arXiv subjects

Jeff Huang

Publications and source records attributed to Jeff Huang.

26 records · Page 2Linked to original sources

arXiv2024

A Nançay Radio Telescope study of the hyperactive repeating FRB 20220912A

The repeating fast radio burst source FRB 20220912A was remarkably active in the weeks after its discovery. Here we report 696 bursts detected with the Nançay Radio Telescope (NRT) as part of the Extragalactic Coherent Light from Astrophysical Transients (ÉCLAT) monitoring campaign. We present 68 observations, conducted from October 2022 to April 2023, with a total duration of 61 hours and an event rate peaking at $75^{+10}_{-9}$ bursts per hour above a fluence threshold of 0.59 Jy ms in the $1.2-1.7$-GHz band. Most bursts in the sample occur towards the bottom of the observing band. They follow a bimodal wait-time distribution, with peaks at 33.4 ms and 67.0 s. We find a roughly constant dispersion measure (DM) over time ($δ$DM $\lesssim$ 2 pc cm$^{-3}$) when taking into account `sad-trombone' drift, with a mean drift rate of $-8.8 $MHz ms$^{-1}$. Nonetheless, we confirm small $\sim0.3$ pc cm$^{-3}$ DM variations using microshot structure, while finding that microstructure is rare in our sample -- despite the 16 $μ$s time resolution of the data. The cumulative spectral energy distribution shows more high-energy bursts ($E_ν\gtrsim 10^{31}$ erg/Hz) than would be expected from a simple power-law distribution. The burst rate per observation appears Poissonian, but the full set of observations is better modelled by a Weibull distribution, showing clustering. We discuss the various observational similarities that FRB 20220912A shares with other (hyper)active repeaters, which as a group are beginning to show a common set of phenomenological traits that provide multiple useful dimensions for their quantitative comparison and modelling.

astro-ph.HE↗

arXiv2024

SGLang: Efficient Execution of Structured Language Model Programs

Large language models (LLMs) are increasingly used for complex tasks that require multiple generation calls, advanced prompting techniques, control flow, and structured inputs/outputs. However, efficient systems are lacking for programming and executing these applications. We introduce SGLang, a system for efficient execution of complex language model programs. SGLang consists of a frontend language and a runtime. The frontend simplifies programming with primitives for generation and parallelism control. The runtime accelerates execution with novel optimizations like RadixAttention for KV cache reuse and compressed finite state machines for faster structured output decoding. Experiments show that SGLang achieves up to 6.4x higher throughput compared to state-of-the-art inference systems on various large language and multi-modal models on tasks including agent control, logical reasoning, few-shot learning benchmarks, JSON decoding, retrieval-augmented generation pipelines, and multi-turn chat. The code is publicly available at https://github.com/sgl-project/sglang

cs.AI↗

arXiv2024

Epigraphics: Message-Driven Infographics Authoring

The message a designer wants to convey plays a pivotal role in directing the design of an infographic, yet most authoring workflows start with creating the visualizations or graphics first without gauging whether they fit the message. To address this gap, we propose Epigraphics, a web-based authoring system that treats an "epigraph" as the first-class object, and uses it to guide infographic asset creation, editing, and syncing. The system uses the text-based message to recommend visualizations, graphics, data filters, color palettes, and animations. It further supports between-asset interactions and fine-tuning such as recoloring, highlighting, and animation syncing that enhance the aesthetic cohesiveness of the assets. A gallery and case studies show that our system can produce infographics inspired by existing popular ones, and a task-based usability study with 10 designers show that a text-sourced workflow can standardize content, empower users to think more about the big picture, and facilitate rapid prototyping.

cs.HC↗

arXiv2022

Using Machine Learning to Identify Extragalactic Globular Cluster Candidates from Ground-Based Photometric Surveys of M87

Globular clusters (GCs) have been at the heart of many longstanding questions in many sub-fields of astronomy and, as such, systematic identification of GCs in external galaxies has immense impacts. In this study, we take advantage of M87's well-studied GC system to implement supervised machine learning (ML) classification algorithms - specifically random forest and neural networks - to identify GCs from foreground stars and background galaxies using ground-based photometry from the Canada-France-Hawai'i Telescope (CFHT). We compare these two ML classification methods to studies of "human-selected" GCs and find that the best performing random forest model can reselect 61.2% $\pm$ 8.0% of GCs selected from HST data (ACSVCS) and the best performing neural network model reselects 95.0% $\pm$ 3.4%. When compared to human-classified GCs and contaminants selected from CFHT data - independent of our training data - the best performing random forest model can correctly classify 91.0% $\pm$ 1.2% and the best performing neural network model can correctly classify 57.3% $\pm$ 1.1%. ML methods in astronomy have been receiving much interest as Vera C. Rubin Observatory prepares for first light. The observables in this study are selected to be directly comparable to early Rubin Observatory data and the prospects for running ML algorithms on the upcoming dataset yields promising results.

