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/2306.15541v1
Unleashing the Power of User Reviews: Exploring Airline Choices at Catania Airport, Italy
Vincenzo Miracula, Antonio Picone
cs.CL, cs.AI, cs.IR
This study aims to investigate the possible relationship between the mechanisms of social influence and the choice of airline, through the use of new tools, with the aim of understanding whether they can contribute to a better understanding of the factors influencing the decisions of consumers in the aviation sector. W...
http://arxiv.org/pdf/2306.15541v1
2023-06-27T15:10:57
2023-06-27T15:10:57
{ "title": [ 0.030319214, -0.009994507, -0.0057029724, 0.027008057, -0.018112183, -0.05621338, 0.023895264, -0.044281006, -0.037322998, 0.02758789, -0.04925537, 0.017837524, -0.03744507, -0.0135650635, -0.01159668, -0.03488159, 0.011497498, 0.030...
http://arxiv.org/abs/2306.15518v1
Paradigm Shift in Sustainability Disclosure Analysis: Empowering Stakeholders with CHATREPORT, a Language Model-Based Tool
Jingwei Ni, Julia Bingler, Chiara Colesanti-Senni, Mathias Kraus, Glen Gostlow, Tobias Schimanski, Dominik Stammbach, Saeid Ashraf Vaghefi, Qian Wang, Nicolas Webersinke, Tobias Wekhof, Tingyu Yu, Markus Leippold
cs.CL
This paper introduces a novel approach to enhance Large Language Models (LLMs) with expert knowledge to automate the analysis of corporate sustainability reports by benchmarking them against the Task Force for Climate-Related Financial Disclosures (TCFD) recommendations. Corporate sustainability reports are crucial in ...
http://arxiv.org/pdf/2306.15518v1
2023-06-27T14:46:47
2023-06-27T14:46:47
{ "title": [ -0.026611328, -0.02468872, -0.08123779, -0.07336426, -0.016647339, -0.0045661926, -0.06640625, 0.045410156, 0.002275467, 0.028274536, -0.030075073, -0.010894775, -0.000016272068, -0.01448822, -0.04925537, -0.033203125, -0.032165527, ...
http://arxiv.org/abs/2306.15498v1
Using Large Language Models to Provide Explanatory Feedback to Human Tutors
Jionghao Lin, Danielle R. Thomas, Feifei Han, Shivang Gupta, Wei Tan, Ngoc Dang Nguyen, Kenneth R. Koedinger
cs.CL, cs.AI, cs.HC
Research demonstrates learners engaging in the process of producing explanations to support their reasoning, can have a positive impact on learning. However, providing learners real-time explanatory feedback often presents challenges related to classification accuracy, particularly in domain-specific environments, cont...
http://arxiv.org/pdf/2306.15498v1
2023-06-27T14:19:12
2023-06-27T14:19:12
{ "title": [ -0.029327393, -0.00756073, -0.01776123, -0.03289795, -0.024108887, -0.004383087, -0.04711914, -0.024475098, -0.0014533997, 0.028762817, -0.02583313, -0.0030078888, 0.010910034, -0.0058021545, -0.037384033, -0.0068969727, -0.010429382, ...
http://arxiv.org/abs/2306.15430v1
KnowPrefix-Tuning: A Two-Stage Prefix-Tuning Framework for Knowledge-Grounded Dialogue Generation
Jiaqi Bai, Zhao Yan, Jian Yang, Xinnian Liang, Hongcheng Guo, Zhoujun Li
cs.CL
Existing knowledge-grounded conversation systems generate responses typically in a retrieve-then-generate manner. They require a large knowledge base and a strong knowledge retrieval component, which is time- and resource-consuming. In this paper, we address the challenge by leveraging the inherent knowledge encoded in...
http://arxiv.org/pdf/2306.15430v1
2023-06-27T12:38:49
2023-06-27T12:38:49
{ "title": [ -0.034820557, -0.031311035, -0.05041504, -0.08917236, -0.019302368, 0.0064926147, -0.07885742, 0.026977539, -0.0026226044, 0.017669678, 0.012886047, 0.008224487, -0.022872925, -0.03387451, -0.0423584, -0.015960693, -0.012062073, 0.03...
http://arxiv.org/abs/2306.15425v1
Phase Space Analysis of Cardiac Spectra
Onder Pekcan, Taner Arsan
physics.med-ph, cs.CL, cs.CV
Cardiac diseases are one of the main reasons of mortality in modern, industrialized societies, and they cause high expenses in public health systems. Therefore, it is important to develop analytical methods to improve cardiac diagnostics. Electric activity of heart was first modeled by using a set of nonlinear differen...
http://arxiv.org/pdf/2306.15425v1
2023-06-27T12:37:21
2023-06-27T12:37:21
{ "title": [ 0.009712219, 0.008995056, -0.015213013, -0.011871338, 0.011367798, -0.012626648, 0.008720398, -0.040374756, 0.006843567, 0.058013916, 0.00137043, 0.013305664, 0.03842163, -0.021652222, -0.052825928, 0.029037476, 0.038970947, 0.010330...
http://arxiv.org/abs/2306.15399v1
Quality Estimation of Machine Translated Texts based on Direct Evidence from Training Data
Vibhuti Kumari, Narayana Murthy Kavi
cs.CL
Current Machine Translation systems achieve very good results on a growing variety of language pairs and data sets. However, it is now well known that they produce fluent translation outputs that often can contain important meaning errors. Quality Estimation task deals with the estimation of quality of translations pro...
http://arxiv.org/pdf/2306.15399v1
2023-06-27T11:52:28
2023-06-27T11:52:28
{ "title": [ -0.042175293, -0.0027694702, 0.0006842613, -0.010192871, -0.024932861, 0.004753113, -0.03050232, -0.032348633, -0.008995056, 0.037963867, -0.008758545, 0.0065078735, 0.012481689, -0.017166138, -0.045562744, 0.00642395, -0.017349243, ...
http://arxiv.org/abs/2307.00021v1
SAHAAYAK 2023 -- the Multi Domain Bilingual Parallel Corpus of Sanskrit to Hindi for Machine Translation
Vishvajitsinh Bakrola, Jitendra Nasariwala
cs.CL
The data article presents the large bilingual parallel corpus of low-resourced language pair Sanskrit-Hindi, named SAHAAYAK 2023. The corpus contains total of 1.5M sentence pairs between Sanskrit and Hindi. To make the universal usability of the corpus and to make it balanced, data from multiple domain has been incorpo...
http://arxiv.org/pdf/2307.00021v1
2023-06-27T11:06:44
2023-06-27T11:06:44
{ "title": [ -0.032562256, -0.01600647, -0.05911255, -0.03555298, -0.03265381, -0.013801575, -0.049713135, 0.020339966, -0.030166626, 0.046295166, 0.040618896, 0.01838684, 0.0028018951, -0.046691895, 0.008766174, 0.0033912659, 0.013977051, 0.0017...
http://arxiv.org/abs/2306.15376v1
Exploiting Pseudo Future Contexts for Emotion Recognition in Conversations
Yinyi Wei, Shuaipeng Liu, Hailei Yan, Wei Ye, Tong Mo, Guanglu Wan
cs.CL
With the extensive accumulation of conversational data on the Internet, emotion recognition in conversations (ERC) has received increasing attention. Previous efforts of this task mainly focus on leveraging contextual and speaker-specific features, or integrating heterogeneous external commonsense knowledge. Among them...
http://arxiv.org/pdf/2306.15376v1
2023-06-27T10:51:02
2023-06-27T10:51:02
{ "title": [ 0.020584106, -0.0052108765, -0.03527832, -0.07989502, 0.019592285, -0.01398468, -0.01512146, 0.011306763, 0.040740967, 0.035614014, -0.009887695, -0.02708435, 0.019058228, -0.029510498, -0.0748291, -0.0008049011, -0.013160706, 0.0217...
http://arxiv.org/abs/2306.15364v1
The Architecture of a Biologically Plausible Language Organ
Daniel Mitropolsky, Christos H. Papadimitriou
cs.CL, q-bio.NC
We present a simulated biologically plausible language organ, made up of stylized but realistic neurons, synapses, brain areas, plasticity, and a simplified model of sensory perception. We show through experiments that this model succeeds in an important early step in language acquisition: the learning of nouns, verbs,...
http://arxiv.org/pdf/2306.15364v1
2023-06-27T10:25:22
2023-06-27T10:25:22
{ "title": [ -0.0037517548, -0.023986816, -0.004432678, -0.044921875, -0.0647583, 0.00068330765, -0.053649902, -0.00008326769, -0.01398468, 0.031341553, -0.019805908, -0.02798462, 0.00026893616, -0.0446167, -0.023498535, -0.0025177002, 0.016311646, ...
http://arxiv.org/abs/2306.15354v3
3D-Speaker: A Large-Scale Multi-Device, Multi-Distance, and Multi-Dialect Corpus for Speech Representation Disentanglement
Siqi Zheng, Luyao Cheng, Yafeng Chen, Hui Wang, Qian Chen
cs.CL, cs.SD, eess.AS
Disentangling uncorrelated information in speech utterances is a crucial research topic within speech community. Different speech-related tasks focus on extracting distinct speech representations while minimizing the affects of other uncorrelated information. We present a large-scale speech corpus to facilitate the res...
http://arxiv.org/pdf/2306.15354v3
2023-09-25T02:36:41
2023-06-27T10:09:43
{ "title": [ 0.013015747, 0.025543213, -0.07348633, -0.078552246, 0.028869629, -0.009788513, -0.044158936, 0.012809753, -0.003042221, 0.04864502, -0.027633667, 0.0050735474, 0.056030273, -0.027252197, -0.016159058, 0.009895325, -0.006351471, 0.04...
http://arxiv.org/abs/2306.15334v1
Understanding Client Reactions in Online Mental Health Counseling
Anqi Li, Lizhi Ma, Yaling Mei, Hongliang He, Shuai Zhang, Huachuan Qiu, Zhenzhong Lan
cs.CL
Communication success relies heavily on reading participants' reactions. Such feedback is especially important for mental health counselors, who must carefully consider the client's progress and adjust their approach accordingly. However, previous NLP research on counseling has mainly focused on studying counselors' in...
http://arxiv.org/pdf/2306.15334v1
2023-06-27T09:39:54
2023-06-27T09:39:54
{ "title": [ 0.028823853, 0.012649536, 0.0025482178, -0.02229309, 0.0066719055, -0.018356323, 0.0047836304, -0.02772522, 0.008453369, 0.020019531, -0.01852417, -0.018371582, 0.032318115, -0.045013428, -0.0043907166, -0.0154418945, 0.024230957, -0...
http://arxiv.org/abs/2306.15298v1
Gender Bias in BERT -- Measuring and Analysing Biases through Sentiment Rating in a Realistic Downstream Classification Task
Sophie Jentzsch, Cigdem Turan
cs.CL, cs.AI, cs.CY, cs.LG
Pretrained language models are publicly available and constantly finetuned for various real-life applications. As they become capable of grasping complex contextual information, harmful biases are likely increasingly intertwined with those models. This paper analyses gender bias in BERT models with two main contributio...
http://arxiv.org/pdf/2306.15298v1
2023-06-27T08:36:35
2023-06-27T08:36:35
{ "title": [ -0.01461792, -0.008026123, -0.04763794, -0.016281128, -0.019805908, -0.038146973, -0.018493652, -0.0033550262, 0.017913818, 0.040405273, -0.04296875, -0.025482178, 0.023269653, -0.014831543, -0.059143066, -0.027893066, -0.03994751, 0...
http://arxiv.org/abs/2306.15273v1
IDOL: Indicator-oriented Logic Pre-training for Logical Reasoning
Zihang Xu, Ziqing Yang, Yiming Cui, Shijin Wang
cs.CL, cs.AI
In the field of machine reading comprehension (MRC), existing systems have surpassed the average performance of human beings in many tasks like SQuAD. However, there is still a long way to go when it comes to logical reasoning. Although some methods for it have been put forward, they either are designed in a quite comp...
http://arxiv.org/pdf/2306.15273v1
2023-06-27T07:57:42
2023-06-27T07:57:42
{ "title": [ -0.017623901, -0.012496948, -0.043823242, -0.07373047, 0.0033988953, 0.00059461594, -0.03704834, -0.0005836487, -0.0010147095, 0.017730713, -0.029159546, 0.005722046, -0.009140015, -0.022659302, -0.022735596, -0.011566162, -0.0141067505,...
http://arxiv.org/abs/2306.15268v1
Can Pretrained Language Models Derive Correct Semantics from Corrupt Subwords under Noise?
