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/2307.16709v1
Multilingual context-based pronunciation learning for Text-to-Speech
Giulia Comini, Manuel Sam Ribeiro, Fan Yang, Heereen Shim, Jaime Lorenzo-Trueba
cs.CL, eess.AS
Phonetic information and linguistic knowledge are an essential component of a Text-to-speech (TTS) front-end. Given a language, a lexicon can be collected offline and Grapheme-to-Phoneme (G2P) relationships are usually modeled in order to predict the pronunciation for out-of-vocabulary (OOV) words. Additionally, post-l...
http://arxiv.org/pdf/2307.16709v1
2023-07-31T14:29:06
2023-07-31T14:29:06
{ "title": [ 0.0051345825, 0.00434494, -0.03302002, -0.081970215, 0.009864807, 0.009178162, -0.039001465, 0.033721924, -0.01626587, 0.06414795, -0.034210205, 0.028839111, 0.019989014, -0.0053863525, -0.03744507, 0.002937317, -0.0082473755, 0.0387...
http://arxiv.org/abs/2307.16699v1
Ontology engineering with Large Language Models
Patricia Mateiu, Adrian Groza
cs.AI, cs.CL
We tackle the task of enriching ontologies by automatically translating natural language sentences into Description Logic. Since Large Language Models (LLMs) are the best tools for translations, we fine-tuned a GPT-3 model to convert Natural Language sentences into OWL Functional Syntax. We employ objective and concise...
http://arxiv.org/pdf/2307.16699v1
2023-07-31T14:18:23
2023-07-31T14:18:23
{ "title": [ -0.024917603, -0.018203735, -0.03955078, -0.083740234, 0.014022827, 0.012527466, -0.04660034, 0.0071868896, 0.026535034, 0.015579224, -0.010887146, -0.022369385, -0.00031256676, -0.025177002, -0.037994385, 0.010765076, -0.012641907, ...
http://arxiv.org/abs/2307.16689v1
No that's not what I meant: Handling Third Position Repair in Conversational Question Answering
Vevake Balaraman, Arash Eshghi, Ioannis Konstas, Ioannis Papaioannou
cs.CL
The ability to handle miscommunication is crucial to robust and faithful conversational AI. People usually deal with miscommunication immediately as they detect it, using highly systematic interactional mechanisms called repair. One important type of repair is Third Position Repair (TPR) whereby a speaker is initially ...
http://arxiv.org/pdf/2307.16689v1
2023-07-31T14:02:45
2023-07-31T14:02:45
{ "title": [ 0.021362305, 0.0053520203, -0.0047721863, -0.02444458, -0.005344391, -0.014053345, -0.01928711, -0.028518677, 0.024719238, 0.03955078, -0.040649414, 0.0036029816, 0.0017814636, -0.0075263977, -0.0715332, -0.026428223, 0.00070381165, ...
http://arxiv.org/abs/2307.16680v4
On the Trustworthiness Landscape of State-of-the-art Generative Models: A Comprehensive Survey
Mingyuan Fan, Cen Chen, Chengyu Wang, Jun Huang
cs.LG, cs.AI, cs.CL, cs.CR, cs.CV
Diffusion models and large language models have emerged as leading-edge generative models and have sparked a revolutionary impact on various aspects of human life. However, the practical implementation of these models has also exposed inherent risks, highlighting their dual nature and raising concerns regarding their t...
http://arxiv.org/pdf/2307.16680v4
2023-08-23T03:28:30
2023-07-31T13:57:05
{ "title": [ -0.013465881, -0.013061523, -0.0519104, -0.074035645, 0.011940002, -0.0016126633, -0.04159546, 0.0015096664, 0.006576538, 0.020828247, -0.00050497055, 0.022506714, -0.020706177, -0.029846191, -0.034179688, 0.021530151, -0.023208618, ...
http://arxiv.org/abs/2307.16679v1
Comparing normalizing flows and diffusion models for prosody and acoustic modelling in text-to-speech
Guangyan Zhang, Thomas Merritt, Manuel Sam Ribeiro, Biel Tura-Vecino, Kayoko Yanagisawa, Kamil Pokora, Abdelhamid Ezzerg, Sebastian Cygert, Ammar Abbas, Piotr Bilinski, Roberto Barra-Chicote, Daniel Korzekwa, Jaime Lorenzo-Trueba
eess.AS, cs.CL, cs.LG
Neural text-to-speech systems are often optimized on L1/L2 losses, which make strong assumptions about the distributions of the target data space. Aiming to improve those assumptions, Normalizing Flows and Diffusion Probabilistic Models were recently proposed as alternatives. In this paper, we compare traditional L1/L2...
http://arxiv.org/pdf/2307.16679v1
2023-07-31T13:57:04
2023-07-31T13:57:04
{ "title": [ -0.0107421875, 0.008361816, -0.022415161, -0.06304932, 0.002336502, 0.00881958, -0.039489746, 0.020889282, 0.011138916, 0.062042236, -0.0015459061, 0.009239197, 0.034942627, -0.0178833, -0.05706787, 0.0211792, -0.011833191, 0.0247955...
