id
stringlengths
33
34
title
stringlengths
6
218
authors
stringlengths
5
6.98k
categories
stringlengths
5
188
summary
stringlengths
34
3.26k
link_pdf
stringlengths
33
34
updated
timestamp[s]
published
timestamp[s]
embeddings
dict
http://arxiv.org/abs/2304.02328v1
Enhancing Multimodal Entity and Relation Extraction with Variational Information Bottleneck
Shiyao Cui, Jiangxia Cao, Xin Cong, Jiawei Sheng, Quangang Li, Tingwen Liu, Jinqiao Shi
cs.MM, cs.CL
This paper studies the multimodal named entity recognition (MNER) and multimodal relation extraction (MRE), which are important for multimedia social platform analysis. The core of MNER and MRE lies in incorporating evident visual information to enhance textual semantics, where two issues inherently demand investigatio...
http://arxiv.org/pdf/2304.02328v1
2023-04-05T09:32:25
2023-04-05T09:32:25
{ "title": [ -0.020141602, 0.009811401, -0.078430176, -0.06530762, 0.014465332, -0.01727295, -0.05026245, 0.010284424, 0.038360596, 0.031829834, -0.013084412, 0.016525269, 0.01159668, 0.00021541119, -0.050048828, 0.021209717, -0.020568848, 0.0562...
http://arxiv.org/abs/2304.02313v1
Personality-aware Human-centric Multimodal Reasoning: A New Task
Yaochen Zhu, Xiangqing Shen, Rui Xia
cs.CL
Multimodal reasoning, an area of artificial intelligence that aims at make inferences from multimodal signals such as vision, language and speech, has drawn more and more attention in recent years. People with different personalities may respond differently to the same situation. However, such individual personalities ...
http://arxiv.org/pdf/2304.02313v1
2023-04-05T09:09:10
2023-04-05T09:09:10
{ "title": [ -0.019363403, -0.00289917, -0.044952393, -0.053863525, -0.0014467239, -0.0061950684, -0.038116455, 0.009109497, 0.0016183853, 0.028671265, 0.010345459, 0.004722595, 0.0178833, -0.033355713, -0.019622803, 0.02281189, -0.009757996, 0.0...
http://arxiv.org/abs/2304.03158v1
CoT-MAE v2: Contextual Masked Auto-Encoder with Multi-view Modeling for Passage Retrieval
Xing Wu, Guangyuan Ma, Peng Wang, Meng Lin, Zijia Lin, Fuzheng Zhang, Songlin Hu
cs.CL, cs.AI, cs.IR, cs.LG
Growing techniques have been emerging to improve the performance of passage retrieval. As an effective representation bottleneck pretraining technique, the contextual masked auto-encoder utilizes contextual embedding to assist in the reconstruction of passages. However, it only uses a single auto-encoding pre-task for ...
http://arxiv.org/pdf/2304.03158v1
2023-04-05T08:00:38
2023-04-05T08:00:38
{ "title": [ -0.0116119385, 0.009033203, -0.04559326, -0.041870117, -0.0053100586, -0.014144897, -0.047332764, -0.019866943, 0.019134521, 0.06347656, -0.020370483, 0.0064582825, 0.04711914, -0.011550903, -0.033050537, 0.024597168, -0.0044288635, ...
http://arxiv.org/abs/2304.02247v2
Disentangling Structure and Style: Political Bias Detection in News by Inducing Document Hierarchy
Jiwoo Hong, Yejin Cho, Jaemin Jung, Jiyoung Han, James Thorne
cs.CL, cs.LG
We address an important gap in detecting political bias in news articles. Previous works that perform document classification can be influenced by the writing style of each news outlet, leading to overfitting and limited generalizability. Our approach overcomes this limitation by considering both the sentence-level sem...
http://arxiv.org/pdf/2304.02247v2
2023-10-27T11:35:04
2023-04-05T06:35:41
{ "title": [ -0.0010309219, 0.006290436, -0.041809082, -0.042419434, 0.010055542, -0.02041626, -0.0073432922, -0.03125, -0.015007019, 0.027328491, -0.034362793, 0.012458801, 0.003419876, 0.0075531006, -0.037902832, -0.020706177, -0.028823853, 0.0...
http://arxiv.org/abs/2304.02233v1
Ericson: An Interactive Open-Domain Conversational Search Agent
Zihao Wang, Ali Ahmadvand, Jason Choi, Payam Karisani, Eugene Agichtein
cs.CL, cs.AI, 68T50, 68T07
Open-domain conversational search (ODCS) aims to provide valuable, up-to-date information, while maintaining natural conversations to help users refine and ultimately answer information needs. However, creating an effective and robust ODCS agent is challenging. In this paper, we present a fully functional ODCS system, ...
http://arxiv.org/pdf/2304.02233v1
2023-04-05T05:28:31
2023-04-05T05:28:31
{ "title": [ -0.0007457733, -0.010902405, -0.026306152, -0.046661377, 0.00409317, -0.0048675537, -0.023925781, -0.016815186, 0.016571045, 0.038604736, -0.02835083, -0.0026245117, 0.0141067505, -0.010604858, -0.046691895, -0.0062179565, -0.032684326, ...
http://arxiv.org/abs/2304.02213v5
Large Language Models as Master Key: Unlocking the Secrets of Materials Science with GPT
Tong Xie, Yuwei Wan, Wei Huang, Yufei Zhou, Yixuan Liu, Qingyuan Linghu, Shaozhou Wang, Chunyu Kit, Clara Grazian, Wenjie Zhang, Bram Hoex
cs.CL, cs.AI
The amount of data has growing significance in exploring cutting-edge materials and a number of datasets have been generated either by hand or automated approaches. However, the materials science field struggles to effectively utilize the abundance of data, especially in applied disciplines where materials are evaluate...
http://arxiv.org/pdf/2304.02213v5
2023-04-12T14:06:02
2023-04-05T04:01:52
{ "title": [ -0.040649414, -0.02180481, -0.039611816, -0.044708252, -0.021575928, -0.00016796589, -0.06323242, -0.006965637, -0.0063476562, 0.01737976, 0.00983429, 0.0016326904, -0.0034980774, -0.04156494, -0.020126343, 0.023117065, -0.020599365, ...
http://arxiv.org/abs/2304.02210v2
Document-Level Machine Translation with Large Language Models
Longyue Wang, Chenyang Lyu, Tianbo Ji, Zhirui Zhang, Dian Yu, Shuming Shi, Zhaopeng Tu
cs.CL, cs.AI
Large language models (LLMs) such as ChatGPT can produce coherent, cohesive, relevant, and fluent answers for various natural language processing (NLP) tasks. Taking document-level machine translation (MT) as a testbed, this paper provides an in-depth evaluation of LLMs' ability on discourse modeling. The study focuses...
http://arxiv.org/pdf/2304.02210v2
2023-10-24T14:00:21
2023-04-05T03:49:06
{ "title": [ -0.04046631, -0.016326904, -0.01637268, -0.06463623, -0.0039787292, 0.021514893, -0.060302734, 0.004650116, 0.0158844, 0.043029785, -0.014251709, -0.0075912476, 0.031982422, -0.017440796, -0.04156494, 0.019927979, -0.0015058517, 0.02...
http://arxiv.org/abs/2304.02182v2
How to Design Translation Prompts for ChatGPT: An Empirical Study
Yuan Gao, Ruili Wang, Feng Hou
cs.CL
The recently released ChatGPT has demonstrated surprising abilities in natural language understanding and natural language generation. Machine translation relies heavily on the abilities of language understanding and generation. Thus, in this paper, we explore how to assist machine translation with ChatGPT. We adopt se...
http://arxiv.org/pdf/2304.02182v2
2023-04-21T09:35:44
2023-04-05T01:17:59
{ "title": [ -0.031921387, -0.029510498, 0.008430481, 0.013427734, -0.03463745, 0.010513306, -0.021499634, -0.030029297, -0.007972717, 0.020629883, -0.015838623, 0.029571533, -0.00045108795, -0.017288208, -0.037750244, 0.014923096, -0.006175995, ...
http://arxiv.org/abs/2304.02181v1
On the Impact of Voice Anonymization on Speech-Based COVID-19 Detection
Yi Zhu, Mohamed Imoussaïne-Aïkous, Carolyn Côté-Lussier, Tiago H. Falk
cs.CL, cs.SD, eess.AS
With advances seen in deep learning, voice-based applications are burgeoning, ranging from personal assistants, affective computing, to remote disease diagnostics. As the voice contains both linguistic and paralinguistic information (e.g., vocal pitch, intonation, speech rate, loudness), there is growing interest in vo...
http://arxiv.org/pdf/2304.02181v1
2023-04-05T01:09:58
2023-04-05T01:09:58
{ "title": [ 0.0048561096, 0.0047836304, -0.07244873, -0.099731445, 0.023620605, -0.02017212, -0.043395996, 0.022872925, -0.03982544, 0.03643799, -0.0027313232, 0.0015068054, 0.031463623, -0.0440979, 0.00066804886, -0.0047454834, -0.01260376, 0.0...
http://arxiv.org/abs/2304.02168v2
I2I: Initializing Adapters with Improvised Knowledge
Tejas Srinivasan, Furong Jia, Mohammad Rostami, Jesse Thomason
cs.CL, cs.AI, cs.CV, cs.LG
Adapters present a promising solution to the catastrophic forgetting problem in continual learning. However, training independent Adapter modules for every new task misses an opportunity for cross-task knowledge transfer. We propose Improvise to Initialize (I2I), a continual learning algorithm that initializes Adapters...
http://arxiv.org/pdf/2304.02168v2
2023-07-10T20:41:34
2023-04-04T23:51:48
{ "title": [ -0.05102539, -0.008621216, -0.046691895, -0.015899658, -0.030715942, 0.006385803, -0.03652954, -0.033172607, -0.005428314, 0.035186768, -0.02243042, -0.016418457, -0.0058250427, -0.029663086, -0.049316406, -0.00806427, 0.020431519, 0...
http://arxiv.org/abs/2304.03228v1
FedBot: Enhancing Privacy in Chatbots with Federated Learning
Addi Ait-Mlouk, Sadi Alawadi, Salman Toor, Andreas Hellander
cs.CL, cs.AI, cs.CR, cs.LG
Chatbots are mainly data-driven and usually based on utterances that might be sensitive. However, training deep learning models on shared data can violate user privacy. Such issues have commonly existed in chatbots since their inception. In the literature, there have been many approaches to deal with privacy, such as d...
http://arxiv.org/pdf/2304.03228v1
2023-04-04T23:13:52
2023-04-04T23:13:52
{ "title": [ -0.0054969788, -0.00749588, -0.04611206, -0.062805176, 0.02558899, -0.007156372, 0.0016183853, 0.020553589, -0.024246216, 0.035186768, -0.040893555, 0.0289917, 0.003786087, -0.006336212, -0.028167725, -0.013031006, -0.02067566, 0.039...
http://arxiv.org/abs/2304.02144v1
A Data Fusion Framework for Multi-Domain Morality Learning
Siyi Guo, Negar Mokhberian, Kristina Lerman
cs.CL
Language models can be trained to recognize the moral sentiment of text, creating new opportunities to study the role of morality in human life. As interest in language and morality has grown, several ground truth datasets with moral annotations have been released. However, these datasets vary in the method of data col...
http://arxiv.org/pdf/2304.02144v1
2023-04-04T22:05:02
2023-04-04T22:05:02
{ "title": [ 0.011917114, 0.024719238, -0.04940796, -0.030136108, 0.027435303, -0.051361084, 0.012145996, -0.018630981, -0.0024089813, 0.033355713, -0.017425537, -0.0030822754, -0.019454956, -0.016174316, -0.05404663, -0.017868042, -0.041748047, ...
http://arxiv.org/abs/2304.02138v3
Geotechnical Parrot Tales (GPT): Harnessing Large Language Models in geotechnical engineering
Krishna Kumar
cs.CL, physics.geo-ph, I.2.7; J.2.6
The widespread adoption of large language models (LLMs), such as OpenAI's ChatGPT, could revolutionize various industries, including geotechnical engineering. However, GPT models can sometimes generate plausible-sounding but false outputs, leading to hallucinations. In this article, we discuss the importance of prompt ...