Xinzhe Li, Ming Liu, Shang Gao
cs.CL
For Pretrained Language Models (PLMs), their susceptibility to noise has recently been linked to subword segmentation. However, it is unclear which aspects of segmentation affect their understanding. This study assesses the robustness of PLMs against various disrupted segmentation caused by noise. An evaluation framewo...
http://arxiv.org/pdf/2306.15268v1
2023-06-27T07:51:01
2023-06-27T07:51:01
{ "title": [ -0.0146484375, 0.004863739, -0.017349243, -0.030776978, -0.03363037, -0.0027046204, -0.057434082, -0.011871338, -0.007751465, 0.03652954, -0.039276123, -0.010383606, 0.031173706, -0.019332886, -0.023590088, -0.021469116, 0.0026454926, ...
http://arxiv.org/abs/2306.15261v1
A Survey on Out-of-Distribution Evaluation of Neural NLP Models
Xinzhe Li, Ming Liu, Shang Gao, Wray Buntine
cs.CL
Adversarial robustness, domain generalization and dataset biases are three active lines of research contributing to out-of-distribution (OOD) evaluation on neural NLP models. However, a comprehensive, integrated discussion of the three research lines is still lacking in the literature. In this survey, we 1) compare the...
http://arxiv.org/pdf/2306.15261v1
2023-06-27T07:44:25
2023-06-27T07:44:25
{ "title": [ -0.016113281, -0.0011539459, -0.014289856, -0.011505127, -0.0016613007, -0.005657196, -0.0118255615, -0.022644043, 0.016357422, 0.022598267, -0.029754639, -0.0012025833, -0.011688232, -0.0046577454, -0.063964844, -0.0012903214, -0.037567...
http://arxiv.org/abs/2306.15255v1
GroundNLQ @ Ego4D Natural Language Queries Challenge 2023
Zhijian Hou, Lei Ji, Difei Gao, Wanjun Zhong, Kun Yan, Chao Li, Wing-Kwong Chan, Chong-Wah Ngo, Nan Duan, Mike Zheng Shou
cs.CV, cs.CL
In this report, we present our champion solution for Ego4D Natural Language Queries (NLQ) Challenge in CVPR 2023. Essentially, to accurately ground in a video, an effective egocentric feature extractor and a powerful grounding model are required. Motivated by this, we leverage a two-stage pre-training strategy to train...
http://arxiv.org/pdf/2306.15255v1
2023-06-27T07:27:52
2023-06-27T07:27:52
{ "title": [ -0.019073486, -0.01751709, -0.03125, 0.0013723373, -0.018829346, -0.00308609, -0.01878357, -0.0013980865, -0.0060195923, 0.04650879, -0.005455017, -0.008102417, 0.045135498, -0.047027588, 0.00066280365, -0.0074005127, -0.00089359283, ...
http://arxiv.org/abs/2306.15253v3
MindDial: Belief Dynamics Tracking with Theory-of-Mind Modeling for Situated Neural Dialogue Generation
Shuwen Qiu, Song-Chun Zhu, Zilong Zheng
cs.CL
Humans talk in free-form while negotiating the expressed meanings or common ground. Despite the impressive conversational abilities of the large generative language models, they do not consider the individual differences in contextual understanding in a shared situated environment. In this work, we propose MindDial, a ...
http://arxiv.org/pdf/2306.15253v3
2023-07-03T06:29:41
2023-06-27T07:24:32
{ "title": [ -0.007980347, -0.017990112, -0.026504517, -0.035736084, -0.014602661, -0.0005507469, -0.019729614, -0.00667572, -0.017303467, 0.013328552, 0.002494812, -0.01637268, 0.00674057, -0.048980713, -0.023345947, -0.013885498, -0.019927979, ...
http://arxiv.org/abs/2306.15245v3
C-PMI: Conditional Pointwise Mutual Information for Turn-level Dialogue Evaluation
Liliang Ren, Mankeerat Sidhu, Qi Zeng, Revanth Gangi Reddy, Heng Ji, ChengXiang Zhai
cs.CL
Existing reference-free turn-level evaluation metrics for chatbots inadequately capture the interaction between the user and the system. Consequently, they often correlate poorly with human evaluations. To address this issue, we propose a novel model-agnostic approach that leverages Conditional Pointwise Mutual Informa...
http://arxiv.org/pdf/2306.15245v3
2023-09-01T16:11:40
2023-06-27T06:58:03
{ "title": [ -0.01033783, -0.011367798, -0.02835083, -0.051727295, -0.0158844, -0.0048179626, -0.029220581, 0.0009698868, -0.005809784, 0.017028809, -0.0158844, 0.0018548965, -0.025878906, -0.015045166, -0.047546387, -0.03111267, -0.025665283, 0....
http://arxiv.org/abs/2306.15231v1
Emulating Reader Behaviors for Fake News Detection
Junwei Yin, Min Gao, Kai Shu, Zehua Zhao, Yinqiu Huang, Jia Wang
cs.CL
The wide dissemination of fake news has affected our lives in many aspects, making fake news detection important and attracting increasing attention. Existing approaches make substantial contributions in this field by modeling news from a single-modal or multi-modal perspective. However, these modal-based methods can r...
http://arxiv.org/pdf/2306.15231v1
2023-06-27T06:14:24
2023-06-27T06:14:24
{ "title": [ 0.0025157928, 0.0068130493, -0.012634277, -0.012931824, 0.008323669, -0.01399231, 0.019424438, -0.03149414, -0.023880005, 0.03414917, -0.021987915, 0.0017910004, -0.0002207756, -0.0043144226, -0.040130615, -0.0074386597, -0.0289917, ...
http://arxiv.org/abs/2306.15222v1
Learning to Rank in Generative Retrieval
Yongqi Li, Nan Yang, Liang Wang, Furu Wei, Wenjie Li
cs.CL, cs.AI, cs.IR
Generative retrieval is a promising new paradigm in text retrieval that generates identifier strings of relevant passages as the retrieval target. This paradigm leverages powerful generation models and represents a new paradigm distinct from traditional learning-to-rank methods. However, despite its rapid development, ...
http://arxiv.org/pdf/2306.15222v1
2023-06-27T05:48:14
2023-06-27T05:48:14
{ "title": [ 0.014274597, 0.005821228, -0.026397705, -0.04611206, 0.018112183, -0.010307312, -0.0070991516, -0.023162842, 0.026672363, 0.026885986, -0.01625061, 0.012176514, -0.024505615, -0.010643005, -0.051239014, 0.012084961, -0.007247925, 0.0...
http://arxiv.org/abs/2306.15171v1
Reducing the gap between streaming and non-streaming Transducer-based ASR by adaptive two-stage knowledge distillation
Haitao Tang, Yu Fu, Lei Sun, Jiabin Xue, Dan Liu, Yongchao Li, Zhiqiang Ma, Minghui Wu, Jia Pan, Genshun Wan, Ming'en Zhao
cs.CL
Transducer is one of the mainstream frameworks for streaming speech recognition. There is a performance gap between the streaming and non-streaming transducer models due to limited context. To reduce this gap, an effective way is to ensure that their hidden and output distributions are consistent, which can be achieved...
http://arxiv.org/pdf/2306.15171v1
2023-06-27T03:11:21
2023-06-27T03:11:21
{ "title": [ 0.00025081635, 0.0013189316, -0.059906006, -0.10632324, 0.012138367, -0.004043579, -0.060699463, 0.029296875, -0.0046043396, 0.028717041, -0.030593872, -0.012748718, 0.0015821457, -0.02482605, -0.017852783, -0.015640259, 0.023086548, ...
http://arxiv.org/abs/2306.15705v1
On the Universal Adversarial Perturbations for Efficient Data-free Adversarial Detection
Songyang Gao, Shihan Dou, Qi Zhang, Xuanjing Huang, Jin Ma, Ying Shan
cs.CL, cs.LG
Detecting adversarial samples that are carefully crafted to fool the model is a critical step to socially-secure applications. However, existing adversarial detection methods require access to sufficient training data, which brings noteworthy concerns regarding privacy leakage and generalizability. In this work, we val...
http://arxiv.org/pdf/2306.15705v1
2023-06-27T02:54:07
2023-06-27T02:54:07
{ "title": [ 0.000101447105, 0.0082473755, -0.018692017, -0.04058838, 0.030059814, -0.012870789, 0.009025574, -0.036102295, -0.02243042, 0.023590088, 0.0058059692, 0.033294678, -0.028808594, -0.019180298, -0.02015686, 0.008285522, 0.002532959, 0....
http://arxiv.org/abs/2306.15164v1
DSRM: Boost Textual Adversarial Training with Distribution Shift Risk Minimization
Songyang Gao, Shihan Dou, Yan Liu, Xiao Wang, Qi Zhang, Zhongyu Wei, Jin Ma, Ying Shan
cs.CL, cs.LG
Adversarial training is one of the best-performing methods in improving the robustness of deep language models. However, robust models come at the cost of high time consumption, as they require multi-step gradient ascents or word substitutions to obtain adversarial samples. In addition, these generated samples are defi...
http://arxiv.org/pdf/2306.15164v1
2023-06-27T02:46:08
2023-06-27T02:46:08
{ "title": [ -0.029205322, -0.0010766983, -0.038024902, -0.03970337, 0.011978149, 0.008712769, -0.021774292, -0.015731812, -0.022598267, 0.026641846, -0.00617218, 0.03024292, -0.010215759, -0.010650635, -0.030059814, 0.0015249252, -0.00995636, 0....
http://arxiv.org/abs/2306.15162v2
YouTube-ASL: A Large-Scale, Open-Domain American Sign Language-English Parallel Corpus
David Uthus, Garrett Tanzer, Manfred Georg
cs.CL, cs.CV
Machine learning for sign languages is bottlenecked by data. In this paper, we present YouTube-ASL, a large-scale, open-domain corpus of American Sign Language (ASL) videos and accompanying English captions drawn from YouTube. With ~1000 hours of videos and >2500 unique signers, YouTube-ASL is ~3x as large and has ~10x...
http://arxiv.org/pdf/2306.15162v2
2023-10-26T22:57:49
2023-06-27T02:44:07
{ "title": [ -0.03555298, 0.0129776, -0.06854248, -0.061462402, -0.00033211708, 0.0009036064, -0.05368042, 0.004589081, -0.007209778, 0.03475952, -0.026916504, 0.01751709, 0.010269165, -0.031799316, -0.012763977, -0.011978149, -0.013679504, 0.049...
http://arxiv.org/abs/2306.15123v2
Investigating Cross-Domain Behaviors of BERT in Review Understanding
Albert Lu, Meng Jiang
cs.CL, cs.CE
Review score prediction requires review text understanding, a critical real-world application of natural language processing. Due to dissimilar text domains in product reviews, a common practice is fine-tuning BERT models upon reviews of differing domains. However, there has not yet been an empirical study of cross-dom...
http://arxiv.org/pdf/2306.15123v2
2023-06-28T01:26:40
2023-06-27T00:23:35
{ "title": [ -0.0029678345, -0.02407837, 0.002790451, -0.033203125, -0.028289795, -0.008102417, -0.008476257, -0.0073394775, 0.0008049011, 0.032043457, -0.042541504, 0.017959595, 0.030853271, -0.0018587112, -0.026000977, 0.021713257, -0.022140503, ...