http://arxiv.org/abs/2307.16659v1
The World Literature Knowledge Graph
Marco Antonio Stranisci, Eleonora Bernasconi, Viviana Patti, Stefano Ferilli, Miguel Ceriani, Rossana Damiano
cs.DL, cs.CL
Digital media have enabled the access to unprecedented literary knowledge. Authors, readers, and scholars are now able to discover and share an increasing amount of information about books and their authors. However, these sources of knowledge are fragmented and do not adequately represent non-Western writers and their...
http://arxiv.org/pdf/2307.16659v1
2023-07-31T13:41:31
2023-07-31T13:41:31
{ "title": [ -0.0044555664, -0.0079956055, -0.05117798, -0.057525635, -0.03543091, -0.025466919, -0.039215088, 0.021560669, -0.022644043, 0.03692627, -0.01965332, -0.0020713806, 0.025268555, -0.045654297, 0.011276245, -0.0035037994, 0.02619934, 0...
http://arxiv.org/abs/2307.16648v2
LLMs4OL: Large Language Models for Ontology Learning
Hamed Babaei Giglou, Jennifer D'Souza, Sören Auer
cs.AI, cs.CL, cs.IT, cs.LG, math.IT
We propose the LLMs4OL approach, which utilizes Large Language Models (LLMs) for Ontology Learning (OL). LLMs have shown significant advancements in natural language processing, demonstrating their ability to capture complex language patterns in different knowledge domains. Our LLMs4OL paradigm investigates the followi...
http://arxiv.org/pdf/2307.16648v2
2023-08-02T07:47:26
2023-07-31T13:27:21
{ "title": [ -0.03277588, -0.017715454, -0.04901123, -0.07611084, 0.0051651, 0.007461548, -0.048675537, 0.014762878, 0.01360321, 0.014854431, -0.020065308, 0.0042304993, -0.007675171, -0.020629883, -0.02178955, 0.006690979, -0.017486572, 0.037780...
http://arxiv.org/abs/2307.16645v1
Scaling Sentence Embeddings with Large Language Models
Ting Jiang, Shaohan Huang, Zhongzhi Luan, Deqing Wang, Fuzhen Zhuang
cs.CL
Large language models (LLMs) have recently garnered significant interest. With in-context learning, LLMs achieve impressive results in various natural language tasks. However, the application of LLMs to sentence embeddings remains an area of ongoing research. In this work, we propose an in-context learning-based method...
http://arxiv.org/pdf/2307.16645v1
2023-07-31T13:26:03
2023-07-31T13:26:03
{ "title": [ -0.00434494, -0.003250122, -0.031280518, -0.06100464, 0.010635376, 0.002878189, -0.03982544, 0.00092220306, 0.026031494, 0.015357971, -0.028396606, 0.005332947, 0.0029907227, -0.009254456, -0.036193848, 0.009559631, -0.014350891, 0.0...
http://arxiv.org/abs/2307.16643v1
Improving grapheme-to-phoneme conversion by learning pronunciations from speech recordings
Manuel Sam Ribeiro, Giulia Comini, Jaime Lorenzo-Trueba
eess.AS, cs.CL
The Grapheme-to-Phoneme (G2P) task aims to convert orthographic input into a discrete phonetic representation. G2P conversion is beneficial to various speech processing applications, such as text-to-speech and speech recognition. However, these tend to rely on manually-annotated pronunciation dictionaries, which are of...
http://arxiv.org/pdf/2307.16643v1
2023-07-31T13:25:38
2023-07-31T13:25:38
{ "title": [ -0.011314392, -0.008201599, -0.018753052, -0.031097412, -0.0071258545, -0.0043945312, -0.020736694, 0.017456055, -0.017486572, 0.040618896, -0.01638794, 0.0060157776, 0.0027332306, -0.020751953, -0.037963867, -0.006629944, -0.023147583, ...
http://arxiv.org/abs/2307.16638v1
VacancySBERT: the approach for representation of titles and skills for semantic similarity search in the recruitment domain
Maiia Bocharova, Eugene Malakhov, Vitaliy Mezhuyev
cs.CL, 68T50
The paper focuses on deep learning semantic search algorithms applied in the HR domain. The aim of the article is developing a novel approach to training a Siamese network to link the skills mentioned in the job ad with the title. It has been shown that the title normalization process can be based either on classificat...
http://arxiv.org/pdf/2307.16638v1
2023-07-31T13:21:15
2023-07-31T13:21:15
{ "title": [ 0.024047852, -0.003458023, -0.0044517517, -0.009735107, -0.0022525787, -0.026916504, 0.024291992, -0.04663086, 0.025924683, 0.04849243, -0.02796936, 0.0052604675, 0.058776855, -0.0090789795, -0.033691406, 0.005317688, -0.026641846, 0...
http://arxiv.org/abs/2307.16630v1
Text-CRS: A Generalized Certified Robustness Framework against Textual Adversarial Attacks
Xinyu Zhang, Hanbin Hong, Yuan Hong, Peng Huang, Binghui Wang, Zhongjie Ba, Kui Ren
cs.CR, cs.CL, cs.LG
The language models, especially the basic text classification models, have been shown to be susceptible to textual adversarial attacks such as synonym substitution and word insertion attacks. To defend against such attacks, a growing body of research has been devoted to improving the model robustness. However, providin...
