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/2306.03652v1
Injecting knowledge into language generation: a case study in auto-charting after-visit care instructions from medical dialogue
Maksim Eremeev, Ilya Valmianski, Xavier Amatriain, Anitha Kannan
cs.CL
Factual correctness is often the limiting factor in practical applications of natural language generation in high-stakes domains such as healthcare. An essential requirement for maintaining factuality is the ability to deal with rare tokens. This paper focuses on rare tokens that appear in both the source and the refer...
http://arxiv.org/pdf/2306.03652v1
2023-06-06T13:13:27
2023-06-06T13:13:27
{ "title": [ -0.011978149, -0.045013428, -0.039855957, -0.07446289, -0.029296875, 0.0025043488, -0.06750488, 0.029830933, -0.021209717, 0.03564453, -0.0096588135, 0.023162842, -0.011154175, -0.027816772, -0.025405884, -0.007160187, 0.018722534, -...
http://arxiv.org/abs/2306.03650v1
A Quantum Probability Driven Framework for Joint Multi-Modal Sarcasm, Sentiment and Emotion Analysis
Yaochen Liu, Yazhou Zhang, Dawei Song
cs.CL, cs.AI
Sarcasm, sentiment, and emotion are three typical kinds of spontaneous affective responses of humans to external events and they are tightly intertwined with each other. Such events may be expressed in multiple modalities (e.g., linguistic, visual and acoustic), e.g., multi-modal conversations. Joint analysis of humans...
http://arxiv.org/pdf/2306.03650v1
2023-06-06T13:08:22
2023-06-06T13:08:22
{ "title": [ 0.0010404587, 0.0066070557, -0.019866943, -0.022460938, 0.0018119812, -0.013427734, -0.015411377, -0.04425049, 0.025146484, 0.04849243, 0.012428284, -0.036224365, 0.041229248, -0.043273926, -0.06591797, 0.013938904, -0.021148682, 0.0...
http://arxiv.org/abs/2306.05370v1
Detecting Human Rights Violations on Social Media during Russia-Ukraine War
Poli Nemkova, Solomon Ubani, Suleyman Olcay Polat, Nayeon Kim, Rodney D. Nielsen
cs.CY, cs.CL
The present-day Russia-Ukraine military conflict has exposed the pivotal role of social media in enabling the transparent and unbridled sharing of information directly from the frontlines. In conflict zones where freedom of expression is constrained and information warfare is pervasive, social media has emerged as an i...
http://arxiv.org/pdf/2306.05370v1
2023-06-06T12:59:03
2023-06-06T12:59:03
{ "title": [ 0.031677246, 0.022277832, -0.031463623, -0.023284912, 0.02960205, -0.032714844, 0.037841797, -0.009765625, -0.044433594, 0.024047852, -0.03378296, 0.05065918, 0.022338867, -0.015823364, 0.02029419, 0.0023784637, -0.008712769, 0.04492...
http://arxiv.org/abs/2306.03628v1
Convergence and Diversity in the Control Hierarchy
Alexandra Butoi, Ryan Cotterell, David Chiang
cs.FL, cs.CL
Weir has defined a hierarchy of language classes whose second member ($\mathcal{L}_2$) is generated by tree-adjoining grammars (TAG), linear indexed grammars (LIG), combinatory categorial grammars, and head grammars. The hierarchy is obtained using the mechanism of control, and $\mathcal{L}_2$ is obtained using a conte...
http://arxiv.org/pdf/2306.03628v1
2023-06-06T12:30:29
2023-06-06T12:30:29
{ "title": [ 0.043914795, 0.04248047, -0.05606079, -0.076049805, 0.038757324, -0.054992676, -0.0062217712, -0.015716553, -0.047973633, 0.03189087, -0.011421204, 0.0042152405, -0.026809692, -0.0345459, -0.01084137, -0.0078086853, 0.015151978, 0.03...
http://arxiv.org/abs/2306.03608v1
A Survey of Quantum-Cognitively Inspired Sentiment Analysis Models
Yaochen Liu, Qiuchi Li, Benyou Wang, Yazhou Zhang, Dawei Song
cs.CL, cs.AI
Quantum theory, originally proposed as a physical theory to describe the motions of microscopic particles, has been applied to various non-physics domains involving human cognition and decision-making that are inherently uncertain and exhibit certain non-classical, quantum-like characteristics. Sentiment analysis is a ...
http://arxiv.org/pdf/2306.03608v1
2023-06-06T11:54:48
2023-06-06T11:54:48
{ "title": [ -0.00970459, 0.00957489, -0.000790596, 0.0037441254, -0.000044226646, -0.0115356445, 0.014533997, -0.043273926, 0.011192322, 0.058013916, 0.0043792725, -0.013648987, 0.017471313, -0.013305664, -0.08312988, 0.0073051453, -0.02053833, ...
http://arxiv.org/abs/2306.03598v1
CUE: An Uncertainty Interpretation Framework for Text Classifiers Built on Pre-Trained Language Models
Jiazheng Li, Zhaoyue Sun, Bin Liang, Lin Gui, Yulan He
cs.CL
Text classifiers built on Pre-trained Language Models (PLMs) have achieved remarkable progress in various tasks including sentiment analysis, natural language inference, and question-answering. However, the occurrence of uncertain predictions by these classifiers poses a challenge to their reliability when deployed in ...
http://arxiv.org/pdf/2306.03598v1
2023-06-06T11:37:46
2023-06-06T11:37:46
{ "title": [ -0.04623413, 0.0007982254, -0.042755127, -0.0345459, -0.003189087, -0.0025405884, -0.05996704, -0.022750854, 0.007637024, 0.013389587, -0.03765869, 0.0107421875, -0.0070495605, -0.009666443, -0.044158936, -0.01701355, -0.009849548, 0...
http://arxiv.org/abs/2306.03586v1
Language acquisition: do children and language models follow similar learning stages?
Linnea Evanson, Yair Lakretz, Jean-Rémi King
cs.CL, cs.AI
During language acquisition, children follow a typical sequence of learning stages, whereby they first learn to categorize phonemes before they develop their lexicon and eventually master increasingly complex syntactic structures. However, the computational principles that lead to this learning trajectory remain largel...
http://arxiv.org/pdf/2306.03586v1
2023-06-06T11:08:20
2023-06-06T11:08:20
{ "title": [ -0.050567627, 0.008796692, -0.034606934, -0.025817871, -0.014320374, 0.0022678375, -0.027709961, -0.0016918182, -0.04849243, 0.020767212, -0.015213013, 0.022567749, -0.022079468, -0.03475952, -0.0067481995, -0.01108551, -0.013084412, ...
http://arxiv.org/abs/2306.03557v2
Take the Hint: Improving Arabic Diacritization with Partially-Diacritized Text
Parnia Bahar, Mattia Di Gangi, Nick Rossenbach, Mohammad Zeineldeen
cs.CL, cs.AI
Automatic Arabic diacritization is useful in many applications, ranging from reading support for language learners to accurate pronunciation predictor for downstream tasks like speech synthesis. While most of the previous works focused on models that operate on raw non-diacritized text, production systems can gain accu...
http://arxiv.org/pdf/2306.03557v2
2023-07-31T12:29:10
2023-06-06T10:18:17
{ "title": [ -0.013061523, 0.0011310577, -0.0073547363, -0.051330566, -0.003698349, 0.0012311935, -0.02355957, -0.0029315948, -0.040252686, 0.029159546, -0.020965576, 0.021362305, -0.0047912598, -0.031402588, -0.0071907043, -0.020080566, 0.013954163,...
http://arxiv.org/abs/2306.03535v2
SciLit: A Platform for Joint Scientific Literature Discovery, Summarization and Citation Generation
Nianlong Gu, Richard H. R. Hahnloser
cs.CL
Scientific writing involves retrieving, summarizing, and citing relevant papers, which can be time-consuming processes in large and rapidly evolving fields. By making these processes inter-operable, natural language processing (NLP) provides opportunities for creating end-to-end assistive writing tools. We propose SciL...
http://arxiv.org/pdf/2306.03535v2
2023-11-06T15:53:23
2023-06-06T09:34:45
{ "title": [ -0.03475952, -0.0050239563, -0.05606079, -0.0713501, 0.010475159, 0.008720398, -0.055633545, -0.014755249, -0.016967773, 0.043304443, 0.016647339, 0.021118164, -0.0054397583, -0.0128479, -0.050201416, 0.017044067, -0.015975952, 0.048...
http://arxiv.org/abs/2306.03507v1
"A Little is Enough": Few-Shot Quality Estimation based Corpus Filtering improves Machine Translation
Akshay Batheja, Pushpak Bhattacharyya
cs.CL
Quality Estimation (QE) is the task of evaluating the quality of a translation when reference translation is not available. The goal of QE aligns with the task of corpus filtering, where we assign the quality score to the sentence pairs present in the pseudo-parallel corpus. We propose a Quality Estimation based Filter...
http://arxiv.org/pdf/2306.03507v1
2023-06-06T08:53:01
2023-06-06T08:53:01
{ "title": [ -0.0110321045, -0.011245728, 0.02130127, -0.027572632, -0.022079468, 0.0033035278, -0.016098022, -0.015777588, 0.008132935, 0.03488159, -0.01991272, -0.0076026917, 0.036010742, -0.0309906, -0.051605225, 0.005153656, -0.016525269, 0.0...
http://arxiv.org/abs/2306.03503v2
Applying Standards to Advance Upstream & Downstream Ethics in Large Language Models
Jose Berengueres, Marybeth Sandell
cs.CY, cs.AI, cs.CL, K.4.1; I.2.0
This paper explores how AI-owners can develop safeguards for AI-generated content by drawing from established codes of conduct and ethical standards in other content-creation industries. It delves into the current state of ethical awareness on Large Language Models (LLMs). By dissecting the mechanism of content generat...
http://arxiv.org/pdf/2306.03503v2
2023-06-11T10:45:57
2023-06-06T08:47:42
{ "title": [ -0.02507019, 0.006416321, -0.070495605, -0.057647705, -0.0040512085, -0.027236938, -0.055419922, 0.013793945, -0.012565613, 0.04547119, -0.025314331, 0.00017249584, 0.04034424, -0.02458191, -0.018127441, 0.01449585, -0.02609253, 0.04...
http://arxiv.org/abs/2306.03500v1
Towards Adaptable and Interactive Image Captioning with Data Augmentation and Episodic Memory
Aliki Anagnostopoulou, Mareike Hartmann, Daniel Sonntag
cs.CL, cs.CV
Interactive machine learning (IML) is a beneficial learning paradigm in cases of limited data availability, as human feedback is incrementally integrated into the training process. In this paper, we present an IML pipeline for image captioning which allows us to incrementally adapt a pre-trained image captioning model ...
http://arxiv.org/pdf/2306.03500v1
2023-06-06T08:38:10
2023-06-06T08:38:10
{ "title": [ 0.026626587, -0.0025749207, -0.019973755, -0.048614502, 0.0014333725, -0.029937744, 0.017837524, 0.0019702911, 0.021636963, 0.03604126, -0.043945312, 0.0076522827, -0.0045280457, 0.005569458, -0.05328369, -0.008644104, 0.0051345825, ...
http://arxiv.org/abs/2306.03491v1
SciCap+: A Knowledge Augmented Dataset to Study the Challenges of Scientific Figure Captioning
Zhishen Yang, Raj Dabre, Hideki Tanaka, Naoaki Okazaki
cs.CV, cs.CL
In scholarly documents, figures provide a straightforward way of communicating scientific findings to readers. Automating figure caption generation helps move model understandings of scientific documents beyond text and will help authors write informative captions that facilitate communicating scientific findings. Unli...
http://arxiv.org/pdf/2306.03491v1
2023-06-06T08:16:16
2023-06-06T08:16:16
{ "title": [ -0.008018494, -0.0055160522, -0.042388916, -0.031402588, -0.003944397, -0.029525757, -0.0036449432, -0.009490967, 0.029327393, 0.05340576, -0.025131226, -0.00059604645, 0.009155273, -0.0038967133, -0.07055664, 0.007949829, 0.0013113022, ...