http://arxiv.org/abs/2306.15112v1
FeedbackMap: a tool for making sense of open-ended survey responses
Doug Beeferman, Nabeel Gillani
cs.CL
Analyzing open-ended survey responses is a crucial yet challenging task for social scientists, non-profit organizations, and educational institutions, as they often face the trade-off between obtaining rich data and the burden of reading and coding textual responses. This demo introduces FeedbackMap, a web-based tool t...
http://arxiv.org/pdf/2306.15112v1
2023-06-26T23:38:24
2023-06-26T23:38:24
{ "title": [ -0.020751953, 0.006919861, -0.009109497, -0.0044784546, -0.024398804, -0.010520935, -0.03086853, -0.05883789, 0.0024204254, -0.0031032562, -0.03427124, 0.013847351, -0.061553955, -0.020568848, -0.042114258, -0.031219482, 0.02760315, ...
http://arxiv.org/abs/2306.15103v1
Structured Dialogue Discourse Parsing
Ta-Chung Chi, Alexander I. Rudnicky
cs.CL
Dialogue discourse parsing aims to uncover the internal structure of a multi-participant conversation by finding all the discourse~\emph{links} and corresponding~\emph{relations}. Previous work either treats this task as a series of independent multiple-choice problems, in which the link existence and relations are dec...
http://arxiv.org/pdf/2306.15103v1
2023-06-26T22:51:01
2023-06-26T22:51:01
{ "title": [ -0.018005371, -0.006839752, -0.037475586, -0.07910156, -0.018997192, -0.0060272217, -0.06878662, 0.027389526, -0.015090942, 0.02368164, -0.027755737, -0.012649536, -0.0016307831, -0.03274536, -0.010375977, -0.011627197, -0.014892578, ...
http://arxiv.org/abs/2306.15091v1
Understanding In-Context Learning via Supportive Pretraining Data
Xiaochuang Han, Daniel Simig, Todor Mihaylov, Yulia Tsvetkov, Asli Celikyilmaz, Tianlu Wang
cs.CL
In-context learning (ICL) improves language models' performance on a variety of NLP tasks by simply demonstrating a handful of examples at inference time. It is not well understood why ICL ability emerges, as the model has never been specifically trained on such demonstrations. Unlike prior work that explores implicit ...
http://arxiv.org/pdf/2306.15091v1
2023-06-26T22:14:04
2023-06-26T22:14:04
{ "title": [ 0.023025513, 0.0034561157, -0.04156494, -0.067993164, 0.018478394, -0.033050537, -0.013130188, -0.017837524, -0.018814087, 0.014137268, -0.035095215, 0.018127441, -0.03125, -0.023971558, -0.018859863, -0.01689148, 0.002292633, 0.0410...
http://arxiv.org/abs/2306.15087v1
WinoQueer: A Community-in-the-Loop Benchmark for Anti-LGBTQ+ Bias in Large Language Models
Virginia K. Felkner, Ho-Chun Herbert Chang, Eugene Jang, Jonathan May
cs.CL, cs.CY
We present WinoQueer: a benchmark specifically designed to measure whether large language models (LLMs) encode biases that are harmful to the LGBTQ+ community. The benchmark is community-sourced, via application of a novel method that generates a bias benchmark from a community survey. We apply our benchmark to several...
http://arxiv.org/pdf/2306.15087v1
2023-06-26T22:07:33
2023-06-26T22:07:33
{ "title": [ 0.011329651, 0.005584717, -0.025299072, -0.032440186, 0.00737381, -0.024291992, 0.008987427, -0.010635376, 0.018615723, 0.049224854, -0.036895752, 0.006061554, 0.0579834, -0.016601562, -0.03475952, 0.0018987656, -0.033721924, 0.05242...
http://arxiv.org/abs/2306.15063v2
Pretraining task diversity and the emergence of non-Bayesian in-context learning for regression
Allan Raventós, Mansheej Paul, Feng Chen, Surya Ganguli
cs.LG, cs.AI, cs.CL
Pretrained transformers exhibit the remarkable ability of in-context learning (ICL): they can learn tasks from just a few examples provided in the prompt without updating any weights. This raises a foundational question: can ICL solve fundamentally $\textit{new}$ tasks that are very different from those seen during pre...
http://arxiv.org/pdf/2306.15063v2
2023-11-08T18:12:03
2023-06-26T21:05:20
{ "title": [ 0.016983032, 0.016235352, -0.05392456, -0.048339844, 0.02508545, -0.04321289, -0.007587433, -0.012023926, -0.041381836, 0.014968872, -0.01625061, 0.03945923, -0.043426514, -0.024902344, -0.005214691, -0.019882202, 0.01701355, 0.01062...
http://arxiv.org/abs/2306.15006v1
DNABERT-2: Efficient Foundation Model and Benchmark For Multi-Species Genome
Zhihan Zhou, Yanrong Ji, Weijian Li, Pratik Dutta, Ramana Davuluri, Han Liu
q-bio.GN, cs.AI, cs.CE, cs.CL
Decoding the linguistic intricacies of the genome is a crucial problem in biology, and pre-trained foundational models such as DNABERT and Nucleotide Transformer have made significant strides in this area. Existing works have largely hinged on k-mer, fixed-length permutations of A, T, C, and G, as the token of the geno...
http://arxiv.org/pdf/2306.15006v1
2023-06-26T18:43:46
2023-06-26T18:43:46
{ "title": [ -0.0670166, -0.013580322, -0.05050659, -0.07409668, 0.0012750626, 0.009277344, -0.07098389, 0.019439697, 0.010231018, 0.028152466, -0.00075531006, 0.012451172, -0.0014162064, -0.014167786, -0.04510498, 0.024505615, -0.0053482056, 0.0...
http://arxiv.org/abs/2306.14899v1
FunQA: Towards Surprising Video Comprehension
Binzhu Xie, Sicheng Zhang, Zitang Zhou, Bo Li, Yuanhan Zhang, Jack Hessel, Jingkang Yang, Ziwei Liu
cs.CV, cs.AI, cs.CL, cs.MM
Surprising videos, e.g., funny clips, creative performances, or visual illusions, attract significant attention. Enjoyment of these videos is not simply a response to visual stimuli; rather, it hinges on the human capacity to understand (and appreciate) commonsense violations depicted in these videos. We introduce FunQ...
http://arxiv.org/pdf/2306.14899v1
2023-06-26T17:59:55
2023-06-26T17:59:55
{ "title": [ -0.036376953, -0.015914917, -0.028198242, -0.04159546, -0.008529663, -0.0027389526, -0.020599365, -0.014167786, 0.024475098, 0.018600464, -0.02355957, -0.01965332, -0.011367798, -0.012962341, -0.05154419, -0.008987427, -0.00491333, 0...
http://arxiv.org/abs/2306.14898v3
InterCode: Standardizing and Benchmarking Interactive Coding with Execution Feedback
John Yang, Akshara Prabhakar, Karthik Narasimhan, Shunyu Yao
cs.CL, cs.LG, cs.SE
Humans write code in a fundamentally interactive manner and rely on constant execution feedback to correct errors, resolve ambiguities, and decompose tasks. While LLMs have recently exhibited promising coding capabilities, current coding benchmarks mostly consider a static instruction-to-code sequence transduction proc...
http://arxiv.org/pdf/2306.14898v3
2023-10-30T17:52:18
2023-06-26T17:59:50
{ "title": [ -0.045562744, -0.018035889, -0.021652222, -0.04800415, -0.007797241, 0.004257202, -0.045288086, -0.014717102, -0.00007534027, 0.028579712, -0.02961731, 0.023956299, -0.02758789, -0.0070228577, -0.04525757, -0.009384155, 0.007865906, ...
http://arxiv.org/abs/2306.14893v1
LongCoder: A Long-Range Pre-trained Language Model for Code Completion
Daya Guo, Canwen Xu, Nan Duan, Jian Yin, Julian McAuley
cs.SE, cs.AI, cs.CL, cs.LG
In this paper, we introduce a new task for code completion that focuses on handling long code input and propose a sparse Transformer model, called LongCoder, to address this task. LongCoder employs a sliding window mechanism for self-attention and introduces two types of globally accessible tokens - bridge tokens and m...
http://arxiv.org/pdf/2306.14893v1
2023-06-26T17:59:24
2023-06-26T17:59:24
{ "title": [ -0.014907837, -0.022842407, -0.032409668, -0.07775879, -0.0006632805, 0.0034694672, -0.054718018, 0.022521973, -0.002937317, 0.036193848, -0.017944336, 0.010269165, 0.015312195, -0.0126571655, -0.03552246, 0.0039596558, -0.01927185, ...
http://arxiv.org/abs/2306.14870v1
Composing Parameter-Efficient Modules with Arithmetic Operations
Jinghan Zhang, Shiqi Chen, Junteng Liu, Junxian He
cs.CL
As an efficient alternative to conventional full finetuning, parameter-efficient finetuning (PEFT) is becoming the prevailing method to adapt pretrained language models. In PEFT, a lightweight module is learned on each dataset while the underlying pretrained language model remains unchanged, resulting in multiple compa...
http://arxiv.org/pdf/2306.14870v1
2023-06-26T17:33:21
2023-06-26T17:33:21
{ "title": [ -0.018295288, -0.0068359375, -0.054718018, -0.123413086, 0.010787964, 0.020507812, -0.080078125, 0.045684814, -0.015464783, 0.029342651, 0.01058197, -0.0028896332, -0.033233643, -0.027679443, -0.024276733, 0.0044937134, -0.009017944, ...
http://arxiv.org/abs/2306.14939v2
The Art of Embedding Fusion: Optimizing Hate Speech Detection
Mohammad Aflah Khan, Neemesh Yadav, Mohit Jain, Sanyam Goyal
cs.CL, cs.LG
Hate speech detection is a challenging natural language processing task that requires capturing linguistic and contextual nuances. Pre-trained language models (PLMs) offer rich semantic representations of text that can improve this task. However there is still limited knowledge about ways to effectively combine represe...
http://arxiv.org/pdf/2306.14939v2
2023-10-08T07:11:44
2023-06-26T17:30:35
{ "title": [ 0.018966675, 0.011703491, -0.028656006, -0.043670654, 0.03286743, -0.023361206, 0.010971069, -0.021011353, -0.018722534, 0.033569336, -0.026046753, 0.01890564, 0.0041007996, -0.0079574585, -0.038635254, -0.0105896, -0.016204834, 0.06...
http://arxiv.org/abs/2306.14866v1
Enriching the NArabizi Treebank: A Multifaceted Approach to Supporting an Under-Resourced Language
Riabi Arij, Mahamdi Menel, Seddah Djamé
cs.CL
In this paper we address the scarcity of annotated data for NArabizi, a Romanized form of North African Arabic used mostly on social media, which poses challenges for Natural Language Processing (NLP). We introduce an enriched version of NArabizi Treebank (Seddah et al., 2020) with three main contributions: the additio...
http://arxiv.org/pdf/2306.14866v1
2023-06-26T17:27:31
2023-06-26T17:27:31
{ "title": [ -0.016159058, -0.015731812, -0.03427124, -0.10235596, -0.018325806, 0.026412964, -0.11456299, 0.0007004738, 0.010475159, 0.052764893, -0.04006958, 0.001452446, 0.01979065, -0.017288208, -0.06488037, 0.0041389465, -0.004272461, 0.0363...
http://arxiv.org/abs/2306.15447v1
Are aligned neural networks adversarially aligned?
