-
The Task-oriented Queries Benchmark (ToQB)
Authors:
Keun Soo Yim
Abstract:
Task-oriented queries (e.g., one-shot queries to play videos, order food, or call a taxi) are crucial for assessing the quality of virtual assistants, chatbots, and other large language model (LLM)-based services. However, a standard benchmark for task-oriented queries is not yet available, as existing benchmarks in the relevant NLP (Natural Language Processing) fields have primarily focused on ta…
▽ More
Task-oriented queries (e.g., one-shot queries to play videos, order food, or call a taxi) are crucial for assessing the quality of virtual assistants, chatbots, and other large language model (LLM)-based services. However, a standard benchmark for task-oriented queries is not yet available, as existing benchmarks in the relevant NLP (Natural Language Processing) fields have primarily focused on task-oriented dialogues. Thus, we present a new methodology for efficiently generating the Task-oriented Queries Benchmark (ToQB) using existing task-oriented dialogue datasets and an LLM service. Our methodology involves formulating the underlying NLP task to summarize the original intent of a speaker in each dialogue, detailing the key steps to perform the devised NLP task using an LLM service, and outlining a framework for automating a major part of the benchmark generation process. Through a case study encompassing three domains (i.e., two single-task domains and one multi-task domain), we demonstrate how to customize the LLM prompts (e.g., omitting system utterances or speaker labels) for those three domains and characterize the generated task-oriented queries. The generated ToQB dataset is made available to the public. We further discuss new domains that can be added to ToQB by community contributors and its practical applications.
△ Less
Submitted 5 June, 2024;
originally announced June 2024.
-
Predicting Likely-Vulnerable Code Changes: Machine Learning-based Vulnerability Protections for Android Open Source Project
Authors:
Keun Soo Yim
Abstract:
This paper presents a framework that selectively triggers security reviews for incoming source code changes. Functioning as a review bot within a code review service, the framework can automatically request additional security reviews at pre-submit time before the code changes are submitted to a source code repository. Because performing such secure code reviews add cost, the framework employs a c…
▽ More
This paper presents a framework that selectively triggers security reviews for incoming source code changes. Functioning as a review bot within a code review service, the framework can automatically request additional security reviews at pre-submit time before the code changes are submitted to a source code repository. Because performing such secure code reviews add cost, the framework employs a classifier trained to identify code changes with a high likelihood of vulnerabilities. The online classifier leverages various types of input features to analyze the review patterns, track the software engineering process, and mine specific text patterns within given code changes. The classifier and its features are meticulously chosen and optimized using data from the submitted code changes and reported vulnerabilities in Android Open Source Project (AOSP). The evaluation results demonstrate that our Vulnerability Prevention (VP) framework identifies approximately 80% of the vulnerability-inducing code changes in the dataset with a precision ratio of around 98% and a false positive rate of around 1.7%. We discuss the implications of deploying the VP framework in multi-project settings and future directions for Android security research. This paper explores and validates our approach to code change-granularity vulnerability prediction, offering a preventive technique for software security by preemptively detecting vulnerable code changes before submission.
△ Less
Submitted 26 May, 2024;
originally announced May 2024.
-
Multitask Extension of Geometrically Aligned Transfer Encoder
Authors:
Sung Moon Ko,
Sumin Lee,
Dae-Woong Jeong,
Hyunseung Kim,
Chanhui Lee,
Soorin Yim,
Sehui Han
Abstract:
Molecular datasets often suffer from a lack of data. It is well-known that gathering data is difficult due to the complexity of experimentation or simulation involved. Here, we leverage mutual information across different tasks in molecular data to address this issue. We extend an algorithm that utilizes the geometric characteristics of the encoding space, known as the Geometrically Aligned Transf…
▽ More
Molecular datasets often suffer from a lack of data. It is well-known that gathering data is difficult due to the complexity of experimentation or simulation involved. Here, we leverage mutual information across different tasks in molecular data to address this issue. We extend an algorithm that utilizes the geometric characteristics of the encoding space, known as the Geometrically Aligned Transfer Encoder (GATE), to a multi-task setup. Thus, we connect multiple molecular tasks by aligning the curved coordinates onto locally flat coordinates, ensuring the flow of information from source tasks to support performance on target data.