http://arxiv.org/pdf/2304.02138v3
2023-06-21T15:55:24
2023-04-04T21:47:41
{ "title": [ -0.036987305, -0.029434204, -0.033294678, -0.037872314, -0.012275696, 0.0006184578, -0.04837036, 0.018417358, 0.0013446808, 0.024597168, -0.019119263, -0.005722046, 0.005722046, -0.033721924, -0.032470703, 0.005466461, -0.011070251, ...
http://arxiv.org/abs/2304.02080v1
Scalable and Accurate Self-supervised Multimodal Representation Learning without Aligned Video and Text Data
Vladislav Lialin, Stephen Rawls, David Chan, Shalini Ghosh, Anna Rumshisky, Wael Hamza
cs.CV, cs.CL
Scaling up weakly-supervised datasets has shown to be highly effective in the image-text domain and has contributed to most of the recent state-of-the-art computer vision and multimodal neural networks. However, existing large-scale video-text datasets and mining techniques suffer from several limitations, such as the ...
http://arxiv.org/pdf/2304.02080v1
2023-04-04T19:11:05
2023-04-04T19:11:05
{ "title": [ -0.011833191, 0.02394104, -0.06616211, -0.05407715, 0.020843506, -0.022277832, -0.039276123, 0.00020158291, -0.002313614, 0.029571533, -0.009407043, 0.012001038, -0.019683838, -0.018478394, -0.029571533, 0.0009279251, 0.0034236908, 0...
http://arxiv.org/abs/2304.02034v1
Effective Theory of Transformers at Initialization
Emily Dinan, Sho Yaida, Susan Zhang
cs.LG, cs.CL, hep-th, stat.ML
We perform an effective-theory analysis of forward-backward signal propagation in wide and deep Transformers, i.e., residual neural networks with multi-head self-attention blocks and multilayer perceptron blocks. This analysis suggests particular width scalings of initialization and training hyperparameters for these m...
http://arxiv.org/pdf/2304.02034v1
2023-04-04T18:00:01
2023-04-04T18:00:01
{ "title": [ 0.004337311, 0.033416748, -0.0021400452, -0.044525146, 0.030151367, 0.019973755, 0.0015907288, -0.039733887, -0.02784729, 0.05996704, 0.0017623901, 0.01499939, 0.004753113, -0.025268555, -0.024917603, 0.022262573, 0.00061655045, 0.05...
http://arxiv.org/abs/2304.01982v2
Rethinking the Role of Token Retrieval in Multi-Vector Retrieval
Jinhyuk Lee, Zhuyun Dai, Sai Meher Karthik Duddu, Tao Lei, Iftekhar Naim, Ming-Wei Chang, Vincent Y. Zhao
cs.CL, cs.IR
Multi-vector retrieval models such as ColBERT [Khattab and Zaharia, 2020] allow token-level interactions between queries and documents, and hence achieve state of the art on many information retrieval benchmarks. However, their non-linear scoring function cannot be scaled to millions of documents, necessitating a three...
http://arxiv.org/pdf/2304.01982v2
2023-05-23T17:13:52
2023-04-04T17:37:06
{ "title": [ 0.015129089, -0.00944519, -0.02659607, -0.05154419, 0.007575989, -0.01838684, -0.008781433, -0.021911621, 0.009506226, 0.022354126, -0.038269043, -0.0070343018, -0.0030956268, -0.011062622, -0.046295166, -0.006996155, -0.0107040405, ...
http://arxiv.org/abs/2304.01974v1
Dialogue-Contextualized Re-ranking for Medical History-Taking
Jian Zhu, Ilya Valmianski, Anitha Kannan
cs.CL, cs.IR
AI-driven medical history-taking is an important component in symptom checking, automated patient intake, triage, and other AI virtual care applications. As history-taking is extremely varied, machine learning models require a significant amount of data to train. To overcome this challenge, existing systems are develop...
http://arxiv.org/pdf/2304.01974v1
2023-04-04T17:31:32
2023-04-04T17:31:32
{ "title": [ 0.03778076, 0.005252838, -0.015129089, -0.09222412, 0.00466156, 0.0020618439, -0.037994385, -0.016479492, -0.0020523071, 0.033813477, -0.020629883, 0.049346924, -0.016647339, -0.012489319, -0.022750854, 0.011161804, -0.0038471222, -0...
http://arxiv.org/abs/2304.01969v2
MEGClass: Extremely Weakly Supervised Text Classification via Mutually-Enhancing Text Granularities
Priyanka Kargupta, Tanay Komarlu, Susik Yoon, Xuan Wang, Jiawei Han
cs.CL, cs.AI
Text classification is essential for organizing unstructured text. Traditional methods rely on human annotations or, more recently, a set of class seed words for supervision, which can be costly, particularly for specialized or emerging domains. To address this, using class surface names alone as extremely weak supervi...
http://arxiv.org/pdf/2304.01969v2
2023-10-29T21:03:54
2023-04-04T17:26:11
{ "title": [ 0.005996704, 0.0065689087, -0.0317688, -0.064208984, 0.015541077, -0.00024187565, -0.034423828, -0.03112793, 0.0048103333, 0.023910522, -0.04031372, 0.028915405, 0.0039482117, -0.0023155212, -0.042175293, -0.0076522827, -0.013214111, ...
http://arxiv.org/abs/2304.01961v1
AToMiC: An Image/Text Retrieval Test Collection to Support Multimedia Content Creation
Jheng-Hong Yang, Carlos Lassance, Rafael Sampaio de Rezende, Krishna Srinivasan, Miriam Redi, Stéphane Clinchant, Jimmy Lin
cs.IR, cs.CL, cs.CV
This paper presents the AToMiC (Authoring Tools for Multimedia Content) dataset, designed to advance research in image/text cross-modal retrieval. While vision-language pretrained transformers have led to significant improvements in retrieval effectiveness, existing research has relied on image-caption datasets that fe...
http://arxiv.org/pdf/2304.01961v1
2023-04-04T17:11:34
2023-04-04T17:11:34
{ "title": [ -0.005466461, 0.0013017654, -0.03552246, -0.03414917, 0.009063721, -0.0128479, -0.015525818, -0.016815186, 0.012237549, 0.047576904, -0.027862549, 0.010559082, 0.004550934, -0.0027065277, -0.06262207, -0.0018863678, 0.00013244152, 0....
http://arxiv.org/abs/2304.01933v3
LLM-Adapters: An Adapter Family for Parameter-Efficient Fine-Tuning of Large Language Models
Zhiqiang Hu, Lei Wang, Yihuai Lan, Wanyu Xu, Ee-Peng Lim, Lidong Bing, Xing Xu, Soujanya Poria, Roy Ka-Wei Lee
cs.CL
The success of large language models (LLMs), like GPT-4 and ChatGPT, has led to the development of numerous cost-effective and accessible alternatives that are created by finetuning open-access LLMs with task-specific data (e.g., ChatDoctor) or instruction data (e.g., Alpaca). Among the various fine-tuning methods, ada...
http://arxiv.org/pdf/2304.01933v3
2023-10-09T15:38:46
2023-04-04T16:31:37
{ "title": [ -0.033691406, -0.015281677, -0.04837036, -0.089538574, 0.0024909973, 0.008644104, -0.07354736, 0.019577026, 0.0008010864, 0.035461426, 0.0031852722, -0.004802704, 0.016830444, -0.022109985, -0.043182373, 0.024047852, -0.009742737, 0....
http://arxiv.org/abs/2304.01922v1
Resources and Few-shot Learners for In-context Learning in Slavic Languages
Michal Štefánik, Marek Kadlčík, Piotr Gramacki, Petr Sojka
cs.CL
Despite the rapid recent progress in creating accurate and compact in-context learners, most recent work focuses on in-context learning (ICL) for tasks in English. However, the ability to interact with users of languages outside English presents a great potential for broadening the applicability of language technologie...
http://arxiv.org/pdf/2304.01922v1
2023-04-04T16:16:25
2023-04-04T16:16:25
{ "title": [ 0.051635742, -0.015563965, -0.0072021484, -0.015220642, -0.015510559, -0.02670288, 0.019165039, -0.006542206, -0.05645752, 0.04144287, -0.04714966, 0.03387451, 0.008148193, -0.021255493, 0.012016296, -0.032714844, 0.007572174, 0.0170...
http://arxiv.org/abs/2304.01904v1
REFINER: Reasoning Feedback on Intermediate Representations
Debjit Paul, Mete Ismayilzada, Maxime Peyrard, Beatriz Borges, Antoine Bosselut, Robert West, Boi Faltings
cs.CL
Language models (LMs) have recently shown remarkable performance on reasoning tasks by explicitly generating intermediate inferences, e.g., chain-of-thought prompting. However, these intermediate inference steps may be inappropriate deductions from the initial context and lead to incorrect final predictions. Here we in...
http://arxiv.org/pdf/2304.01904v1
2023-04-04T15:57:28
2023-04-04T15:57:28
{ "title": [ -0.017364502, 0.0064926147, -0.030700684, -0.06976318, 0.001786232, 0.009780884, -0.05706787, -0.028213501, 0.027175903, 0.021514893, -0.053131104, -0.0025138855, -0.0024795532, -0.00932312, -0.037139893, -0.004295349, 0.0035305023, ...
http://arxiv.org/abs/2304.01894v1
San-BERT: Extractive Summarization for Sanskrit Documents using BERT and it's variants
Kartik Bhatnagar, Sampath Lonka, Jammi Kunal, Mahabala Rao M G
cs.CL
In this work, we develop language models for the Sanskrit language, namely Bidirectional Encoder Representations from Transformers (BERT) and its variants: A Lite BERT (ALBERT), and Robustly Optimized BERT (RoBERTa) using Devanagari Sanskrit text corpus. Then we extracted the features for the given text from these mode...
http://arxiv.org/pdf/2304.01894v1
2023-04-04T15:47:26
2023-04-04T15:47:26
{ "title": [ -0.025177002, -0.013900757, -0.018569946, -0.05581665, -0.035247803, 0.0103302, -0.06378174, -0.009605408, 0.005466461, 0.04244995, -0.036743164, 0.023529053, 0.0289917, -0.014427185, -0.04168701, 0.0067863464, -0.007118225, 0.026550...
http://arxiv.org/abs/2304.01890v4
Sociocultural knowledge is needed for selection of shots in hate speech detection tasks
Antonis Maronikolakis, Abdullatif Köksal, Hinrich Schütze
cs.CL, cs.AI, cs.LG
We introduce HATELEXICON, a lexicon of slurs and targets of hate speech for the countries of Brazil, Germany, India and Kenya, to aid training and interpretability of models. We demonstrate how our lexicon can be used to interpret model predictions, showing that models developed to classify extreme speech rely heavily ...
http://arxiv.org/pdf/2304.01890v4
2023-05-17T23:01:20
2023-04-04T15:42:08
{ "title": [ 0.0602417, -0.011940002, -0.0029850006, -0.008903503, -0.014373779, -0.056488037, 0.0368042, -0.013839722, -0.017593384, 0.014183044, -0.036590576, -0.0061302185, 0.011413574, -0.035247803, 0.016342163, -0.02909851, 0.022232056, 0.03...
http://arxiv.org/abs/2304.01852v4
Summary of ChatGPT-Related Research and Perspective Towards the Future of Large Language Models
Yiheng Liu, Tianle Han, Siyuan Ma, Jiayue Zhang, Yuanyuan Yang, Jiaming Tian, Hao He, Antong Li, Mengshen He, Zhengliang Liu, Zihao Wu, Lin Zhao, Dajiang Zhu, Xiang Li, Ning Qiang, Dingang Shen, Tianming Liu, Bao Ge
cs.CL
This paper presents a comprehensive survey of ChatGPT-related (GPT-3.5 and GPT-4) research, state-of-the-art large language models (LLM) from the GPT series, and their prospective applications across diverse domains. Indeed, key innovations such as large-scale pre-training that captures knowledge across the entire worl...