http://arxiv.org/abs/2306.03476v1
Putting Humans in the Image Captioning Loop
Aliki Anagnostopoulou, Mareike Hartmann, Daniel Sonntag
cs.CL, cs.CV
Image Captioning (IC) models can highly benefit from human feedback in the training process, especially in cases where data is limited. We present work-in-progress on adapting an IC system to integrate human feedback, with the goal to make it easily adaptable to user-specific data. Our approach builds on a base IC mode...
http://arxiv.org/pdf/2306.03476v1
2023-06-06T07:50:46
2023-06-06T07:50:46
{ "title": [ 0.057006836, 0.0077171326, 0.003774643, 0.003353119, -0.03253174, -0.027908325, 0.036712646, -0.029159546, -0.0036773682, 0.06964111, -0.042755127, 0.010101318, 0.006931305, 0.008895874, -0.013717651, -0.01309967, 0.005180359, 0.0075...
http://arxiv.org/abs/2306.03469v1
Joint Event Extraction via Structural Semantic Matching
Haochen Li, Tianhao Gao, Jingkun Wang, Weiping Li
cs.CL
Event Extraction (EE) is one of the essential tasks in information extraction, which aims to detect event mentions from text and find the corresponding argument roles. The EE task can be abstracted as a process of matching the semantic definitions and argument structures of event types with the target text. This paper ...
http://arxiv.org/pdf/2306.03469v1
2023-06-06T07:42:39
2023-06-06T07:42:39
{ "title": [ -0.008010864, 0.010795593, -0.041503906, -0.11224365, 0.035858154, 0.02268982, -0.06555176, -0.01600647, 0.0028858185, 0.027282715, 0.014099121, -0.003835678, -0.00969696, -0.031341553, -0.033569336, 0.015205383, -0.01361084, 0.01065...
http://arxiv.org/abs/2306.03460v1
Natural Language Commanding via Program Synthesis
Apurva Gandhi, Thong Q. Nguyen, Huitian Jiao, Robert Steen, Ameya Bhatawdekar
cs.LG, cs.CL, cs.HC
We present Semantic Interpreter, a natural language-friendly AI system for productivity software such as Microsoft Office that leverages large language models (LLMs) to execute user intent across application features. While LLMs are excellent at understanding user intent expressed as natural language, they are not suff...
http://arxiv.org/pdf/2306.03460v1
2023-06-06T07:28:49
2023-06-06T07:28:49
{ "title": [ -0.031036377, -0.006580353, -0.025482178, -0.0692749, -0.0006914139, 0.01725769, -0.05819702, -0.003873825, -0.0028076172, 0.020690918, -0.033111572, -0.009613037, -0.013595581, -0.015625, -0.030548096, -0.010360718, -0.016220093, 0....
http://arxiv.org/abs/2306.03457v2
TwistList: Resources and Baselines for Tongue Twister Generation
Tyler Loakman, Chen Tang, Chenghua Lin
cs.CL, cs.AI
Previous work in phonetically-grounded language generation has mainly focused on domains such as lyrics and poetry. In this paper, we present work on the generation of tongue twisters - a form of language that is required to be phonetically conditioned to maximise sound overlap, whilst maintaining semantic consistency ...
http://arxiv.org/pdf/2306.03457v2
2023-06-07T05:24:25
2023-06-06T07:20:51
{ "title": [ -0.018844604, 0.004497528, -0.05407715, -0.030685425, 0.0020503998, -0.010177612, -0.022872925, -0.008972168, -0.00182724, 0.054534912, -0.013656616, 0.019851685, 0.04827881, -0.031585693, -0.014030457, 0.026153564, 0.0066223145, 0.0...
http://arxiv.org/abs/2306.03444v1
Automatic Assessment of Oral Reading Accuracy for Reading Diagnostics
Bo Molenaar, Cristian Tejedor-Garcia, Helmer Strik, Catia Cucchiarini
cs.CL, cs.SD, eess.AS
Automatic assessment of reading fluency using automatic speech recognition (ASR) holds great potential for early detection of reading difficulties and subsequent timely intervention. Precise assessment tools are required, especially for languages other than English. In this study, we evaluate six state-of-the-art ASR-b...
http://arxiv.org/pdf/2306.03444v1
2023-06-06T06:49:58
2023-06-06T06:49:58
{ "title": [ -0.021148682, 0.010215759, -0.034729004, -0.031585693, -0.013755798, 0.00004285574, -0.03387451, -0.021118164, -0.004436493, 0.057434082, -0.029800415, 0.010147095, 0.013725281, -0.007724762, -0.066223145, 0.0032749176, -0.0060920715, ...
http://arxiv.org/abs/2306.03443v1
Alzheimer Disease Classification through ASR-based Transcriptions: Exploring the Impact of Punctuation and Pauses
Lucía Gómez-Zaragozá, Simone Wills, Cristian Tejedor-Garcia, Javier Marín-Morales, Mariano Alcañiz, Helmer Strik
cs.CL, cs.SD, eess.AS
Alzheimer's Disease (AD) is the world's leading neurodegenerative disease, which often results in communication difficulties. Analysing speech can serve as a diagnostic tool for identifying the condition. The recent ADReSS challenge provided a dataset for AD classification and highlighted the utility of manual transcri...
http://arxiv.org/pdf/2306.03443v1
2023-06-06T06:49:41
2023-06-06T06:49:41
{ "title": [ -0.028152466, -0.010002136, -0.062927246, -0.078186035, -0.013114929, -0.012840271, -0.0692749, 0.0317688, -0.017730713, 0.022277832, -0.023773193, -0.01802063, 0.00806427, -0.03237915, -0.025634766, -0.027282715, -0.007972717, 0.018...
http://arxiv.org/abs/2306.03438v1
Large Language Models of Code Fail at Completing Code with Potential Bugs
Tuan Dinh, Jinman Zhao, Samson Tan, Renato Negrinho, Leonard Lausen, Sheng Zha, George Karypis
cs.LG, cs.AI, cs.CL, cs.SE
Large language models of code (Code-LLMs) have recently brought tremendous advances to code completion, a fundamental feature of programming assistance and code intelligence. However, most existing works ignore the possible presence of bugs in the code context for generation, which are inevitable in software developmen...
http://arxiv.org/pdf/2306.03438v1
2023-06-06T06:35:27
2023-06-06T06:35:27
{ "title": [ -0.002922058, -0.04977417, 0.0006403923, -0.035583496, -0.06878662, 0.014587402, -0.003074646, 0.027542114, 0.014381409, 0.046661377, -0.05895996, -0.032958984, 0.010948181, -0.017822266, -0.018081665, -0.025482178, 0.014305115, 0.02...
http://arxiv.org/abs/2306.03435v1
On the Role of Attention in Prompt-tuning
Samet Oymak, Ankit Singh Rawat, Mahdi Soltanolkotabi, Christos Thrampoulidis
cs.LG, cs.CL, stat.ML
Prompt-tuning is an emerging strategy to adapt large language models (LLM) to downstream tasks by learning a (soft-)prompt parameter from data. Despite its success in LLMs, there is limited theoretical understanding of the power of prompt-tuning and the role of the attention mechanism in prompting. In this work, we exp...
http://arxiv.org/pdf/2306.03435v1
2023-06-06T06:23:38
2023-06-06T06:23:38
{ "title": [ 0.017410278, 0.0035190582, -0.007987976, -0.039855957, -0.017562866, -0.016983032, -0.02684021, -0.023803711, -0.010238647, 0.02835083, -0.021118164, -0.03540039, 0.0058631897, -0.030349731, -0.031219482, -0.010192871, -0.00040197372, ...
http://arxiv.org/abs/2306.03415v1
Efficient and Interpretable Compressive Text Summarisation with Unsupervised Dual-Agent Reinforcement Learning
Peggy Tang, Junbin Gao, Lei Zhang, Zhiyong Wang
cs.CL
Recently, compressive text summarisation offers a balance between the conciseness issue of extractive summarisation and the factual hallucination issue of abstractive summarisation. However, most existing compressive summarisation methods are supervised, relying on the expensive effort of creating a new training datase...
http://arxiv.org/pdf/2306.03415v1
2023-06-06T05:30:49
2023-06-06T05:30:49
{ "title": [ -0.030227661, 0.005256653, -0.01739502, -0.03540039, -0.013938904, 0.0053100586, -0.04321289, -0.012512207, -0.008911133, 0.02671814, -0.009086609, 0.0052986145, -0.027404785, -0.012901306, -0.04324341, -0.03289795, -0.0049972534, 0....
http://arxiv.org/abs/2306.03411v1
Generate-then-Retrieve: Intent-Aware FAQ Retrieval in Product Search
Zhiyu Chen, Jason Choi, Besnik Fetahu, Oleg Rokhlenko, Shervin Malmasi
cs.CL, cs.AI, cs.IR
Customers interacting with product search engines are increasingly formulating information-seeking queries. Frequently Asked Question (FAQ) retrieval aims to retrieve common question-answer pairs for a user query with question intent. Integrating FAQ retrieval in product search can not only empower users to make more i...
http://arxiv.org/pdf/2306.03411v1
2023-06-06T05:18:21
2023-06-06T05:18:21
{ "title": [ -0.0063171387, -0.0075187683, -0.05307007, -0.04083252, 0.0062828064, -0.030151367, -0.0154418945, 0.00047683716, -0.010734558, 0.050842285, -0.01676941, -0.001912117, 0.017990112, -0.013168335, -0.0541687, -0.018463135, -0.0015325546, ...
http://arxiv.org/abs/2306.03377v1
TextFormer: A Query-based End-to-End Text Spotter with Mixed Supervision
Yukun Zhai, Xiaoqiang Zhang, Xiameng Qin, Sanyuan Zhao, Xingping Dong, Jianbing Shen
cs.CV, cs.CL
End-to-end text spotting is a vital computer vision task that aims to integrate scene text detection and recognition into a unified framework. Typical methods heavily rely on Region-of-Interest (RoI) operations to extract local features and complex post-processing steps to produce final predictions. To address these li...
http://arxiv.org/pdf/2306.03377v1
2023-06-06T03:37:41
2023-06-06T03:37:41
{ "title": [ -0.0029029846, 0.004146576, -0.01474762, -0.07373047, 0.009887695, 0.0037670135, -0.04135132, -0.022155762, 0.01335144, 0.03564453, -0.031066895, 0.0031814575, 0.030227661, -0.012886047, -0.029953003, 0.0013465881, -0.012077332, 0.02...
http://arxiv.org/abs/2306.03361v3
WHAT, WHEN, and HOW to Ground: Designing User Persona-Aware Conversational Agents for Engaging Dialogue
Deuksin Kwon, Sunwoo Lee, Ki Hyun Kim, Seojin Lee, Taeyoon Kim, Eric Davis
cs.CL, cs.AI, I.2.1, I.2.7
This paper presents a method for building a personalized open-domain dialogue system to address the WWH (WHAT, WHEN, and HOW) problem for natural response generation in a commercial setting, where personalized dialogue responses are heavily interleaved with casual response turns. The proposed approach involves weighted...
http://arxiv.org/pdf/2306.03361v3
2023-07-03T22:44:12
2023-06-06T02:28:38
{ "title": [ 0.009315491, 0.004310608, -0.034698486, -0.04397583, 0.0124053955, -0.019454956, -0.013847351, -0.006477356, 0.006214142, 0.037750244, -0.010864258, 0.0068359375, 0.012260437, -0.015670776, -0.06451416, 0.003074646, -0.036254883, 0.0...
