by quasiparticle tunneling, and these inter-wire interactions collectively provide an effective description of the surface state. In this paper, we study a general class of symmetry-preserving or breaking SPT/SET surface states that admit finite excitation energy gaps and Abelian topological orders via the coupled wire construction. In particular, we focus on the prototype Abelian surface topological orders that fall under the $ADE$ classification of simply-laced Lie algebras. We also elaborate on the emergent symmetry and duality properties of the coupled wire models. △ Less

Submitted 3 June, 2019; v1 submitted 26 July, 2018; originally announced July 2018.

Comments: 28 pages, 6 figures. Published version

Journal ref: Phys. Rev. B 99, 235102 (2019)

Coupled Wire Models of Interacting Dirac Nodal Superconductors

Authors: Moon Jip Park, Syed Raza, Matthew J. Gilbert, Jeffrey C. Y. Teo

Abstract: Topological nodal superconductors possess gapless low energy excitations that are characterized by point or line nodal Fermi surfaces. In this work, using a coupled wire construction, we study topological nodal superconductors that have protected Dirac nodal points. In this construction, the low-energy electronic degrees of freedom are confined in a three dimensional array of wires, which emerge a… ▽ More Topological nodal superconductors possess gapless low energy excitations that are characterized by point or line nodal Fermi surfaces. In this work, using a coupled wire construction, we study topological nodal superconductors that have protected Dirac nodal points. In this construction, the low-energy electronic degrees of freedom are confined in a three dimensional array of wires, which emerge as pairing vortices of a microscopic superconducting system. The vortex array harbors an antiferromagnetic time-reversal and a mirror glide symmetry that protect the massless Dirac fermion in the single-body non-interacting limit. Within this model, we demonstrate exact-solvable many-body interactions that preserve the underlying symmetries and introduce a finite excitation energy gap. These gap** interactions support fractionalization and generically lead to non-trivial topological order. We also construct a special case of $N=16$ Dirac fermions where corresponding the gap** interaction leads to a trivial $E_8$ topological order that is closely related to the cancellation of the large gravitational anomaly. △ Less

Submitted 25 June, 2018; originally announced June 2018.

Comments: 25 pages, 9 figures

Journal ref: Phys. Rev. B 98, 184514 (2018)

From Dirac semimetals to topological phases in three dimensions: a coupled wire construction

Authors: Syed Raza, Alexander Sirota, Jeffrey C. Y. Teo

Abstract: Weyl and Dirac (semi)metals in three dimensions have robust gapless electronic band structures. Their massless single-body energy spectra are protected by symmetries such as lattice translation, (screw) rotation and time reversal. In this manuscript, we discuss many-body interactions in these systems. We focus on strong interactions that preserve symmetries and are outside the single-body mean-fie… ▽ More Weyl and Dirac (semi)metals in three dimensions have robust gapless electronic band structures. Their massless single-body energy spectra are protected by symmetries such as lattice translation, (screw) rotation and time reversal. In this manuscript, we discuss many-body interactions in these systems. We focus on strong interactions that preserve symmetries and are outside the single-body mean-field regime. By map** a Dirac (semi)metal to a model based on a three dimensional array of coupled Dirac wires, we show (1) the Dirac (semi)metal can acquire a many-body excitation energy gap without breaking the relevant symmetries, and (2) interaction can enable an anomalous Weyl (semi)metallic phase that is otherwise forbidden by symmetries in the single-body setting and can only be present holographically on the boundary of a four dimensional weak topological insulator. Both of these topological states support fractional gapped (gapless) bulk (resp. boundary) quasiparticle excitations. △ Less

Submitted 4 March, 2019; v1 submitted 15 November, 2017; originally announced November 2017.