http://arxiv.org/pdf/2304.01852v4
2023-08-22T03:18:43
2023-04-04T15:01:06
{ "title": [ -0.019454956, -0.015594482, -0.02897644, -0.040374756, -0.019744873, -0.0066719055, -0.05029297, 0.010818481, -0.018463135, 0.045562744, -0.0057373047, 0.0059928894, 0.050689697, -0.030044556, 0.0045204163, 0.03488159, -0.01914978, 0...
http://arxiv.org/abs/2304.01774v1
A User-Centered, Interactive, Human-in-the-Loop Topic Modelling System
Zheng Fang, Lama Alqazlan, Du Liu, Yulan He, Rob Procter
cs.CL, cs.HC, cs.LG
Human-in-the-loop topic modelling incorporates users' knowledge into the modelling process, enabling them to refine the model iteratively. Recent research has demonstrated the value of user feedback, but there are still issues to consider, such as the difficulty in tracking changes, comparing different models and the l...
http://arxiv.org/pdf/2304.01774v1
2023-04-04T13:05:10
2023-04-04T13:05:10
{ "title": [ -0.002029419, 0.009399414, -0.026641846, -0.036315918, 0.00061416626, -0.012069702, -0.020858765, -0.039916992, 0.029510498, 0.03201294, -0.014274597, 0.0015058517, -0.013298035, -0.021026611, -0.054473877, 0.00086688995, -0.0007162094, ...
http://arxiv.org/abs/2304.01752v3
Black Box Few-Shot Adaptation for Vision-Language models
Yassine Ouali, Adrian Bulat, Brais Martinez, Georgios Tzimiropoulos
cs.CV, cs.CL, cs.LG
Vision-Language (V-L) models trained with contrastive learning to align the visual and language modalities have been shown to be strong few-shot learners. Soft prompt learning is the method of choice for few-shot downstream adaptation aiming to bridge the modality gap caused by the distribution shift induced by the new...
http://arxiv.org/pdf/2304.01752v3
2023-08-17T17:22:41
2023-04-04T12:42:29
{ "title": [ -0.025848389, -0.01096344, -0.035339355, -0.036865234, -0.01474762, -0.01374054, -0.03933716, -0.004501343, 0.021087646, 0.030426025, -0.032073975, -0.0032806396, -0.0047073364, -0.0040359497, -0.05291748, -0.0031757355, 0.02557373, ...
http://arxiv.org/abs/2304.01746v1
Is ChatGPT a Highly Fluent Grammatical Error Correction System? A Comprehensive Evaluation
Tao Fang, Shu Yang, Kaixin Lan, Derek F. Wong, Jinpeng Hu, Lidia S. Chao, Yue Zhang
cs.CL
ChatGPT, a large-scale language model based on the advanced GPT-3.5 architecture, has shown remarkable potential in various Natural Language Processing (NLP) tasks. However, there is currently a dearth of comprehensive study exploring its potential in the area of Grammatical Error Correction (GEC). To showcase its capa...
http://arxiv.org/pdf/2304.01746v1
2023-04-04T12:33:40
2023-04-04T12:33:40
{ "title": [ 0.0051498413, -0.0047454834, -0.011550903, -0.036712646, 0.0010728836, 0.0064964294, -0.014587402, -0.018707275, -0.014259338, 0.034301758, -0.017227173, 0.017028809, 0.002986908, -0.0072898865, -0.040130615, 0.014579773, -0.05493164, ...
http://arxiv.org/abs/2304.01712v1
Rumour Detection and Analysis on Twitter
Yaohou Fan
cs.CL, cs.AI
In recent years people have become increasingly reliant on social media to read news and get information, and some social media users post unsubstantiated information to gain attention. Such information is known as rumours. Nowadays, rumour detection is receiving a growing amount of attention because of the pandemic of...
http://arxiv.org/pdf/2304.01712v1
2023-04-04T11:20:37
2023-04-04T11:20:37
{ "title": [ 0.004600525, 0.0042648315, -0.020080566, -0.048736572, 0.005180359, 0.0070762634, -0.01348114, -0.019943237, -0.0036640167, 0.04650879, -0.03845215, 0.021560669, 0.012084961, -0.0004925728, -0.043029785, 0.00072956085, -0.024429321, ...
http://arxiv.org/abs/2304.01680v1
Can BERT eat RuCoLA? Topological Data Analysis to Explain
Irina Proskurina, Irina Piontkovskaya, Ekaterina Artemova
cs.CL
This paper investigates how Transformer language models (LMs) fine-tuned for acceptability classification capture linguistic features. Our approach uses the best practices of topological data analysis (TDA) in NLP: we construct directed attention graphs from attention matrices, derive topological features from them, an...
http://arxiv.org/pdf/2304.01680v1
2023-04-04T10:11:06
2023-04-04T10:11:06
{ "title": [ -0.013832092, -0.040008545, -0.03503418, -0.022232056, -0.054992676, -0.02519226, -0.023529053, 0.015403748, -0.0211792, 0.013015747, -0.01828003, -0.019226074, 0.04663086, -0.036712646, 0.011360168, -0.032318115, 0.021942139, 0.0422...
http://arxiv.org/abs/2304.01666v1
A Survey on Contextualised Semantic Shift Detection
Stefano Montanelli, Francesco Periti
cs.CL
Semantic Shift Detection (SSD) is the task of identifying, interpreting, and assessing the possible change over time in the meanings of a target word. Traditionally, SSD has been addressed by linguists and social scientists through manual and time-consuming activities. In the recent years, computational approaches base...
http://arxiv.org/pdf/2304.01666v1
2023-04-04T09:50:19
2023-04-04T09:50:19
{ "title": [ -0.021347046, 0.027389526, -0.05621338, -0.02859497, 0.041168213, -0.010665894, -0.03024292, -0.053955078, 0.010848999, 0.04647827, -0.010597229, -0.008338928, 0.017578125, -0.019119263, -0.041931152, 0.010116577, -0.0118255615, 0.03...
http://arxiv.org/abs/2304.01665v2
Mastering Symbolic Operations: Augmenting Language Models with Compiled Neural Networks
Yixuan Weng, Minjun Zhu, Fei Xia, Bin Li, Shizhu He, Kang Liu, Jun Zhao
cs.CL
Language models (LMs) proficiency in handling deterministic symbolic reasoning and rule-based tasks remains limited due to their dependency implicit learning on textual data. To enable fully rule comprehension ability, we explore how to incorporate compiled neural networks (CoNNs) which weight is specially designed int...
http://arxiv.org/pdf/2304.01665v2
2023-05-18T08:48:43
2023-04-04T09:50:07
{ "title": [ -0.01576233, -0.010955811, -0.017929077, -0.059661865, 0.00058174133, 0.010665894, -0.060455322, -0.00013697147, 0.00007271767, 0.036743164, -0.00009483099, -0.015991211, -0.011978149, -0.017181396, -0.06689453, 0.0053520203, -0.00061368...
http://arxiv.org/abs/2304.01662v1
Cross-Domain Image Captioning with Discriminative Finetuning
Roberto Dessì, Michele Bevilacqua, Eleonora Gualdoni, Nathanael Carraz Rakotonirina, Francesca Franzon, Marco Baroni
cs.CV, cs.AI, cs.CL
Neural captioners are typically trained to mimic human-generated references without optimizing for any specific communication goal, leading to problems such as the generation of vague captions. In this paper, we show that fine-tuning an out-of-the-box neural captioner with a self-supervised discriminative communication...
http://arxiv.org/pdf/2304.01662v1
2023-04-04T09:33:16
2023-04-04T09:33:16
{ "title": [ 0.01638794, 0.017608643, -0.009712219, -0.04949951, 0.023544312, -0.01902771, 0.010047913, -0.0284729, 0.0056877136, 0.044921875, -0.026733398, 0.027069092, -0.006454468, -0.0027885437, -0.042633057, 0.003200531, -0.019119263, 0.0575...
http://arxiv.org/abs/2304.01638v1
Multidimensional Perceptron for Efficient and Explainable Long Text Classification
Yexiang Wang, Yating Zhang, Xiaozhong Liu, Changlong Sun
cs.CL, cs.AI
Because of the inevitable cost and complexity of transformer and pre-trained models, efficiency concerns are raised for long text classification. Meanwhile, in the highly sensitive domains, e.g., healthcare and legal long-text mining, potential model distrust, yet underrated and underexplored, may hatch vital apprehens...
http://arxiv.org/pdf/2304.01638v1
2023-04-04T08:49:39
2023-04-04T08:49:39
{ "title": [ -0.017654419, 0.0039367676, -0.030593872, -0.050323486, 0.0039749146, -0.006134033, -0.00642395, 0.001698494, -0.00045132637, 0.023712158, -0.022155762, 0.0119018555, -0.017974854, -0.003955841, -0.04244995, -0.024993896, -0.024856567, ...
http://arxiv.org/abs/2304.01622v2
An interpretability framework for Similar case matching
Nankai Lin, Haonan Liu, Jiajun Fang, Dong Zhou, Aimin Yang
cs.CL
Similar Case Matching (SCM) plays a pivotal role in the legal system by facilitating the efficient identification of similar cases for legal professionals. While previous research has primarily concentrated on enhancing the performance of SCM models, the aspect of interpretability has been neglected. To bridge the gap,...
http://arxiv.org/pdf/2304.01622v2
2023-08-16T11:58:38
2023-04-04T08:25:12
{ "title": [ -0.0069007874, 0.027114868, -0.029907227, -0.08099365, 0.032287598, 0.022720337, -0.06097412, -0.040924072, 0.010284424, 0.013801575, -0.036132812, -0.008155823, -0.001329422, -0.013282776, -0.03652954, -0.018310547, -0.019058228, 0....
http://arxiv.org/abs/2304.01621v1
SimCSum: Joint Learning of Simplification and Cross-lingual Summarization for Cross-lingual Science Journalism
Mehwish Fatima, Tim Kolber, Katja Markert, Michael Strube
cs.CL
Cross-lingual science journalism generates popular science stories of scientific articles different from the source language for a non-expert audience. Hence, a cross-lingual popular summary must contain the salient content of the input document, and the content should be coherent, comprehensible, and in a local langua...
http://arxiv.org/pdf/2304.01621v1
2023-04-04T08:24:22
2023-04-04T08:24:22
{ "title": [ -0.036346436, -0.028076172, -0.02142334, -0.05697632, -0.023895264, 0.024139404, -0.064086914, -0.00919342, 0.008468628, 0.04788208, -0.0004041195, 0.03161621, -0.00957489, -0.010063171, -0.043548584, 0.020309448, -0.012992859, 0.044...
http://arxiv.org/abs/2304.01612v1
EDeR: A Dataset for Exploring Dependency Relations Between Events
Ruiqi Li, Patrik Haslum, Leyang Cui
cs.CL
Relation extraction is a central task in natural language processing (NLP) and information retrieval (IR) research. We argue that an important type of relation not explored in NLP or IR research to date is that of an event being an argument - required or optional - of another event. We introduce the human-annotated Eve...
http://arxiv.org/pdf/2304.01612v1
2023-04-04T08:07:07
2023-04-04T08:07:07
{ "title": [ 0.014389038, -0.002149582, -0.062561035, -0.13293457, 0.031585693, -0.00051927567, -0.07684326, 0.0079193115, 0.02558899, 0.0131073, 0.0005979538, -0.0069122314, -0.0051116943, -0.030685425, -0.035064697, 0.013381958, -0.0072135925, ...
http://arxiv.org/abs/2304.01597v1
Unsupervised Improvement of Factual Knowledge in Language Models
Nafis Sadeq, Byungkyu Kang, Prarit Lamba, Julian McAuley
cs.CL
Masked language modeling (MLM) plays a key role in pretraining large language models. But the MLM objective is often dominated by high-frequency words that are sub-optimal for learning factual knowledge. In this work, we propose an approach for influencing MLM pretraining in a way that can improve language model perfor...