http://arxiv.org/abs/2306.03355v1
BatchSampler: Sampling Mini-Batches for Contrastive Learning in Vision, Language, and Graphs
Zhen Yang, Tinglin Huang, Ming Ding, Yuxiao Dong, Rex Ying, Yukuo Cen, Yangliao Geng, Jie Tang
cs.LG, cs.CL, cs.CV
In-Batch contrastive learning is a state-of-the-art self-supervised method that brings semantically-similar instances close while pushing dissimilar instances apart within a mini-batch. Its key to success is the negative sharing strategy, in which every instance serves as a negative for the others within the mini-batch...
http://arxiv.org/pdf/2306.03355v1
2023-06-06T02:13:27
2023-06-06T02:13:27
{ "title": [ 0.011650085, -0.006263733, -0.034179688, -0.04046631, 0.0025024414, -0.027709961, -0.005493164, -0.011688232, 0.01020813, 0.023620605, -0.04824829, 0.021102905, -0.0028152466, -0.014907837, -0.032989502, -0.015106201, 0.008888245, 0....
http://arxiv.org/abs/2306.03350v1
Click: Controllable Text Generation with Sequence Likelihood Contrastive Learning
Chujie Zheng, Pei Ke, Zheng Zhang, Minlie Huang
cs.CL
It has always been an important yet challenging problem to control language models to avoid generating texts with undesirable attributes, such as toxic language and unnatural repetition. We introduce Click for controllable text generation, which needs no modification to the model architecture and facilitates out-of-the...
http://arxiv.org/pdf/2306.03350v1
2023-06-06T01:56:44
2023-06-06T01:56:44
{ "title": [ -0.028503418, -0.014282227, -0.030334473, -0.048217773, -0.015083313, 0.003917694, -0.038513184, -0.014007568, 0.0046157837, 0.027526855, -0.021881104, 0.03010559, -0.005508423, -0.014289856, -0.033355713, 0.0027694702, -0.024475098, ...
http://arxiv.org/abs/2306.03341v5
Inference-Time Intervention: Eliciting Truthful Answers from a Language Model
Kenneth Li, Oam Patel, Fernanda Viégas, Hanspeter Pfister, Martin Wattenberg
cs.LG, cs.AI, cs.CL
We introduce Inference-Time Intervention (ITI), a technique designed to enhance the "truthfulness" of large language models (LLMs). ITI operates by shifting model activations during inference, following a set of directions across a limited number of attention heads. This intervention significantly improves the performa...
http://arxiv.org/pdf/2306.03341v5
2023-10-20T01:37:21
2023-06-06T01:26:53
{ "title": [ -0.030532837, -0.014724731, -0.041778564, -0.052734375, -0.014442444, -0.013374329, -0.052124023, -0.014083862, -0.010017395, 0.005203247, -0.022750854, 0.0006799698, -0.011459351, -0.020843506, -0.03414917, -0.011878967, -0.020828247, ...
http://arxiv.org/abs/2306.03316v1
CoSiNES: Contrastive Siamese Network for Entity Standardization
Jiaqing Yuan, Michele Merler, Mihir Choudhury, Raju Pavuluri, Munindar P. Singh, Maja Vukovic
cs.CL
Entity standardization maps noisy mentions from free-form text to standard entities in a knowledge base. The unique challenge of this task relative to other entity-related tasks is the lack of surrounding context and numerous variations in the surface form of the mentions, especially when it comes to generalization acr...
http://arxiv.org/pdf/2306.03316v1
2023-06-05T23:58:40
2023-06-05T23:58:40
{ "title": [ -0.014122009, 0.0016031265, -0.05633545, -0.07086182, 0.025604248, -0.016494751, -0.04006958, 0.0005774498, 0.0051116943, 0.035308838, -0.017456055, 0.033111572, 0.019577026, -0.008659363, -0.019683838, 0.010795593, -0.016616821, 0.0...
http://arxiv.org/abs/2306.03315v1
Few Shot Rationale Generation using Self-Training with Dual Teachers
Aditya Srikanth Veerubhotla, Lahari Poddar, Jun Yin, György Szarvas, Sharanya Eswaran
cs.CL, cs.AI
Self-rationalizing models that also generate a free-text explanation for their predicted labels are an important tool to build trustworthy AI applications. Since generating explanations for annotated labels is a laborious and costly pro cess, recent models rely on large pretrained language models (PLMs) as their backbo...
http://arxiv.org/pdf/2306.03315v1
2023-06-05T23:57:52
2023-06-05T23:57:52
{ "title": [ 0.028839111, -0.028717041, 0.002916336, -0.03842163, -0.040740967, -0.0060310364, 0.008583069, 0.013626099, -0.031829834, 0.01109314, 0.024536133, 0.016067505, -0.057159424, -0.04916382, 0.0152282715, -0.019363403, 0.0066719055, 0.03...
http://arxiv.org/abs/2306.03313v1
A Scalable and Adaptive System to Infer the Industry Sectors of Companies: Prompt + Model Tuning of Generative Language Models
Lele Cao, Vilhelm von Ehrenheim, Astrid Berghult, Cecilia Henje, Richard Anselmo Stahl, Joar Wandborg, Sebastian Stan, Armin Catovic, Erik Ferm, Hannes Ingelhag
cs.CL, cs.AI, 68T50, 68T05, I.2.7; I.2.1
The Private Equity (PE) firms operate investment funds by acquiring and managing companies to achieve a high return upon selling. Many PE funds are thematic, meaning investment professionals aim to identify trends by covering as many industry sectors as possible, and picking promising companies within these sectors. So...
http://arxiv.org/pdf/2306.03313v1
2023-06-05T23:55:09
2023-06-05T23:55:09
{ "title": [ -0.053833008, 0.002111435, -0.058532715, -0.030059814, -0.028030396, -0.005683899, -0.0836792, 0.006038666, -0.0030002594, -0.0019550323, -0.037200928, -0.020553589, -0.037475586, -0.03817749, -0.013465881, -0.053375244, -0.02229309, ...
http://arxiv.org/abs/2306.03268v1
Stack Over-Flowing with Results: The Case for Domain-Specific Pre-Training Over One-Size-Fits-All Models
Manisha Mukherjee, Vincent J. Hellendoorn
cs.CL, cs.SE
Large pre-trained neural language models have brought immense progress to both NLP and software engineering. Models in OpenAI's GPT series now dwarf Google's BERT and Meta's RoBERTa, which previously set new benchmarks on a wide range of NLP applications. These models are trained on massive corpora of heterogeneous dat...
http://arxiv.org/pdf/2306.03268v1
2023-06-05T21:38:30
2023-06-05T21:38:30
{ "title": [ 0.0011701584, -0.0104904175, -0.04348755, -0.066589355, -0.00017547607, -0.018814087, -0.03656006, -0.0115356445, -0.023468018, 0.024215698, 0.0069503784, 0.017990112, -0.030349731, -0.037139893, -0.02760315, -0.00712204, 0.005569458, ...
http://arxiv.org/abs/2306.03264v1
shs-nlp at RadSum23: Domain-Adaptive Pre-training of Instruction-tuned LLMs for Radiology Report Impression Generation
Sanjeev Kumar Karn, Rikhiya Ghosh, Kusuma P, Oladimeji Farri
cs.CL
Instruction-tuned generative Large language models (LLMs) like ChatGPT and Bloomz possess excellent generalization abilities, but they face limitations in understanding radiology reports, particularly in the task of generating the IMPRESSIONS section from the FINDINGS section. They tend to generate either verbose or in...
http://arxiv.org/pdf/2306.03264v1
2023-06-05T21:33:04
2023-06-05T21:33:04
{ "title": [ -0.024368286, -0.033966064, -0.044921875, -0.043273926, -0.056488037, -0.00844574, -0.047454834, 0.019943237, -0.0069618225, -0.013893127, -0.028366089, 0.017089844, -0.05328369, -0.042999268, 0.017684937, -0.030212402, -0.012069702, ...
http://arxiv.org/abs/2306.03241v1
Understanding the Effectiveness of Early Weight Averaging for Training Large Language Models
Sunny Sanyal, Jean Kaddour, Abhishek Kumar, Sujay Sanghavi
cs.LG, cs.AI, cs.CL
Training LLMs is expensive, and recent evidence indicates training all the way to convergence is inefficient. In this paper, we investigate the ability of a simple idea, checkpoint averaging along the trajectory of a training run to improve the quality of models before they have converged. This approach incurs no extra...
http://arxiv.org/pdf/2306.03241v1
2023-06-05T20:51:44
2023-06-05T20:51:44
{ "title": [ -0.031585693, -0.008033752, -0.042419434, -0.04336548, -0.004047394, -0.0072364807, -0.05783081, 0.0022678375, 0.0002309084, 0.026443481, -0.00907135, -0.00274086, 0.011817932, -0.035186768, -0.034179688, 0.016052246, 0.00010037422, ...
http://arxiv.org/abs/2306.03208v1
NLU on Data Diets: Dynamic Data Subset Selection for NLP Classification Tasks
Jean-Michel Attendu, Jean-Philippe Corbeil
cs.CL, cs.AI
Finetuning large language models inflates the costs of NLU applications and remains the bottleneck of development cycles. Recent works in computer vision use data pruning to reduce training time. Pruned data selection with static methods is based on a score calculated for each training example prior to finetuning, whic...
http://arxiv.org/pdf/2306.03208v1
2023-06-05T19:30:41
2023-06-05T19:30:41
{ "title": [ -0.026626587, 0.00038075447, -0.06939697, -0.02897644, 0.01537323, -0.03353882, -0.0007824898, -0.00012338161, -0.017654419, 0.008682251, -0.04574585, 0.036010742, -0.04309082, -0.0051994324, -0.031051636, -0.025177002, -0.010902405, ...
http://arxiv.org/abs/2306.03203v1
A Static Evaluation of Code Completion by Large Language Models
Hantian Ding, Varun Kumar, Yuchen Tian, Zijian Wang, Rob Kwiatkowski, Xiaopeng Li, Murali Krishna Ramanathan, Baishakhi Ray, Parminder Bhatia, Sudipta Sengupta, Dan Roth, Bing Xiang
cs.CL, cs.SE
Large language models trained on code have shown great potential to increase productivity of software developers. Several execution-based benchmarks have been proposed to evaluate functional correctness of model-generated code on simple programming problems. Nevertheless, it is expensive to perform the same evaluation ...
http://arxiv.org/pdf/2306.03203v1
2023-06-05T19:23:34
2023-06-05T19:23:34
{ "title": [ -0.019882202, -0.020111084, -0.033843994, -0.082214355, 0.0071029663, 0.004207611, -0.06137085, 0.028549194, 0.008010864, 0.039398193, -0.033172607, 0.009521484, 0.002210617, -0.000113487244, -0.05407715, 0.003162384, -0.02545166, 0....
http://arxiv.org/abs/2306.03197v1
AutoScrum: Automating Project Planning Using Large Language Models
Martin Schroder
cs.AI, cs.CL
Recent advancements in the field of large language models have made it possible to use language models for advanced reasoning. In this paper we leverage this ability for designing complex project plans based only on knowing the current state and the desired state. Two approaches are demonstrated - a scrum based approac...
http://arxiv.org/pdf/2306.03197v1
2023-06-05T19:16:37
2023-06-05T19:16:37
{ "title": [ -0.062805176, -0.027496338, -0.044555664, -0.049224854, -0.0053749084, 0.027511597, -0.07110596, -0.0024299622, -0.0016841888, 0.017501831, -0.045532227, 0.0003273487, -0.009689331, -0.00957489, -0.016098022, -0.016418457, -0.006038666, ...
http://arxiv.org/abs/2306.03189v1
Easy-to-Read in Germany: A Survey on its Current State and Available Resources
Margot Madina, Itziar Gonzalez-Dios, Melanie Siegel
cs.CL
Easy-to-Read Language (E2R) is a controlled language variant that makes any written text more accessible through the use of clear, direct and simple language. It is mainly aimed at people with cognitive or intellectual disabilities, among other target users. Plain Language (PL), on the other hand, is a variant of a giv...