http://arxiv.org/pdf/2307.16630v1
2023-07-31T13:08:16
2023-07-31T13:08:16
{ "title": [ -0.0052375793, 0.010978699, -0.029327393, -0.06530762, 0.035308838, 0.010368347, -0.030715942, -0.022613525, -0.031158447, 0.036010742, -0.0012311935, 0.03753662, -0.0053520203, -0.0127334595, -0.04244995, -0.010986328, -0.022384644, ...
http://arxiv.org/abs/2307.16609v1
Noisy Self-Training with Data Augmentations for Offensive and Hate Speech Detection Tasks
João A. Leite, Carolina Scarton, Diego F. Silva
cs.CL, cs.LG, cs.SI
Online social media is rife with offensive and hateful comments, prompting the need for their automatic detection given the sheer amount of posts created every second. Creating high-quality human-labelled datasets for this task is difficult and costly, especially because non-offensive posts are significantly more frequ...
http://arxiv.org/pdf/2307.16609v1
2023-07-31T12:35:54
2023-07-31T12:35:54
{ "title": [ 0.02458191, 0.022369385, -0.03704834, -0.038879395, 0.026382446, -0.02998352, 0.022918701, -0.0046920776, -0.009483337, 0.02909851, -0.026855469, 0.01687622, -0.0033168793, -0.006881714, -0.043518066, -0.0040779114, -0.00484848, 0.06...
http://arxiv.org/abs/2308.04889v1
NLLG Quarterly arXiv Report 06/23: What are the most influential current AI Papers?
Steffen Eger, Christoph Leiter, Jonas Belouadi, Ran Zhang, Aida Kostikova, Daniil Larionov, Yanran Chen, Vivian Fresen
cs.CY, cs.AI, cs.CL, cs.DL, cs.LG
The rapid growth of information in the field of Generative Artificial Intelligence (AI), particularly in the subfields of Natural Language Processing (NLP) and Machine Learning (ML), presents a significant challenge for researchers and practitioners to keep pace with the latest developments. To address the problem of i...
http://arxiv.org/pdf/2308.04889v1
2023-07-31T11:53:52
2023-07-31T11:53:52
{ "title": [ -0.055389404, -0.014968872, -0.03729248, -0.0067481995, -0.029312134, -0.022506714, 0.0035743713, 0.021484375, -0.03604126, 0.015304565, -0.0027446747, 0.0016441345, -0.000047922134, -0.037506104, 0.013046265, -0.05355835, 0.007820129, ...
http://arxiv.org/abs/2308.02524v1
Chatbot Application to Support Smart Agriculture in Thailand
Paweena Suebsombut, Pradorn Sureephong, Aicha Sekhari, Suepphong Chernbumroong, Abdelaziz Bouras
cs.CL, cs.AI
A chatbot is a software developed to help reply to text or voice conversations automatically and quickly in real time. In the agriculture sector, the existing smart agriculture systems just use data from sensing and internet of things (IoT) technologies that exclude crop cultivation knowledge to support decision-making...
http://arxiv.org/pdf/2308.02524v1
2023-07-31T11:42:44
2023-07-31T11:42:44
{ "title": [ 0.049224854, -0.030044556, 0.014717102, -0.03955078, -0.03265381, 0.0011320114, 0.0079193115, 0.0004005432, -0.03427124, 0.06317139, -0.034454346, -0.010223389, 0.044921875, -0.027130127, -0.015556335, -0.034576416, -0.0014839172, 0....
http://arxiv.org/abs/2307.16576v1
Toward Quantum Machine Translation of Syntactically Distinct Languages
Mina Abbaszade, Mariam Zomorodi, Vahid Salari, Philip Kurian
cs.CL, cs.AI
The present study aims to explore the feasibility of language translation using quantum natural language processing algorithms on noisy intermediate-scale quantum (NISQ) devices. Classical methods in natural language processing (NLP) struggle with handling large-scale computations required for complex language tasks, b...
http://arxiv.org/pdf/2307.16576v1
2023-07-31T11:24:54
2023-07-31T11:24:54
{ "title": [ -0.047546387, -0.0015010834, -0.008293152, -0.029846191, -0.009544373, 0.01158905, -0.031951904, -0.0154418945, -0.026687622, 0.058807373, 0.026763916, 0.014328003, -0.0011577606, -0.017333984, -0.057006836, 0.017623901, -0.020111084, ...
http://arxiv.org/abs/2307.16573v2
Deep Dive into the Language of International Relations: NLP-based Analysis of UNESCO's Summary Records
Joanna Wojciechowska, Mateusz Sypniewski, Maria Śmigielska, Igor Kamiński, Emilia Wiśnios, Hanna Schreiber, Bartosz Pieliński
cs.CL
Cultural heritage is an arena of international relations that interests all states worldwide. The inscription process on the UNESCO World Heritage List and the UNESCO Representative List of the Intangible Cultural Heritage of Humanity often leads to tensions and conflicts among states. This research addresses these cha...