http://arxiv.org/pdf/2304.01597v1
2023-04-04T07:37:06
2023-04-04T07:37:06
{ "title": [ -0.024154663, -0.00035309792, -0.045440674, -0.0635376, 0.00045490265, -0.011909485, -0.062683105, 0.0013303757, 0.019317627, 0.033355713, -0.022521973, 0.0031814575, 0.020935059, -0.009033203, -0.042541504, 0.025131226, -0.012298584, ...
http://arxiv.org/abs/2304.01563v1
Attribute-Consistent Knowledge Graph Representation Learning for Multi-Modal Entity Alignment
Qian Li, Shu Guo, Yangyifei Luo, Cheng Ji, Lihong Wang, Jiawei Sheng, Jianxin Li
cs.CL
The multi-modal entity alignment (MMEA) aims to find all equivalent entity pairs between multi-modal knowledge graphs (MMKGs). Rich attributes and neighboring entities are valuable for the alignment task, but existing works ignore contextual gap problems that the aligned entities have different numbers of attributes on...
http://arxiv.org/pdf/2304.01563v1
2023-04-04T06:39:36
2023-04-04T06:39:36
{ "title": [ -0.030151367, 0.020324707, -0.08581543, -0.068115234, 0.017654419, -0.013839722, -0.07312012, 0.0049934387, 0.018981934, 0.0519104, 0.008346558, 0.0063819885, 0.05630493, -0.022994995, -0.034698486, 0.02268982, -0.004306793, 0.026596...
http://arxiv.org/abs/2304.06488v1
One Small Step for Generative AI, One Giant Leap for AGI: A Complete Survey on ChatGPT in AIGC Era
Chaoning Zhang, Chenshuang Zhang, Chenghao Li, Yu Qiao, Sheng Zheng, Sumit Kumar Dam, Mengchun Zhang, Jung Uk Kim, Seong Tae Kim, Jinwoo Choi, Gyeong-Moon Park, Sung-Ho Bae, Lik-Hang Lee, Pan Hui, In So Kweon, Choong Seon Hong
cs.CY, cs.AI, cs.CL, cs.CV, cs.LG
OpenAI has recently released GPT-4 (a.k.a. ChatGPT plus), which is demonstrated to be one small step for generative AI (GAI), but one giant leap for artificial general intelligence (AGI). Since its official release in November 2022, ChatGPT has quickly attracted numerous users with extensive media coverage. Such unprec...
http://arxiv.org/pdf/2304.06488v1
2023-04-04T06:22:09
2023-04-04T06:22:09
{ "title": [ -0.0141067505, -0.0087509155, -0.04815674, -0.014350891, -0.0060653687, -0.023361206, -0.0234375, -0.013191223, -0.0017814636, 0.025817871, -0.03250122, 0.0064811707, 0.023101807, -0.025848389, -0.01852417, -0.0046157837, -0.0034484863, ...
http://arxiv.org/abs/2304.01515v1
Text-Conditioned Sampling Framework for Text-to-Image Generation with Masked Generative Models
Jaewoong Lee, Sangwon Jang, Jaehyeong Jo, Jaehong Yoon, Yunji Kim, Jin-Hwa Kim, Jung-Woo Ha, Sung Ju Hwang
cs.LG, cs.CL, cs.CV, I.5.4; I.2.10; I.4.m
Token-based masked generative models are gaining popularity for their fast inference time with parallel decoding. While recent token-based approaches achieve competitive performance to diffusion-based models, their generation performance is still suboptimal as they sample multiple tokens simultaneously without consider...
http://arxiv.org/pdf/2304.01515v1
2023-04-04T03:52:49
2023-04-04T03:52:49
{ "title": [ -0.022827148, 0.0032920837, -0.024139404, -0.025131226, -0.020889282, 0.007888794, -0.03540039, -0.029174805, 0.030456543, 0.042510986, -0.039978027, 0.030426025, 0.017059326, -0.0103302, -0.031158447, 0.0074157715, -0.0071983337, 0....
http://arxiv.org/abs/2304.01492v5
A Unified Contrastive Transfer Framework with Propagation Structure for Boosting Low-Resource Rumor Detection
Hongzhan Lin, Jing Ma, Ruichao Yang, Zhiwei Yang, Mingfei Cheng
cs.CL
The truth is significantly hampered by massive rumors that spread along with breaking news or popular topics. Since there is sufficient corpus gathered from the same domain for model training, existing rumor detection algorithms show promising performance on yesterday's news. However, due to a lack of substantial train...
http://arxiv.org/pdf/2304.01492v5
2023-10-17T02:37:46
2023-04-04T03:13:03
{ "title": [ 0.017974854, -0.0012493134, -0.018081665, -0.038238525, -0.004512787, -0.016433716, -0.02709961, -0.03741455, 0.024658203, 0.047668457, -0.042816162, 0.0063285828, 0.021606445, -0.0079956055, -0.043884277, -0.00050115585, -0.02708435, ...
http://arxiv.org/abs/2304.01487v2
To ChatGPT, or not to ChatGPT: That is the question!
Alessandro Pegoraro, Kavita Kumari, Hossein Fereidooni, Ahmad-Reza Sadeghi
cs.LG, cs.AI, cs.CL
ChatGPT has become a global sensation. As ChatGPT and other Large Language Models (LLMs) emerge, concerns of misusing them in various ways increase, such as disseminating fake news, plagiarism, manipulating public opinion, cheating, and fraud. Hence, distinguishing AI-generated from human-generated becomes increasingly...
http://arxiv.org/pdf/2304.01487v2
2023-04-05T04:28:41
2023-04-04T03:04:28
{ "title": [ -0.01550293, -0.022155762, -0.0138168335, -0.024337769, -0.012359619, 0.004173279, -0.0013446808, -0.015899658, -0.021026611, 0.010528564, -0.015197754, 0.01777649, -0.04043579, -0.025146484, -0.014533997, -0.0020999908, -0.019851685, ...
http://arxiv.org/abs/2304.01483v2
Blockwise Compression of Transformer-based Models without Retraining
Gaochen Dong, Wei Chen
cs.CL, cs.LG
Transformer-based models, exemplified by GPT-3, ChatGPT, and GPT-4, have recently garnered considerable attention in both academia and industry due to their promising performance in general language tasks. Nevertheless, these models typically involve computationally encoding processes, and in some cases, decoding proce...
http://arxiv.org/pdf/2304.01483v2
2023-09-17T22:47:50
2023-04-04T02:55:40
{ "title": [ -0.022033691, 0.014350891, -0.032684326, -0.05569458, 0.01361084, 0.029541016, -0.074157715, -0.03390503, 0.005596161, 0.04107666, 0.0012016296, -0.0050354004, 0.0053977966, -0.024398804, -0.033172607, 0.020614624, 0.026947021, -0.00...
http://arxiv.org/abs/2304.01481v1
The Vector Grounding Problem
Dimitri Coelho Mollo, Raphaël Millière
cs.CL
The remarkable performance of large language models (LLMs) on complex linguistic tasks has sparked a lively debate on the nature of their capabilities. Unlike humans, these models learn language exclusively from textual data, without direct interaction with the real world. Nevertheless, they can generate seemingly mean...
http://arxiv.org/pdf/2304.01481v1
2023-04-04T02:54:04
2023-04-04T02:54:04
{ "title": [ 0.049987793, 0.015609741, -0.0063667297, -0.0043182373, -0.032684326, -0.035339355, -0.003856659, -0.05444336, 0.021347046, 0.033935547, -0.006427765, -0.03555298, 0.012359619, -0.050079346, -0.013168335, 0.008735657, 0.026397705, 0....
http://arxiv.org/abs/2304.01424v1
Polarity based Sarcasm Detection using Semigraph
Swapnil Mane, Vaibhav Khatavkar
cs.CL
Sarcasm is an advanced linguistic expression often found on various online platforms. Sarcasm detection is challenging in natural language processing tasks that affect sentiment analysis. This article presents the inventive method of the semigraph, including semigraph construction and sarcasm detection processes. A var...
http://arxiv.org/pdf/2304.01424v1
2023-04-04T00:13:55
2023-04-04T00:13:55
{ "title": [ -0.018737793, 0.01676941, -0.040405273, -0.025802612, 0.002828598, -0.007858276, -0.03466797, -0.027999878, 0.014785767, 0.040649414, -0.010910034, -0.04876709, 0.03591919, -0.026290894, -0.072143555, -0.014167786, -0.033081055, 0.05...
http://arxiv.org/abs/2304.01423v1
Thematic context vector association based on event uncertainty for Twitter
Vaibhav Khatavkar, Swapnil Mane, Parag Kulkarni
cs.CL
Keyword extraction is a crucial process in text mining. The extraction of keywords with respective contextual events in Twitter data is a big challenge. The challenging issues are mainly because of the informality in the language used. The use of misspelled words, acronyms, and ambiguous terms causes informality. The e...
http://arxiv.org/pdf/2304.01423v1
2023-04-04T00:13:47
2023-04-04T00:13:47
{ "title": [ 0.032226562, 0.015419006, -0.016067505, -0.06933594, 0.029510498, -0.00076150894, -0.019500732, -0.032287598, 0.025527954, 0.055847168, -0.023513794, 0.029464722, 0.013450623, -0.0076789856, -0.050933838, 0.01939392, -0.0042686462, 0...
http://arxiv.org/abs/2304.01412v2
The StatCan Dialogue Dataset: Retrieving Data Tables through Conversations with Genuine Intents
Xing Han Lu, Siva Reddy, Harm de Vries
cs.CL
We introduce the StatCan Dialogue Dataset consisting of 19,379 conversation turns between agents working at Statistics Canada and online users looking for published data tables. The conversations stem from genuine intents, are held in English or French, and lead to agents retrieving one of over 5000 complex data tables...
http://arxiv.org/pdf/2304.01412v2
2023-04-05T01:20:51
2023-04-03T23:18:30
{ "title": [ -0.013412476, -0.018600464, -0.059906006, -0.05935669, 0.012413025, -0.021408081, -0.044799805, -0.0071983337, 0.009941101, 0.002336502, -0.03189087, 0.0032024384, -0.039215088, -0.021942139, -0.03451538, -0.024414062, -0.0110321045, ...
http://arxiv.org/abs/2304.02020v1
A Bibliometric Review of Large Language Models Research from 2017 to 2023
Lizhou Fan, Lingyao Li, Zihui Ma, Sanggyu Lee, Huizi Yu, Libby Hemphill
cs.DL, cs.CL, cs.CY, cs.SI
Large language models (LLMs) are a class of language models that have demonstrated outstanding performance across a range of natural language processing (NLP) tasks and have become a highly sought-after research area, because of their ability to generate human-like language and their potential to revolutionize science ...
http://arxiv.org/pdf/2304.02020v1
2023-04-03T21:46:41
2023-04-03T21:46:41
{ "title": [ -0.052886963, -0.013404846, -0.054107666, -0.05517578, -0.015197754, -0.0054512024, -0.081848145, 0.0038967133, -0.013031006, 0.043792725, -0.004436493, 0.008483887, 0.026138306, -0.02796936, -0.014205933, 0.03289795, -0.0074653625, ...
http://arxiv.org/abs/2304.01373v2
Pythia: A Suite for Analyzing Large Language Models Across Training and Scaling
Stella Biderman, Hailey Schoelkopf, Quentin Anthony, Herbie Bradley, Kyle O'Brien, Eric Hallahan, Mohammad Aflah Khan, Shivanshu Purohit, USVSN Sai Prashanth, Edward Raff, Aviya Skowron, Lintang Sutawika, Oskar van der Wal
cs.CL
How do large language models (LLMs) develop and evolve over the course of training? How do these patterns change as models scale? To answer these questions, we introduce \textit{Pythia}, a suite of 16 LLMs all trained on public data seen in the exact same order and ranging in size from 70M to 12B parameters. We provide...