http://arxiv.org/pdf/2306.03189v1
2023-06-05T19:00:25
2023-06-05T19:00:25
{ "title": [ -0.02583313, -0.0039634705, -0.02432251, 0.0015172958, -0.033843994, -0.003440857, -0.014213562, -0.034332275, -0.00869751, 0.049682617, -0.0028381348, -0.00015974045, 0.027679443, -0.05090332, -0.015930176, 0.024291992, 0.014816284, ...
http://arxiv.org/abs/2306.03168v1
Composition and Deformance: Measuring Imageability with a Text-to-Image Model
Si Wu, David A. Smith
cs.CL, cs.CV
Although psycholinguists and psychologists have long studied the tendency of linguistic strings to evoke mental images in hearers or readers, most computational studies have applied this concept of imageability only to isolated words. Using recent developments in text-to-image generation models, such as DALLE mini, we ...
http://arxiv.org/pdf/2306.03168v1
2023-06-05T18:22:23
2023-06-05T18:22:23
{ "title": [ -0.01612854, 0.020080566, -0.006996155, -0.035125732, 0.014595032, 0.009757996, -0.030014038, -0.053222656, 0.032348633, 0.024276733, -0.030792236, 0.005859375, -0.0014858246, -0.013999939, -0.050476074, -0.005836487, -0.002872467, 0...
http://arxiv.org/abs/2306.03166v1
Unsupervised Dense Retrieval with Relevance-Aware Contrastive Pre-Training
Yibin Lei, Liang Ding, Yu Cao, Changtong Zan, Andrew Yates, Dacheng Tao
cs.IR, cs.CL
Dense retrievers have achieved impressive performance, but their demand for abundant training data limits their application scenarios. Contrastive pre-training, which constructs pseudo-positive examples from unlabeled data, has shown great potential to solve this problem. However, the pseudo-positive examples crafted b...
http://arxiv.org/pdf/2306.03166v1
2023-06-05T18:20:27
2023-06-05T18:20:27
{ "title": [ 0.029174805, 0.024536133, -0.016952515, -0.060028076, 0.030456543, -0.006931305, -0.010284424, -0.033721924, 0.039031982, 0.033081055, -0.04019165, 0.009315491, 0.005596161, 0.0032348633, -0.050109863, 0.0031528473, -0.00047707558, 0...
http://arxiv.org/abs/2306.03091v2
RepoBench: Benchmarking Repository-Level Code Auto-Completion Systems
Tianyang Liu, Canwen Xu, Julian McAuley
cs.CL, cs.AI, cs.SE
Large Language Models (LLMs) have greatly advanced code auto-completion systems, with a potential for substantial productivity enhancements for developers. However, current benchmarks mainly focus on single-file tasks, leaving an assessment gap for more complex, real-world, multi-file programming scenarios. To fill thi...
http://arxiv.org/pdf/2306.03091v2
2023-10-04T01:13:49
2023-06-05T17:59:41
{ "title": [ -0.03338623, -0.01979065, -0.037872314, -0.08404541, 0.00008362532, 0.013381958, -0.04611206, 0.024261475, 0.01423645, 0.04937744, -0.057159424, 0.042816162, 0.018692017, 0.01423645, -0.03466797, 0.0082473755, -0.008094788, 0.0028820...
http://arxiv.org/abs/2306.03090v1
Is ChatGPT a Good Teacher Coach? Measuring Zero-Shot Performance For Scoring and Providing Actionable Insights on Classroom Instruction
Rose E. Wang, Dorottya Demszky
cs.CL, cs.AI
Coaching, which involves classroom observation and expert feedback, is a widespread and fundamental part of teacher training. However, the majority of teachers do not have access to consistent, high quality coaching due to limited resources and access to expertise. We explore whether generative AI could become a cost-e...
http://arxiv.org/pdf/2306.03090v1
2023-06-05T17:59:21
2023-06-05T17:59:21
{ "title": [ 0.010375977, -0.01977539, 0.009742737, 0.026672363, -0.0335083, -0.02268982, 0.028121948, -0.02029419, -0.015991211, 0.018859863, -0.024902344, 0.013870239, -0.013000488, -0.018188477, 0.011833191, -0.030654907, -0.000269413, 0.01016...
http://arxiv.org/abs/2306.03081v1
Sequential Monte Carlo Steering of Large Language Models using Probabilistic Programs
Alexander K. Lew, Tan Zhi-Xuan, Gabriel Grand, Vikash K. Mansinghka
cs.AI, cs.CL, cs.PL, stat.CO
Even after fine-tuning and reinforcement learning, large language models (LLMs) can be difficult, if not impossible, to control reliably with prompts alone. We propose a new inference-time approach to enforcing syntactic and semantic constraints on the outputs of LLMs, called sequential Monte Carlo (SMC) steering. The ...
http://arxiv.org/pdf/2306.03081v1
2023-06-05T17:55:05
2023-06-05T17:55:05
{ "title": [ -0.05545044, -0.0018320084, -0.039886475, -0.07861328, -0.015144348, 0.019332886, -0.09869385, 0.0005335808, -0.009773254, 0.006668091, -0.0006289482, 0.008712769, 0.011764526, -0.04244995, -0.0005455017, 0.013725281, -0.02583313, 0....
http://arxiv.org/abs/2306.03079v1
Machine Learning and Statistical Approaches to Measuring Similarity of Political Parties
Daria Boratyn, Damian Brzyski, Beata Kosowska-Gąstoł, Jan Rybicki, Wojciech Słomczyński, Dariusz Stolicki
cs.CL, 91F10 (Primary) 68T50 (Secondary), J.4; I.2.7
Mapping political party systems to metric policy spaces is one of the major methodological problems in political science. At present, in most political science project this task is performed by domain experts relying on purely qualitative assessments, with all the attendant problems of subjectivity and labor intensiven...
http://arxiv.org/pdf/2306.03079v1
2023-06-05T17:53:41
2023-06-05T17:53:41
{ "title": [ -0.00447464, 0.017730713, -0.05343628, -0.028381348, 0.0062789917, -0.036468506, -0.0014410019, -0.026931763, -0.019943237, 0.03567505, -0.033477783, 0.0140686035, -0.00919342, -0.025939941, -0.01133728, -0.010398865, 0.007030487, 0....
http://arxiv.org/abs/2306.03078v1
SpQR: A Sparse-Quantized Representation for Near-Lossless LLM Weight Compression
Tim Dettmers, Ruslan Svirschevski, Vage Egiazarian, Denis Kuznedelev, Elias Frantar, Saleh Ashkboos, Alexander Borzunov, Torsten Hoefler, Dan Alistarh
cs.CL, cs.LG
Recent advances in large language model (LLM) pretraining have led to high-quality LLMs with impressive abilities. By compressing such LLMs via quantization to 3-4 bits per parameter, they can fit into memory-limited devices such as laptops and mobile phones, enabling personalized use. However, quantization down to 3-4...
http://arxiv.org/pdf/2306.03078v1
2023-06-05T17:53:28
2023-06-05T17:53:28
{ "title": [ -0.093811035, -0.03744507, -0.059814453, -0.062316895, -0.04248047, 0.0154418945, -0.08288574, 0.04550171, -0.0034770966, 0.05105591, 0.019866943, 0.0152282715, -0.016799927, -0.033447266, -0.040039062, 0.017562866, 0.01776123, -0.01...
http://arxiv.org/abs/2306.03067v1
Interactive Editing for Text Summarization
Yujia Xie, Xun Wang, Si-Qing Chen, Wayne Xiong, Pengcheng He
cs.CL, cs.AI
Summarizing lengthy documents is a common and essential task in our daily lives. Although recent advancements in neural summarization models can assist in crafting general-purpose summaries, human writers often have specific requirements that call for a more customized approach. To address this need, we introduce REVIS...
http://arxiv.org/pdf/2306.03067v1
2023-06-05T17:43:53
2023-06-05T17:43:53
{ "title": [ -0.011161804, 0.001750946, -0.016403198, -0.05859375, -0.025314331, 0.008399963, -0.038909912, -0.0078086853, -0.0025520325, 0.054901123, -0.016281128, 0.005001068, -0.0033798218, -0.0011548996, -0.06500244, -0.00044202805, -0.007217407,...
http://arxiv.org/abs/2306.03061v1
Structured Voronoi Sampling
Afra Amini, Li Du, Ryan Cotterell
cs.CL, cs.AI
Recently, there has been a growing interest in the development of gradient-based sampling algorithms for text generation, especially in the context of controlled generation. However, there exists a lack of theoretically grounded and principled approaches for this task. In this paper, we take an important step toward bu...
http://arxiv.org/pdf/2306.03061v1
2023-06-05T17:32:35
2023-06-05T17:32:35
{ "title": [ 0.003396988, 0.016098022, -0.022735596, -0.028045654, -0.02998352, -0.012931824, -0.027297974, -0.04751587, 0.021728516, 0.015777588, -0.08117676, 0.017181396, -0.03604126, -0.020217896, -0.0038661957, -0.0075798035, 0.032989502, -0....
http://arxiv.org/abs/2306.03055v1
Analyzing Syntactic Generalization Capacity of Pre-trained Language Models on Japanese Honorific Conversion
Ryo Sekizawa, Hitomi Yanaka
cs.CL
Using Japanese honorifics is challenging because it requires not only knowledge of the grammatical rules but also contextual information, such as social relationships. It remains unclear whether pre-trained large language models (LLMs) can flexibly handle Japanese honorifics like humans. To analyze this, we introduce a...
http://arxiv.org/pdf/2306.03055v1
2023-06-05T17:27:48
2023-06-05T17:27:48
{ "title": [ -0.000027656555, 0.015296936, -0.021499634, -0.05706787, -0.0047035217, 0.005077362, -0.06121826, -0.018478394, 0.0049743652, 0.049713135, 0.015144348, -0.00932312, 0.01525116, -0.026931763, -0.06124878, 0.0007481575, -0.019561768, 0...
http://arxiv.org/abs/2306.06077v2
Visually-Grounded Descriptions Improve Zero-Shot Image Classification
Michael Ogezi, Bradley Hauer, Grzegorz Kondrak
cs.CV, cs.AI, cs.CL
Language-vision models like CLIP have made significant progress in zero-shot vision tasks, such as zero-shot image classification (ZSIC). However, generating specific and expressive class descriptions remains a major challenge. Existing approaches suffer from granularity and label ambiguity issues. To tackle these chal...
http://arxiv.org/pdf/2306.06077v2
2023-06-23T16:29:51
2023-06-05T17:22:54
{ "title": [ 0.024398804, 0.009384155, -0.03414917, -0.03201294, 0.0035495758, -0.046447754, 0.007255554, -0.017837524, 0.027770996, 0.035308838, -0.029663086, 0.015548706, -0.012062073, -0.007972717, -0.059631348, -0.000023126602, 0.0046577454, ...
http://arxiv.org/abs/2306.03030v3
Benchmarking Large Language Models on CMExam -- A Comprehensive Chinese Medical Exam Dataset
Junling Liu, Peilin Zhou, Yining Hua, Dading Chong, Zhongyu Tian, Andrew Liu, Helin Wang, Chenyu You, Zhenhua Guo, Lei Zhu, Michael Lingzhi Li
cs.CL
Recent advancements in large language models (LLMs) have transformed the field of question answering (QA). However, evaluating LLMs in the medical field is challenging due to the lack of standardized and comprehensive datasets. To address this gap, we introduce CMExam, sourced from the Chinese National Medical Licensin...
http://arxiv.org/pdf/2306.03030v3
2023-10-23T02:55:08
2023-06-05T16:48:41
{ "title": [ -0.026519775, -0.007881165, -0.030532837, -0.0552063, -0.00749588, -0.0023822784, -0.050048828, 0.0018568039, 0.0053215027, 0.037719727, -0.0020999908, -0.0053749084, 0.022705078, -0.025421143, -0.042419434, 0.002111435, -0.009613037, ...