Comments: 29 pages, 19 figures. This version has an expanded 'Summary of Results' and 'Conclusion and Discussion' section to make it more accessible to a broader audience

Journal ref: Phys. Rev. X 9, 011039 (2019)

Majorana stripe order on the surface of a three-dimensional topological insulator

Authors: Y. Kamiya, A. Furusaki, J. C. Y. Teo, G. -W. Chern

Abstract: The issue on the effect of interactions in topological states concerns not only interacting topological phases but also novel symmetry-breaking phases and phase transitions. Here we study the interaction effect on Majorana zero modes (MZMs) bound to a square vortex lattice in two-dimensional (2D) topological superconductors. Under the neutrality condition, where single-body hybridization between M… ▽ More The issue on the effect of interactions in topological states concerns not only interacting topological phases but also novel symmetry-breaking phases and phase transitions. Here we study the interaction effect on Majorana zero modes (MZMs) bound to a square vortex lattice in two-dimensional (2D) topological superconductors. Under the neutrality condition, where single-body hybridization between MZMs is prohibited by an emergent symmetry, a minimal square-lattice model for MZMs can be faithfully mapped to a quantum spin model, which has no sign problem in the world-line quantum Monte Carlo simulation. Guided by an insight from a further duality map**, we demonstrate that the interaction induces a Majorana stripe state, a gapped state spontaneously breaking lattice translational and rotational symmetries, as opposed to the previously conjectured topological quantum criticality. Away from neutrality, a mean-field theory suggests a quantum critical point induced by hybridization. △ Less

Submitted 28 October, 2018; v1 submitted 9 November, 2017; originally announced November 2017.

Comments: 4 pages, 5 figures + supplemental material

Journal ref: Phys. Rev. B 98, 161409 (2018)

Twofold twist defect chains at criticality

Authors: Xiongjie Yu, Xiao Chen, Abhishek Roy, Jeffrey C. Y. Teo

Abstract: The twofold twist defects in the $D(\mathbb{Z}_k)$ quantum double model (abelian topological phase) carry non-abelian fractional Majorana-like characteristics. We align these twist defects in a line and construct a one dimensional Hamiltonian which only includes the pairwise interaction. For the defect chain with even number of twist defects, it is equivalent to the $\mathbb{Z}_k$ clock model with… ▽ More The twofold twist defects in the $D(\mathbb{Z}_k)$ quantum double model (abelian topological phase) carry non-abelian fractional Majorana-like characteristics. We align these twist defects in a line and construct a one dimensional Hamiltonian which only includes the pairwise interaction. For the defect chain with even number of twist defects, it is equivalent to the $\mathbb{Z}_k$ clock model with periodic boundary condition (up to some phase factor for boundary term), while for odd number case, it maps to $\mathbb{Z}_k$ clock model with duality twisted boundary condition. At critical point, for both cases, the twist defect chain enjoys an additional translation symmetry, which corresponds to the Kramers-Wannier duality symmetry in the $\mathbb{Z}_k$ clock model and can be generated by a series of braiding operators for twist defects. We further numerically investigate the low energy excitation spectrum for $k=3,~4,~5$ and $6$. For even-defect chain, the critical points are the same as the $\mathbb{Z}_k$ clock conformal field theories (CFTs), while for odd-defect chain, when $k\neq 4$, the critical points correspond to orbifolding a $\mathbb{Z}_2$ symmetry of CFTs of the even-defect chain. For $k=4$ case, we numerically observe some similarity to the $\mathbb{Z}_4$ twist fields in $SU(2)_1/D_4$ orbifold CFT. △ Less

Submitted 16 September, 2017; originally announced September 2017.

Comments: 18 pages, 15 figures

Journal ref: Phys. Rev. B 96, 205435 (2017)

Optimal Photon Generation from Spontaneous Raman Processes in Cold Atoms

Authors: Melvyn Ho, Colin Teo, Hugues de Riedmatten, Nicolas Sangouard

Abstract: Spontaneous Raman processes in cold atoms have been widely used in the past decade for generating single photons. Here, we present a method to optimize their efficiencies for given atomic coherences and optical depths. We give a simple and complete recipe that can be used in present-day experiments, attaining near-optimal single photon emission while preserving the photon purity. Spontaneous Raman processes in cold atoms have been widely used in the past decade for generating single photons. Here, we present a method to optimize their efficiencies for given atomic coherences and optical depths. We give a simple and complete recipe that can be used in present-day experiments, attaining near-optimal single photon emission while preserving the photon purity. △ Less

Submitted 11 July, 2017; originally announced July 2017.