http://arxiv.org/pdf/2304.01373v2
2023-05-31T17:54:07
2023-04-03T20:58:15
{ "title": [ -0.024368286, -0.01638794, -0.049713135, -0.07147217, 0.014549255, -0.0047836304, -0.056884766, 0.0063972473, -0.0020580292, 0.015808105, -0.012825012, 0.00484848, 0.0236969, -0.030899048, -0.022445679, 0.017669678, -0.010910034, 0.0...
http://arxiv.org/abs/2304.01352v2
A Simple and Effective Method of Cross-Lingual Plagiarism Detection
Karen Avetisyan, Arthur Malajyan, Tsolak Ghukasyan, Arutyun Avetisyan
cs.CL, cs.IR
We present a simple cross-lingual plagiarism detection method applicable to a large number of languages. The presented approach leverages open multilingual thesauri for candidate retrieval task and pre-trained multilingual BERT-based language models for detailed analysis. The method does not rely on machine translation...
http://arxiv.org/pdf/2304.01352v2
2023-04-05T09:23:17
2023-04-03T20:27:10
{ "title": [ -0.008880615, 0.00007277727, -0.027572632, -0.030654907, 0.012641907, -0.010192871, -0.010925293, -0.034820557, -0.0068130493, 0.039794922, -0.012466431, 0.016937256, 0.019592285, -0.01689148, -0.029006958, 0.0098724365, -0.015792847, ...
http://arxiv.org/abs/2304.01344v1
End-to-End Models for Chemical-Protein Interaction Extraction: Better Tokenization and Span-Based Pipeline Strategies
Xuguang Ai, Ramakanth Kavuluru
cs.CL
End-to-end relation extraction (E2ERE) is an important task in information extraction, more so for biomedicine as scientific literature continues to grow exponentially. E2ERE typically involves identifying entities (or named entity recognition (NER)) and associated relations, while most RE tasks simply assume that the ...
http://arxiv.org/pdf/2304.01344v1
2023-04-03T20:20:22
2023-04-03T20:20:22
{ "title": [ -0.007423401, -0.0016107559, -0.038970947, -0.08569336, 0.012992859, 0.011154175, -0.046905518, 0.01979065, 0.0004940033, 0.009552002, -0.007396698, 0.016357422, -0.035858154, -0.024383545, -0.02305603, 0.0035877228, -0.015144348, 0....
http://arxiv.org/abs/2304.01331v1
Creating Custom Event Data Without Dictionaries: A Bag-of-Tricks
Andrew Halterman, Philip A. Schrodt, Andreas Beger, Benjamin E. Bagozzi, Grace I. Scarborough
cs.CL
Event data, or structured records of ``who did what to whom'' that are automatically extracted from text, is an important source of data for scholars of international politics. The high cost of developing new event datasets, especially using automated systems that rely on hand-built dictionaries, means that most resear...
http://arxiv.org/pdf/2304.01331v1
2023-04-03T19:51:00
2023-04-03T19:51:00
{ "title": [ 0.0113220215, 0.0011777878, -0.045318604, -0.119262695, 0.00630188, 0.0036697388, -0.07513428, 0.0049324036, -0.023742676, 0.019241333, 0.018554688, -0.029174805, -0.030380249, -0.05218506, -0.01360321, -0.006248474, 0.023651123, 0.0...
http://arxiv.org/abs/2304.01330v1
A Comparison of Document Similarity Algorithms
Nicholas Gahman, Vinayak Elangovan
cs.CL, cs.AI
Document similarity is an important part of Natural Language Processing and is most commonly used for plagiarism-detection and text summarization. Thus, finding the overall most effective document similarity algorithm could have a major positive impact on the field of Natural Language Processing. This report sets out t...
http://arxiv.org/pdf/2304.01330v1
2023-04-03T19:50:55
2023-04-03T19:50:55
{ "title": [ -0.022994995, 0.005317688, -0.027114868, -0.04840088, 0.018722534, -0.0042381287, -0.017700195, -0.020248413, 0.007801056, 0.04547119, -0.032592773, 0.016296387, 0.007648468, -0.0019483566, -0.03491211, 0.009895325, 0.0049095154, 0.0...
http://arxiv.org/abs/2304.01328v1
Grand Challenge On Detecting Cheapfakes
Duc-Tien Dang-Nguyen, Sohail Ahmed Khan, Cise Midoglu, Michael Riegler, Pål Halvorsen, Minh-Son Dao
cs.CV, cs.CL
Cheapfake is a recently coined term that encompasses non-AI ("cheap") manipulations of multimedia content. Cheapfakes are known to be more prevalent than deepfakes. Cheapfake media can be created using editing software for image/video manipulations, or even without using any software, by simply altering the context of ...
http://arxiv.org/pdf/2304.01328v1
2023-04-03T19:50:26
2023-04-03T19:50:26
{ "title": [ 0.06555176, 0.026809692, 0.017913818, -0.0059890747, -0.014587402, -0.002664566, 0.02494812, -0.049926758, -0.054870605, 0.030090332, -0.016601562, 0.008384705, 0.010383606, -0.047973633, 0.030303955, 0.0010662079, -0.011482239, 0.07...
http://arxiv.org/abs/2304.01322v1
PALI: A Language Identification Benchmark for Perso-Arabic Scripts
Sina Ahmadi, Milind Agarwal, Antonios Anastasopoulos
cs.CL
The Perso-Arabic scripts are a family of scripts that are widely adopted and used by various linguistic communities around the globe. Identifying various languages using such scripts is crucial to language technologies and challenging in low-resource setups. As such, this paper sheds light on the challenges of detectin...
http://arxiv.org/pdf/2304.01322v1
2023-04-03T19:40:14
2023-04-03T19:40:14
{ "title": [ -0.031402588, 0.009094238, -0.03326416, -0.04840088, 0.00088500977, -0.00217247, -0.04196167, -0.0013179779, -0.008010864, 0.010498047, -0.008201599, 0.023513794, -0.031097412, -0.03652954, -0.030059814, -0.009735107, 0.015113831, 0....
http://arxiv.org/abs/2304.01319v1
Approaches to Corpus Creation for Low-Resource Language Technology: the Case of Southern Kurdish and Laki
Sina Ahmadi, Zahra Azin, Sara Belelli, Antonios Anastasopoulos
cs.CL
One of the major challenges that under-represented and endangered language communities face in language technology is the lack or paucity of language data. This is also the case of the Southern varieties of the Kurdish and Laki languages for which very limited resources are available with insubstantial progress in tool...
http://arxiv.org/pdf/2304.01319v1
2023-04-03T19:36:32
2023-04-03T19:36:32
{ "title": [ -0.030654907, 0.0035419464, -0.039215088, -0.030288696, -0.017089844, 0.0023231506, -0.048797607, -0.0032596588, 0.004470825, 0.03427124, -0.034179688, -0.018249512, 0.01184845, -0.018951416, -0.057525635, -0.017532349, -0.016403198, ...
http://arxiv.org/abs/2304.01295v3
Efficiently Aligned Cross-Lingual Transfer Learning for Conversational Tasks using Prompt-Tuning
Lifu Tu, Jin Qu, Semih Yavuz, Shafiq Joty, Wenhao Liu, Caiming Xiong, Yingbo Zhou
cs.CL, cs.AI
Cross-lingual transfer of language models trained on high-resource languages like English has been widely studied for many NLP tasks, but focus on conversational tasks has been rather limited. This is partly due to the high cost of obtaining non-English conversational data, which results in limited coverage. In this wo...
http://arxiv.org/pdf/2304.01295v3
2023-10-31T05:20:14
2023-04-03T18:46:01
{ "title": [ -0.008338928, -0.0058784485, -0.033233643, -0.04498291, -0.0056152344, -0.0018262863, -0.05444336, -0.008102417, 0.0017786026, 0.019348145, -0.009971619, 0.013023376, 0.013328552, -0.017074585, -0.0118255615, 0.004962921, -0.008735657, ...
http://arxiv.org/abs/2304.01282v2
PEACH: Pre-Training Sequence-to-Sequence Multilingual Models for Translation with Semi-Supervised Pseudo-Parallel Document Generation
Alireza Salemi, Amirhossein Abaskohi, Sara Tavakoli, Yadollah Yaghoobzadeh, Azadeh Shakery
cs.CL
Multilingual pre-training significantly improves many multilingual NLP tasks, including machine translation. Most existing methods are based on some variants of masked language modeling and text-denoising objectives on monolingual data. Multilingual pre-training on monolingual data ignores the availability of parallel ...
http://arxiv.org/pdf/2304.01282v2
2023-04-14T17:54:58
2023-04-03T18:19:26
{ "title": [ -0.060577393, -0.0077934265, -0.02305603, -0.055755615, -0.029678345, 0.027755737, -0.09295654, 0.0023975372, -0.0078125, 0.04449463, 0.0063934326, 0.012275696, 0.042053223, -0.027648926, -0.021347046, 0.013496399, -0.010231018, 0.02...
http://arxiv.org/abs/2304.01196v3
Baize: An Open-Source Chat Model with Parameter-Efficient Tuning on Self-Chat Data
Canwen Xu, Daya Guo, Nan Duan, Julian McAuley
cs.CL, cs.AI
Chat models, such as ChatGPT, have shown impressive capabilities and have been rapidly adopted across numerous domains. However, these models are only accessible through a restricted API, creating barriers for new research and progress in the field. We propose a pipeline that can automatically generate a high-quality m...
http://arxiv.org/pdf/2304.01196v3
2023-05-23T19:40:03
2023-04-03T17:59:09
{ "title": [ -0.022460938, 0.0059280396, -0.023742676, -0.068115234, -0.013771057, 0.006580353, -0.054656982, -0.005695343, -0.004764557, 0.04550171, 0.0016059875, -0.0024738312, 0.0036716461, -0.021408081, -0.05886841, 0.00030374527, -0.004333496, ...
http://arxiv.org/abs/2304.01179v1
Hate Speech Targets Detection in Parler using BERT
Nadav Schneider, Shimon Shouei, Saleem Ghantous, Elad Feldman
cs.CL, cs.AI
Online social networks have become a fundamental component of our everyday life. Unfortunately, these platforms are also a stage for hate speech. Popular social networks have regularized rules against hate speech. Consequently, social networks like Parler and Gab advocating and claiming to be free speech platforms have...
http://arxiv.org/pdf/2304.01179v1
2023-04-03T17:49:04
2023-04-03T17:49:04
{ "title": [ 0.01828003, -0.00617218, -0.031036377, -0.0440979, -0.0128479, -0.011726379, -0.008460999, 0.0067825317, -0.04107666, 0.04220581, -0.015090942, 0.021942139, 0.06738281, -0.033966064, 0.018936157, 0.014465332, -0.029937744, 0.06860351...
http://arxiv.org/abs/2304.01246v2
Safety Analysis in the Era of Large Language Models: A Case Study of STPA using ChatGPT
Yi Qi, Xingyu Zhao, Siddartha Khastgir, Xiaowei Huang
cs.CL, cs.AI, cs.CY, cs.SE
Can safety analysis make use of Large Language Models (LLMs)? A case study explores Systems Theoretic Process Analysis (STPA) applied to Automatic Emergency Brake (AEB) and Electricity Demand Side Management (DSM) systems using ChatGPT. We investigate how collaboration schemes, input semantic complexity, and prompt gui...
http://arxiv.org/pdf/2304.01246v2
2023-11-05T21:58:36
2023-04-03T16:46:49
{ "title": [ -0.0146865845, -0.012504578, -0.042907715, -0.060546875, 0.0044784546, -0.0010995865, -0.045288086, 0.009338379, -0.050109863, 0.0016508102, -0.0039138794, 0.021987915, -0.007331848, -0.04168701, 0.027069092, -0.0046157837, -0.027908325,...
http://arxiv.org/abs/2304.01106v1
Crossword: A Semantic Approach to Data Compression via Masking
Mingxiao Li, Rui Jin, Liyao Xiang, Kaiming Shen, Shuguang Cui
cs.CL, cs.IT, math.IT
The traditional methods for data compression are typically based on the symbol-level statistics, with the information source modeled as a long sequence of i.i.d. random variables or a stochastic process, thus establishing the fundamental limit as entropy for lossless compression and as mutual information for lossy comp...