http://arxiv.org/abs/2306.03024v1
PokemonChat: Auditing ChatGPT for Pokémon Universe Knowledge
Laura Cabello, Jiaang Li, Ilias Chalkidis
cs.CL
The recently released ChatGPT model demonstrates unprecedented capabilities in zero-shot question-answering. In this work, we probe ChatGPT for its conversational understanding and introduce a conversational framework (protocol) that can be adopted in future studies. The Pok\'emon universe serves as an ideal testing gr...
http://arxiv.org/pdf/2306.03024v1
2023-06-05T16:44:27
2023-06-05T16:44:27
{ "title": [ -0.012504578, -0.055603027, -0.017990112, -0.023742676, -0.058013916, 0.010696411, -0.02268982, 0.0049819946, 0.01108551, 0.022903442, -0.06365967, 0.03353882, 0.0019950867, -0.005722046, -0.032714844, -0.0099487305, -0.017868042, 0....
http://arxiv.org/abs/2306.02982v2
PolyVoice: Language Models for Speech to Speech Translation
Qianqian Dong, Zhiying Huang, Qiao Tian, Chen Xu, Tom Ko, Yunlong Zhao, Siyuan Feng, Tang Li, Kexin Wang, Xuxin Cheng, Fengpeng Yue, Ye Bai, Xi Chen, Lu Lu, Zejun Ma, Yuping Wang, Mingxuan Wang, Yuxuan Wang
cs.CL, eess.AS
We propose PolyVoice, a language model-based framework for speech-to-speech translation (S2ST) system. Our framework consists of two language models: a translation language model and a speech synthesis language model. We use discretized speech units, which are generated in a fully unsupervised way, and thus our framewo...
http://arxiv.org/pdf/2306.02982v2
2023-06-13T15:15:17
2023-06-05T15:53:15
{ "title": [ -0.03274536, -0.025405884, -0.020492554, -0.07104492, 0.0042762756, 0.022888184, -0.0552063, 0.012702942, 0.0054512024, 0.039978027, -0.005748749, 0.0017166138, 0.035003662, -0.030136108, -0.041534424, 0.019821167, -0.0054473877, 0.0...
http://arxiv.org/abs/2306.02980v1
KNOW How to Make Up Your Mind! Adversarially Detecting and Alleviating Inconsistencies in Natural Language Explanations
Myeongjun Jang, Bodhisattwa Prasad Majumder, Julian McAuley, Thomas Lukasiewicz, Oana-Maria Camburu
cs.CL, cs.AI
While recent works have been considerably improving the quality of the natural language explanations (NLEs) generated by a model to justify its predictions, there is very limited research in detecting and alleviating inconsistencies among generated NLEs. In this work, we leverage external knowledge bases to significant...
http://arxiv.org/pdf/2306.02980v1
2023-06-05T15:51:58
2023-06-05T15:51:58
{ "title": [ -0.01689148, -0.013916016, -0.036193848, -0.054107666, -0.0021858215, -0.009841919, -0.042541504, -0.006214142, -0.03918457, 0.0017433167, -0.036468506, 0.016860962, -0.02192688, -0.025604248, -0.0087509155, -0.05630493, -0.0033340454, ...
http://arxiv.org/abs/2306.02978v1
Which Argumentative Aspects of Hate Speech in Social Media can be reliably identified?
Damián Furman, Pablo Torres, José A. Rodríguez, Diego Letzen, Vanina Martínez, Laura Alonso Alemany
cs.CL, cs.AI
With the increasing diversity of use cases of large language models, a more informative treatment of texts seems necessary. An argumentative analysis could foster a more reasoned usage of chatbots, text completion mechanisms or other applications. However, it is unclear which aspects of argumentation can be reliably id...
http://arxiv.org/pdf/2306.02978v1
2023-06-05T15:50:57
2023-06-05T15:50:57
{ "title": [ 0.07977295, 0.011451721, -0.003156662, -0.018814087, 0.0084991455, -0.03616333, 0.01966858, -0.036834717, -0.02720642, 0.018829346, 0.024551392, 0.0060043335, 0.019638062, -0.061462402, 0.038085938, -0.010055542, 0.0020008087, 0.0764...
http://arxiv.org/abs/2306.02955v1
A Simple and Flexible Modeling for Mental Disorder Detection by Learning from Clinical Questionnaires
Hoyun Song, Jisu Shin, Huije Lee, Jong C. Park
cs.CL, cs.LG
Social media is one of the most highly sought resources for analyzing characteristics of the language by its users. In particular, many researchers utilized various linguistic features of mental health problems from social media. However, existing approaches to detecting mental disorders face critical challenges, such ...
http://arxiv.org/pdf/2306.02955v1
2023-06-05T15:23:55
2023-06-05T15:23:55
{ "title": [ 0.003967285, 0.012184143, -0.031433105, -0.052642822, 0.03817749, -0.014846802, -0.01939392, -0.028961182, 0.012321472, 0.013061523, -0.00044369698, -0.027664185, -0.021652222, -0.047943115, -0.050720215, -0.013198853, -0.0015926361, ...
http://arxiv.org/abs/2306.02923v2
MidMed: Towards Mixed-Type Dialogues for Medical Consultation
Xiaoming Shi, Zeming Liu, Chuan Wang, Haitao Leng, Kui Xue, Xiaofan Zhang, Shaoting Zhang
cs.CL
Most medical dialogue systems assume that patients have clear goals (medicine querying, surgical operation querying, etc.) before medical consultation. However, in many real scenarios, due to the lack of medical knowledge, it is usually difficult for patients to determine clear goals with all necessary slots. In this p...
http://arxiv.org/pdf/2306.02923v2
2023-06-14T02:56:06
2023-06-05T14:36:31
{ "title": [ -0.0003604889, -0.0121536255, -0.039855957, -0.08001709, 0.025741577, -0.00023174286, -0.032287598, 0.009887695, -0.0093307495, 0.047180176, 0.009780884, 0.0034542084, 0.0026512146, -0.025863647, -0.052215576, 0.0034561157, 0.008377075, ...
http://arxiv.org/abs/2306.02920v1
Second Language Acquisition of Neural Language Models
Miyu Oba, Tatsuki Kuribayashi, Hiroki Ouchi, Taro Watanabe
cs.CL
With the success of neural language models (LMs), their language acquisition has gained much attention. This work sheds light on the second language (L2) acquisition of LMs, while previous work has typically explored their first language (L1) acquisition. Specifically, we trained bilingual LMs with a scenario similar t...
http://arxiv.org/pdf/2306.02920v1
2023-06-05T14:32:41
2023-06-05T14:32:41
{ "title": [ -0.030349731, -0.0014753342, -0.013450623, -0.0054130554, -0.014976501, 0.0008330345, -0.020401001, -0.011917114, -0.025375366, 0.04083252, -0.0030879974, 0.0030117035, -0.004005432, -0.031951904, -0.036499023, 0.011459351, -0.020904541,...
http://arxiv.org/abs/2306.02907v1
SelfEvolve: A Code Evolution Framework via Large Language Models
Shuyang Jiang, Yuhao Wang, Yu Wang
cs.CL, cs.SE
Large language models (LLMs) have already revolutionized code generation, after being pretrained on publicly available code data. However, while various methods have been proposed to augment LLMs with retrieved knowledge and enhance the quality of code generation, the performance of these retrieval-based methods is lim...
http://arxiv.org/pdf/2306.02907v1
2023-06-05T14:12:46
2023-06-05T14:12:46
{ "title": [ -0.048187256, -0.02507019, -0.03857422, -0.060913086, -0.0018119812, 0.009429932, -0.043395996, 0.019241333, -0.00095272064, 0.03479004, 0.0027694702, 0.0034046173, -0.010185242, -0.023620605, -0.035339355, 0.0050735474, -0.009864807, ...
http://arxiv.org/abs/2306.02902v1
N-Shot Benchmarking of Whisper on Diverse Arabic Speech Recognition
Bashar Talafha, Abdul Waheed, Muhammad Abdul-Mageed
cs.CL, cs.SD, eess.AS
Whisper, the recently developed multilingual weakly supervised model, is reported to perform well on multiple speech recognition benchmarks in both monolingual and multilingual settings. However, it is not clear how Whisper would fare under diverse conditions even on languages it was evaluated on such as Arabic. In thi...
http://arxiv.org/pdf/2306.02902v1
2023-06-05T14:09:25
2023-06-05T14:09:25
{ "title": [ -0.020889282, -0.020843506, -0.01991272, -0.066467285, -0.0259552, 0.002117157, -0.047180176, 0.016021729, -0.03527832, 0.03967285, -0.012336731, 0.0074920654, -0.030441284, -0.037017822, -0.028808594, -0.02355957, 0.016830444, 0.007...
http://arxiv.org/abs/2306.02887v2
Gen-IR @ SIGIR 2023: The First Workshop on Generative Information Retrieval
Gabriel Bénédict, Ruqing Zhang, Donald Metzler
cs.IR, cs.CL
Generative information retrieval (IR) has experienced substantial growth across multiple research communities (e.g., information retrieval, computer vision, natural language processing, and machine learning), and has been highly visible in the popular press. Theoretical, empirical, and actual user-facing products have ...
http://arxiv.org/pdf/2306.02887v2
2023-06-13T15:20:13
2023-06-05T13:56:36
{ "title": [ 0.0042686462, -0.008628845, -0.046539307, -0.049560547, 0.011276245, -0.0079422, -0.009216309, -0.016189575, 0.026168823, 0.0209198, -0.038269043, 0.02684021, -0.008865356, -0.019073486, -0.024765015, 0.020141602, -0.009147644, 0.026...
http://arxiv.org/abs/2306.02873v1
DecompX: Explaining Transformers Decisions by Propagating Token Decomposition
Ali Modarressi, Mohsen Fayyaz, Ehsan Aghazadeh, Yadollah Yaghoobzadeh, Mohammad Taher Pilehvar
cs.CL
An emerging solution for explaining Transformer-based models is to use vector-based analysis on how the representations are formed. However, providing a faithful vector-based explanation for a multi-layer model could be challenging in three aspects: (1) Incorporating all components into the analysis, (2) Aggregating th...
http://arxiv.org/pdf/2306.02873v1
2023-06-05T13:46:31
2023-06-05T13:46:31
{ "title": [ -0.0050735474, -0.007686615, -0.0063323975, -0.06323242, -0.0011329651, 0.017654419, -0.032989502, -0.026382446, -0.021621704, 0.0055999756, -0.047698975, 0.021255493, -0.013214111, -0.016815186, 0.0030136108, -0.017547607, -0.011856079,...
http://arxiv.org/abs/2306.02871v1
Text-To-KG Alignment: Comparing Current Methods on Classification Tasks
Sondre Wold, Lilja Øvrelid, Erik Velldal
cs.CL
In contrast to large text corpora, knowledge graphs (KG) provide dense and structured representations of factual information. This makes them attractive for systems that supplement or ground the knowledge found in pre-trained language models with an external knowledge source. This has especially been the case for class...
http://arxiv.org/pdf/2306.02871v1
2023-06-05T13:45:45
2023-06-05T13:45:45
{ "title": [ -0.07220459, 0.0011358261, -0.044647217, 0.0033359528, -0.019104004, -0.0065841675, -0.046722412, -0.027832031, 0.006752014, 0.05407715, -0.0061073303, 0.00014436245, 0.0054473877, -0.020385742, -0.06732178, -0.0029335022, 0.0007004738, ...
http://arxiv.org/abs/2306.02870v1
On "Scientific Debt" in NLP: A Case for More Rigour in Language Model Pre-Training Research
Made Nindyatama Nityasya, Haryo Akbarianto Wibowo, Alham Fikri Aji, Genta Indra Winata, Radityo Eko Prasojo, Phil Blunsom, Adhiguna Kuncoro
cs.CL
This evidence-based position paper critiques current research practices within the language model pre-training literature. Despite rapid recent progress afforded by increasingly better pre-trained language models (PLMs), current PLM research practices often conflate different possible sources of model improvement, with...