Nicholas Carlini, Milad Nasr, Christopher A. Choquette-Choo, Matthew Jagielski, Irena Gao, Anas Awadalla, Pang Wei Koh, Daphne Ippolito, Katherine Lee, Florian Tramer, Ludwig Schmidt
cs.CL, cs.AI, cs.CR, cs.LG
Large language models are now tuned to align with the goals of their creators, namely to be "helpful and harmless." These models should respond helpfully to user questions, but refuse to answer requests that could cause harm. However, adversarial users can construct inputs which circumvent attempts at alignment. In thi...
http://arxiv.org/pdf/2306.15447v1
2023-06-26T17:18:44
2023-06-26T17:18:44
{ "title": [ -0.024963379, 0.024673462, -0.03564453, 0.010147095, -0.0023555756, -0.018218994, 0.0021457672, -0.035339355, -0.049865723, 0.050079346, 0.001789093, 0.0021438599, -0.014320374, -0.022613525, -0.026351929, -0.012458801, 0.0126571655, ...
http://arxiv.org/abs/2306.14828v1
HonestBait: Forward References for Attractive but Faithful Headline Generation
Chih-Yao Chen, Dennis Wu, Lun-Wei Ku
cs.CL, cs.AI
Current methods for generating attractive headlines often learn directly from data, which bases attractiveness on the number of user clicks and views. Although clicks or views do reflect user interest, they can fail to reveal how much interest is raised by the writing style and how much is due to the event or topic its...
http://arxiv.org/pdf/2306.14828v1
2023-06-26T16:34:37
2023-06-26T16:34:37
{ "title": [ -0.009216309, -0.0066947937, -0.024475098, -0.061035156, 0.0019340515, -0.002401352, -0.02658081, -0.035003662, -0.024490356, 0.035583496, -0.009002686, 0.030822754, -0.0132369995, -0.013725281, -0.03744507, 0.0023727417, -0.020248413, ...
http://arxiv.org/abs/2306.14827v1
Vietnamese multi-document summary using subgraph selection approach -- VLSP 2022 AbMuSu Shared Task
Huu-Thin Nguyen, Tam Doan Thanh, Cam-Van Thi Nguyen
cs.CL, cs.AI
Document summarization is a task to generate afluent, condensed summary for a document, andkeep important information. A cluster of documents serves as the input for multi-document summarizing (MDS), while the cluster summary serves as the output. In this paper, we focus on transforming the extractive MDS problem into ...
http://arxiv.org/pdf/2306.14827v1
2023-06-26T16:34:02
2023-06-26T16:34:02
{ "title": [ -0.022079468, -0.02659607, -0.055114746, -0.039916992, -0.039764404, -0.008255005, -0.074279785, 0.029220581, -0.021987915, 0.024520874, 0.0075683594, 0.0013608932, 0.028152466, -0.04736328, 0.017669678, -0.014678955, 0.0056610107, 0...
http://arxiv.org/abs/2306.14824v3
Kosmos-2: Grounding Multimodal Large Language Models to the World
Zhiliang Peng, Wenhui Wang, Li Dong, Yaru Hao, Shaohan Huang, Shuming Ma, Furu Wei
cs.CL, cs.CV
We introduce Kosmos-2, a Multimodal Large Language Model (MLLM), enabling new capabilities of perceiving object descriptions (e.g., bounding boxes) and grounding text to the visual world. Specifically, we represent refer expressions as links in Markdown, i.e., ``[text span](bounding boxes)'', where object descriptions ...
http://arxiv.org/pdf/2306.14824v3
2023-07-13T05:41:34
2023-06-26T16:32:47
{ "title": [ -0.04449463, -0.00944519, -0.047454834, -0.06842041, 0.001745224, 0.009521484, -0.07305908, 0.008293152, 0.005455017, 0.032592773, -0.0076904297, -0.006526947, 0.018447876, -0.01626587, -0.04147339, 0.013435364, -0.013618469, 0.05337...
http://arxiv.org/abs/2306.14822v1
Label-Aware Hyperbolic Embeddings for Fine-grained Emotion Classification
Chih-Yao Chen, Tun-Min Hung, Yi-Li Hsu, Lun-Wei Ku
cs.CL, cs.AI
Fine-grained emotion classification (FEC) is a challenging task. Specifically, FEC needs to handle subtle nuance between labels, which can be complex and confusing. Most existing models only address text classification problem in the euclidean space, which we believe may not be the optimal solution as labels of close s...
http://arxiv.org/pdf/2306.14822v1
2023-06-26T16:28:33
2023-06-26T16:28:33
{ "title": [ 0.017440796, 0.0016336441, -0.036346436, -0.0473938, 0.016418457, -0.018844604, 0.0060310364, -0.0031166077, 0.037902832, 0.022964478, -0.021820068, -0.008415222, 0.009002686, -0.017105103, -0.050231934, 0.0012578964, 0.0067977905, 0...
http://arxiv.org/abs/2306.14806v1
A Positive-Unlabeled Metric Learning Framework for Document-Level Relation Extraction with Incomplete Labeling
Ye Wang, Huazheng Pan, Tao Zhang, Wen Wu, Wenxin Hu
cs.CL, cs.AI
The goal of document-level relation extraction (RE) is to identify relations between entities that span multiple sentences. Recently, incomplete labeling in document-level RE has received increasing attention, and some studies have used methods such as positive-unlabeled learning to tackle this issue, but there is stil...
http://arxiv.org/pdf/2306.14806v1
2023-06-26T16:05:59
2023-06-26T16:05:59
{ "title": [ -0.0038261414, 0.001039505, -0.057739258, -0.08111572, 0.0029888153, 0.0009584427, -0.061462402, 0.0054244995, 0.02859497, 0.016586304, -0.005256653, 0.006832123, 0.0009703636, -0.016601562, -0.035949707, 0.010543823, 0.0096206665, 0...
http://arxiv.org/abs/2306.14795v2
MotionGPT: Human Motion as a Foreign Language
Biao Jiang, Xin Chen, Wen Liu, Jingyi Yu, Gang Yu, Tao Chen
cs.CV, cs.CL, cs.GR
Though the advancement of pre-trained large language models unfolds, the exploration of building a unified model for language and other multi-modal data, such as motion, remains challenging and untouched so far. Fortunately, human motion displays a semantic coupling akin to human language, often perceived as a form of ...
http://arxiv.org/pdf/2306.14795v2
2023-07-20T03:39:19
2023-06-26T15:53:02
{ "title": [ 0.0026607513, -0.018600464, -0.029464722, -0.018615723, -0.0068588257, -0.025039673, -0.0134887695, -0.030212402, -0.017730713, 0.034851074, -0.016021729, 0.009162903, 0.027328491, -0.04348755, -0.0009775162, 0.005760193, 0.011100769, ...
http://arxiv.org/abs/2306.14790v2
Automatic Assessment of Divergent Thinking in Chinese Language with TransDis: A Transformer-Based Language Model Approach
Tianchen Yang, Qifan Zhang, Zhaoyang Sun, Yubo Hou
cs.CL, stat.AP
Language models have been increasingly popular for automatic creativity assessment, generating semantic distances to objectively measure the quality of creative ideas. However, there is currently a lack of an automatic assessment system for evaluating creative ideas in the Chinese language. To address this gap, we deve...
http://arxiv.org/pdf/2306.14790v2
2023-10-19T00:55:28
2023-06-26T15:48:05
{ "title": [ -0.0049934387, -0.00207901, 0.0107803345, 0.005962372, -0.033355713, -0.0013542175, -0.02444458, -0.030044556, 0.002506256, 0.014785767, -0.040924072, -0.012840271, 0.021118164, -0.031311035, -0.050231934, -0.027252197, -0.030014038, ...
http://arxiv.org/abs/2306.14764v2
Uncovering Political Hate Speech During Indian Election Campaign: A New Low-Resource Dataset and Baselines
Farhan Ahmad Jafri, Mohammad Aman Siddiqui, Surendrabikram Thapa, Kritesh Rauniyar, Usman Naseem, Imran Razzak
cs.CL
The detection of hate speech in political discourse is a critical issue, and this becomes even more challenging in low-resource languages. To address this issue, we introduce a new dataset named IEHate, which contains 11,457 manually annotated Hindi tweets related to the Indian Assembly Election Campaign from November ...
http://arxiv.org/pdf/2306.14764v2
2023-06-27T16:55:14
2023-06-26T15:17:54
{ "title": [ 0.031219482, -0.002210617, -0.04458618, -0.070495605, 0.030349731, -0.031311035, -0.0059776306, 0.015472412, -0.028640747, 0.022949219, -0.014549255, 0.018295288, -0.0050621033, -0.01802063, -0.03543091, -0.013694763, -0.014503479, 0...
http://arxiv.org/abs/2306.14754v1
Représentation graphique de la langue des signes française et édition logicielle
Michael Filhol, Thomas von Ascheberg
cs.CL
Cet article propose une m\'ethode pour d\'efinir une forme graphique \'editable standardis\'ee pour les langues des signes, ainsi qu'une proposition "AZVD" et un \'editeur logiciel associ\'e. Inspir\'ee d'une part par les r\'egularit\'es observ\'ees dans les pratiques spontan\'ees de locuteurs pratiquant la sch\'ematis...
http://arxiv.org/pdf/2306.14754v1
2023-06-26T15:09:51
2023-06-26T15:09:51
{ "title": [ -0.027328491, -0.00043845177, -0.033294678, -0.038635254, -0.017089844, 0.018981934, -0.06512451, 0.006614685, -0.024017334, 0.03692627, 0.00046491623, -0.025390625, 0.025238037, -0.050689697, 0.0073547363, -0.001783371, 0.0118637085, ...
http://arxiv.org/abs/2306.14728v1
Learn over Past, Evolve for Future: Forecasting Temporal Trends for Fake News Detection
Beizhe Hu, Qiang Sheng, Juan Cao, Yongchun Zhu, Danding Wang, Zhengjia Wang, Zhiwei Jin
cs.CL, cs.AI, cs.SI
Fake news detection has been a critical task for maintaining the health of the online news ecosystem. However, very few existing works consider the temporal shift issue caused by the rapidly-evolving nature of news data in practice, resulting in significant performance degradation when training on past data and testing...
http://arxiv.org/pdf/2306.14728v1
2023-06-26T14:29:05
2023-06-26T14:29:05
{ "title": [ 0.011886597, -0.006942749, -0.042633057, -0.05706787, 0.015655518, -0.020126343, 0.00075674057, -0.003730774, -0.018707275, 0.040802002, -0.026412964, 0.020019531, -0.012687683, -0.0025291443, -0.05380249, -0.02406311, -0.028900146, ...
http://arxiv.org/abs/2306.14704v3
Ontology Enrichment from Texts: A Biomedical Dataset for Concept Discovery and Placement
Hang Dong, Jiaoyan Chen, Yuan He, Ian Horrocks
cs.CL, I.2.7
Mentions of new concepts appear regularly in texts and require automated approaches to harvest and place them into Knowledge Bases (KB), e.g., ontologies and taxonomies. Existing datasets suffer from three issues, (i) mostly assuming that a new concept is pre-discovered and cannot support out-of-KB mention discovery; (...
http://arxiv.org/pdf/2306.14704v3
2023-09-01T15:26:45
2023-06-26T13:54:47
{ "title": [ -0.0072402954, 0.012252808, -0.04977417, -0.068237305, 0.026733398, -0.0031700134, -0.043060303, -0.021148682, 0.007091522, 0.025253296, -0.022369385, 0.016082764, -0.0012016296, -0.009513855, -0.032226562, 0.0011901855, -0.024551392, ...
http://arxiv.org/abs/2306.14696v1
How About Kind of Generating Hedges using End-to-End Neural Models?