Comments: 6+6 pages, 3 figures, 1 table

Journal ref: New J. Phys., 20, 123018 (2018)

From orbifolding conformal field theories to gauging topological phases

Authors: Xiao Chen, Abhishek Roy, Jeffrey C. Y. Teo, Shinsei Ryu

Abstract: Topological phases of matter in (2+1) dimensions are commonly equipped with global symmetries, such as electric-magnetic duality in gauge theories and bilayer symmetry in fractional quantum Hall states. Gauging these symmetries into local dynamical ones is one way of obtaining exotic phases from conventional systems. We study this using the bulk-boundary correspondence and applying the orbifold co… ▽ More Topological phases of matter in (2+1) dimensions are commonly equipped with global symmetries, such as electric-magnetic duality in gauge theories and bilayer symmetry in fractional quantum Hall states. Gauging these symmetries into local dynamical ones is one way of obtaining exotic phases from conventional systems. We study this using the bulk-boundary correspondence and applying the orbifold construction to the (1+1) dimensional edge described by a conformal field theory (CFT). Our procedure puts twisted boundary conditions into the partition function, and predicts the fusion, spin and braiding behavior of anyonic excitations after gauging. We demonstrate this for the electric-magnetic self-dual $\mathbb{Z}_N$ gauge theory, the twofold symmetric $SU(3)_1$, and the $S_3$-symmetric $SO(8)_1$ Wess-Zumino-Witten theories. △ Less

Submitted 2 June, 2017; originally announced June 2017.

Comments: 23 pages, 1 figure

Journal ref: Phys. Rev. B 96, 115447 (2017)

Symmetric-Gapped Surface States of Fractional Topological Insulators

Authors: Gil Young Cho, Jeffrey C. Y. Teo, Eduardo Fradkin

Abstract: We construct the symmetric-gapped surface states of a fractional topological insulator with electromagnetic $θ$-angle $θ_{em} = \fracπ{3}$ and a discrete $\mathbb{Z}_3$ gauge field. They are the proper generalizations of the T-pfaffian state and pfaffian/anti-semion state and feature an extended periodicity compared with their of "integer" topological band insulators counterparts. We demonstrate t… ▽ More We construct the symmetric-gapped surface states of a fractional topological insulator with electromagnetic $θ$-angle $θ_{em} = \fracπ{3}$ and a discrete $\mathbb{Z}_3$ gauge field. They are the proper generalizations of the T-pfaffian state and pfaffian/anti-semion state and feature an extended periodicity compared with their of "integer" topological band insulators counterparts. We demonstrate that the surface states have the correct anomalies associated with time-reversal symmetry and charge conservation. △ Less

Submitted 1 June, 2017; originally announced June 2017.

Comments: 5 pages, 33 references and 2 pages of supplemental material

Journal ref: Phys. Rev. B 96, 195105 (2017)

Topological strings linking with quasi-particle exchange in superconducting Dirac semimetals

Authors: Pedro L. S. Lopes, Jeffrey C. Y. Teo, Shinsei Ryu

Abstract: We demonstrate a topological classification of vortices in three dimensional time-reversal invariant topological superconductors based on superconducting Dirac semimetals with an s-wave superconducting order parameter by means of a pair of numbers $(N_Φ,N)$, accounting how many units $N_Φ$ of magnetic fluxes $hc/4e$ and how many $N$ chiral Majorana modes the vortex carries. From these quantities,… ▽ More We demonstrate a topological classification of vortices in three dimensional time-reversal invariant topological superconductors based on superconducting Dirac semimetals with an s-wave superconducting order parameter by means of a pair of numbers $(N_Φ,N)$, accounting how many units $N_Φ$ of magnetic fluxes $hc/4e$ and how many $N$ chiral Majorana modes the vortex carries. From these quantities, we introduce a topological invariant which further classifies the properties of such vortices under linking processes. While such processes are known to be related to instanton processes in a field theoretic description, we demonstrate here that they are, in fact, also equivalent to the fractional Josephson effect on junctions based at the edges of quantum spin Hall systems. This allows one to consider microscopically the effects of interactions in the linking problem. We therefore demonstrate that associated to links between vortices, one has the exchange of quasi-particles, either Majorana zero-modes or $e/2$ quasi-particles, which allows for a topological classification of vortices in these systems, seen to be $\mathbb{Z}_8$ classified. While $N_Φ$ and $N$ are shown to be both even or odd in the weakly-interacting limit, in the strongly interacting scenario one loosens this constraint. In this case, one may have further fractionalization possibilities for the vortices, whose excitations are described by $SO(3)_3$-like conformal field theories with quasi-particle exchanges of more exotic types. △ Less