http://arxiv.org/pdf/2306.02870v1
2023-06-05T13:43:50
2023-06-05T13:43:50
{ "title": [ -0.0209198, -0.014083862, -0.033996582, -0.046447754, -0.019134521, -0.010269165, -0.043395996, -0.011726379, -0.028823853, 0.014511108, -0.02420044, 0.01335144, 0.0006918907, -0.019897461, -0.023452759, -0.0061569214, -0.022415161, ...
http://arxiv.org/abs/2306.02864v2
Leveraging Large Language Models for Topic Classification in the Domain of Public Affairs
Alejandro Peña, Aythami Morales, Julian Fierrez, Ignacio Serna, Javier Ortega-Garcia, Iñigo Puente, Jorge Cordova, Gonzalo Cordova
cs.AI, cs.CL
The analysis of public affairs documents is crucial for citizens as it promotes transparency, accountability, and informed decision-making. It allows citizens to understand government policies, participate in public discourse, and hold representatives accountable. This is crucial, and sometimes a matter of life or deat...
http://arxiv.org/pdf/2306.02864v2
2023-08-08T09:48:36
2023-06-05T13:35:01
{ "title": [ -0.005050659, 0.0013046265, -0.037872314, -0.04711914, 0.0045776367, -0.010238647, -0.0395813, -0.01576233, 0.0058631897, 0.01574707, -0.032562256, 0.01222229, -0.021636963, -0.011772156, -0.028839111, -0.014793396, -0.018737793, 0.0...
http://arxiv.org/abs/2306.02858v4
Video-LLaMA: An Instruction-tuned Audio-Visual Language Model for Video Understanding
Hang Zhang, Xin Li, Lidong Bing
cs.CL, cs.CV, cs.SD, eess.AS
We present Video-LLaMA a multi-modal framework that empowers Large Language Models (LLMs) with the capability of understanding both visual and auditory content in the video. Video-LLaMA bootstraps cross-modal training from the frozen pre-trained visual and audio encoders and the frozen LLMs. Unlike previous works that ...
http://arxiv.org/pdf/2306.02858v4
2023-10-25T06:23:31
2023-06-05T13:17:27
{ "title": [ -0.030776978, -0.0061073303, -0.021835327, -0.051757812, -0.012458801, -0.00076436996, -0.04849243, -0.017990112, 0.005077362, 0.023513794, -0.03375244, 0.006000519, -0.010238647, -0.011016846, -0.049621582, -0.018920898, -0.008049011, ...
http://arxiv.org/abs/2306.02842v1
Improving Conversational Recommendation Systems via Counterfactual Data Simulation
Xiaolei Wang, Kun Zhou, Xinyu Tang, Wayne Xin Zhao, Fan Pan, Zhao Cao, Ji-Rong Wen
cs.CL, cs.IR
Conversational recommender systems (CRSs) aim to provide recommendation services via natural language conversations. Although a number of approaches have been proposed for developing capable CRSs, they typically rely on sufficient training data for training. Since it is difficult to annotate recommendation-oriented dia...
http://arxiv.org/pdf/2306.02842v1
2023-06-05T12:48:56
2023-06-05T12:48:56
{ "title": [ 0.036224365, -0.0043258667, -0.020263672, -0.03488159, -0.0016956329, -0.020263672, -0.011428833, -0.020462036, 0.0046653748, 0.02507019, -0.029052734, 0.019378662, -0.017593384, -0.013748169, -0.041992188, -0.020568848, -0.024032593, ...
http://arxiv.org/abs/2306.02840v1
Learning to Substitute Spans towards Improving Compositional Generalization
Zhaoyi Li, Ying Wei, Defu Lian
cs.CL, cs.LG
Despite the rising prevalence of neural sequence models, recent empirical evidences suggest their deficiency in compositional generalization. One of the current de-facto solutions to this problem is compositional data augmentation, aiming to incur additional compositional inductive bias. Nonetheless, the improvement of...
http://arxiv.org/pdf/2306.02840v1
2023-06-05T12:44:18
2023-06-05T12:44:18
{ "title": [ -0.018432617, -0.011413574, -0.020812988, -0.06591797, -0.0071487427, 0.008399963, -0.05831909, 0.020706177, -0.014015198, 0.008842468, -0.026992798, 0.013725281, -0.038269043, -0.013092041, -0.02130127, -0.031433105, -0.036987305, 0...
http://arxiv.org/abs/2306.02827v2
UNIDECOR: A Unified Deception Corpus for Cross-Corpus Deception Detection
Aswathy Velutharambath, Roman Klinger
cs.CL
Verbal deception has been studied in psychology, forensics, and computational linguistics for a variety of reasons, like understanding behaviour patterns, identifying false testimonies, and detecting deception in online communication. Varying motivations across research fields lead to differences in the domain choices ...
http://arxiv.org/pdf/2306.02827v2
2023-06-07T23:07:26
2023-06-05T12:23:04
{ "title": [ -0.009735107, 0.0017442703, -0.03778076, -0.032226562, 0.014541626, -0.012413025, -0.017227173, -0.028533936, 0.0029850006, 0.026824951, -0.029220581, 0.003829956, 0.019958496, -0.006134033, -0.03741455, -0.0001026392, -0.04876709, 0...
http://arxiv.org/abs/2306.02819v1
Enhancing Language Representation with Constructional Information for Natural Language Understanding
Lvxiaowei Xu, Jianwang Wu, Jiawei Peng, Zhilin Gong, Ming Cai, Tianxiang Wang
cs.CL
Natural language understanding (NLU) is an essential branch of natural language processing, which relies on representations generated by pre-trained language models (PLMs). However, PLMs primarily focus on acquiring lexico-semantic information, while they may be unable to adequately handle the meaning of constructions....
http://arxiv.org/pdf/2306.02819v1
2023-06-05T12:15:12
2023-06-05T12:15:12
{ "title": [ -0.011695862, -0.010108948, -0.026290894, -0.07879639, -0.01083374, 0.011787415, -0.072509766, -0.004432678, 0.013198853, 0.006542206, -0.027175903, 0.0054092407, -0.014984131, -0.017929077, -0.034454346, -0.0025615692, -0.01436615, ...
http://arxiv.org/abs/2306.02797v3
Human-like Few-Shot Learning via Bayesian Reasoning over Natural Language
Kevin Ellis
cs.CL, cs.AI, cs.LG
A core tension in models of concept learning is that the model must carefully balance the tractability of inference against the expressivity of the hypothesis class. Humans, however, can efficiently learn a broad range of concepts. We introduce a model of inductive learning that seeks to be human-like in that sense. It...
http://arxiv.org/pdf/2306.02797v3
2023-09-29T16:45:41
2023-06-05T11:46:45
{ "title": [ -0.0063209534, -0.022750854, -0.026031494, -0.036895752, -0.01197052, -0.01701355, -0.008171082, 0.009262085, -0.0024929047, 0.007888794, -0.044128418, 0.015022278, -0.029037476, -0.0036773682, -0.03567505, -0.016952515, 0.0031204224, ...
http://arxiv.org/abs/2306.02796v1
MCTS: A Multi-Reference Chinese Text Simplification Dataset
Ruining Chong, Luming Lu, Liner Yang, Jinran Nie, Shuhan Zhou, Yaoxin Li, Erhong Yang
cs.CL
Text simplification aims to make the text easier to understand by applying rewriting transformations. There has been very little research on Chinese text simplification for a long time. The lack of generic evaluation data is an essential reason for this phenomenon. In this paper, we introduce MCTS, a multi-reference Ch...
http://arxiv.org/pdf/2306.02796v1
2023-06-05T11:46:36
2023-06-05T11:46:36
{ "title": [ -0.034606934, 0.0015420914, -0.039123535, -0.06695557, 0.016143799, 0.007987976, -0.06427002, -0.0032997131, 0.021484375, 0.07006836, -0.005340576, 0.024902344, 0.040649414, -0.008056641, -0.053863525, 0.026824951, -0.001917839, 0.02...
http://arxiv.org/abs/2306.02790v1
Exploring the Relationship between Alignment and Cross-lingual Transfer in Multilingual Transformers
Félix Gaschi, Patricio Cerda, Parisa Rastin, Yannick Toussaint
cs.CL, cs.AI
Without any explicit cross-lingual training data, multilingual language models can achieve cross-lingual transfer. One common way to improve this transfer is to perform realignment steps before fine-tuning, i.e., to train the model to build similar representations for pairs of words from translated sentences. But such ...
http://arxiv.org/pdf/2306.02790v1
2023-06-05T11:35:40
2023-06-05T11:35:40
{ "title": [ -0.020812988, 0.004386902, -0.040618896, -0.035125732, 0.009422302, -0.009117126, -0.03265381, -0.01878357, -0.012680054, 0.061645508, -0.0035305023, 0.019744873, 0.04333496, -0.019866943, -0.03274536, 0.019241333, -0.005264282, 0.04...
http://arxiv.org/abs/2306.02777v1
German CheXpert Chest X-ray Radiology Report Labeler
Alessandro Wollek, Sardi Hyska, Thomas Sedlmeyr, Philip Haitzer, Johannes Rueckel, Bastian O. Sabel, Michael Ingrisch, Tobias Lasser
cs.CL
This study aimed to develop an algorithm to automatically extract annotations for chest X-ray classification models from German thoracic radiology reports. An automatic label extraction model was designed based on the CheXpert architecture, and a web-based annotation interface was created for iterative improvements. Re...
http://arxiv.org/pdf/2306.02777v1
2023-06-05T11:01:58
2023-06-05T11:01:58
{ "title": [ -0.0017347336, -0.016021729, -0.047912598, -0.023086548, -0.021408081, -0.023971558, -0.051971436, -0.012672424, -0.004055023, 0.03289795, -0.0052108765, -0.0006132126, 0.03137207, -0.02923584, 0.012260437, -0.005554199, 0.009651184, ...
http://arxiv.org/abs/2306.02771v1
Identifying the style by a qualified reader on a short fragment of generated poetry
Boris Orekhov
cs.CL, cs.AI, cs.LG
Style is an important concept in today's challenges in natural language generating. After the success in the field of image style transfer, the task of text style transfer became actual and attractive. Researchers are also interested in the tasks of style reproducing in generation of the poetic text. Evaluation of styl...
http://arxiv.org/pdf/2306.02771v1
2023-06-05T10:55:15
2023-06-05T10:55:15
{ "title": [ 0.005783081, -0.010002136, -0.052734375, -0.0013971329, -0.030334473, -0.03164673, -0.011016846, -0.001821518, -0.027832031, 0.04159546, 0.009475708, -0.027023315, 0.031341553, -0.06890869, 0.017364502, -0.017364502, 0.013053894, 0.0...
http://arxiv.org/abs/2306.02767v1
Cross-Lingual Transfer with Target Language-Ready Task Adapters
Marinela Parović, Alan Ansell, Ivan Vulić, Anna Korhonen
cs.CL
Adapters have emerged as a modular and parameter-efficient approach to (zero-shot) cross-lingual transfer. The established MAD-X framework employs separate language and task adapters which can be arbitrarily combined to perform the transfer of any task to any target language. Subsequently, BAD-X, an extension of the MA...
http://arxiv.org/pdf/2306.02767v1
2023-06-05T10:46:33
2023-06-05T10:46:33
{ "title": [ -0.022094727, -0.0096588135, -0.033599854, -0.052703857, 0.0036849976, -0.00017344952, -0.030731201, -0.014335632, 0.004299164, 0.0446167, 0.016418457, -0.0013694763, 0.027572632, -0.025604248, -0.032470703, 0.031585693, 0.008377075, ...
http://arxiv.org/abs/2306.02754v1
PULSAR: Pre-training with Extracted Healthcare Terms for Summarising Patients' Problems and Data Augmentation with Black-box Large Language Models
Hao Li, Yuping Wu, Viktor Schlegel, Riza Batista-Navarro, Thanh-Tung Nguyen, Abhinav Ramesh Kashyap, Xiaojun Zeng, Daniel Beck, Stefan Winkler, Goran Nenadic
cs.CL
Medical progress notes play a crucial role in documenting a patient's hospital journey, including his or her condition, treatment plan, and any updates for healthcare providers. Automatic summarisation of a patient's problems in the form of a problem list can aid stakeholders in understanding a patient's condition, red...