Alafate Abulimiti, Chloé Clavel, Justine Cassell
cs.CL, cs.AI, 68T50, I.2.7
Hedging is a strategy for softening the impact of a statement in conversation. In reducing the strength of an expression, it may help to avoid embarrassment (more technically, ``face threat'') to one's listener. For this reason, it is often found in contexts of instruction, such as tutoring. In this work, we develop a ...
http://arxiv.org/pdf/2306.14696v1
2023-06-26T13:43:06
2023-06-26T13:43:06
{ "title": [ 0.027069092, 0.005256653, -0.023513794, 0.012641907, -0.002937317, -0.022583008, 0.036834717, -0.020370483, -0.027435303, 0.02255249, 0.036102295, 0.0011358261, -0.013786316, -0.052337646, 0.0022697449, -0.0042037964, -0.010002136, 0...
http://arxiv.org/abs/2306.14633v1
JSEEGraph: Joint Structured Event Extraction as Graph Parsing
Huiling You, Samia Touileb, Lilja Øvrelid
cs.CL
We propose a graph-based event extraction framework JSEEGraph that approaches the task of event extraction as general graph parsing in the tradition of Meaning Representation Parsing. It explicitly encodes entities and events in a single semantic graph, and further has the flexibility to encode a wider range of additio...
http://arxiv.org/pdf/2306.14633v1
2023-06-26T12:12:54
2023-06-26T12:12:54
{ "title": [ -0.03277588, -0.013626099, -0.052856445, -0.10699463, 0.017700195, 0.021514893, -0.08288574, -0.0026111603, -0.0029449463, 0.0046691895, 0.0029582977, 0.0070495605, -0.030456543, -0.032043457, -0.036895752, 0.0033988953, -0.013389587, ...
http://arxiv.org/abs/2306.14933v1
Integrating Bidirectional Long Short-Term Memory with Subword Embedding for Authorship Attribution
Abiodun Modupe, Turgay Celik, Vukosi Marivate, Oludayo O. Olugbara
cs.CL, cs.AI, cs.LG
The problem of unveiling the author of a given text document from multiple candidate authors is called authorship attribution. Manifold word-based stylistic markers have been successfully used in deep learning methods to deal with the intrinsic problem of authorship attribution. Unfortunately, the performance of word-b...
http://arxiv.org/pdf/2306.14933v1
2023-06-26T11:35:47
2023-06-26T11:35:47
{ "title": [ 0.011306763, 0.010063171, -0.027633667, -0.041229248, 0.003025055, -0.009521484, -0.016998291, -0.032806396, -0.011978149, 0.01663208, -0.012939453, 0.028320312, 0.0055732727, -0.018737793, 0.0016908646, -0.008720398, -0.023345947, 0...
http://arxiv.org/abs/2306.14610v1
SugarCrepe: Fixing Hackable Benchmarks for Vision-Language Compositionality
Cheng-Yu Hsieh, Jieyu Zhang, Zixian Ma, Aniruddha Kembhavi, Ranjay Krishna
cs.CV, cs.CL, cs.LG
In the last year alone, a surge of new benchmarks to measure compositional understanding of vision-language models have permeated the machine learning ecosystem. Given an image, these benchmarks probe a model's ability to identify its associated caption amongst a set of compositional distractors. Surprisingly, we find ...
http://arxiv.org/pdf/2306.14610v1
2023-06-26T11:35:22
2023-06-26T11:35:22
{ "title": [ -0.022766113, -0.015159607, -0.029785156, -0.06854248, -0.0021305084, -0.0070495605, -0.059020996, -0.0028572083, 0.0029773712, 0.036071777, -0.020507812, 0.014251709, -0.0060310364, 0.0009899139, -0.05657959, -0.0027198792, 0.011230469,...
http://arxiv.org/abs/2306.14608v1
Factorised Speaker-environment Adaptive Training of Conformer Speech Recognition Systems
Jiajun Deng, Guinan Li, Xurong Xie, Zengrui Jin, Mingyu Cui, Tianzi Wang, Shujie Hu, Mengzhe Geng, Xunying Liu
eess.AS, cs.CL
Rich sources of variability in natural speech present significant challenges to current data intensive speech recognition technologies. To model both speaker and environment level diversity, this paper proposes a novel Bayesian factorised speaker-environment adaptive training and test time adaptation approach for Confo...
http://arxiv.org/pdf/2306.14608v1
2023-06-26T11:32:05
2023-06-26T11:32:05
{ "title": [ -0.015808105, 0.014602661, -0.06793213, -0.067871094, 0.007286072, -0.011711121, -0.03955078, 0.024337769, -0.03173828, 0.033111572, -0.011764526, 0.016311646, -0.014175415, -0.030090332, -0.025848389, -0.009880066, -0.016311646, 0.0...
http://arxiv.org/abs/2306.14592v1
Transfer Learning across Several Centuries: Machine and Historian Integrated Method to Decipher Royal Secretary's Diary
Sojung Lucia Kim, Taehong Jang, Joonmo Ahn, Hyungil Lee, Jaehyuk Lee
cs.CL, cs.DL
A named entity recognition and classification plays the first and foremost important role in capturing semantics in data and anchoring in translation as well as downstream study for history. However, NER in historical text has faced challenges such as scarcity of annotated corpus, multilanguage variety, various noise, ...
http://arxiv.org/pdf/2306.14592v1
2023-06-26T11:00:35
2023-06-26T11:00:35
{ "title": [ -0.020431519, 0.005355835, -0.0423584, 0.006465912, -0.018508911, -0.036590576, -0.0043182373, -0.01914978, 0.00022876263, 0.019882202, -0.034606934, 0.006450653, 0.015914917, -0.02178955, -0.01108551, -0.022888184, 0.0067253113, 0.0...
http://arxiv.org/abs/2306.14583v1
Exploring the Robustness of Large Language Models for Solving Programming Problems
Atsushi Shirafuji, Yutaka Watanobe, Takumi Ito, Makoto Morishita, Yuki Nakamura, Yusuke Oda, Jun Suzuki
cs.CL, cs.AI, cs.SE
Using large language models (LLMs) for source code has recently gained attention. LLMs, such as Transformer-based models like Codex and ChatGPT, have been shown to be highly capable of solving a wide range of programming problems. However, the extent to which LLMs understand problem descriptions and generate programs a...
http://arxiv.org/pdf/2306.14583v1
2023-06-26T10:48:50
2023-06-26T10:48:50
{ "title": [ -0.028747559, -0.014503479, -0.030593872, -0.058807373, 0.008460999, 0.00019109249, -0.065979004, 0.0026226044, 0.0072669983, 0.018707275, -0.012893677, -0.0021648407, -0.0140686035, -0.016647339, -0.046325684, -0.0015392303, -0.01707458...
http://arxiv.org/abs/2306.14580v1
TransERR: Translation-based Knowledge Graph Completion via Efficient Relation Rotation
Jiang Li, Xiangdong Su
cs.CL, cs.AI
This paper presents translation-based knowledge graph completion method via efficient relation rotation (TransERR), a straightforward yet effective alternative to traditional translation-based knowledge graph completion models. Different from the previous translation-based models, TransERR encodes knowledge graphs in t...
http://arxiv.org/pdf/2306.14580v1
2023-06-26T10:45:16
2023-06-26T10:45:16
{ "title": [ -0.037231445, -0.03164673, -0.061523438, -0.10839844, -0.004951477, 0.029067993, -0.08282471, 0.02947998, 0.0022068024, 0.034942627, -0.03491211, 0.03253174, 0.013702393, 0.009132385, -0.03967285, 0.001531601, 0.004699707, 0.02574157...
http://arxiv.org/abs/2306.14565v3
Mitigating Hallucination in Large Multi-Modal Models via Robust Instruction Tuning
Fuxiao Liu, Kevin Lin, Linjie Li, Jianfeng Wang, Yaser Yacoob, Lijuan Wang
cs.CV, cs.AI, cs.CE, cs.CL, cs.MM
Despite the promising progress in multi-modal tasks, current large multi-modal models (LMMs) are prone to hallucinating inconsistent descriptions with respect to the associated image and human instructions. This paper addresses this issue by introducing the first large and diverse visual instruction tuning dataset, nam...
http://arxiv.org/pdf/2306.14565v3
2023-09-29T16:02:28
2023-06-26T10:26:33
{ "title": [ -0.021636963, 0.02116394, -0.010147095, -0.031402588, -0.013053894, 0.0050811768, -0.060455322, -0.04647827, 0.010948181, 0.045654297, -0.014221191, 0.015701294, 0.01096344, -0.010414124, -0.048339844, 0.0074806213, -0.005027771, 0.0...
http://arxiv.org/abs/2306.14517v1
Cross-Lingual Cross-Age Group Adaptation for Low-Resource Elderly Speech Emotion Recognition
Samuel Cahyawijaya, Holy Lovenia, Willy Chung, Rita Frieske, Zihan Liu, Pascale Fung
cs.CL, cs.SD, eess.AS
Speech emotion recognition plays a crucial role in human-computer interactions. However, most speech emotion recognition research is biased toward English-speaking adults, which hinders its applicability to other demographic groups in different languages and age groups. In this work, we analyze the transferability of e...
http://arxiv.org/pdf/2306.14517v1
2023-06-26T08:48:08
2023-06-26T08:48:08
{ "title": [ 0.0015096664, 0.006942749, -0.047668457, -0.0592041, 0.015342712, -0.016021729, -0.044281006, 0.007663727, 0.021224976, 0.04135132, -0.032318115, -0.01689148, 0.023529053, -0.012809753, -0.05709839, -0.0035762787, 0.003206253, 0.0047...
http://arxiv.org/abs/2306.14514v2
Data-Driven Approach for Formality-Sensitive Machine Translation: Language-Specific Handling and Synthetic Data Generation
Seugnjun Lee, Hyeonseok Moon, Chanjun Park, Heuiseok Lim
cs.CL, cs.AI
In this paper, we introduce a data-driven approach for Formality-Sensitive Machine Translation (FSMT) that caters to the unique linguistic properties of four target languages. Our methodology centers on two core strategies: 1) language-specific data handling, and 2) synthetic data generation using large-scale language ...
http://arxiv.org/pdf/2306.14514v2
2023-06-27T06:59:28
2023-06-26T08:45:47
{ "title": [ -0.032104492, -0.0066223145, -0.043701172, -0.07836914, -0.009994507, -0.002023697, -0.06347656, 0.019332886, -0.018539429, 0.039031982, 0.0011224747, 0.013717651, 0.017486572, -0.025161743, -0.027664185, 0.018371582, -0.0073394775, ...
http://arxiv.org/abs/2306.14470v1
Knowledge Graph-Augmented Korean Generative Commonsense Reasoning
Dahyun Jung, Jaehyung Seo, Jaewook Lee, Chanjun Park, Heuiseok Lim
cs.CL, cs.AI
Generative commonsense reasoning refers to the task of generating acceptable and logical assumptions about everyday situations based on commonsense understanding. By utilizing an existing dataset such as Korean CommonGen, language generation models can learn commonsense reasoning specific to the Korean language. Howeve...
http://arxiv.org/pdf/2306.14470v1
2023-06-26T07:23:47
2023-06-26T07:23:47
{ "title": [ -0.01612854, -0.032409668, -0.04373169, -0.053497314, -0.051116943, -0.0029277802, -0.05697632, 0.009681702, 0.0062065125, 0.028625488, -0.02406311, 0.008049011, -0.0084991455, -0.024841309, -0.015388489, -0.0110321045, -0.007255554, ...
http://arxiv.org/abs/2306.14457v1
Fauno: The Italian Large Language Model that will leave you senza parole!