http://arxiv.org/pdf/2307.16573v2
2023-08-01T10:17:28
2023-07-31T11:06:08
{ "title": [ -0.001373291, -0.018463135, -0.035858154, -0.018218994, -0.036193848, -0.022323608, -0.019454956, 0.01727295, -0.044921875, 0.010101318, -0.04437256, 0.02243042, 0.0027809143, -0.024475098, 0.018066406, -0.047943115, -0.011978149, 0....
http://arxiv.org/abs/2307.16549v1
DiffProsody: Diffusion-based Latent Prosody Generation for Expressive Speech Synthesis with Prosody Conditional Adversarial Training
Hyung-Seok Oh, Sang-Hoon Lee, Seong-Whan Lee
cs.SD, cs.CL, eess.AS
Expressive text-to-speech systems have undergone significant advancements owing to prosody modeling, but conventional methods can still be improved. Traditional approaches have relied on the autoregressive method to predict the quantized prosody vector; however, it suffers from the issues of long-term dependency and sl...
http://arxiv.org/pdf/2307.16549v1
2023-07-31T10:28:45
2023-07-31T10:28:45
{ "title": [ -0.026870728, 0.009033203, -0.06744385, -0.081970215, -0.0041275024, 0.0118637085, -0.06842041, 0.04171753, 0.010063171, 0.028900146, -0.030517578, 0.0064086914, 0.021850586, -0.049987793, -0.0024738312, -0.01638794, -0.019210815, 0....
http://arxiv.org/abs/2307.16548v2
Specification of MiniDemographicABM.jl: A simplified agent-based demographic model of the UK
Atiyah Elsheikh
cs.CL
This documentation specifies a simplified non-calibrated demographic agent-based model of the UK, a largely simplified version of the Lone Parent Model presented in [Gostolil and Silverman 2020]. In the presented model, individuals of an initial population are subject to ageing, deaths, births, divorces and marriages t...
http://arxiv.org/pdf/2307.16548v2
2023-10-20T11:41:12
2023-07-31T10:28:23
{ "title": [ 0.05053711, -0.0044555664, -0.013336182, -0.022949219, 0.008430481, -0.02798462, 0.019348145, -0.018173218, -0.046081543, 0.029251099, 0.015266418, -0.0023460388, -0.013412476, -0.048614502, -0.0057144165, -0.025726318, 0.018035889, ...
http://arxiv.org/abs/2307.16544v1
Utilisation of open intent recognition models for customer support intent detection
Rasheed Mohammad, Oliver Favell, Shariq Shah, Emmett Cooper, Edlira Vakaj
cs.CL
Businesses have sought out new solutions to provide support and improve customer satisfaction as more products and services have become interconnected digitally. There is an inherent need for businesses to provide or outsource fast, efficient and knowledgeable support to remain competitive. Support solutions are also a...
http://arxiv.org/pdf/2307.16544v1
2023-07-31T10:20:16
2023-07-31T10:20:16
{ "title": [ 0.0042037964, 0.020492554, -0.039489746, -0.038360596, 0.01133728, -0.02168274, -0.03692627, -0.043762207, 0.015464783, 0.030776978, -0.023117065, -0.024429321, 0.019821167, -0.02482605, -0.05065918, -0.0031757355, -0.0013437271, 0.0...
http://arxiv.org/abs/2307.16525v1
Transferable Decoding with Visual Entities for Zero-Shot Image Captioning
Junjie Fei, Teng Wang, Jinrui Zhang, Zhenyu He, Chengjie Wang, Feng Zheng
cs.CV, cs.CL
Image-to-text generation aims to describe images using natural language. Recently, zero-shot image captioning based on pre-trained vision-language models (VLMs) and large language models (LLMs) has made significant progress. However, we have observed and empirically demonstrated that these methods are susceptible to mo...
http://arxiv.org/pdf/2307.16525v1
2023-07-31T09:47:06
2023-07-31T09:47:06
{ "title": [ -0.0007996559, -0.015777588, -0.02468872, -0.0418396, -0.01826477, -0.0209198, -0.001991272, 0.00089502335, 0.02798462, 0.066589355, -0.0008878708, 0.0154953, 0.030532837, -0.011100769, -0.058013916, 0.03552246, 0.006919861, 0.035858...
http://arxiv.org/abs/2307.16513v1
Deception Abilities Emerged in Large Language Models
Thilo Hagendorff
cs.CL, cs.AI, cs.LG
Large language models (LLMs) are currently at the forefront of intertwining artificial intelligence (AI) systems with human communication and everyday life. Thus, aligning them with human values is of great importance. However, given the steady increase in reasoning abilities, future LLMs are under suspicion of becomin...
http://arxiv.org/pdf/2307.16513v1
2023-07-31T09:27:01
2023-07-31T09:27:01
{ "title": [ 0.01889038, -0.0124053955, -0.010215759, -0.0063705444, -0.05706787, -0.031143188, 0.0005683899, -0.0023078918, -0.011680603, 0.032287598, -0.021102905, -0.025177002, 0.028030396, -0.04296875, 0.029418945, 0.0056991577, -0.010948181, ...