http://arxiv.org/pdf/2306.02754v1
2023-06-05T10:17:50
2023-06-05T10:17:50
{ "title": [ 0.029907227, -0.014297485, -0.039093018, -0.076416016, -0.023925781, -0.0077438354, -0.06896973, -0.017288208, -0.009628296, 0.027145386, -0.025299072, 0.016204834, -0.00082731247, -0.01096344, -0.004962921, -0.03125, 0.0061836243, 0...
http://arxiv.org/abs/2306.02719v1
Multiple output samples for each input in a single-output Gaussian process
Jeremy H. M. Wong, Huayun Zhang, Nancy F. Chen
cs.CL, cs.LG, cs.SD, eess.AS
The standard Gaussian Process (GP) only considers a single output sample per input in the training set. Datasets for subjective tasks, such as spoken language assessment, may be annotated with output labels from multiple human raters per input. This paper proposes to generalise the GP to allow for these multiple output...
http://arxiv.org/pdf/2306.02719v1
2023-06-05T09:12:34
2023-06-05T09:12:34
{ "title": [ 0.033081055, -0.02154541, -0.011909485, -0.03378296, -0.046173096, -0.046722412, -0.011833191, -0.019973755, -0.05380249, 0.014640808, -0.027374268, 0.03527832, -0.017288208, -0.04119873, 0.056640625, -0.0024642944, 0.031234741, -0.0...
http://arxiv.org/abs/2306.02707v1
Orca: Progressive Learning from Complex Explanation Traces of GPT-4
Subhabrata Mukherjee, Arindam Mitra, Ganesh Jawahar, Sahaj Agarwal, Hamid Palangi, Ahmed Awadallah
cs.CL, cs.LG
Recent research has focused on enhancing the capability of smaller models through imitation learning, drawing on the outputs generated by large foundation models (LFMs). A number of issues impact the quality of these models, ranging from limited imitation signals from shallow LFM outputs; small scale homogeneous traini...
http://arxiv.org/pdf/2306.02707v1
2023-06-05T08:58:39
2023-06-05T08:58:39
{ "title": [ -0.0009651184, -0.005680084, -0.021911621, -0.05795288, 0.007259369, -0.011184692, -0.008300781, -0.016906738, -0.010253906, -0.01272583, 0.0047683716, 0.010986328, -0.036895752, -0.057250977, -0.00573349, 0.0045051575, -0.010528564, ...
http://arxiv.org/abs/2306.05431v1
LexGPT 0.1: pre-trained GPT-J models with Pile of Law
Jieh-Sheng Lee
cs.CL
This research aims to build generative language models specialized for the legal domain. The manuscript presents the development of LexGPT models based on GPT-J models and pre-trained with Pile of Law. The foundation model built in this manuscript is the initial step for the development of future applications in the le...
http://arxiv.org/pdf/2306.05431v1
2023-06-05T08:42:59
2023-06-05T08:42:59
{ "title": [ -0.0118255615, -0.027069092, -0.02822876, -0.06774902, -0.026885986, 0.0024433136, -0.07348633, -0.012969971, -0.027679443, 0.015037537, -0.028503418, 0.005355835, -0.0049819946, -0.042114258, 0.0009288788, -0.008811951, -0.02255249, ...
http://arxiv.org/abs/2306.02693v2
CELDA: Leveraging Black-box Language Model as Enhanced Classifier without Labels
Hyunsoo Cho, Youna Kim, Sang-goo Lee
cs.CL
Utilizing language models (LMs) without internal access is becoming an attractive paradigm in the field of NLP as many cutting-edge LMs are released through APIs and boast a massive scale. The de-facto method in this type of black-box scenario is known as prompting, which has shown progressive performance enhancements ...
http://arxiv.org/pdf/2306.02693v2
2023-06-09T05:16:21
2023-06-05T08:35:31
{ "title": [ -0.018722534, -0.013206482, -0.044952393, -0.0637207, 0.014709473, -0.005847931, -0.03756714, -0.0090408325, 0.019744873, 0.011550903, -0.012863159, 0.005317688, -0.016403198, -0.0121154785, -0.042510986, 0.0009665489, -0.0118255615, ...
http://arxiv.org/abs/2306.02682v1
End-to-End Word-Level Pronunciation Assessment with MASK Pre-training
Yukang Liang, Kaitao Song, Shaoguang Mao, Huiqiang Jiang, Luna Qiu, Yuqing Yang, Dongsheng Li, Linli Xu, Lili Qiu
cs.CL, eess.AS
Pronunciation assessment is a major challenge in the computer-aided pronunciation training system, especially at the word (phoneme)-level. To obtain word (phoneme)-level scores, current methods usually rely on aligning components to obtain acoustic features of each word (phoneme), which limits the performance of assess...
http://arxiv.org/pdf/2306.02682v1
2023-06-05T08:18:01
2023-06-05T08:18:01
{ "title": [ 0.010383606, 0.0067977905, -0.06970215, -0.072509766, -0.022079468, -0.024658203, -0.044006348, 0.039916992, -0.030883789, 0.02053833, -0.014732361, 0.021057129, 0.02911377, -0.06402588, 0.031311035, 0.0076942444, -0.003698349, 0.043...
http://arxiv.org/abs/2306.02680v1
BeAts: Bengali Speech Acts Recognition using Multimodal Attention Fusion
Ahana Deb, Sayan Nag, Ayan Mahapatra, Soumitri Chattopadhyay, Aritra Marik, Pijush Kanti Gayen, Shankha Sanyal, Archi Banerjee, Samir Karmakar
cs.CL, cs.LG, cs.SD, eess.AS
Spoken languages often utilise intonation, rhythm, intensity, and structure, to communicate intention, which can be interpreted differently depending on the rhythm of speech of their utterance. These speech acts provide the foundation of communication and are unique in expression to the language. Recent advancements in...
http://arxiv.org/pdf/2306.02680v1
2023-06-05T08:12:17
2023-06-05T08:12:17
{ "title": [ 0.0002424717, 0.0023040771, -0.03314209, -0.050445557, -0.0074272156, -0.008796692, -0.026763916, 0.013252258, 0.028427124, 0.029373169, -0.05267334, 0.00044560432, 0.02255249, -0.009780884, -0.047546387, -0.020507812, -0.01537323, 0...
http://arxiv.org/abs/2306.02679v1
Joint Pre-training and Local Re-training: Transferable Representation Learning on Multi-source Knowledge Graphs
Zequn Sun, Jiacheng Huang, Jinghao Lin, Xiaozhou Xu, Qijin Chen, Wei Hu
cs.CL, cs.LG
In this paper, we present the ``joint pre-training and local re-training'' framework for learning and applying multi-source knowledge graph (KG) embeddings. We are motivated by the fact that different KGs contain complementary information to improve KG embeddings and downstream tasks. We pre-train a large teacher KG em...
http://arxiv.org/pdf/2306.02679v1
2023-06-05T08:11:59
2023-06-05T08:11:59
{ "title": [ -0.006549835, 0.010353088, -0.047180176, -0.088134766, 0.011131287, 0.008872986, -0.061767578, 0.0024852753, 0.00944519, 0.02168274, -0.01134491, 0.032043457, -0.035888672, -0.005241394, -0.020980835, 0.0060653687, 0.014678955, 0.035...
http://arxiv.org/abs/2306.02671v1
Improving Grammar-based Sequence-to-Sequence Modeling with Decomposition and Constraints
Chao Lou, Kewei Tu
cs.CL
Neural QCFG is a grammar-based sequence-tosequence (seq2seq) model with strong inductive biases on hierarchical structures. It excels in interpretability and generalization but suffers from expensive inference. In this paper, we study two low-rank variants of Neural QCFG for faster inference with different trade-offs b...
http://arxiv.org/pdf/2306.02671v1
2023-06-05T08:05:05
2023-06-05T08:05:05
{ "title": [ -0.034118652, -0.005142212, -0.040283203, -0.107543945, 0.021713257, 0.034332275, -0.095336914, 0.031921387, 0.010910034, 0.020385742, -0.012031555, 0.0045433044, 0.01499176, -0.017150879, -0.026168823, 0.015197754, -0.010566711, 0.0...
http://arxiv.org/abs/2306.02646v1
Colexifications for Bootstrapping Cross-lingual Datasets: The Case of Phonology, Concreteness, and Affectiveness
Yiyi Chen, Johannes Bjerva
cs.CL
Colexification refers to the linguistic phenomenon where a single lexical form is used to convey multiple meanings. By studying cross-lingual colexifications, researchers have gained valuable insights into fields such as psycholinguistics and cognitive sciences [Jackson et al.,2019]. While several multilingual colexifi...
http://arxiv.org/pdf/2306.02646v1
2023-06-05T07:32:21
2023-06-05T07:32:21
{ "title": [ -0.041412354, -0.0046539307, -0.070495605, -0.031036377, -0.010925293, -0.023269653, -0.050323486, 0.014419556, 0.016479492, 0.021957397, -0.027862549, -0.006046295, 0.012374878, -0.028564453, -0.038269043, -0.010368347, -0.031021118, ...
http://arxiv.org/abs/2306.02623v1
Do-GOOD: Towards Distribution Shift Evaluation for Pre-Trained Visual Document Understanding Models
Jiabang He, Yi Hu, Lei Wang, Xing Xu, Ning Liu, Hui Liu, Heng Tao Shen
cs.CV, cs.CL, cs.MM
Numerous pre-training techniques for visual document understanding (VDU) have recently shown substantial improvements in performance across a wide range of document tasks. However, these pre-trained VDU models cannot guarantee continued success when the distribution of test data differs from the distribution of trainin...
http://arxiv.org/pdf/2306.02623v1
2023-06-05T06:50:42
2023-06-05T06:50:42
{ "title": [ -0.006576538, 0.006587982, -0.02456665, -0.05117798, -0.008071899, -0.0053215027, -0.047790527, -0.025238037, 0.030563354, 0.015350342, -0.053833008, -0.014076233, 0.0075569153, -0.0035324097, -0.038391113, -0.021606445, -0.00068330765, ...
http://arxiv.org/abs/2306.02622v1
What Makes Entities Similar? A Similarity Flooding Perspective for Multi-sourced Knowledge Graph Embeddings
Zequn Sun, Jiacheng Huang, Xiaozhou Xu, Qijin Chen, Weijun Ren, Wei Hu
cs.LG, cs.AI, cs.CL
Joint representation learning over multi-sourced knowledge graphs (KGs) yields transferable and expressive embeddings that improve downstream tasks. Entity alignment (EA) is a critical step in this process. Despite recent considerable research progress in embedding-based EA, how it works remains to be explored. In this...
http://arxiv.org/pdf/2306.02622v1
2023-06-05T06:50:09
2023-06-05T06:50:09
{ "title": [ -0.0096206665, 0.008857727, -0.04876709, -0.07092285, 0.019821167, -0.0051193237, -0.036743164, 0.0056648254, 0.0042648315, 0.024902344, -0.02760315, 0.03262329, -0.00071668625, -0.005355835, -0.019714355, 0.010848999, -0.00033831596, ...
http://arxiv.org/abs/2306.04459v1
Uncertainty in Natural Language Processing: Sources, Quantification, and Applications
Mengting Hu, Zhen Zhang, Shiwan Zhao, Minlie Huang, Bingzhe Wu
cs.CL
As a main field of artificial intelligence, natural language processing (NLP) has achieved remarkable success via deep neural networks. Plenty of NLP tasks have been addressed in a unified manner, with various tasks being associated with each other through sharing the same paradigm. However, neural networks are black b...