Andrea Bacciu, Giovanni Trappolini, Andrea Santilli, Emanuele Rodolà, Fabrizio Silvestri
cs.CL
This paper presents Fauno, the first and largest open-source Italian conversational Large Language Model (LLM). Our goal with Fauno is to democratize the study of LLMs in Italian, demonstrating that obtaining a fine-tuned conversational bot with a single GPU is possible. In addition, we release a collection of datasets...
http://arxiv.org/pdf/2306.14457v1
2023-06-26T07:00:38
2023-06-26T07:00:38
{ "title": [ -0.03451538, -0.02746582, -0.045166016, 0.015655518, -0.056365967, -0.0037517548, -0.031463623, -0.009590149, -0.009346008, 0.020614624, 0.0020980835, -0.04135132, 0.010803223, -0.053619385, -0.021728516, -0.053771973, 0.0029754639, ...
http://arxiv.org/abs/2306.14422v2
The Singing Voice Conversion Challenge 2023
Wen-Chin Huang, Lester Phillip Violeta, Songxiang Liu, Jiatong Shi, Tomoki Toda
cs.SD, cs.CL, eess.AS
We present the latest iteration of the voice conversion challenge (VCC) series, a bi-annual scientific event aiming to compare and understand different voice conversion (VC) systems based on a common dataset. This year we shifted our focus to singing voice conversion (SVC), thus named the challenge the Singing Voice Co...
http://arxiv.org/pdf/2306.14422v2
2023-07-06T08:17:31
2023-06-26T05:04:58
{ "title": [ -0.021972656, -0.0085372925, -0.031173706, -0.02255249, -0.029327393, -0.004333496, -0.006778717, 0.008682251, -0.005142212, 0.048919678, 0.015556335, -0.022445679, 0.034973145, -0.054748535, 0.027938843, -0.0134887695, 0.010284424, ...
http://arxiv.org/abs/2306.14393v1
Constraint-aware and Ranking-distilled Token Pruning for Efficient Transformer Inference
Junyan Li, Li Lyna Zhang, Jiahang Xu, Yujing Wang, Shaoguang Yan, Yunqing Xia, Yuqing Yang, Ting Cao, Hao Sun, Weiwei Deng, Qi Zhang, Mao Yang
cs.CL
Deploying pre-trained transformer models like BERT on downstream tasks in resource-constrained scenarios is challenging due to their high inference cost, which grows rapidly with input sequence length. In this work, we propose a constraint-aware and ranking-distilled token pruning method ToP, which selectively removes ...
http://arxiv.org/pdf/2306.14393v1
2023-06-26T03:06:57
2023-06-26T03:06:57
{ "title": [ -0.0070343018, -0.007751465, -0.032562256, -0.08935547, 0.0038261414, 0.029251099, -0.08001709, 0.010726929, 0.009277344, 0.0021305084, -0.044830322, 0.002128601, 0.04849243, -0.026367188, 0.015701294, -0.015464783, -0.00083065033, 0...
http://arxiv.org/abs/2306.14377v1
Synthetic Alone: Exploring the Dark Side of Synthetic Data for Grammatical Error Correction
Chanjun Park, Seonmin Koo, Seolhwa Lee, Jaehyung Seo, Sugyeong Eo, Hyeonseok Moon, Heuiseok Lim
cs.CL, cs.AI
Data-centric AI approach aims to enhance the model performance without modifying the model and has been shown to impact model performance positively. While recent attention has been given to data-centric AI based on synthetic data, due to its potential for performance improvement, data-centric AI has long been exclusiv...
http://arxiv.org/pdf/2306.14377v1
2023-06-26T01:40:28
2023-06-26T01:40:28
{ "title": [ -0.022476196, 0.010147095, -0.036956787, -0.05493164, 0.015594482, -0.0052490234, -0.03253174, -0.0061416626, -0.011917114, 0.022216797, -0.046142578, 0.013130188, -0.016342163, -0.003736496, -0.026138306, -0.010772705, -0.03866577, ...
http://arxiv.org/abs/2306.14374v1
Transcending Traditional Boundaries: Leveraging Inter-Annotator Agreement (IAA) for Enhancing Data Management Operations (DMOps)
Damrin Kim, NamHyeok Kim, Chanjun Park, Harksoo Kim
cs.CL, cs.AI
This paper presents a novel approach of leveraging Inter-Annotator Agreement (IAA), traditionally used for assessing labeling consistency, to optimize Data Management Operations (DMOps). We advocate for the use of IAA in predicting the labeling quality of individual annotators, leading to cost and time efficiency in da...
http://arxiv.org/pdf/2306.14374v1
2023-06-26T01:33:58
2023-06-26T01:33:58
{ "title": [ 0.016525269, 0.023376465, -0.055023193, -0.0836792, 0.017150879, -0.022399902, -0.040496826, 0.0025939941, 0.007030487, 0.02079773, -0.029510498, -0.00258255, 0.011405945, -0.018478394, -0.02204895, -0.03201294, -0.0072402954, 0.0436...
http://arxiv.org/abs/2306.14373v1
Inter-Annotator Agreement in the Wild: Uncovering Its Emerging Roles and Considerations in Real-World Scenarios
NamHyeok Kim, Chanjun Park
cs.CL, cs.AI
Inter-Annotator Agreement (IAA) is commonly used as a measure of label consistency in natural language processing tasks. However, in real-world scenarios, IAA has various roles and implications beyond its traditional usage. In this paper, we not only consider IAA as a measure of consistency but also as a versatile tool...
http://arxiv.org/pdf/2306.14373v1
2023-06-26T01:28:58
2023-06-26T01:28:58
{ "title": [ -0.019241333, -0.0032157898, -0.057525635, -0.06323242, -0.0021533966, -0.0143585205, -0.05368042, 0.004211426, -0.019973755, 0.022628784, -0.013587952, 0.004550934, -0.0074653625, -0.025436401, -0.029083252, -0.00983429, -0.007217407, ...
http://arxiv.org/abs/2306.14321v1
RobuT: A Systematic Study of Table QA Robustness Against Human-Annotated Adversarial Perturbations
Yilun Zhao, Chen Zhao, Linyong Nan, Zhenting Qi, Wenlin Zhang, Xiangru Tang, Boyu Mi, Dragomir Radev
cs.CL, cs.AI
Despite significant progress having been made in question answering on tabular data (Table QA), it's unclear whether, and to what extent existing Table QA models are robust to task-specific perturbations, e.g., replacing key question entities or shuffling table columns. To systematically study the robustness of Table Q...
http://arxiv.org/pdf/2306.14321v1
2023-06-25T19:23:21
2023-06-25T19:23:21
{ "title": [ 0.018310547, 0.01322937, -0.02029419, -0.034576416, 0.013374329, -0.026657104, -0.017074585, -0.024108887, -0.024551392, 0.054473877, -0.012168884, 0.033843994, -0.0041275024, -0.009292603, -0.053833008, -0.01247406, -0.0020332336, 0...
http://arxiv.org/abs/2306.14310v1
Addressing Cold Start Problem for End-to-end Automatic Speech Scoring
Jungbae Park, Seungtaek Choi
cs.CL, cs.SD, eess.AS
Integrating automatic speech scoring/assessment systems has become a critical aspect of second-language speaking education. With self-supervised learning advancements, end-to-end speech scoring approaches have exhibited promising results. However, this study highlights the significant decrease in the performance of spe...
http://arxiv.org/pdf/2306.14310v1
2023-06-25T18:48:21
2023-06-25T18:48:21
{ "title": [ -0.012229919, 0.011367798, -0.023712158, -0.06604004, 0.013999939, 0.015213013, -0.04534912, -0.0041122437, -0.008522034, 0.031463623, -0.010559082, 0.0033283234, -0.0181427, -0.030380249, -0.045776367, 0.00006908178, -0.010108948, 0...
http://arxiv.org/abs/2306.14308v1
Let's Do a Thought Experiment: Using Counterfactuals to Improve Moral Reasoning
Xiao Ma, Swaroop Mishra, Ahmad Beirami, Alex Beutel, Jilin Chen
cs.CL, cs.AI
Language models still struggle on moral reasoning, despite their impressive performance in many other tasks. In particular, the Moral Scenarios task in MMLU (Multi-task Language Understanding) is among the worst performing tasks for many language models, including GPT-3. In this work, we propose a new prompting framewo...
http://arxiv.org/pdf/2306.14308v1
2023-06-25T18:40:43
2023-06-25T18:40:43
{ "title": [ 0.034484863, 0.0034313202, -0.019012451, 0.0068092346, -0.007003784, -0.042541504, 0.026779175, -0.047088623, -0.016540527, 0.009216309, -0.021514893, -0.014587402, -0.016647339, -0.04046631, -0.006034851, -0.025787354, -0.010871887, ...
http://arxiv.org/abs/2306.14264v1
Visual Question Answering in Remote Sensing with Cross-Attention and Multimodal Information Bottleneck
Jayesh Songara, Shivam Pande, Shabnam Choudhury, Biplab Banerjee, Rajbabu Velmurugan
cs.CV, cs.CL, cs.LG
In this research, we deal with the problem of visual question answering (VQA) in remote sensing. While remotely sensed images contain information significant for the task of identification and object detection, they pose a great challenge in their processing because of high dimensionality, volume and redundancy. Furthe...
http://arxiv.org/pdf/2306.14264v1
2023-06-25T15:09:21
2023-06-25T15:09:21
{ "title": [ -0.020980835, 0.0021018982, -0.025131226, -0.021011353, -0.012084961, -0.008110046, -0.013717651, -0.0064086914, 0.04321289, 0.012077332, -0.064453125, -0.006160736, -0.018569946, -0.0017681122, -0.05493164, -0.024749756, 0.015975952, ...
http://arxiv.org/abs/2306.14222v1
Unveiling the Potential of Sentiment: Can Large Language Models Predict Chinese Stock Price Movements?
Haohan Zhang, Fengrui Hua, Chengjin Xu, Jian Guo, Hao Kong, Ruiting Zuo
cs.CL, cs.AI, q-fin.ST
The rapid advancement of Large Language Models (LLMs) has led to extensive discourse regarding their potential to boost the return of quantitative stock trading strategies. This discourse primarily revolves around harnessing the remarkable comprehension capabilities of LLMs to extract sentiment factors which facilitate...
http://arxiv.org/pdf/2306.14222v1
2023-06-25T12:08:44
2023-06-25T12:08:44
{ "title": [ -0.027862549, 0.008804321, -0.030075073, -0.0046463013, -0.0026607513, -0.018341064, -0.009468079, -0.026046753, 0.00014770031, 0.040985107, -0.030075073, -0.025680542, 0.017730713, -0.015731812, -0.052764893, -0.016281128, 0.003063202, ...
http://arxiv.org/abs/2306.14203v2
Stance Prediction and Analysis of Twitter data : A case study of Ghana 2020 Presidential Elections
Shester Gueuwou, Rose-Mary Owusuaa Mensah Gyening
cs.CL
On December 7, 2020, Ghanaians participated in the polls to determine their president for the next four years. To gain insights from this presidential election, we conducted stance analysis (which is not always equivalent to sentiment analysis) to understand how Twitter, a popular social media platform, reflected the o...
http://arxiv.org/pdf/2306.14203v2
2023-06-27T18:39:53
2023-06-25T10:54:24
{ "title": [ 0.019165039, -0.012809753, -0.010269165, -0.076538086, -0.0034542084, 0.008453369, -0.032592773, 0.00472641, -0.0121536255, 0.042938232, -0.044403076, 0.025863647, 0.005771637, -0.0037727356, -0.04525757, -0.0026454926, -0.0034255981, ...
http://arxiv.org/abs/2306.14192v3
$α$-$β$-Factorization and the Binary Case of Simon's Congruence
Pamela Fleischmann, Jonas Höfer, Annika Huch, Dirk Nowotka
math.CO, cs.CL
In 1991 H\'ebrard introduced a factorization of words that turned out to be a powerful tool for the investigation of a word's scattered factors (also known as (scattered) subwords or subsequences). Based on this, first Karandikar and Schnoebelen introduced the notion of $k$-richness and later on Barker et al. the notio...
http://arxiv.org/pdf/2306.14192v3
2023-09-11T06:47:27
2023-06-25T10:16:49
{ "title": [ -0.038024902, 0.015571594, -0.060333252, -0.111328125, 0.036743164, 0.01763916, -0.053253174, 0.029174805, -0.0032577515, 0.03515625, 0.03137207, -0.017120361, 0.008514404, -0.039031982, -0.061401367, 0.028915405, -0.024215698, 0.047...
http://arxiv.org/abs/2306.14176v1
Sentence-level Event Detection without Triggers via Prompt Learning and Machine Reading Comprehension
Tongtao Ling, Lei Chen, Huangxu Sheng, Zicheng Cai, Hai-Lin Liu
cs.CL
The traditional way of sentence-level event detection involves two important subtasks: trigger identification and trigger classifications, where the identified event trigger words are used to classify event types from sentences. However, trigger classification highly depends on abundant annotated trigger words and the ...
http://arxiv.org/pdf/2306.14176v1
2023-06-25T09:03:56
2023-06-25T09:03:56
{ "title": [ 0.0048294067, 0.00061273575, -0.0256958, -0.07385254, 0.0074424744, -0.0008020401, -0.05239868, -0.016036987, 0.025512695, 0.008804321, -0.011207581, 0.0006747246, 0.0072746277, -0.012542725, -0.043182373, 0.02029419, -0.010185242, 0...