△ Less
Submitted 3 May, 2024;
originally announced May 2024.
-
Laser mode-hop** assisted all-optical single beam pulsed atomic magnetometer
Authors:
Ji Hoon Yoon,
Sang Hyuk Hong,
Taek Jeong,
Sin Hyuk Yim,
Kyu Min Shim,
Sangkyung Lee
Abstract:
We demonstrate an all-optical single beam pulsed atomic magnetometer assisted by laser mode-hop** in a distributed Bragg reflector (DBR) laser. We implement a temporal sequence of the laser current; sinusoidal current modulation including the laser mode-hop current for synchronous optical pum** and the following constant current for paramagnetic Faraday rotation measurements to probe the free…
▽ More
We demonstrate an all-optical single beam pulsed atomic magnetometer assisted by laser mode-hop** in a distributed Bragg reflector (DBR) laser. We implement a temporal sequence of the laser current; sinusoidal current modulation including the laser mode-hop current for synchronous optical pum** and the following constant current for paramagnetic Faraday rotation measurements to probe the free induction decay (FID) of transverse $^{87}$Rb spin polarization. Repetitive sudden frequency shifts of 20 GHz around the pressure-broadened $^{87}$Rb spectra originating from laser mode-hop** enables discontinuous optical pum** modulation with a large depth which enhances transverse spin polarization. We achieve a sensitivity of 3.77 pT/Hz$^{1/2}$ in a magnetic field of 14 $μ$T, limited by the performance of the frequency counter. The Cramer-Rao lower bound (CRLB) of the sensitivity due to the non-magnetic noise such as photon shot-noise is 191 fT/Hz$^{1/2}$. Our approach based on laser mode-hop** can be applied to miniaturization of all-optical atomic magnetometers with sub-pT/Hz$^{1/2}$ sensitivities.
△ Less
Submitted 2 April, 2024;
originally announced April 2024.
-
An in silico drug repurposing pipeline to identify drugs with the potential to inhibit SARS-CoV-2 replication
Authors:
Méabh MacMahon,
Woochang Hwang,
Soorin Yim,
Eoghan MacMahon,
Alexandre Abraham,
Justin Barton,
Mukunthan Tharmakulasingam,
Paul Bilokon,
Vasanthi Priyadarshini Gaddi,
Namshik Han
Abstract:
Drug repurposing provides an opportunity to redeploy drugs, which ideally are already approved for use in humans, for the treatment of other diseases. For example, the repurposing of dexamethasone and baricitinib has played a crucial role in saving patient lives during the ongoing SARS-CoV-2 pandemic. There remains a need to expand therapeutic approaches to prevent life-threatening complications i…
▽ More
Drug repurposing provides an opportunity to redeploy drugs, which ideally are already approved for use in humans, for the treatment of other diseases. For example, the repurposing of dexamethasone and baricitinib has played a crucial role in saving patient lives during the ongoing SARS-CoV-2 pandemic. There remains a need to expand therapeutic approaches to prevent life-threatening complications in patients with COVID-19. Using an in silico approach based on structural similarity to drugs already in clinical trials for COVID-19, potential drugs were predicted for repurposing. For a subset of identified drugs with different targets to their corresponding COVID-19 clinical trial drug, a mechanism of action analysis was applied to establish whether they might have a role in inhibiting the replication of SARS-CoV-2. Of sixty drugs predicted in this study, two with the potential to inhibit SARS-CoV-2 replication were identified using mechanism of action analysis. Triamcinolone is a corticosteroid that is structurally similar to dexamethasone; gallopamil is a calcium channel blocker that is structurally similar to verapamil. In silico approaches indicate possible mechanisms of action for both drugs in inhibiting SARS-CoV-2 replication. The identification of these drugs as potentially useful for patients with COVID-19 who are at a higher risk of develo** severe disease supports the use of in silico approaches to facilitate quick and cost-effective drug repurposing. Such drugs could expand the number of treatments available to patients who are not protected by vaccination.