Submitted 19 June, 2017; v1 submitted 23 March, 2017; originally announced March 2017.

Comments: 25 pages, 9 figures

Journal ref: Phys. Rev. B 95, 235134 (2017)

Ultrasensitive Inertial and Force Sensors with Diamagnetically Levitated Magnets

Authors: J. Prat-Camps, C. Teo, C. C. Rusconi, W. Wieczorek, O. Romero-Isart

Abstract: We theoretically show that a magnet can be stably levitated on top of a punctured superconductor sheet in the Meissner state without applying any external field. The trap** potential created by such induced-only superconducting currents is characterized for magnetic spheres ranging from tens of nanometers to tens of millimeters. Such a diamagnetically levitated magnet is predicted to be extremel… ▽ More We theoretically show that a magnet can be stably levitated on top of a punctured superconductor sheet in the Meissner state without applying any external field. The trap** potential created by such induced-only superconducting currents is characterized for magnetic spheres ranging from tens of nanometers to tens of millimeters. Such a diamagnetically levitated magnet is predicted to be extremely well isolated from the environment. We therefore propose to use it as an ultrasensitive force and inertial sensor. A magnetomechanical read-out of its displacement can be performed by using superconducting quantum interference devices. An analysis using current technology shows that force and acceleration sensitivities on the order of $10^{-23}\text{N}/\sqrt{\text{Hz}}$ (for a 100 nm magnet) and $10^{-14}g/\sqrt{\text{Hz}}$ (for a 10 mm magnet) might be within reach in a cryogenic environment. Such unprecedented sensitivities can be used for a variety of purposes, from designing ultra-sensitive inertial sensors for technological applications (i.e. gravimetry, avionics, and space industry), to scientific investigations on measuring Casimir forces of magnetic origin and gravitational physics. △ Less

Submitted 1 March, 2017; originally announced March 2017.

Comments: 10 pages, 6 figures, article + supplemental material

Journal ref: Phys. Rev. Applied 8, 034002 (2017)

Surfaces and Slabs of Fractional Topological Insulator Heterostructures

Authors: Sharmistha Sahoo, Alexander Sirota, Gil Young Cho, Jeffrey C. Y. Teo

Abstract: Fractional topological insulators (FTI) are electronic topological phases in $(3+1)$ dimensions enriched by time reversal (TR) and charge $U(1)$ conservation symmetries. We focus on the simplest series of fermionic FTI, whose bulk quasiparticles consist of deconfined partons that carry fractional electric charges in integral units of $e^\ast=e/(2n+1)$ and couple to a discrete $\mathbb{Z}_{2n+1}$ g… ▽ More Fractional topological insulators (FTI) are electronic topological phases in $(3+1)$ dimensions enriched by time reversal (TR) and charge $U(1)$ conservation symmetries. We focus on the simplest series of fermionic FTI, whose bulk quasiparticles consist of deconfined partons that carry fractional electric charges in integral units of $e^\ast=e/(2n+1)$ and couple to a discrete $\mathbb{Z}_{2n+1}$ gauge theory. We propose massive symmetry preserving or breaking FTI surface states. Combining the long-ranged entangled bulk with these topological surface states, we deduce the novel topological order of quasi-$(2+1)$ dimensional FTI slabs as well as their corresponding edge conformal field theories. △ Less

Submitted 23 October, 2017; v1 submitted 30 January, 2017; originally announced January 2017.