http://arxiv.org/abs/2307.16511v1
Classifying multilingual party manifestos: Domain transfer across country, time, and genre
Matthias Aßenmacher, Nadja Sauter, Christian Heumann
cs.CL, cs.CY, cs.LG, stat.ML
Annotating costs of large corpora are still one of the main bottlenecks in empirical social science research. On the one hand, making use of the capabilities of domain transfer allows re-using annotated data sets and trained models. On the other hand, it is not clear how well domain transfer works and how reliable the ...
http://arxiv.org/pdf/2307.16511v1
2023-07-31T09:16:13
2023-07-31T09:16:13
{ "title": [ -0.008934021, 0.0037460327, -0.01965332, -0.028274536, 0.009727478, -0.0029296875, -0.023666382, -0.025802612, -0.032043457, 0.02885437, 0.022521973, 0.025314331, -0.019500732, -0.029251099, -0.017578125, 0.0065956116, -0.013130188, ...
http://arxiv.org/abs/2308.01430v1
FinVis-GPT: A Multimodal Large Language Model for Financial Chart Analysis
Ziao Wang, Yuhang Li, Junda Wu, Jaehyeon Soon, Xiaofeng Zhang
cs.CL
In this paper, we propose FinVis-GPT, a novel multimodal large language model (LLM) specifically designed for financial chart analysis. By leveraging the power of LLMs and incorporating instruction tuning and multimodal capabilities, FinVis-GPT is capable of interpreting financial charts and providing valuable analysis...
http://arxiv.org/pdf/2308.01430v1
2023-07-31T07:44:15
2023-07-31T07:44:15
{ "title": [ -0.03515625, -0.013122559, -0.062927246, -0.057617188, -0.009475708, -0.0039596558, -0.0625, 0.0045204163, 0.03918457, 0.033111572, -0.012489319, -0.017578125, 0.004261017, -0.014228821, -0.06793213, 0.0071258545, 0.0037765503, 0.022...
http://arxiv.org/abs/2308.00014v1
A new mapping of technological interdependence
A. Fronzetti Colladon, B. Guardabascio, F. Venturini
econ.EM, cs.CL, cs.SI, I.7; I.2.7; J.4
Which technological linkages affect the sector's ability to innovate? How do these effects transmit through the technology space? This paper answers these two key questions using novel methods of text mining and network analysis. We examine technological interdependence across sectors over a period of half a century (f...
http://arxiv.org/pdf/2308.00014v1
2023-07-31T07:37:37
2023-07-31T07:37:37
{ "title": [ 0.03112793, 0.016983032, 0.003452301, 0.00016224384, -0.032073975, -0.04232788, 0.020431519, -0.030883789, -0.044403076, 0.012931824, -0.033172607, 0.015777588, -0.0524292, -0.028335571, 0.014831543, -0.04888916, 0.022827148, -0.0025...
http://arxiv.org/abs/2307.16457v1
A Benchmark for Understanding Dialogue Safety in Mental Health Support
Huachuan Qiu, Tong Zhao, Anqi Li, Shuai Zhang, Hongliang He, Zhenzhong Lan
cs.CL
Dialogue safety remains a pervasive challenge in open-domain human-machine interaction. Existing approaches propose distinctive dialogue safety taxonomies and datasets for detecting explicitly harmful responses. However, these taxonomies may not be suitable for analyzing response safety in mental health support. In rea...
http://arxiv.org/pdf/2307.16457v1
2023-07-31T07:33:16
2023-07-31T07:33:16
{ "title": [ 0.03225708, 0.0056610107, -0.026535034, -0.09716797, 0.01751709, -0.0051498413, -0.05041504, 0.0031166077, -0.0154800415, 0.03149414, 0.01361084, 0.0041503906, 0.0060157776, -0.053771973, -0.019210815, 0.006790161, -0.003030777, -0.0...
http://arxiv.org/abs/2307.16456v1
Camoscio: an Italian Instruction-tuned LLaMA
Andrea Santilli, Emanuele Rodolà
cs.CL
In recent years Large Language Models (LLMs) have increased the state of the art on several natural language processing tasks. However, their accessibility is often limited to paid API services, posing challenges for researchers in conducting extensive investigations. On the other hand, while some open-source models ha...
http://arxiv.org/pdf/2307.16456v1
2023-07-31T07:31:48
2023-07-31T07:31:48
{ "title": [ -0.045043945, -0.051330566, -0.019226074, 0.051696777, -0.0871582, -0.009880066, -0.0048980713, -0.016098022, -0.049316406, 0.06604004, 0.02848816, -0.04989624, 0.016479492, -0.04220581, -0.04763794, -0.03781128, 0.00566864, 0.044250...
http://arxiv.org/abs/2308.01415v1
An Effective Data Creation Pipeline to Generate High-quality Financial Instruction Data for Large Language Model
Ziao Wang, Jianning Wang, Junda Wu, Xiaofeng Zhang
cs.CL, cs.AI, cs.LG
At the beginning era of large language model, it is quite critical to generate a high-quality financial dataset to fine-tune a large language model for financial related tasks. Thus, this paper presents a carefully designed data creation pipeline for this purpose. Particularly, we initiate a dialogue between an AI inve...