http://arxiv.org/pdf/2304.01106v1
2023-04-03T16:04:06
2023-04-03T16:04:06
{ "title": [ -0.08331299, -0.012519836, -0.066589355, -0.023666382, -0.006500244, 0.0049324036, -0.01940918, 0.023498535, 0.006793976, 0.049713135, -0.017227173, 0.0033893585, 0.029312134, -0.021774292, -0.028884888, 0.009246826, 0.015045166, 0.0...
http://arxiv.org/abs/2304.01097v2
DoctorGLM: Fine-tuning your Chinese Doctor is not a Herculean Task
Honglin Xiong, Sheng Wang, Yitao Zhu, Zihao Zhao, Yuxiao Liu, Linlin Huang, Qian Wang, Dinggang Shen
cs.CL
The recent progress of large language models (LLMs), including ChatGPT and GPT-4, in comprehending and responding to human instructions has been remarkable. Nevertheless, these models typically perform better in English and have not been explicitly trained for the medical domain, resulting in suboptimal precision in di...
http://arxiv.org/pdf/2304.01097v2
2023-04-17T17:06:29
2023-04-03T15:57:51
{ "title": [ 0.0118637085, -0.013961792, -0.00007545948, -0.007320404, -0.024124146, -0.0030136108, -0.017669678, -0.04510498, -0.038391113, 0.028244019, 0.02456665, 0.012107849, -0.027877808, -0.054260254, 0.0037059784, -0.038482666, 0.015281677, ...
http://arxiv.org/abs/2304.01089v4
RPTQ: Reorder-based Post-training Quantization for Large Language Models
Zhihang Yuan, Lin Niu, Jiawei Liu, Wenyu Liu, Xinggang Wang, Yuzhang Shang, Guangyu Sun, Qiang Wu, Jiaxiang Wu, Bingzhe Wu
cs.CL
Large-scale language models (LLMs) have demonstrated impressive performance, but their deployment presents challenges due to their significant memory usage. This issue can be alleviated through quantization. In this paper, we identify that the challenge in quantizing activations in LLMs arises from varying ranges acros...
http://arxiv.org/pdf/2304.01089v4
2023-05-17T10:07:33
2023-04-03T15:46:15
{ "title": [ -0.03479004, 0.0006866455, -0.03894043, -0.05657959, 0.004840851, 0.00484848, -0.070007324, -0.011054993, -0.0061683655, 0.054229736, 0.004737854, 0.0038700104, 0.026382446, -0.022567749, -0.042236328, 0.018692017, -0.013496399, 0.03...
http://arxiv.org/abs/2304.01083v1
Can the Inference Logic of Large Language Models be Disentangled into Symbolic Concepts?
Wen Shen, Lei Cheng, Yuxiao Yang, Mingjie Li, Quanshi Zhang
cs.CL, cs.AI, cs.CV, cs.LG
In this paper, we explain the inference logic of large language models (LLMs) as a set of symbolic concepts. Many recent studies have discovered that traditional DNNs usually encode sparse symbolic concepts. However, because an LLM has much more parameters than traditional DNNs, whether the LLM also encodes sparse symb...
http://arxiv.org/pdf/2304.01083v1
2023-04-03T15:39:35
2023-04-03T15:39:35
{ "title": [ -0.010093689, -0.032470703, -0.02192688, -0.039245605, -0.04119873, 0.0010108948, -0.050476074, 0.0070152283, 0.0060920715, 0.025268555, -0.036224365, -0.024505615, 0.03164673, -0.02696228, 0.0020179749, -0.024917603, 0.0032138824, 0...
http://arxiv.org/abs/2304.01046v3
Deep Manifold Learning for Reading Comprehension and Logical Reasoning Tasks with Polytuplet Loss
Jeffrey Lu, Ivan Rodriguez
cs.CL, cs.AI, cs.LG, 68T50, I.2.7
The current trend in developing machine learning models for reading comprehension and logical reasoning tasks is focused on improving the models' abilities to understand and utilize logical rules. This work focuses on providing a novel loss function and accompanying model architecture that has more interpretable compon...
http://arxiv.org/pdf/2304.01046v3
2023-10-15T16:11:05
2023-04-03T14:48:34
{ "title": [ -0.025253296, -0.027130127, -0.03100586, -0.074401855, -0.026321411, 0.0009737015, -0.06427002, 0.014884949, 0.009567261, 0.03277588, -0.024124146, -0.009185791, 0.033294678, -0.015640259, -0.016342163, 0.007511139, -0.0014419556, 0....
http://arxiv.org/abs/2304.01019v1
Simple Yet Effective Neural Ranking and Reranking Baselines for Cross-Lingual Information Retrieval
Jimmy Lin, David Alfonso-Hermelo, Vitor Jeronymo, Ehsan Kamalloo, Carlos Lassance, Rodrigo Nogueira, Odunayo Ogundepo, Mehdi Rezagholizadeh, Nandan Thakur, Jheng-Hong Yang, Xinyu Zhang
cs.IR, cs.CL
The advent of multilingual language models has generated a resurgence of interest in cross-lingual information retrieval (CLIR), which is the task of searching documents in one language with queries from another. However, the rapid pace of progress has led to a confusing panoply of methods and reproducibility has lagge...
http://arxiv.org/pdf/2304.01019v1
2023-04-03T14:17:00
2023-04-03T14:17:00
{ "title": [ -0.0011539459, 0.007621765, -0.018844604, -0.033325195, 0.022460938, -0.001115799, -0.015449524, -0.036590576, 0.021987915, 0.05331421, -0.010040283, 0.0059661865, 0.019836426, -0.0050086975, -0.050598145, 0.019500732, -0.01058197, 0...
http://arxiv.org/abs/2304.01002v3
Does Human Collaboration Enhance the Accuracy of Identifying LLM-Generated Deepfake Texts?
Adaku Uchendu, Jooyoung Lee, Hua Shen, Thai Le, Ting-Hao 'Kenneth' Huang, Dongwon Lee
cs.CL, cs.AI, cs.HC
Advances in Large Language Models (e.g., GPT-4, LLaMA) have improved the generation of coherent sentences resembling human writing on a large scale, resulting in the creation of so-called deepfake texts. However, this progress poses security and privacy concerns, necessitating effective solutions for distinguishing dee...
http://arxiv.org/pdf/2304.01002v3
2023-10-09T21:57:47
2023-04-03T14:06:47
{ "title": [ -0.0184021, 0.02999878, -0.04748535, 0.011558533, -0.03805542, -0.01576233, -0.037384033, -0.027938843, -0.032928467, 0.02861023, -0.0126953125, 0.03375244, -0.0033435822, -0.020904541, 0.0079956055, -0.016418457, -0.03994751, 0.0338...
http://arxiv.org/abs/2304.00964v1
Robust Text-driven Image Editing Method that Adaptively Explores Directions in Latent Spaces of StyleGAN and CLIP
Tsuyoshi Baba, Kosuke Nishida, Kyosuke Nishida
cs.CV, cs.CL
Automatic image editing has great demands because of its numerous applications, and the use of natural language instructions is essential to achieving flexible and intuitive editing as the user imagines. A pioneering work in text-driven image editing, StyleCLIP, finds an edit direction in the CLIP space and then edits ...
http://arxiv.org/pdf/2304.00964v1
2023-04-03T13:30:48
2023-04-03T13:30:48
{ "title": [ -0.013557434, 0.0061531067, -0.05328369, 0.009773254, -0.02230835, -0.04421997, 0.0015192032, -0.0043563843, -0.0099105835, 0.029953003, 0.00349617, 0.027526855, 0.028701782, -0.041290283, 0.0231781, 0.00983429, 0.027175903, -0.03372...
http://arxiv.org/abs/2304.00958v2
DrBERT: A Robust Pre-trained Model in French for Biomedical and Clinical domains
Yanis Labrak, Adrien Bazoge, Richard Dufour, Mickael Rouvier, Emmanuel Morin, Béatrice Daille, Pierre-Antoine Gourraud
cs.CL
In recent years, pre-trained language models (PLMs) achieve the best performance on a wide range of natural language processing (NLP) tasks. While the first models were trained on general domain data, specialized ones have emerged to more effectively treat specific domains. In this paper, we propose an original study o...
http://arxiv.org/pdf/2304.00958v2
2023-05-04T19:59:38
2023-04-03T13:25:53
{ "title": [ -0.024475098, -0.011428833, -0.038604736, -0.07080078, -0.0009889603, 0.014633179, -0.076660156, -0.016479492, -0.013885498, 0.033813477, 0.01737976, -0.011489868, 0.009742737, -0.054656982, -0.030258179, 0.007621765, -0.0093307495, ...
http://arxiv.org/abs/2304.01242v1
Enhancing Clinical Evidence Recommendation with Multi-Channel Heterogeneous Learning on Evidence Graphs
Maolin Luo, Xiang Zhang
cs.CL, cs.AI
Clinical evidence encompasses the associations and impacts between patients, interventions (such as drugs or physiotherapy), problems, and outcomes. The goal of recommending clinical evidence is to provide medical practitioners with relevant information to support their decision-making processes and to generate new evi...
http://arxiv.org/pdf/2304.01242v1
2023-04-03T12:15:53
2023-04-03T12:15:53
{ "title": [ -0.010383606, 0.002588272, -0.059173584, -0.076293945, 0.004096985, -0.011352539, -0.042663574, 0.018844604, -0.026504517, 0.022613525, -0.025238037, 0.034240723, -0.018478394, -0.030731201, -0.0013818741, -0.017349243, -0.0041618347, ...
http://arxiv.org/abs/2304.01241v1
Detection of Homophobia & Transphobia in Dravidian Languages: Exploring Deep Learning Methods
Deepawali Sharma, Vedika Gupta, Vivek Kumar Singh
cs.CL, cs.LG
The increase in abusive content on online social media platforms is impacting the social life of online users. Use of offensive and hate speech has been making so-cial media toxic. Homophobia and transphobia constitute offensive comments against LGBT+ community. It becomes imperative to detect and handle these comments...
http://arxiv.org/pdf/2304.01241v1
2023-04-03T12:15:27
2023-04-03T12:15:27
{ "title": [ 0.010597229, 0.016113281, -0.046875, -0.029251099, 0.0062942505, -0.036224365, -0.00059080124, 0.004470825, -0.025146484, 0.041809082, -0.019470215, -0.0181427, 0.041656494, -0.034973145, -0.041625977, -0.020706177, -0.01158905, 0.02...
http://arxiv.org/abs/2304.00913v1
LAHM : Large Annotated Dataset for Multi-Domain and Multilingual Hate Speech Identification
Ankit Yadav, Shubham Chandel, Sushant Chatufale, Anil Bandhakavi
cs.CL, cs.AI
Current research on hate speech analysis is typically oriented towards monolingual and single classification tasks. In this paper, we present a new multilingual hate speech analysis dataset for English, Hindi, Arabic, French, German and Spanish languages for multiple domains across hate speech - Abuse, Racism, Sexism, ...
http://arxiv.org/pdf/2304.00913v1
2023-04-03T12:03:45
2023-04-03T12:03:45
{ "title": [ 0.020507812, 0.008583069, -0.041992188, -0.0647583, 0.03387451, -0.020187378, -0.01309967, -0.0074882507, -0.017562866, 0.036956787, 0.009986877, 0.017654419, 0.013938904, -0.020980835, -0.026290894, 0.0060195923, -0.01675415, 0.0717...
http://arxiv.org/abs/2304.01240v2
Identifying Mentions of Pain in Mental Health Records Text: A Natural Language Processing Approach
Jaya Chaturvedi, Sumithra Velupillai, Robert Stewart, Angus Roberts
cs.CL, cs.LG
Pain is a common reason for accessing healthcare resources and is a growing area of research, especially in its overlap with mental health. Mental health electronic health records are a good data source to study this overlap. However, much information on pain is held in the free text of these records, where mentions of...