http://arxiv.org/abs/2306.14152v1
Low-Rank Prune-And-Factorize for Language Model Compression
Siyu Ren, Kenny Q. Zhu
cs.CL
The components underpinning PLMs -- large weight matrices -- were shown to bear considerable redundancy. Matrix factorization, a well-established technique from matrix theory, has been utilized to reduce the number of parameters in PLM. However, it fails to retain satisfactory performance under moderate to high compres...
http://arxiv.org/pdf/2306.14152v1
2023-06-25T07:38:43
2023-06-25T07:38:43
{ "title": [ -0.05899048, -0.01638794, -0.047912598, -0.07006836, -0.008239746, 0.03036499, -0.07965088, 0.036895752, 0.0050811768, 0.02494812, 0.0075149536, -0.0017652512, 0.008178711, -0.032958984, -0.03756714, 0.015167236, 0.012023926, -0.0149...
http://arxiv.org/abs/2306.14149v1
SciMRC: Multi-perspective Scientific Machine Reading Comprehension
Xiao Zhang, Heqi Zheng, Yuxiang Nie, Heyan Huang, Xian-Ling Mao
cs.CL
Scientific machine reading comprehension (SMRC) aims to understand scientific texts through interactions with humans by given questions. As far as we know, there is only one dataset focused on exploring full-text scientific machine reading comprehension. However, the dataset has ignored the fact that different readers ...
http://arxiv.org/pdf/2306.14149v1
2023-06-25T07:25:14
2023-06-25T07:25:14
{ "title": [ -0.050231934, -0.019439697, -0.03137207, -0.04437256, -0.023712158, 0.010749817, -0.052978516, 0.0017318726, 0.021102905, 0.04043579, -0.019683838, -0.009002686, 0.0031909943, -0.017562866, -0.066223145, 0.011940002, -0.023773193, 0....
http://arxiv.org/abs/2306.14145v1
DSE-TTS: Dual Speaker Embedding for Cross-Lingual Text-to-Speech
Sen Liu, Yiwei Guo, Chenpeng Du, Xie Chen, Kai Yu
cs.SD, cs.CL, eess.AS
Although high-fidelity speech can be obtained for intralingual speech synthesis, cross-lingual text-to-speech (CTTS) is still far from satisfactory as it is difficult to accurately retain the speaker timbres(i.e. speaker similarity) and eliminate the accents from their first language(i.e. nativeness). In this paper, we...
http://arxiv.org/pdf/2306.14145v1
2023-06-25T06:46:36
2023-06-25T06:46:36
{ "title": [ -0.0036640167, 0.0022907257, -0.024215698, -0.06384277, 0.020263672, 0.0070724487, -0.033416748, 0.008201599, 0.0056152344, 0.050598145, -0.015945435, 0.015419006, 0.027862549, -0.005344391, -0.040740967, 0.018341064, -0.013496399, 0...
http://arxiv.org/abs/2306.14135v1
Interpretable Neural Embeddings with Sparse Self-Representation
Minxue Xia, Hao Zhu
cs.CL, cs.LG
Interpretability benefits the theoretical understanding of representations. Existing word embeddings are generally dense representations. Hence, the meaning of latent dimensions is difficult to interpret. This makes word embeddings like a black-box and prevents them from being human-readable and further manipulation. M...
http://arxiv.org/pdf/2306.14135v1
2023-06-25T05:57:01
2023-06-25T05:57:01
{ "title": [ -0.009429932, 0.010002136, -0.037384033, -0.045166016, 0.010154724, -0.0019464493, -0.032836914, -0.005748749, 0.01751709, 0.0057868958, -0.009742737, -0.013404846, -0.029724121, -0.02319336, -0.04434204, -0.0062789917, 0.020996094, ...
http://arxiv.org/abs/2306.14122v3
Chain-of-Thought Prompt Distillation for Multimodal Named Entity Recognition and Multimodal Relation Extraction
Feng Chen, Yujian Feng
cs.CL, cs.CV
Multimodal Named Entity Recognition (MNER) and Multimodal Relation Extraction (MRE) necessitate the fundamental reasoning capacity for intricate linguistic and multimodal comprehension. In this study, we explore distilling the reasoning ability of large language models (LLMs) into a more compact student model by genera...
http://arxiv.org/pdf/2306.14122v3
2023-08-23T05:04:58
2023-06-25T04:33:56
{ "title": [ 0.0031967163, 0.00945282, -0.060272217, -0.074401855, 0.014167786, -0.009841919, -0.05618286, 0.008964539, 0.03201294, 0.007446289, -0.03668213, 0.00085163116, 0.0035648346, -0.022949219, -0.032043457, -0.0034217834, -0.013076782, 0....
http://arxiv.org/abs/2306.14115v2
Towards Trustworthy Explanation: On Causal Rationalization
Wenbo Zhang, Tong Wu, Yunlong Wang, Yong Cai, Hengrui Cai
cs.LG, cs.AI, cs.CL, stat.ME, stat.ML
With recent advances in natural language processing, rationalization becomes an essential self-explaining diagram to disentangle the black box by selecting a subset of input texts to account for the major variation in prediction. Yet, existing association-based approaches on rationalization cannot identify true rationa...
http://arxiv.org/pdf/2306.14115v2
2023-09-09T03:50:04
2023-06-25T03:34:06
{ "title": [ -0.0087890625, -0.004337311, -0.03540039, -0.075927734, 0.008811951, 0.002380371, -0.021652222, -0.0065231323, -0.04788208, 0.00034689903, 0.019317627, 0.0064048767, -0.052490234, -0.043792725, -0.01777649, -0.016448975, -0.012794495, ...
http://arxiv.org/abs/2306.14096v5
Chinese Fine-Grained Financial Sentiment Analysis with Large Language Models
Yinyu Lan, Yanru Wu, Wang Xu, Weiqiang Feng, Youhao Zhang
cs.CL, cs.AI
Entity-level fine-grained sentiment analysis in the financial domain is a crucial subtask of sentiment analysis and currently faces numerous challenges. The primary challenge stems from the lack of high-quality and large-scale annotated corpora specifically designed for financial text sentiment analysis, which in turn ...
http://arxiv.org/pdf/2306.14096v5
2023-09-15T08:19:44
2023-06-25T02:24:30
{ "title": [ -0.021102905, -0.0028591156, -0.030838013, -0.03062439, 0.007083893, -0.0063552856, -0.016311646, -0.010284424, 0.020904541, 0.046020508, -0.032470703, -0.02255249, 0.012138367, -0.0033187866, -0.07977295, -0.008682251, -0.010955811, ...
http://arxiv.org/abs/2306.14067v1
UAlberta at SemEval-2023 Task 1: Context Augmentation and Translation for Multilingual Visual Word Sense Disambiguation
Michael Ogezi, Bradley Hauer, Talgat Omarov, Ning Shi, Grzegorz Kondrak
cs.CL
We describe the systems of the University of Alberta team for the SemEval-2023 Visual Word Sense Disambiguation (V-WSD) Task. We present a novel algorithm that leverages glosses retrieved from BabelNet, in combination with text and image encoders. Furthermore, we compare language-specific encoders against the applicati...
http://arxiv.org/pdf/2306.14067v1
2023-06-24T22:00:06
2023-06-24T22:00:06
{ "title": [ 0.0033912659, -0.013977051, -0.023590088, -0.035858154, -0.008712769, 0.0053367615, -0.030822754, -0.006942749, 0.033843994, 0.05670166, -0.042175293, 0.022949219, 0.0335083, 0.0076904297, -0.045684814, 0.022537231, 0.0055160522, 0.0...
http://arxiv.org/abs/2306.14060v1
DesCo: Learning Object Recognition with Rich Language Descriptions
Liunian Harold Li, Zi-Yi Dou, Nanyun Peng, Kai-Wei Chang
cs.CV, cs.CL, cs.LG
Recent development in vision-language approaches has instigated a paradigm shift in learning visual recognition models from language supervision. These approaches align objects with language queries (e.g. "a photo of a cat") and improve the models' adaptability to identify novel objects and domains. Recently, several s...
http://arxiv.org/pdf/2306.14060v1
2023-06-24T21:05:02
2023-06-24T21:05:02
{ "title": [ -0.026824951, -0.033935547, -0.020935059, -0.050994873, -0.009674072, 0.003042221, -0.033569336, -0.008857727, 0.015449524, -0.0031528473, -0.049438477, 0.010406494, -0.03173828, -0.00843811, -0.02949524, -0.013893127, 0.0076293945, ...
http://arxiv.org/abs/2306.14050v1
Symbolic Chain-of-Thought Distillation: Small Models Can Also "Think" Step-by-Step
Liunian Harold Li, Jack Hessel, Youngjae Yu, Xiang Ren, Kai-Wei Chang, Yejin Choi
cs.CL
Chain-of-thought prompting (e.g., "Let's think step-by-step") primes large language models to verbalize rationalization for their predictions. While chain-of-thought can lead to dramatic performance gains, benefits appear to emerge only for sufficiently large models (beyond 50B parameters). We show that orders-of-magni...
http://arxiv.org/pdf/2306.14050v1
2023-06-24T20:15:07
2023-06-24T20:15:07
{ "title": [ 0.0038452148, -0.006576538, -0.043670654, -0.0104599, -0.036132812, -0.019348145, -0.033355713, -0.027542114, -0.010223389, 0.011886597, -0.025970459, -0.017822266, 0.002450943, -0.04650879, 0.00017392635, -0.043640137, 0.03125, -0.0...
http://arxiv.org/abs/2306.14040v1
Weighted Automata Extraction and Explanation of Recurrent Neural Networks for Natural Language Tasks
Zeming Wei, Xiyue Zhang, Yihao Zhang, Meng Sun
cs.CL, cs.LG
Recurrent Neural Networks (RNNs) have achieved tremendous success in processing sequential data, yet understanding and analyzing their behaviours remains a significant challenge. To this end, many efforts have been made to extract finite automata from RNNs, which are more amenable for analysis and explanation. However,...
http://arxiv.org/pdf/2306.14040v1
2023-06-24T19:16:56
2023-06-24T19:16:56
{ "title": [ -0.020431519, -0.017028809, -0.03765869, -0.07873535, -0.0031223297, 0.009101868, -0.05807495, 0.026855469, -0.0067367554, 0.013053894, 0.0019512177, -0.012641907, -0.03250122, -0.022964478, -0.049102783, -0.0011978149, -0.014808655, ...
http://arxiv.org/abs/2306.14030v2
My Boli: Code-mixed Marathi-English Corpora, Pretrained Language Models and Evaluation Benchmarks
Tanmay Chavan, Omkar Gokhale, Aditya Kane, Shantanu Patankar, Raviraj Joshi
cs.CL, cs.LG
The research on code-mixed data is limited due to the unavailability of dedicated code-mixed datasets and pre-trained language models. In this work, we focus on the low-resource Indian language Marathi which lacks any prior work in code-mixing. We present L3Cube-MeCorpus, a large code-mixed Marathi-English (Mr-En) corp...
http://arxiv.org/pdf/2306.14030v2
2023-07-20T13:54:05
2023-06-24T18:17:38
{ "title": [ -0.031311035, -0.0055503845, -0.05014038, -0.032806396, -0.032562256, 0.0032176971, -0.07421875, 0.008659363, -0.04901123, 0.027648926, -0.021820068, 0.014892578, 0.028625488, -0.051940918, 0.018844604, -0.019332886, -0.0075569153, 0...
http://arxiv.org/abs/2306.14003v1
Weakly Supervised Multi-Label Classification of Full-Text Scientific Papers
Yu Zhang, Bowen Jin, Xiusi Chen, Yanzhen Shen, Yunyi Zhang, Yu Meng, Jiawei Han
cs.CL, cs.DL, cs.LG
Instead of relying on human-annotated training samples to build a classifier, weakly supervised scientific paper classification aims to classify papers only using category descriptions (e.g., category names, category-indicative keywords). Existing studies on weakly supervised paper classification are less concerned wit...