http://arxiv.org/pdf/2308.01415v1
2023-07-31T07:23:11
2023-07-31T07:23:11
{ "title": [ -0.022994995, -0.01939392, -0.043884277, -0.047210693, -0.032165527, -0.015731812, -0.05731201, 0.011528015, -0.023345947, 0.021743774, -0.037628174, 0.008781433, 0.004512787, -0.013031006, -0.0099105835, -0.0151901245, -0.023925781, ...
http://arxiv.org/abs/2308.01414v1
HouYi: An open-source large language model specially designed for renewable energy and carbon neutrality field
Mingliang Bai, Zhihao Zhou, Ruidong Wang, Yusheng Yang, Zizhen Qin, Yunxiao Chen, Chunjin Mu, Jinfu Liu, Daren Yu
cs.CL, cs.AI
Renewable energy is important for achieving carbon neutrality goal. With the great success of Large Language Models (LLMs) like ChatGPT in automatic content generation, LLMs are playing an increasingly important role. However, there has not been a specially designed LLM for renewable energy. Meanwhile, there has not be...
http://arxiv.org/pdf/2308.01414v1
2023-07-31T06:59:36
2023-07-31T06:59:36
{ "title": [ -0.039916992, -0.0046806335, -0.07330322, -0.022583008, -0.013458252, -0.022949219, -0.03857422, 0.021408081, -0.03302002, 0.04434204, -0.009979248, -0.025054932, 0.045532227, -0.052612305, 0.0069007874, -0.025299072, -0.01184845, 0....
http://arxiv.org/abs/2308.01966v1
DCTM: Dilated Convolutional Transformer Model for Multimodal Engagement Estimation in Conversation
Vu Ngoc Tu, Van Thong Huynh, Hyung-Jeong Yang, M. Zaigham Zaheer, Shah Nawaz, Karthik Nandakumar, Soo-Hyung Kim
cs.MM, cs.CL, cs.LG, cs.SD, eess.AS
Conversational engagement estimation is posed as a regression problem, entailing the identification of the favorable attention and involvement of the participants in the conversation. This task arises as a crucial pursuit to gain insights into human's interaction dynamics and behavior patterns within a conversation. In...
http://arxiv.org/pdf/2308.01966v1
2023-07-31T06:02:35
2023-07-31T06:02:35
{ "title": [ 0.009307861, 0.0211792, -0.054840088, -0.07409668, 0.011398315, -0.021728516, -0.06298828, -0.004962921, 0.018463135, 0.025039673, -0.023025513, -0.0050086975, 0.029708862, -0.027130127, -0.034332275, 0.0024662018, -0.03451538, 0.056...
http://arxiv.org/abs/2307.16400v1
SelfSeg: A Self-supervised Sub-word Segmentation Method for Neural Machine Translation
Haiyue Song, Raj Dabre, Chenhui Chu, Sadao Kurohashi, Eiichiro Sumita
cs.CL
Sub-word segmentation is an essential pre-processing step for Neural Machine Translation (NMT). Existing work has shown that neural sub-word segmenters are better than Byte-Pair Encoding (BPE), however, they are inefficient as they require parallel corpora, days to train and hours to decode. This paper introduces SelfS...
http://arxiv.org/pdf/2307.16400v1
2023-07-31T04:38:47
2023-07-31T04:38:47
{ "title": [ -0.026672363, -0.0047454834, -0.0013332367, -0.038146973, -0.01423645, 0.015991211, -0.030929565, -0.0022411346, -0.021957397, 0.054534912, 0.0013713837, 0.018630981, 0.007095337, -0.018539429, -0.04147339, 0.018707275, 0.0004734993, ...
http://arxiv.org/abs/2307.16395v1
Bridging the Gap: Exploring the Capabilities of Bridge-Architectures for Complex Visual Reasoning Tasks
Kousik Rajesh, Mrigank Raman, Mohammed Asad Karim, Pranit Chawla
cs.CV, cs.AI, cs.CL, cs.LG
In recent times there has been a surge of multi-modal architectures based on Large Language Models, which leverage the zero shot generation capabilities of LLMs and project image embeddings into the text space and then use the auto-regressive capacity to solve tasks such as VQA, captioning, and image retrieval. We name...
http://arxiv.org/pdf/2307.16395v1
2023-07-31T03:57:31
2023-07-31T03:57:31
{ "title": [ -0.025161743, -0.0039634705, -0.027770996, -0.040893555, 0.009010315, 0.004119873, -0.018005371, -0.0047416687, 0.020874023, 0.023986816, -0.042388916, -0.0129470825, -0.004512787, -0.009353638, -0.0546875, -0.01576233, 0.011413574, ...
http://arxiv.org/abs/2307.16389v1
STL: A Signed and Truncated Logarithm Activation Function for Neural Networks
Yuanhao Gong
cs.LG, cs.AI, cs.CE, cs.CL, cs.CV, cs.NE
Activation functions play an essential role in neural networks. They provide the non-linearity for the networks. Therefore, their properties are important for neural networks' accuracy and running performance. In this paper, we present a novel signed and truncated logarithm function as activation function. The proposed...
http://arxiv.org/pdf/2307.16389v1
2023-07-31T03:41:14
2023-07-31T03:41:14
{ "title": [ -0.0018367767, 0.002576828, -0.051635742, -0.044158936, 0.025558472, -0.018356323, -0.021575928, -0.025802612, 0.0026569366, 0.02079773, -0.0075263977, 0.009506226, -0.040771484, -0.022888184, -0.053100586, 0.0006990433, 0.0022640228, ...
http://arxiv.org/abs/2307.16382v1
Does fine-tuning GPT-3 with the OpenAI API leak personally-identifiable information?