△ Less
Submitted 23 November, 2022; v1 submitted 5 July, 2021;
originally announced July 2021.
-
Coherent multi-mode conversion from microwave to optical wave via a magnon-cavity hybrid system
Authors:
Yong Sup Ihn,
Su-Yong Lee,
Dongkyu Kim,
Sin Hyuk Yim,
Zaeill Kim
Abstract:
Coherent conversion from microwave to optical wave opens new research avenues towards long distant quantum network covering quantum communication, computing, and sensing out of the laboratory. Especially multi-mode enabled system is essential for practical applications. Here we experimentally demonstrate coherent multi-mode conversion from the microwave to optical wave via collective spin excitati…
▽ More
Coherent conversion from microwave to optical wave opens new research avenues towards long distant quantum network covering quantum communication, computing, and sensing out of the laboratory. Especially multi-mode enabled system is essential for practical applications. Here we experimentally demonstrate coherent multi-mode conversion from the microwave to optical wave via collective spin excitation in a single crystal yttrium iron garnet (YIG, Y3Fe5O12) which is strongly coupled to a microwave cavity mode in a three-dimensional rectangular cavity. Expanding collective spin excitation mode of our magnon-cavity hybrid system from Kittel to multi magnetostatic modes, we verify that the size of YIG sphere predominantly plays a crucial role for the microwave-to-optical multi-mode conversion efficiency at resonant conditions. We also find that the coupling strength between multi magnetostatic modes and a cavity mode is manipulated by the position of a YIG inside the cavity. It is expected to be valuable for designing a magnon hybrid system that can be used for coherent conversion between microwave and optical photons.
△ Less
Submitted 30 July, 2020;
originally announced July 2020.
-
Multi-colour optical monitoring of eight red blazars
Authors:
Minfeng Gu,
C. -U. Lee,
Soojong Pak,
H. S. Yim,
A. B. Fletcher
Abstract:
We present the observational results of multi-colour optical monitoring of eight red blazars from 2003 September to 2004 February. The aim of our monitoring is to investigate the spectral variability as well as the flux variations at short and long time scales. The observations were carried out using the 1.0 m robotic telescope of Mt. Lemmon Optical Astronomy Observatory, in Arizona, USA, the 0.…
▽ More
We present the observational results of multi-colour optical monitoring of eight red blazars from 2003 September to 2004 February. The aim of our monitoring is to investigate the spectral variability as well as the flux variations at short and long time scales. The observations were carried out using the 1.0 m robotic telescope of Mt. Lemmon Optical Astronomy Observatory, in Arizona, USA, the 0.6 m telescope of Sobaeksan Optical Astronomy Observatory and the 1.8 m telescope of Bohyunsan Optical Astronomy Observatory, in the Republic of Korea. During the observations, all sources show strong flux variations with amplitudes of larger than 0.5 mag. Variations with amplitudes of over 1 mag are found in four sources. Intraday variations with amplitudes larger than 0.15 mag, and a rapid brightness increase with a rate of ~0.2 mag per day in four days, are detected in S5 0716+71. We investigate the relationship between the colour index and source brightness for each source. We find that two out of three FSRQs tend to be redder when they are brighter, and, conversely, all BL Lac objects tend to be bluer. In particular, we find a significant anti-correlation between the V-I colour index and R magnitude for 3C 454.3. This implies that the spectrum became steeper when the source was brighter, which is opposite to the common trend for blazars. In contrast, significant positive correlations are found in 3C 66A, S5 0716+71, and BL Lac. However, there are only very weak correlations for PKS 0735+17 and OJ 287. We propose that the different relative contributions of the thermal versus non-thermal radiation to the optical emission may be responsible for the different trends of the colour index with brightness in FSRQs and BL Lac objects.
△ Less
Submitted 8 February, 2006;
originally announced February 2006.