Comments: 7 pages, 1 figure

Journal ref: Phys. Rev. B 96, 161108 (2017)

Minimal motor for powering particle motion from spin imbalance

Authors: Ulf Bissbort, Colin Teo, Chu Guo, Giulio Casati, Giuliano Benenti, Dario Poletti

Abstract: We introduce a minimalistic quantum motor for coupled energy and particle transport. The system is composed of two spins, each coupled to a different bath and to a particle which can move on a ring consisting of three sites. We show that the energy flowing from the baths to the system can be partially converted to perform work against an external driving, even in the presence of moderate dissipati… ▽ More We introduce a minimalistic quantum motor for coupled energy and particle transport. The system is composed of two spins, each coupled to a different bath and to a particle which can move on a ring consisting of three sites. We show that the energy flowing from the baths to the system can be partially converted to perform work against an external driving, even in the presence of moderate dissipation. We also analytically demonstrate the necessity of coupling between the spins. We suggest an experimental realization of our model using trapped ions or quantum dots. △ Less

Submitted 3 July, 2017; v1 submitted 9 September, 2016; originally announced September 2016.

Comments: 6 pages, 5 figures

Journal ref: Phys. Rev. E 95, 062143 (2017)

Converting heat into directed transport on a tilted lattice

Authors: Colin Teo, Ulf Bissbort, Dario Poletti

Abstract: We present a self-contained engine, which is made of one or more two-level systems, each of which is coupled to a single bath, as well as to a common load composed of a particle on a tilted lattice. We show that the energy and the entropy absorbed by the spins are transferred to the particle thus setting it into upward motion at an average constant speed, even when driven by a single spin connecte… ▽ More We present a self-contained engine, which is made of one or more two-level systems, each of which is coupled to a single bath, as well as to a common load composed of a particle on a tilted lattice. We show that the energy and the entropy absorbed by the spins are transferred to the particle thus setting it into upward motion at an average constant speed, even when driven by a single spin connected to a single bath. When considering an ensemble of different spins, the velocity of the particle is larger when the tilt is on resonance with any of the spins' energy splitting. Interestingly, we find regimes where the spins' polarization enters periodic cycles with the oscillation period being determined by the tilt of the lattice. △ Less

Submitted 8 September, 2016; originally announced September 2016.

Comments: 7 pages, 4 figures

Journal ref: Phys. Rev. E 95, 030102 (2017)

Fermion Parity Flips and Majorana Bound States at twist defects in Superconducting Fractional Topological Phases

Authors: Mayukh Nilay Khan, Jeffrey C. Y. Teo, Taylor L. Hughes, Smitha Vishveshwara

Abstract: In this paper we consider a layered heterostructure of an Abelian topologically ordered state (TO), such as a fractional Chern insulator/quantum Hall state with an s-wave superconductor in order to explore the existence of non-Abelian defects. In order to uncover such defects we must augment the original TO by a $\mathbb{Z}_2$ gauge theory sector coming from the s-wave SC. We first determine the e… ▽ More In this paper we consider a layered heterostructure of an Abelian topologically ordered state (TO), such as a fractional Chern insulator/quantum Hall state with an s-wave superconductor in order to explore the existence of non-Abelian defects. In order to uncover such defects we must augment the original TO by a $\mathbb{Z}_2$ gauge theory sector coming from the s-wave SC. We first determine the extended TO for a wide variety of fractional quantum Hall or fractional Chern insulator heterostructures. We prove the existence of a general anyon permutation symmetry (AS) that exists in any fermionic Abelian TO state in contact with an s-wave superconductor. Physically this permutation corresponds to adding a fermion to an odd flux vortices (in units of $h/2e$) as they travel around the associated topological (twist) defect. As such, we call it a fermion parity flip AS. We consider twist defects which mutate anyons according to the fermion parity flip symmetry and show that they can be realized at domain walls between distinct gapped edges or interfaces of the TO superconducting state. We analyze the properties of such defects and show that fermion parity flip twist defects are always associated with Majorana zero modes. Our formalism also reproduces known results such as Majorana/parafermionic bound states at superconducting domain walls of topological/Fractional Chern insulators when twist defects are constructed based on charge conjugation symmetry. Finally, we briefly describe more exotic twist liquid phases obtained by gauging the AS where the twist defects become deconfined anyonic excitations. △ Less

Submitted 14 March, 2016; originally announced March 2016.