Albert Yu Sun, Eliott Zemour, Arushi Saxena, Udith Vaidyanathan, Eric Lin, Christian Lau, Vaikkunth Mugunthan
cs.LG, cs.CL
Machine learning practitioners often fine-tune generative pre-trained models like GPT-3 to improve model performance at specific tasks. Previous works, however, suggest that fine-tuned machine learning models memorize and emit sensitive information from the original fine-tuning dataset. Companies such as OpenAI offer f...
http://arxiv.org/pdf/2307.16382v1
2023-07-31T03:17:51
2023-07-31T03:17:51
{ "title": [ 0.0037441254, 0.021270752, -0.034942627, -0.009155273, -0.00566864, -0.0115356445, 0.0045928955, -0.04208374, -0.06427002, 0.037963867, -0.05230713, 0.036071777, -0.037078857, 0.00016534328, -0.003358841, -0.04284668, 0.014320374, 0....
http://arxiv.org/abs/2307.16376v1
When Large Language Models Meet Personalization: Perspectives of Challenges and Opportunities
Jin Chen, Zheng Liu, Xu Huang, Chenwang Wu, Qi Liu, Gangwei Jiang, Yuanhao Pu, Yuxuan Lei, Xiaolong Chen, Xingmei Wang, Defu Lian, Enhong Chen
cs.IR, cs.AI, cs.CL
The advent of large language models marks a revolutionary breakthrough in artificial intelligence. With the unprecedented scale of training and model parameters, the capability of large language models has been dramatically improved, leading to human-like performances in understanding, language synthesizing, and common...
http://arxiv.org/pdf/2307.16376v1
2023-07-31T02:48:56
2023-07-31T02:48:56
{ "title": [ -0.019363403, -0.010429382, -0.031173706, -0.072387695, -0.00062561035, 0.010719299, -0.052368164, 0.00970459, -0.020553589, 0.04019165, 0.006164551, 0.011306763, 0.009124756, -0.021896362, -0.021438599, 0.017227173, -0.013908386, 0....
http://arxiv.org/abs/2307.16338v1
Distractor generation for multiple-choice questions with predictive prompting and large language models
Semere Kiros Bitew, Johannes Deleu, Chris Develder, Thomas Demeester
cs.CL
Large Language Models (LLMs) such as ChatGPT have demonstrated remarkable performance across various tasks and have garnered significant attention from both researchers and practitioners. However, in an educational context, we still observe a performance gap in generating distractors -- i.e., plausible yet incorrect an...
http://arxiv.org/pdf/2307.16338v1
2023-07-30T23:15:28
2023-07-30T23:15:28
{ "title": [ 0.0063171387, 0.0006380081, -0.0005617142, -0.03729248, -0.020355225, 0.0013227463, -0.024810791, -0.027023315, 0.022003174, 0.0070266724, -0.020355225, -0.0026607513, 0.00004118681, -0.029907227, -0.015289307, -0.010467529, -0.01373291,...
http://arxiv.org/abs/2307.16324v1
Mispronunciation detection using self-supervised speech representations
Jazmin Vidal, Pablo Riera, Luciana Ferrer
cs.CL, cs.SD, eess.AS
In recent years, self-supervised learning (SSL) models have produced promising results in a variety of speech-processing tasks, especially in contexts of data scarcity. In this paper, we study the use of SSL models for the task of mispronunciation detection for second language learners. We compare two downstream approa...
http://arxiv.org/pdf/2307.16324v1
2023-07-30T21:20:58
2023-07-30T21:20:58
{ "title": [ 0.021133423, 0.009933472, -0.047729492, -0.058532715, 0.0041236877, -0.012664795, -0.026687622, 0.01399231, -0.0072669983, 0.0513916, -0.030441284, -0.0007543564, 0.018692017, -0.02508545, -0.03186035, -0.0011434555, -0.0234375, 0.06...
http://arxiv.org/abs/2308.01413v2
LaFiCMIL: Rethinking Large File Classification from the Perspective of Correlated Multiple Instance Learning
Tiezhu Sun, Weiguo Pian, Nadia Daoudi, Kevin Allix, Tegawendé F. Bissyandé, Jacques Klein
cs.CL, cs.AI
Transformer-based models, such as BERT, have revolutionized various language tasks, but still struggle with large file classification due to their input limit (e.g., 512 tokens). Despite several attempts to alleviate this limitation, no method consistently excels across all benchmark datasets, primarily because they ca...
http://arxiv.org/pdf/2308.01413v2
2023-08-15T12:19:56
2023-07-30T18:47:54
{ "title": [ 0.0047798157, -0.010620117, -0.047302246, -0.057495117, 0.01689148, -0.019485474, -0.022247314, -0.014770508, -0.016433716, 0.009681702, -0.028259277, 0.022201538, -0.024398804, -0.02218628, -0.010314941, -0.017105103, -0.015090942, ...
