Publications

This paper presents a production Semi-Supervised Learning (SSL) pipeline based on the student-teacher framework, which leverages millions of unlabeled examples to improve Natural Language Understanding (NLU) tasks. We investigate two questions related to the use of unlabeled data in the production SSL context: 1) how to select samples from a huge unlabeled data pool that are beneficial for SSL training,

Electromagnetic Interference (EMI) failure is a common occurrence in electronic devices. Failing to comply with FCC/CE requirements set by government agencies delays the product time to market. Besides following proper design guidelines of layout, grounding, shielding, filtering, etc., using simulation to predict EMI failures during early design stage will greatly save time and cost. In this paper, EMI

In this paper, a near field scanning based method is utilized to characterize wireless coexistence issues in design of a practical electronic device. This device supports multiple wireless communication radios. Based on near field, the radiation at the intermodulation frequency when two different radios both operate are evaluated. Reduction of scanned near field is proved to be an effective method to predict

Semantic segmentation is a challenging problem due to difficulties in modeling context in complex scenes and class confusions along boundaries. Most literature either focuses on context modeling or boundary refinement, which is less generalizable in open-world scenarios. In this work, we advocate a unified framework (UN-EPT) to segment objects by considering both context information and boundary artifacts

Adversarial attack on question answering systems over tabular data (TableQA) can help evaluate to what extent TableQA systems can understand natural language questions and reason with tables. However, generating natural language adversarial questions is difficult, because even a single character swap could lead to huge semantic difference in human perception. In this paper, we propose SAGE (Semantically

The authors enhance pretrained language models with Securities and Exchange Commission filings data to create better language representations for features used in a predictive model. Specifically, they train RoBERTa class models with additional financial regulatory text, which they denote as a class of RoBERTa-Fin models. Using different datasets, the authors assess whether there is material improvement

The Principle of Insufficient Reason (PIR) assigns equal probabilities to each alternative of a random experiment whenever there is no reason to prefer one over the other. The Maximum Entropy Principle (MaxEnt) generalizes PIR to the case where statistical information like expectations are given. It is known that both principles result in paradoxical probability updates for joint distributions of cause

Pre-trained language models (PLMs) such as BERT and GPT learn general text representations and encode extensive world knowledge; thus, they can efficiently and accurately adapt to various downstream tasks. In this work, we propose to leverage these powerful PLMs as recommender systems and use prompts to reformulate the session-based recommendation task to a multi-token cloze task. We evaluate the proposed

In this article, we explain why an interventional view of recommendation provides a rigorous framework for thinking about recommender systems—enabling new insights both at a technical level for evaluation and learning, as well as at a conceptual level when we reason about the future of recommender systems. In some respects, the view of recommender systems as autonomous systems that act through their recommendations

Time series data are often corrupted by outliers or other kinds of anomalies. Identifying the anomalous points can be a goal on its own (anomaly detection), or a means to improving performance of other time series tasks (e.g. forecasting). Recent deep-learning-based approaches to anomaly detection and forecasting commonly assume that the proportion of anomalies in the training data is small enough to ignore

Open-book question answering is a subset of question answering (QA) tasks where the system aims to find answers in a given set of documents (open-book) and common knowledge about a topic. This article proposes a solution for answering natural language questions from a corpus of Amazon Web Services (AWS) technical documents with no domain-specific labeled data (zero-shot). These questions have a yes–no–none

Quantum simulation is expected to be one of the key applications of future quantum computers. Product formulas, or Trotterization, are the oldest and, still today, an appealing method for quantum simulation. For an accurate product formula approximation in the spectral norm, the state-of-the-art gate complexity depends on the number of Hamiltonian terms and a certain 1-norm of its local terms. This work

The disjointness of a stabilizer code is a quantity used to constrain the level of the logical Clifford hierarchy attainable by transversal gates and constant-depth quantum circuits. We show that for any positive integer constant c, the problem of calculating the c-disjointness, or even approximating it to within a constant multiplicative factor, is NP-complete. We provide bounds on the disjointness for

Quantum many-body systems involving bosonic modes or gauge fields have infinite-dimensional local Hilbert spaces which must be truncated to perform simulations of real-time dynamics on classical or quantum computers. To analyze the truncation error, we develop methods for bounding the rate of growth of local quantum numbers such as the occupation number of a mode at a lattice site, or the electric field

Quantum codes typically rely on large numbers of degrees of freedom to achieve low error rates. However each additional degree of freedom introduces a new set of error mechanisms. Hence minimizing the degrees of freedom that a quantum code utilizes is helpful. One quantum error correction solution is to encode quantum information into one or more bosonic modes. We revisit rotation-invariant bosonic codes