Comments: 14+8 pages. 6 figures

Journal ref: Phys. Rev. B 95, 205112 (2017)

Generation of single photons with highly tunable wave shape from a cold atomic quantum memory

Authors: Pau Farrera, Georg Heinze, Boris Albrecht, Melvyn Ho, Matías Chávez, Colin Teo, Nicolas Sangouard, Hugues de Riedmatten

Abstract: We report on a single photon source with highly tunable photon shape based on a cold ensemble of Rubidium atoms. We follow the DLCZ scheme to implement an emissive quantum memory, which can be operated as a photon pair source with controllable delay. We find that the temporal wave shape of the emitted read photon can be precisely controlled by changing the shape of the driving read pulse. We gener… ▽ More We report on a single photon source with highly tunable photon shape based on a cold ensemble of Rubidium atoms. We follow the DLCZ scheme to implement an emissive quantum memory, which can be operated as a photon pair source with controllable delay. We find that the temporal wave shape of the emitted read photon can be precisely controlled by changing the shape of the driving read pulse. We generate photons with temporal durations varying over three orders of magnitude up to 10 μs without a significant change of the read-out efficiency. We prove the non-classicality of the emitted photons by measuring their antibunching, showing near single photon behavior at low excitation probabilities. We also show that the photons are emitted in a pure state by measuring unconditional autocorrelation functions. Finally, to demonstrate the usability of the source for realistic applications, we create ultra-long single photons with a rising exponential or doubly peaked wave shape which are important for several quantum information tasks. △ Less

Submitted 26 January, 2016; originally announced January 2016.

Comments: 10 pages, 7 figures

Journal ref: Nat. Commun. 7, 13556 (2016)

Globally symmetric topological phase: from anyonic symmetry to twist defect

Authors: Jeffrey C. Y. Teo

Abstract: Topological phases in two dimensions support anyonic quasiparticle excitations that obey neither bosonic nor fermionic statistics. These anyon structures often carry global symmetries that relate distinct anyons with similar fusion and statistical properties. Anyonic symmetries associate topological defects or fluxes in topological phases. As the symmetries are global and static, these extrinsic d… ▽ More Topological phases in two dimensions support anyonic quasiparticle excitations that obey neither bosonic nor fermionic statistics. These anyon structures often carry global symmetries that relate distinct anyons with similar fusion and statistical properties. Anyonic symmetries associate topological defects or fluxes in topological phases. As the symmetries are global and static, these extrinsic defects are semiclassical objects that behave disparately from conventional quantum anyons. Remarkably, even when the topological states supporting them are Abelian, they are generically non-Abelian and powerful enough for topological quantum computation. In this article, I review the most recent theoretical developments on symmetries and defects in topological phases. △ Less

Submitted 3 November, 2015; originally announced November 2015.

Comments: 72 pages, 26 figures. arXiv admin note: text overlap with arXiv:1503.06812

Journal ref: Journal of Physics: Condensed Matter, 28, 143001 (2016)

Coupled wire model of symmetric Majorana surfaces of topological superconductors

Authors: Sharmistha Sahoo, Zhao Zhang, Jeffrey C. Y. Teo

Abstract: Time reversal symmetric topological superconductors in three spatial dimensions carry gapless surface Majorana fermions. They are robust against any time reversal symmetric single-body perturbation weaker than the bulk energy gap. We mimic the massless surface Majorana's by coupled wire models in two spatial dimensions. We introduce explicit many-body interwire interactions that preserve time reve… ▽ More Time reversal symmetric topological superconductors in three spatial dimensions carry gapless surface Majorana fermions. They are robust against any time reversal symmetric single-body perturbation weaker than the bulk energy gap. We mimic the massless surface Majorana's by coupled wire models in two spatial dimensions. We introduce explicit many-body interwire interactions that preserve time reversal symmetry and give energy gaps to all low energy degrees of freedom. We show the gapped models generically carry non-trivial topological order and support anyonic excitations. △ Less

Submitted 23 September, 2015; originally announced September 2015.