http://arxiv.org/pdf/2304.01240v2
2023-04-05T14:02:53
2023-04-03T11:56:11
{ "title": [ -0.0020885468, 0.00091934204, -0.03353882, -0.054107666, 0.01625061, 0.001627922, -0.026367188, -0.014350891, -0.008682251, 0.012870789, -0.023117065, 0.014213562, 0.009155273, -0.026306152, -0.010284424, -0.0060310364, -0.016647339, ...
http://arxiv.org/abs/2304.00906v1
ScandEval: A Benchmark for Scandinavian Natural Language Processing
Dan Saattrup Nielsen
cs.CL, cs.LG
This paper introduces a Scandinavian benchmarking platform, ScandEval, which can benchmark any pretrained model on four different tasks in the Scandinavian languages. The datasets used in two of the tasks, linguistic acceptability and question answering, are new. We develop and release a Python package and command-line...
http://arxiv.org/pdf/2304.00906v1
2023-04-03T11:51:46
2023-04-03T11:51:46
{ "title": [ -0.008453369, -0.021362305, -0.009559631, -0.061645508, 0.017318726, 0.0027561188, -0.03729248, -0.012161255, 0.007255554, 0.03652954, -0.029663086, 0.025390625, 0.012908936, 0.003232956, -0.04244995, 0.017425537, -0.010093689, 0.028...
http://arxiv.org/abs/2304.00884v1
Dialog-to-Actions: Building Task-Oriented Dialogue System via Action-Level Generation
Yuncheng Hua, Xiangyu Xi, Zheng Jiang, Guanwei Zhang, Chaobo Sun, Guanglu Wan, Wei Ye
cs.CL
End-to-end generation-based approaches have been investigated and applied in task-oriented dialogue systems. However, in industrial scenarios, existing methods face the bottlenecks of controllability (e.g., domain-inconsistent responses, repetition problem, etc) and efficiency (e.g., long computation time, etc). In thi...
http://arxiv.org/pdf/2304.00884v1
2023-04-03T11:09:20
2023-04-03T11:09:20
{ "title": [ -0.016204834, -0.0129470825, -0.044769287, -0.06274414, 0.010925293, 0.0012168884, -0.03439331, 0.004142761, 0.0050697327, 0.03213501, 0.00447464, 0.009994507, -0.00945282, -0.021820068, -0.053009033, -0.0019359589, -0.011947632, 0.0...
http://arxiv.org/abs/2304.00869v1
GreekBART: The First Pretrained Greek Sequence-to-Sequence Model
Iakovos Evdaimon, Hadi Abdine, Christos Xypolopoulos, Stamatis Outsios, Michalis Vazirgiannis, Giorgos Stamou
cs.CL
The era of transfer learning has revolutionized the fields of Computer Vision and Natural Language Processing, bringing powerful pretrained models with exceptional performance across a variety of tasks. Specifically, Natural Language Processing tasks have been dominated by transformer-based language models. In Natural ...
http://arxiv.org/pdf/2304.00869v1
2023-04-03T10:48:51
2023-04-03T10:48:51
{ "title": [ 0.0001348257, 0.0006136894, -0.022201538, -0.07147217, 0.0020599365, 0.0074386597, -0.054901123, -0.006439209, 0.014915466, 0.034820557, -0.0101623535, -0.0011205673, 0.03781128, -0.047180176, -0.03152466, 0.030258179, 0.008148193, 0...
http://arxiv.org/abs/2304.01238v3
Spam-T5: Benchmarking Large Language Models for Few-Shot Email Spam Detection
Maxime Labonne, Sean Moran
cs.CL, cs.AI
This paper investigates the effectiveness of large language models (LLMs) in email spam detection by comparing prominent models from three distinct families: BERT-like, Sentence Transformers, and Seq2Seq. Additionally, we examine well-established machine learning techniques for spam detection, such as Na\"ive Bayes and...
http://arxiv.org/pdf/2304.01238v3
2023-05-07T10:57:51
2023-04-03T10:27:53
{ "title": [ 0.008636475, -0.01940918, -0.020568848, -0.04800415, -0.0040626526, -0.01096344, -0.044006348, -0.018188477, -0.0018405914, 0.03942871, -0.012039185, 0.008422852, 0.019836426, -0.011192322, -0.047576904, 0.0009627342, -0.01789856, 0....
http://arxiv.org/abs/2304.00830v2
AUDIT: Audio Editing by Following Instructions with Latent Diffusion Models
Yuancheng Wang, Zeqian Ju, Xu Tan, Lei He, Zhizheng Wu, Jiang Bian, Sheng Zhao
cs.SD, cs.AI, cs.CL, cs.LG, eess.AS
Audio editing is applicable for various purposes, such as adding background sound effects, replacing a musical instrument, and repairing damaged audio. Recently, some diffusion-based methods achieved zero-shot audio editing by using a diffusion and denoising process conditioned on the text description of the output aud...
http://arxiv.org/pdf/2304.00830v2
2023-04-05T12:13:48
2023-04-03T09:15:51
{ "title": [ -0.004196167, -0.023712158, -0.047698975, -0.00086688995, -0.03491211, -0.031433105, -0.014266968, -0.0118637085, -0.020629883, 0.027770996, -0.024475098, 0.0011920929, 0.018569946, -0.04324341, 0.0034275055, -0.031234741, -0.024002075, ...
http://arxiv.org/abs/2304.00824v1
Towards Integration of Discriminability and Robustness for Document-Level Relation Extraction
Jia Guo, Stanley Kok, Lidong Bing
cs.CL
Document-level relation extraction (DocRE) predicts relations for entity pairs that rely on long-range context-dependent reasoning in a document. As a typical multi-label classification problem, DocRE faces the challenge of effectively distinguishing a small set of positive relations from the majority of negative ones....
http://arxiv.org/pdf/2304.00824v1
2023-04-03T09:11:18
2023-04-03T09:11:18
{ "title": [ 0.013206482, 0.012481689, -0.054138184, -0.085998535, 0.032440186, -0.0028495789, -0.05206299, -0.0022201538, 0.044677734, 0.006477356, -0.027069092, -0.007534027, -0.0069351196, 0.000108361244, -0.061157227, -0.008552551, -0.016525269, ...
http://arxiv.org/abs/2304.00815v1
Design Choices for Crowdsourcing Implicit Discourse Relations: Revealing the Biases Introduced by Task Design
Valentina Pyatkin, Frances Yung, Merel C. J. Scholman, Reut Tsarfaty, Ido Dagan, Vera Demberg
cs.CL
Disagreement in natural language annotation has mostly been studied from a perspective of biases introduced by the annotators and the annotation frameworks. Here, we propose to analyze another source of bias: task design bias, which has a particularly strong impact on crowdsourced linguistic annotations where natural l...
http://arxiv.org/pdf/2304.00815v1
2023-04-03T09:04:18
2023-04-03T09:04:18
{ "title": [ 0.00049591064, -0.008003235, -0.031188965, -0.02394104, -0.016967773, -0.016921997, -0.010429382, -0.009811401, 0.0059432983, 0.018234253, -0.022369385, 0.010345459, -0.022949219, -0.0040359497, -0.048706055, -0.026473999, -0.026657104, ...
http://arxiv.org/abs/2304.08174v1
Benchmarking Faithfulness: Towards Accurate Natural Language Explanations in Vision-Language Tasks
Jakob Ambsdorf
cs.CL, cs.AI, cs.CV
With deep neural models increasingly permeating our daily lives comes a need for transparent and comprehensible explanations of their decision-making. However, most explanation methods that have been developed so far are not intuitively understandable for lay users. In contrast, natural language explanations (NLEs) pro...
http://arxiv.org/pdf/2304.08174v1
2023-04-03T08:24:10
2023-04-03T08:24:10
{ "title": [ -0.023803711, -0.015289307, -0.0045700073, -0.062561035, -0.020874023, 0.002292633, -0.057159424, -0.011291504, 0.013175964, 0.023590088, -0.011726379, 0.0061035156, 0.0041656494, -0.00623703, -0.051757812, -0.00283432, 0.006801605, ...
http://arxiv.org/abs/2304.01233v1
Multi-Modal Perceiver Language Model for Outcome Prediction in Emergency Department
Sabri Boughorbel, Fethi Jarray, Abdulaziz Al Homaid, Rashid Niaz, Khalid Alyafei
cs.CL, cs.LG
Language modeling have shown impressive progress in generating compelling text with good accuracy and high semantic coherence. An interesting research direction is to augment these powerful models for specific applications using contextual information. In this work, we explore multi-modal language modeling for healthca...
http://arxiv.org/pdf/2304.01233v1
2023-04-03T06:32:00
2023-04-03T06:32:00
{ "title": [ -0.0087509155, 0.02192688, -0.043701172, -0.057922363, 0.010017395, -0.0066986084, -0.043884277, -0.015525818, -0.009239197, 0.03753662, 0.0006070137, -0.00028824806, 0.022064209, -0.036071777, -0.022705078, 0.005794525, -0.00005799532, ...
http://arxiv.org/abs/2304.00740v2
Inspecting and Editing Knowledge Representations in Language Models
Evan Hernandez, Belinda Z. Li, Jacob Andreas
cs.CL
Neural language models (LMs) represent facts about the world described by text. Sometimes these facts derive from training data (in most LMs, a representation of the word "banana" encodes the fact that bananas are fruits). Sometimes facts derive from input text itself (a representation of the sentence "I poured out the...
http://arxiv.org/pdf/2304.00740v2
2023-05-22T14:53:20
2023-04-03T06:24:10
{ "title": [ -0.039093018, -0.023117065, -0.0423584, -0.07647705, -0.016174316, 0.011001587, -0.06976318, 0.008514404, 0.005870819, 0.036895752, -0.028656006, -0.017211914, 0.011558533, -0.008781433, -0.055847168, 0.0031490326, -0.013633728, 0.03...
http://arxiv.org/abs/2304.00723v3
Exploring the Use of Large Language Models for Reference-Free Text Quality Evaluation: An Empirical Study
Yi Chen, Rui Wang, Haiyun Jiang, Shuming Shi, Ruifeng Xu
cs.CL
Evaluating the quality of generated text is a challenging task in NLP, due to the inherent complexity and diversity of text. Recently, large language models (LLMs) have garnered significant attention due to their impressive performance in various tasks. Therefore, we present this paper to investigate the effectiveness ...
http://arxiv.org/pdf/2304.00723v3
2023-09-18T03:52:18
2023-04-03T05:29:58
{ "title": [ -0.021896362, -0.010131836, -0.022003174, -0.049041748, -0.015914917, 0.007701874, -0.04663086, -0.004322052, -0.0000013709068, 0.06488037, -0.0154418945, 0.0029296875, 0.03387451, 0.0031204224, -0.072021484, 0.007385254, -0.0362854, ...
http://arxiv.org/abs/2304.00717v1
MiniRBT: A Two-stage Distilled Small Chinese Pre-trained Model
Xin Yao, Ziqing Yang, Yiming Cui, Shijin Wang
cs.CL
In natural language processing, pre-trained language models have become essential infrastructures. However, these models often suffer from issues such as large size, long inference time, and challenging deployment. Moreover, most mainstream pre-trained models focus on English, and there are insufficient studies on smal...
http://arxiv.org/pdf/2304.00717v1
2023-04-03T04:45:57
2023-04-03T04:45:57
{ "title": [ 0.0035476685, -0.027252197, -0.021255493, -0.050201416, -0.03387451, -0.010620117, -0.04916382, -0.014701843, -0.025360107, 0.030334473, 0.0016489029, -0.011291504, -0.003145218, -0.056152344, -0.010223389, -0.0040779114, 0.007873535, ...