astro-ph.GA↗

arXiv2021

Toward Taming the Overhead Monster for Data-Flow Integrity

Data-Flow Integrity (DFI) is a well-known approach to effectively detecting a wide range of software attacks. However, its real-world application has been quite limited so far because of the prohibitive performance overhead it incurs. Moreover, the overhead is enormously difficult to overcome without substantially lowering the DFI criterion. In this work, an analysis is performed to understand the main factors contributing to the overhead. Accordingly, a hardware-assisted parallel approach is proposed to tackle the overhead challenge. Simulations on SPEC CPU 2006 benchmark show that the proposed approach can completely enforce the DFI defined in the original seminal work while reducing performance overhead by 4x, on average.

cs.AR↗

arXiv2021

PYInfer: Deep Learning Semantic Type Inference for Python Variables

Python type inference is challenging in practice. Due to its dynamic properties and extensive dependencies on third-party libraries without type annotations, the performance of traditional static analysis techniques is limited. Although semantics in source code can help manifest intended usage for variables (thus help infer types), they are usually ignored by existing tools. In this paper, we propose PYInfer, an end-to-end learning-based type inference tool that automatically generates type annotations for Python variables. The key insight is that contextual code semantics is critical in inferring the type for a variable. For each use of a variable, we collect a few tokens within its contextual scope, and design a neural network to predict its type. One challenge is that it is difficult to collect a high-quality human-labeled training dataset for this purpose. To address this issue, we apply an existing static analyzer to generate the ground truth for variables in source code. Our main contribution is a novel approach to statically infer variable types effectively and efficiently. Formulating the type inference as a classification problem, we can handle user-defined types and predict type probabilities for each variable. Our model achieves 91.2% accuracy on classifying 11 basic types in Python and 81.2% accuracy on classifying 500 most common types. Our results substantially outperform the state-of-the-art type annotators. Moreover, PYInfer achieves 5.2X more code coverage and is 187X faster than a state-of-the-art learning-based tool. With similar time consumption, our model annotates 5X more variables than a state-of-the-art static analysis tool. Our model also outperforms a learning-based function-level annotator on annotating types for variables and function arguments. All our tools and datasets are publicly available to facilitate future research in this direction.

cs.SE↗

arXiv2020

Distributed Networked Real-time Learning

Many machine learning algorithms have been developed under the assumption that data sets are already available in batch form. Yet in many application domains data is only available sequentially overtime via compute nodes in different geographic locations. In this paper, we consider the problem of learning a model when streaming data cannot be transferred to a single location in a timely fashion. In such cases, a distributed architecture for learning relying on a network of interconnected "local" nodes is required. We propose a distributed scheme in which every local node implements stochastic gradient updates based upon a local data stream. To ensure robust estimation, a network regularization penalty is used to maintain a measure of cohesion in the ensemble of models. We show the ensemble average approximates a stationary point and characterize the degree to which individual models differ from the ensemble average. We compare the results with federated learning to conclude the proposed approach is more robust to heterogeneity in data streams (data rates and estimation quality). We illustrate the results with an application to image classification with a deep learning model based upon convolutional neural networks.

math.OC↗

arXiv2010

Intrinsic Reliability improvement in Biaxially Strained SiGe p-MOSFETs

In this letter we not only show improvement in the performance but also in the reliability of 30nm thick biaxially strained SiGe (20%Ge) channel on Si p-MOSFETs. Compared to Si channel, strained SiGe channel allows larger hole mobility (μh) in the transport direction and alleviates charge flow towards the gate oxide. μh enhancement by 40% in SiGe and 100% in Si-cap SiGe is observed compared to the Si hole universal mobility. A ~40% reduction in NBTI degradation, gate leakage and flicker noise (1/f) is observed which is attributed to a 4% increase in the hole-oxide barrier height (ϕ) in SiGe. Similar field acceleration factor (Γ) for threshold voltage shift (ΔVT) and increase in noise (ΔSVG) in Si and SiGe suggests identical degradation mechanisms.

cond-mat.mes-hall↗