http://arxiv.org/pdf/2306.14003v1
2023-06-24T15:27:55
2023-06-24T15:27:55
{ "title": [ -0.01777649, 0.010345459, -0.03427124, -0.048034668, 0.0008292198, -0.0021839142, -0.026672363, -0.015808105, -0.014472961, 0.020950317, -0.000017523766, 0.021896362, -0.01928711, -0.019866943, -0.026306152, -0.0054359436, -0.0088272095,...
http://arxiv.org/abs/2306.13986v1
Large Language Models as Sous Chefs: Revising Recipes with GPT-3
Alyssa Hwang, Bryan Li, Zhaoyi Hou, Dan Roth
cs.CL
With their remarkably improved text generation and prompting capabilities, large language models can adapt existing written information into forms that are easier to use and understand. In our work, we focus on recipes as an example of complex, diverse, and widely used instructions. We develop a prompt grounded in the ...
http://arxiv.org/pdf/2306.13986v1
2023-06-24T14:42:43
2023-06-24T14:42:43
{ "title": [ -0.02532959, -0.0034446716, -0.025054932, -0.03744507, -0.010223389, 0.0031776428, -0.047943115, -0.017715454, -0.002916336, 0.029251099, -0.017059326, -0.0009446144, 0.01309967, -0.011871338, -0.03967285, 0.011871338, -0.019699097, ...
http://arxiv.org/abs/2306.13971v2
Towards Robust Aspect-based Sentiment Analysis through Non-counterfactual Augmentations
Xinyu Liu, Yan Ding, Kaikai An, Chunyang Xiao, Pranava Madhyastha, Tong Xiao, Jingbo Zhu
cs.CL
While state-of-the-art NLP models have demonstrated excellent performance for aspect based sentiment analysis (ABSA), substantial evidence has been presented on their lack of robustness. This is especially manifested as significant degradation in performance when faced with out-of-distribution data. Recent solutions th...
http://arxiv.org/pdf/2306.13971v2
2023-07-21T01:58:13
2023-06-24T13:57:32
{ "title": [ 0.022155762, 0.0017585754, -0.024810791, -0.053985596, 0.01537323, -0.0031280518, -0.016647339, -0.011062622, 0.039886475, 0.055877686, -0.017578125, -0.0027999878, 0.011749268, 0.0025196075, -0.087646484, 0.014015198, -0.012107849, ...
http://arxiv.org/abs/2306.13968v1
Fusing Multimodal Signals on Hyper-complex Space for Extreme Abstractive Text Summarization (TL;DR) of Scientific Contents
Yash Kumar Atri, Vikram Goyal, Tanmoy Chakraborty
cs.CL, cs.AI
The realm of scientific text summarization has experienced remarkable progress due to the availability of annotated brief summaries and ample data. However, the utilization of multiple input modalities, such as videos and audio, has yet to be thoroughly explored. At present, scientific multimodal-input-based text summa...
http://arxiv.org/pdf/2306.13968v1
2023-06-24T13:51:42
2023-06-24T13:51:42
{ "title": [ -0.0020828247, 0.018218994, -0.022033691, -0.034729004, -0.013954163, -0.0085372925, -0.030593872, -0.027145386, 0.028442383, 0.054779053, 0.004371643, 0.010902405, 0.03237915, -0.024291992, -0.039733887, 0.009284973, 0.0022830963, 0...
http://arxiv.org/abs/2306.13959v1
Emotion Flip Reasoning in Multiparty Conversations
Shivani Kumar, Shubham Dudeja, Md Shad Akhtar, Tanmoy Chakraborty
cs.CL, cs.AI
In a conversational dialogue, speakers may have different emotional states and their dynamics play an important role in understanding dialogue's emotional discourse. However, simply detecting emotions is not sufficient to entirely comprehend the speaker-specific changes in emotion that occur during a conversation. To u...
http://arxiv.org/pdf/2306.13959v1
2023-06-24T13:22:02
2023-06-24T13:22:02
{ "title": [ -0.010246277, -0.005077362, -0.022216797, -0.027175903, -0.009643555, -0.020095825, -0.022445679, -0.020507812, 0.01108551, 0.03817749, -0.0023498535, -0.018112183, 0.019134521, -0.036712646, -0.032928467, -0.0023441315, -0.009101868, ...
http://arxiv.org/abs/2306.13954v1
Characterizing the Emotion Carriers of COVID-19 Misinformation and Their Impact on Vaccination Outcomes in India and the United States
Ridam Pal, Sanjana S, Deepak Mahto, Kriti Agrawal, Gopal Mengi, Sargun Nagpal, Akshaya Devadiga, Tavpritesh Sethi
cs.CL, cs.CY
The COVID-19 Infodemic had an unprecedented impact on health behaviors and outcomes at a global scale. While many studies have focused on a qualitative and quantitative understanding of misinformation, including sentiment analysis, there is a gap in understanding the emotion-carriers of misinformation and their differe...
http://arxiv.org/pdf/2306.13954v1
2023-06-24T12:56:56
2023-06-24T12:56:56
{ "title": [ 0.008125305, -0.0026931763, -0.049621582, -0.04852295, 0.006160736, -0.029174805, -0.012306213, 0.017852783, -0.03527832, 0.014289856, 0.007003784, 0.03527832, -0.009346008, -0.053771973, 0.03390503, -0.0025310516, 0.006801605, 0.011...
http://arxiv.org/abs/2306.13947v1
Comparison of Pre-trained Language Models for Turkish Address Parsing
Muhammed Cihat Ünal, Betül Aygün, Aydın Gerek
cs.CL, cs.LG
Transformer based pre-trained models such as BERT and its variants, which are trained on large corpora, have demonstrated tremendous success for natural language processing (NLP) tasks. Most of academic works are based on the English language; however, the number of multilingual and language specific studies increase s...
http://arxiv.org/pdf/2306.13947v1
2023-06-24T12:09:43
2023-06-24T12:09:43
{ "title": [ 0.024291992, -0.0048942566, -0.013717651, -0.05014038, -0.01524353, -0.012664795, -0.049957275, -0.003118515, -0.021011353, 0.04724121, -0.023803711, -0.011978149, 0.050354004, -0.029556274, -0.012565613, -0.004760742, 0.0014619827, ...
http://arxiv.org/abs/2306.13922v1
Unsupervised Mapping of Arguments of Deverbal Nouns to Their Corresponding Verbal Labels
Aviv Weinstein, Yoav Goldberg
cs.CL
Deverbal nouns are nominal forms of verbs commonly used in written English texts to describe events or actions, as well as their arguments. However, many NLP systems, and in particular pattern-based ones, neglect to handle such nominalized constructions. The solutions that do exist for handling arguments of nominalized...
http://arxiv.org/pdf/2306.13922v1
2023-06-24T10:07:01
2023-06-24T10:07:01
{ "title": [ 0.020889282, 0.004119873, 0.009681702, -0.036499023, -0.041809082, 0.0023727417, -0.046203613, -0.043304443, -0.0050849915, 0.009132385, -0.00548172, -0.028564453, 0.038513184, -0.05706787, 0.017547607, -0.0060272217, -0.01890564, 0....
http://arxiv.org/abs/2306.13906v1
Can GPT-4 Support Analysis of Textual Data in Tasks Requiring Highly Specialized Domain Expertise?
Jaromir Savelka, Kevin D. Ashley, Morgan A Gray, Hannes Westermann, Huihui Xu
cs.CL
We evaluated the capability of generative pre-trained transformers~(GPT-4) in analysis of textual data in tasks that require highly specialized domain expertise. Specifically, we focused on the task of analyzing court opinions to interpret legal concepts. We found that GPT-4, prompted with annotation guidelines, perfor...
http://arxiv.org/pdf/2306.13906v1
2023-06-24T08:48:24
2023-06-24T08:48:24
{ "title": [ 0.013381958, 0.013961792, -0.05493164, -0.030532837, -0.008811951, -0.03616333, -0.0050086975, -0.00002604723, -0.0041503906, 0.02670288, 0.004676819, -0.014350891, -0.0012836456, -0.038024902, -0.009155273, -0.011650085, 0.0037727356, ...
http://arxiv.org/abs/2306.13905v1
Spatio-temporal Storytelling? Leveraging Generative Models for Semantic Trajectory Analysis
Shreya Ghosh, Saptarshi Sengupta, Prasenjit Mitra
cs.CL, cs.AI, cs.LG
In this paper, we lay out a vision for analysing semantic trajectory traces and generating synthetic semantic trajectory data (SSTs) using generative language model. Leveraging the advancements in deep learning, as evident by progress in the field of natural language processing (NLP), computer vision, etc. we intend to...
http://arxiv.org/pdf/2306.13905v1
2023-06-24T08:45:47
2023-06-24T08:45:47
{ "title": [ 0.001376152, -0.009513855, -0.055786133, -0.08337402, -0.0030498505, -0.01109314, -0.048736572, -0.003484726, 0.013305664, 0.04147339, -0.014579773, 0.017608643, 0.00006151199, -0.017532349, -0.032836914, 0.0046920776, -0.0011816025, ...
http://arxiv.org/abs/2306.13899v1
Math Word Problem Solving by Generating Linguistic Variants of Problem Statements
Syed Rifat Raiyan, Md. Nafis Faiyaz, Shah Md. Jawad Kabir, Mohsinul Kabir, Hasan Mahmud, Md Kamrul Hasan
cs.CL, cs.AI
The art of mathematical reasoning stands as a fundamental pillar of intellectual progress and is a central catalyst in cultivating human ingenuity. Researchers have recently published a plethora of works centered around the task of solving Math Word Problems (MWP) $-$ a crucial stride towards general AI. These existing...
http://arxiv.org/pdf/2306.13899v1
2023-06-24T08:27:39
2023-06-24T08:27:39
{ "title": [ -0.023483276, -0.020202637, -0.025344849, -0.01058197, -0.03225708, -0.020614624, -0.009567261, -0.0006489754, -0.013153076, 0.020141602, -0.024887085, 0.00056409836, -0.0019073486, -0.026916504, 0.0029907227, -0.009155273, 0.007820129, ...
http://arxiv.org/abs/2306.13891v1
Estimating the Causal Effect of Early ArXiving on Paper Acceptance
Yanai Elazar, Jiayao Zhang, David Wadden, Bo Zhang, Noah A. Smith
cs.CL
What is the effect of releasing a preprint of a paper before it is submitted for peer review? No randomized controlled trial has been conducted, so we turn to observational data to answer this question. We use data from the ICLR conference (2018--2022) and apply methods from causal inference to estimate the effect of a...
http://arxiv.org/pdf/2306.13891v1
2023-06-24T07:45:38
2023-06-24T07:45:38
{ "title": [ -0.008270264, 0.0052948, -0.018966675, -0.02557373, 0.0015029907, -0.009742737, -0.0021877289, -0.025787354, -0.056274414, 0.014320374, 0.010360718, 0.009559631, -0.049621582, -0.035980225, -0.016342163, -0.023025513, -0.0059776306, ...
http://arxiv.org/abs/2306.13888v1
L3Cube-MahaSent-MD: A Multi-domain Marathi Sentiment Analysis Dataset and Transformer Models
Aabha Pingle, Aditya Vyawahare, Isha Joshi, Rahul Tangsali, Raviraj Joshi
cs.CL, cs.LG
The exploration of sentiment analysis in low-resource languages, such as Marathi, has been limited due to the availability of suitable datasets. In this work, we present L3Cube-MahaSent-MD, a multi-domain Marathi sentiment analysis dataset, with four different domains - movie reviews, general tweets, TV show subtitles,...
http://arxiv.org/pdf/2306.13888v1
2023-06-24T07:27:53
2023-06-24T07:27:53
{ "title": [ -0.0017433167, -0.008140564, -0.037475586, -0.06304932, 0.0034255981, 0.0019607544, -0.04626465, -0.0020332336, -0.00026726723, 0.039245605, -0.040222168, -0.0076141357, 0.036987305, -0.009162903, -0.04232788, -0.006679535, -0.0022392273...