http://arxiv.org/abs/2304.01905v2
Dual-Attention Neural Transducers for Efficient Wake Word Spotting in Speech Recognition
Saumya Y. Sahai, Jing Liu, Thejaswi Muniyappa, Kanthashree M. Sathyendra, Anastasios Alexandridis, Grant P. Strimel, Ross McGowan, Ariya Rastrow, Feng-Ju Chang, Athanasios Mouchtaris, Siegfried Kunzmann
cs.SD, cs.CL, cs.LG, eess.AS
We present dual-attention neural biasing, an architecture designed to boost Wake Words (WW) recognition and improve inference time latency on speech recognition tasks. This architecture enables a dynamic switch for its runtime compute paths by exploiting WW spotting to select which branch of its attention networks to e...
http://arxiv.org/pdf/2304.01905v2
2023-04-05T01:22:38
2023-04-03T01:19:39
{ "title": [ 0.017990112, 0.0049934387, -0.031463623, -0.050079346, 0.005519867, -0.020446777, -0.0008716583, 0.0030002594, -0.026824951, 0.058776855, -0.043823242, -0.011947632, 0.025436401, 0.0013751984, -0.05429077, -0.018112183, -0.00907135, ...
http://arxiv.org/abs/2304.00649v1
Multilingual Word Error Rate Estimation: e-WER3
Shammur Absar Chowdhury, Ahmed Ali
cs.CL, cs.LG, cs.SD, eess.AS
The success of the multilingual automatic speech recognition systems empowered many voice-driven applications. However, measuring the performance of such systems remains a major challenge, due to its dependency on manually transcribed speech data in both mono- and multilingual scenarios. In this paper, we propose a nov...
http://arxiv.org/pdf/2304.00649v1
2023-04-02T23:08:11
2023-04-02T23:08:11
{ "title": [ -0.05819702, 0.011482239, -0.070129395, 0.0059318542, -0.012680054, -0.026657104, -0.043304443, -0.028793335, -0.011070251, 0.03918457, -0.02305603, 0.017990112, 0.033843994, -0.041656494, -0.002866745, 0.023666382, -0.010276794, 0.0...
http://arxiv.org/abs/2304.00636v1
Classifying COVID-19 Related Tweets for Fake News Detection and Sentiment Analysis with BERT-based Models
Rabia Bounaama, Mohammed El Amine Abderrahim
cs.CL
The present paper is about the participation of our team "techno" on CERIST'22 shared tasks. We used an available dataset "task1.c" related to covid-19 pandemic. It comprises 4128 tweets for sentiment analysis task and 8661 tweets for fake news detection task. We used natural language processing tools with the combinat...
http://arxiv.org/pdf/2304.00636v1
2023-04-02T22:00:27
2023-04-02T22:00:27
{ "title": [ 0.007820129, -0.0068855286, -0.033416748, -0.06201172, 0.0030345917, 0.0013513565, -0.030670166, -0.005428314, -0.034423828, 0.044647217, -0.03390503, 0.033355713, 0.011787415, -0.012107849, -0.022872925, -0.008392334, -0.03253174, 0...
http://arxiv.org/abs/2304.00634v1
MMT: A Multilingual and Multi-Topic Indian Social Media Dataset
Dwip Dalal, Vivek Srivastava, Mayank Singh
cs.CL, cs.LG, cs.SI
Social media plays a significant role in cross-cultural communication. A vast amount of this occurs in code-mixed and multilingual form, posing a significant challenge to Natural Language Processing (NLP) tools for processing such information, like language identification, topic modeling, and named-entity recognition. ...
http://arxiv.org/pdf/2304.00634v1
2023-04-02T21:39:00
2023-04-02T21:39:00
{ "title": [ -0.0096588135, -0.00050878525, -0.052093506, -0.05621338, 0.026733398, -0.011161804, -0.03591919, 0.0040626526, -0.012924194, 0.036071777, -0.027832031, 0.0435791, -0.0052604675, -0.0023422241, -0.018371582, 0.012359619, -0.010429382, ...
http://arxiv.org/abs/2304.00612v1
Eight Things to Know about Large Language Models
Samuel R. Bowman
cs.CL, cs.AI
The widespread public deployment of large language models (LLMs) in recent months has prompted a wave of new attention and engagement from advocates, policymakers, and scholars from many fields. This attention is a timely response to the many urgent questions that this technology raises, but it can sometimes miss impor...
http://arxiv.org/pdf/2304.00612v1
2023-04-02T20:03:27
2023-04-02T20:03:27
{ "title": [ -0.011314392, 0.0003786087, -0.05432129, -0.07684326, -0.0129470825, -0.022216797, -0.058288574, 0.02758789, -0.03878784, 0.04324341, -0.016052246, -0.008857727, 0.01725769, -0.03741455, -0.016830444, -0.001991272, 0.019332886, 0.035...
http://arxiv.org/abs/2304.00592v1
PK-Chat: Pointer Network Guided Knowledge Driven Generative Dialogue Model
Cheng Deng, Bo Tong, Luoyi Fu, Jiaxin Ding, Dexing Cao, Xinbing Wang, Chenghu Zhou
cs.CL, I.2.7; F.4.1
In the research of end-to-end dialogue systems, using real-world knowledge to generate natural, fluent, and human-like utterances with correct answers is crucial. However, domain-specific conversational dialogue systems may be incoherent and introduce erroneous external information to answer questions due to the out-of...
http://arxiv.org/pdf/2304.00592v1
2023-04-02T18:23:13
2023-04-02T18:23:13
{ "title": [ -0.029647827, -0.016067505, -0.044769287, -0.06719971, -0.015052795, 0.0076179504, -0.058624268, 0.014526367, 0.003440857, 0.027801514, 0.012969971, -0.015388489, -0.009559631, -0.035736084, -0.06713867, -0.00070142746, -0.0056648254, ...
http://arxiv.org/abs/2304.00557v1
Semi-supervised Neural Machine Translation with Consistency Regularization for Low-Resource Languages
Viet H. Pham, Thang M. Pham, Giang Nguyen, Long Nguyen, Dien Dinh
cs.CL, cs.LG
The advent of deep learning has led to a significant gain in machine translation. However, most of the studies required a large parallel dataset which is scarce and expensive to construct and even unavailable for some languages. This paper presents a simple yet effective method to tackle this problem for low-resource l...
http://arxiv.org/pdf/2304.00557v1
2023-04-02T15:24:08
2023-04-02T15:24:08
{ "title": [ -0.028411865, 0.0020637512, -0.025863647, -0.06317139, 0.004737854, 0.015106201, -0.06149292, 0.014183044, -0.016647339, 0.044647217, 0.0042381287, -0.007965088, 0.023162842, -0.0262146, -0.04156494, 0.006000519, -0.0069007874, 0.010...
http://arxiv.org/abs/2304.01228v2
Better Language Models of Code through Self-Improvement
Hung Quoc To, Nghi D. Q. Bui, Jin Guo, Tien N. Nguyen
cs.CL, cs.AI
Pre-trained language models for code (PLMCs) have gained attention in recent research. These models are pre-trained on large-scale datasets using multi-modal objectives. However, fine-tuning them requires extensive supervision and is limited by the size of the dataset provided. We aim to improve this issue by proposing...
http://arxiv.org/pdf/2304.01228v2
2023-05-10T02:36:40
2023-04-02T10:59:19
{ "title": [ -0.0006432533, -0.02709961, -0.053009033, -0.059417725, -0.04626465, -0.021530151, -0.04815674, 0.04660034, -0.018676758, 0.026275635, -0.0002477169, 0.0031337738, -0.0129776, -0.05380249, 0.007091522, 0.006790161, 0.024810791, -0.00...
http://arxiv.org/abs/2304.00483v2
A Data-centric Framework for Improving Domain-specific Machine Reading Comprehension Datasets
Iva Bojic, Josef Halim, Verena Suharman, Sreeja Tar, Qi Chwen Ong, Duy Phung, Mathieu Ravaut, Shafiq Joty, Josip Car
cs.CL
Low-quality data can cause downstream problems in high-stakes applications. Data-centric approach emphasizes on improving dataset quality to enhance model performance. High-quality datasets are needed for general-purpose Large Language Models (LLMs) training, as well as for domain-specific models, which are usually sma...
http://arxiv.org/pdf/2304.00483v2
2023-05-26T05:43:19
2023-04-02T08:26:38
{ "title": [ -0.043518066, -0.013801575, -0.036590576, -0.045410156, -0.018615723, -0.004421234, -0.04925537, 0.006023407, 0.015945435, 0.040985107, -0.035125732, -0.0056037903, 0.0015621185, 0.010513306, -0.07232666, -0.00094652176, -0.025558472, ...
http://arxiv.org/abs/2304.00468v1
Words that Wound: The Impact of Biased Language on News Sentiment and Stock Market Index
Wonseong Kim
cs.CL
This study investigates the impact of biased language, specifically 'Words that Wound,' on sentiment analysis in a dataset of 45,379 South Korean daily economic news articles. Using Word2Vec, cosine similarity, and an expanded lexicon, we analyzed the influence of these words on news titles' sentiment scores. Our findi...
http://arxiv.org/pdf/2304.00468v1
2023-04-02T06:45:18
2023-04-02T06:45:18
{ "title": [ -0.00030732155, 0.008460999, -0.04156494, -0.029129028, 0.024368286, -0.021911621, 0.020904541, -0.020187378, -0.029953003, 0.04324341, -0.035736084, 0.008674622, -0.0003168583, -0.0019054413, -0.05670166, -0.03286743, -0.00944519, 0...
http://arxiv.org/abs/2304.00457v3
LLMMaps -- A Visual Metaphor for Stratified Evaluation of Large Language Models
Patrik Puchert, Poonam Poonam, Christian van Onzenoodt, Timo Ropinski
cs.CL, cs.AI, cs.GR, cs.LG
Large Language Models (LLMs) have revolutionized natural language processing and demonstrated impressive capabilities in various tasks. Unfortunately, they are prone to hallucinations, where the model exposes incorrect or false information in its responses, which renders diligent evaluation approaches mandatory. While ...
http://arxiv.org/pdf/2304.00457v3
2023-10-12T11:35:35
2023-04-02T05:47:09
{ "title": [ -0.033111572, -0.010597229, -0.03390503, -0.03086853, -0.018630981, -0.010032654, -0.032714844, -0.015960693, 0.012054443, 0.010772705, -0.03805542, 0.0013217926, 0.0082473755, -0.017288208, -0.034423828, -0.009994507, -0.018341064, ...
http://arxiv.org/abs/2304.00416v1
Towards Healthy AI: Large Language Models Need Therapists Too
Baihan Lin, Djallel Bouneffouf, Guillermo Cecchi, Kush R. Varshney
cs.AI, cs.CL, cs.CY, cs.HC, cs.LG
Recent advances in large language models (LLMs) have led to the development of powerful AI chatbots capable of engaging in natural and human-like conversations. However, these chatbots can be potentially harmful, exhibiting manipulative, gaslighting, and narcissistic behaviors. We define Healthy AI to be safe, trustwor...
http://arxiv.org/pdf/2304.00416v1
2023-04-02T00:39:12
2023-04-02T00:39:12
{ "title": [ -0.004135132, -0.01146698, -0.03173828, -0.06542969, -0.0044555664, -0.008422852, -0.028717041, 0.008255005, -0.007633209, 0.020141602, -0.015716553, 0.00598526, 0.0072364807, -0.027832031, -0.023239136, 0.008308411, 0.008148193, 0.0...
http://arxiv.org/abs/2304.00363v1
Automatic Authorship Attribution in the Work of Tirso de Molina
Miguel Cavadas, Pablo Gamallo
cs.CL
Automatic Authorship Attribution (AAA) is the result of applying tools and techniques from Digital Humanities to authorship attribution studies. Through a quantitative and statistical approach this discipline can draw further conclusions about renowned authorship issues which traditional critics have been dealing with ...
http://arxiv.org/pdf/2304.00363v1
2023-04-01T18:05:14
2023-04-01T18:05:14
{ "title": [ -0.027191162, -0.0038414001, -0.027328491, -0.012901306, -0.03475952, -0.010292053, -0.02482605, -0.037078857, -0.007621765, 0.05130005, -0.008087158, 0.03253174, 0.017837524, -0.030380249, -0.011070251, 0.019546509, -0.011924744, 0....