http://arxiv.org/pdf/2306.04459v1
2023-06-05T06:46:53
2023-06-05T06:46:53
{ "title": [ -0.039123535, -0.0015306473, -0.049743652, -0.048431396, 0.030059814, 0.0013618469, -0.035858154, -0.01878357, 0.015388489, 0.008460999, -0.026000977, 0.002483368, -0.014343262, -0.02180481, -0.04864502, 0.0012626648, -0.01914978, 0....
http://arxiv.org/abs/2306.02612v1
Building Resilient SMEs: Harnessing Large Language Models for Cyber Security in Australia
Benjamin Kereopa-Yorke
cs.CR, cs.CL
The escalating digitalisation of our lives and enterprises has led to a parallel growth in the complexity and frequency of cyber-attacks. Small and medium-sized enterprises (SMEs), particularly in Australia, are experiencing increased vulnerability to cyber threats, posing a significant challenge to the nation's cyber ...
http://arxiv.org/pdf/2306.02612v1
2023-06-05T06:01:00
2023-06-05T06:01:00
{ "title": [ -0.034179688, -0.018081665, -0.05053711, -0.032440186, -0.003484726, -0.009437561, -0.032196045, 0.013023376, -0.054351807, 0.04196167, -0.0020332336, 0.016998291, 0.013557434, -0.023239136, -0.018920898, -0.02243042, -0.0039863586, ...
http://arxiv.org/abs/2306.02597v1
Early Rumor Detection Using Neural Hawkes Process with a New Benchmark Dataset
Fengzhu Zeng, Wei Gao
cs.CL
Little attention has been paid on \underline{EA}rly \underline{R}umor \underline{D}etection (EARD), and EARD performance was evaluated inappropriately on a few datasets where the actual early-stage information is largely missing. To reverse such situation, we construct BEARD, a new \underline{B}enchmark dataset for \un...
http://arxiv.org/pdf/2306.02597v1
2023-06-05T05:04:28
2023-06-05T05:04:28
{ "title": [ -0.020889282, 0.007801056, -0.018951416, 0.01638794, -0.013305664, -0.013259888, -0.02017212, -0.045562744, -0.02217102, 0.04324341, -0.03479004, 0.0059318542, -0.0104522705, -0.0011348724, -0.050872803, -0.0134887695, -0.026428223, ...
http://arxiv.org/abs/2306.02596v1
A Novel Interpretable and Generalizable Re-synchronization Model for Cued Speech based on a Multi-Cuer Corpus
Lufei Gao, Shan Huang, Li Liu
eess.AS, cs.CL, cs.CV
Cued Speech (CS) is a multi-modal visual coding system combining lip reading with several hand cues at the phonetic level to make the spoken language visible to the hearing impaired. Previous studies solved asynchronous problems between lip and hand movements by a cuer\footnote{The people who perform Cued Speech are ca...
http://arxiv.org/pdf/2306.02596v1
2023-06-05T05:03:11
2023-06-05T05:03:11
{ "title": [ -0.022506714, 0.01828003, -0.03756714, -0.053985596, -0.008781433, 0.010002136, -0.08843994, -0.02658081, 0.011245728, 0.05911255, -0.018417358, -0.0071372986, 0.03591919, -0.0012712479, -0.054382324, 0.01084137, -0.024459839, 0.0339...
http://arxiv.org/abs/2306.02592v1
Graph-Aware Language Model Pre-Training on a Large Graph Corpus Can Help Multiple Graph Applications
Han Xie, Da Zheng, Jun Ma, Houyu Zhang, Vassilis N. Ioannidis, Xiang Song, Qing Ping, Sheng Wang, Carl Yang, Yi Xu, Belinda Zeng, Trishul Chilimbi
cs.CL, cs.AI, cs.LG
Model pre-training on large text corpora has been demonstrated effective for various downstream applications in the NLP domain. In the graph mining domain, a similar analogy can be drawn for pre-training graph models on large graphs in the hope of benefiting downstream graph applications, which has also been explored b...
http://arxiv.org/pdf/2306.02592v1
2023-06-05T04:46:44
2023-06-05T04:46:44
{ "title": [ -0.001871109, -0.009735107, -0.04336548, -0.06726074, -0.03970337, -0.016418457, -0.077819824, 0.024032593, -0.0009756088, 0.025238037, -0.010032654, 0.00061273575, 0.028549194, -0.050842285, 0.018295288, 0.01146698, 0.0058021545, 0....
http://arxiv.org/abs/2306.02579v1
Cross-Lingual Transfer Learning for Phrase Break Prediction with Multilingual Language Model
Hoyeon Lee, Hyun-Wook Yoon, Jong-Hwan Kim, Jae-Min Kim
cs.CL, cs.SD, eess.AS
Phrase break prediction is a crucial task for improving the prosody naturalness of a text-to-speech (TTS) system. However, most proposed phrase break prediction models are monolingual, trained exclusively on a large amount of labeled data. In this paper, we address this issue for low-resource languages with limited lab...
http://arxiv.org/pdf/2306.02579v1
2023-06-05T04:10:04
2023-06-05T04:10:04
{ "title": [ -0.021896362, -0.014343262, -0.02885437, -0.054992676, -0.020004272, 0.015380859, -0.07751465, -0.009155273, -0.016708374, 0.03857422, 0.0036811829, 0.016433716, 0.049438477, -0.024169922, -0.008018494, 0.03353882, -0.011802673, 0.03...
http://arxiv.org/abs/2306.02569v1
Prompt to be Consistent is Better than Self-Consistent? Few-Shot and Zero-Shot Fact Verification with Pre-trained Language Models
Fengzhu Zeng, Wei Gao
cs.CL
Few-shot or zero-shot fact verification only relies on a few or no labeled training examples. In this paper, we propose a novel method called ProToCo, to \underline{Pro}mpt pre-trained language models (PLMs) \underline{To} be \underline{Co}nsistent, for improving the factuality assessment capability of PLMs in the few-...
http://arxiv.org/pdf/2306.02569v1
2023-06-05T03:49:13
2023-06-05T03:49:13
{ "title": [ 0.014297485, -0.013702393, -0.0309906, -0.07269287, -0.01889038, -0.013061523, -0.05206299, 0.009246826, -0.038085938, 0.034698486, -0.009963989, 0.0063171387, 0.008369446, -0.023834229, -0.03277588, -0.017181396, -0.0084991455, 0.04...
http://arxiv.org/abs/2306.02561v3
LLM-Blender: Ensembling Large Language Models with Pairwise Ranking and Generative Fusion
Dongfu Jiang, Xiang Ren, Bill Yuchen Lin
cs.CL, cs.AI, cs.LG
We present LLM-Blender, an ensembling framework designed to attain consistently superior performance by leveraging the diverse strengths of multiple open-source large language models (LLMs). Our framework consists of two modules: PairRanker and GenFuser, addressing the observation that optimal LLMs for different exampl...
http://arxiv.org/pdf/2306.02561v3
2023-06-30T21:39:54
2023-06-05T03:32:26
{ "title": [ -0.035003662, -0.018630981, -0.055114746, -0.07458496, 0.0045051575, 0.0149383545, -0.07458496, 0.014221191, -0.0066947937, 0.026535034, -0.004119873, 0.024017334, 0.00010752678, -0.01701355, -0.023788452, 0.00844574, -0.017593384, 0...
http://arxiv.org/abs/2306.02553v1
Learning to Relate to Previous Turns in Conversational Search
Fengran Mo, Jian-Yun Nie, Kaiyu Huang, Kelong Mao, Yutao Zhu, Peng Li, Yang Liu
cs.IR, cs.CL
Conversational search allows a user to interact with a search system in multiple turns. A query is strongly dependent on the conversation context. An effective way to improve retrieval effectiveness is to expand the current query with historical queries. However, not all the previous queries are related to, and useful ...
http://arxiv.org/pdf/2306.02553v1
2023-06-05T03:00:10
2023-06-05T03:00:10
{ "title": [ 0.02760315, -0.0053977966, 0.009155273, -0.032348633, -0.016647339, -0.015975952, -0.010612488, -0.026275635, 0.019180298, 0.032104492, -0.023010254, 0.009223938, -0.011650085, -0.0022506714, -0.062408447, -0.012741089, -0.02142334, ...
http://arxiv.org/abs/2306.02549v1
Evaluation of AI Chatbots for Patient-Specific EHR Questions
Alaleh Hamidi, Kirk Roberts
cs.CL, cs.AI, cs.IR
This paper investigates the use of artificial intelligence chatbots for patient-specific question answering (QA) from clinical notes using several large language model (LLM) based systems: ChatGPT (versions 3.5 and 4), Google Bard, and Claude. We evaluate the accuracy, relevance, comprehensiveness, and coherence of the...
http://arxiv.org/pdf/2306.02549v1
2023-06-05T02:52:54
2023-06-05T02:52:54
{ "title": [ -0.013031006, -0.0073280334, -0.01902771, -0.027023315, -0.015472412, 0.0042495728, 0.0031642914, -0.004650116, -0.02571106, 0.047668457, -0.045013428, 0.013916016, 0.021820068, -0.014785767, -0.021224976, -0.015129089, 0.010475159, ...
http://arxiv.org/abs/2306.02534v1
Incorporating L2 Phonemes Using Articulatory Features for Robust Speech Recognition
Jisung Wang, Haram Lee, Myungwoo Oh
cs.CL, cs.LG, cs.SD, eess.AS
The limited availability of non-native speech datasets presents a major challenge in automatic speech recognition (ASR) to narrow the performance gap between native and non-native speakers. To address this, the focus of this study is on the efficient incorporation of the L2 phonemes, which in this work refer to Korean ...
http://arxiv.org/pdf/2306.02534v1
2023-06-05T01:55:33
2023-06-05T01:55:33
{ "title": [ -0.025314331, 0.013885498, -0.034851074, -0.04321289, 0.008087158, 0.0004968643, -0.05078125, -0.0042686462, -0.02268982, 0.05407715, -0.0146102905, 0.008155823, 0.004802704, -0.020233154, -0.05557251, -0.0059432983, -0.008468628, 0....
http://arxiv.org/abs/2306.02531v2
PLANNER: Generating Diversified Paragraph via Latent Language Diffusion Model
Yizhe Zhang, Jiatao Gu, Zhuofeng Wu, Shuangfei Zhai, Josh Susskind, Navdeep Jaitly
cs.CL
Autoregressive models for text sometimes generate repetitive and low-quality output because errors accumulate during the steps of generation. This issue is often attributed to exposure bias - the difference between how a model is trained, and how it is used during inference. Denoising diffusion models provide an altern...
http://arxiv.org/pdf/2306.02531v2
2023-10-27T05:53:16
2023-06-05T01:36:39
{ "title": [ -0.03488159, 0.004962921, -0.053100586, -0.026397705, -0.0031108856, -0.009681702, -0.024765015, -0.0113220215, -0.015007019, 0.022003174, -0.014770508, 0.041259766, -0.012283325, -0.01436615, -0.019927979, 0.010147095, -0.013938904, ...
http://arxiv.org/abs/2306.02520v2
A Study of Situational Reasoning for Traffic Understanding
Jiarui Zhang, Filip Ilievski, Kaixin Ma, Aravinda Kollaa, Jonathan Francis, Alessandro Oltramari
cs.CL, cs.AI, cs.LG
Intelligent Traffic Monitoring (ITMo) technologies hold the potential for improving road safety/security and for enabling smart city infrastructure. Understanding traffic situations requires a complex fusion of perceptual information with domain-specific and causal commonsense knowledge. Whereas prior work has provided...
http://arxiv.org/pdf/2306.02520v2
2023-07-15T06:45:20
2023-06-05T01:01:12
{ "title": [ 0.023376465, 0.017486572, -0.029006958, -0.063964844, 0.018432617, -0.016937256, -0.010948181, -0.015266418, 0.0051841736, 0.011726379, -0.059143066, 0.0058174133, -0.04058838, -0.007534027, -0.0317688, -0.02545166, 0.0029239655, -0....