Comments: 31 pages, 8 figures

Journal ref: Phys. Rev. B 94, 165142 (2016)

Topological Phases on Non-orientable Surfaces: Twisting by Parity Symmetry

Authors: AtMa P. O. Chan, Jeffrey C. Y. Teo, Shinsei Ryu

Abstract: We discuss (2+1)D topological phases on non-orientable spatial surfaces, such as Möbius strip, real projective plane and Klein bottle, etc., which are obtained by twisting the parent topological phases by their underlying pairty symmetries through introducing parity defects. We construct the ground states on arbitrary non-orientable closed manifolds and calculate the ground state degeneracy. Such… ▽ More We discuss (2+1)D topological phases on non-orientable spatial surfaces, such as Möbius strip, real projective plane and Klein bottle, etc., which are obtained by twisting the parent topological phases by their underlying pairty symmetries through introducing parity defects. We construct the ground states on arbitrary non-orientable closed manifolds and calculate the ground state degeneracy. Such degeneracy is shown to be robust against continuous deformation of the underlying manifold. We also study the action of the map** class group on the multiplet of ground states on the Klein bottle. The physical properties of the topological states on non-orientable surfaces are deeply related to the parity symmetric anyons which do not have a notion of orientation in their statistics. The number of ground states on the projective plane equals the root of the number of distinguishable parity symmetric anyons, while the ground state degeneracy on the Klein bottle equals the total number of parity symmetric anyons; In deforming the Klein bottle, the Dehn twist encodes the topological spins whereas the Y-homeomorphism tells the particle-hole relation of the parity symmetric anyons. △ Less

Submitted 10 March, 2016; v1 submitted 13 September, 2015; originally announced September 2015.

Comments: 19 pages, 10 figures

Journal ref: New J. Phys. 18, 035005 (2016)

Classification of topological quantum matter with symmetries

Authors: Ching-Kai Chiu, Jeffrey C. Y. Teo, Andreas P. Schnyder, Shinsei Ryu

Abstract: Topological materials have become the focus of intense research in recent years, since they exhibit fundamentally new physical phenomena with potential applications for novel devices and quantum information technology. One of the hallmarks of topological materials is the existence of protected gapless surface states, which arise due to a nontrivial topology of the bulk wave functions. This review… ▽ More Topological materials have become the focus of intense research in recent years, since they exhibit fundamentally new physical phenomena with potential applications for novel devices and quantum information technology. One of the hallmarks of topological materials is the existence of protected gapless surface states, which arise due to a nontrivial topology of the bulk wave functions. This review provides a pedagogical introduction into the field of topological quantum matter with an emphasis on classification schemes. We consider both fully gapped and gapless topological materials and their classification in terms of nonspatial symmetries, such as time-reversal, as well as spatial symmetries, such as reflection. Furthermore, we survey the classification of gapless modes localized on topological defects. The classification of these systems is discussed by use of homotopy groups, Clifford algebras, K-theory, and non-linear sigma models describing the Anderson (de-)localization at the surface or inside a defect of the material. Theoretical model systems and their topological invariants are reviewed together with recent experimental results in order to provide a unified and comprehensive perspective of the field. While the bulk of this article is concerned with the topological properties of noninteracting or mean-field Hamiltonians, we also provide a brief overview of recent results and open questions concerning the topological classifications of interacting systems. △ Less

Submitted 14 April, 2016; v1 submitted 13 May, 2015; originally announced May 2015.

Comments: 10 tables and 14 figures. Accepted for publication in Rev. Mod. Phys

Journal ref: Rev. Mod. Phys. 88, 035005 (2016)