http://arxiv.org/abs/2307.16265v1
Recent Advances in Hierarchical Multi-label Text Classification: A Survey
Rundong Liu, Wenhan Liang, Weijun Luo, Yuxiang Song, He Zhang, Ruohua Xu, Yunfeng Li, Ming Liu
cs.CL, cs.AI
Hierarchical multi-label text classification aims to classify the input text into multiple labels, among which the labels are structured and hierarchical. It is a vital task in many real world applications, e.g. scientific literature archiving. In this paper, we survey the recent progress of hierarchical multi-label te...
http://arxiv.org/pdf/2307.16265v1
2023-07-30T16:13:00
2023-07-30T16:13:00
{ "title": [ -0.017822266, 0.0041160583, -0.04397583, -0.062927246, 0.0110321045, -0.0029525757, -0.029754639, -0.0065956116, -0.015625, 0.042053223, -0.010360718, 0.03652954, -0.0007119179, -0.011070251, -0.03918457, 0.0021820068, -0.029296875, ...
http://arxiv.org/abs/2307.16230v3
A Private Watermark for Large Language Models
Aiwei Liu, Leyi Pan, Xuming Hu, Shu'ang Li, Lijie Wen, Irwin King, Philip S. Yu
cs.CL, 68T50, I.2.7
Recently, text watermarking algorithms for large language models (LLMs) have been mitigating the potential harms of text generated by the LLMs, including fake news and copyright issues. However, the watermark detection of current text algorithms requires the key from the generation process, making them susceptible to b...
http://arxiv.org/pdf/2307.16230v3
2023-10-07T10:04:26
2023-07-30T13:43:27
{ "title": [ -0.024337769, -0.0056037903, -0.0473938, -0.0758667, 0.018661499, 0.0011949539, -0.06011963, 0.012062073, -0.01777649, 0.02949524, -0.009307861, 0.0034923553, -0.00086450577, -0.017120361, -0.028701782, 0.004085541, -0.016723633, 0.0...
http://arxiv.org/abs/2307.16220v1
Optimizing the Neural Network Training for OCR Error Correction of Historical Hebrew Texts
Omri Suissa, Avshalom Elmalech, Maayan Zhitomirsky-Geffet
cs.CL, cs.LG
Over the past few decades, large archives of paper-based documents such as books and newspapers have been digitized using Optical Character Recognition. This technology is error-prone, especially for historical documents. To correct OCR errors, post-processing algorithms have been proposed based on natural language ana...
http://arxiv.org/pdf/2307.16220v1
2023-07-30T12:59:06
2023-07-30T12:59:06
{ "title": [ -0.009002686, 0.020904541, -0.011795044, -0.024932861, 0.009239197, -0.0064468384, -0.008613586, -0.026611328, 0.011116028, 0.059509277, -0.029907227, 0.018188477, 0.048095703, -0.0031433105, -0.04714966, 0.012466431, -0.0135650635, ...
http://arxiv.org/abs/2307.16217v1
Text Analysis Using Deep Neural Networks in Digital Humanities and Information Science
Omri Suissa, Avshalom Elmalech, Maayan Zhitomirsky-Geffet
cs.LG, cs.AI, cs.CL
Combining computational technologies and humanities is an ongoing effort aimed at making resources such as texts, images, audio, video, and other artifacts digitally available, searchable, and analyzable. In recent years, deep neural networks (DNN) dominate the field of automatic text analysis and natural language proc...
http://arxiv.org/pdf/2307.16217v1
2023-07-30T12:54:39
2023-07-30T12:54:39
{ "title": [ -0.026931763, 0.014167786, -0.042297363, -0.023239136, 0.016448975, -0.02255249, 0.0012865067, -0.024795532, 0.010978699, 0.038208008, -0.0067214966, 0.0021018982, 0.0128479, -0.016052246, -0.059417725, 0.013000488, -0.0105896, 0.039...
http://arxiv.org/abs/2307.16214v1
Question Answering with Deep Neural Networks for Semi-Structured Heterogeneous Genealogical Knowledge Graphs
Omri Suissa, Maayan Zhitomirsky-Geffet, Avshalom Elmalech
cs.CL, cs.AI, cs.IR, cs.LG
With the rising popularity of user-generated genealogical family trees, new genealogical information systems have been developed. State-of-the-art natural question answering algorithms use deep neural network (DNN) architecture based on self-attention networks. However, some of these models use sequence-based inputs an...
http://arxiv.org/pdf/2307.16214v1
2023-07-30T12:49:54
2023-07-30T12:49:54
{ "title": [ -0.01486969, 0.007522583, -0.056671143, -0.07519531, 0.01737976, 0.0035858154, -0.059814453, 0.010246277, 0.03857422, 0.023910522, -0.011291504, -0.015090942, 0.014144897, -0.023284912, -0.047058105, 0.008407593, 0.0028762817, -0.006...
http://arxiv.org/abs/2307.16213v1
Toward a Period-Specific Optimized Neural Network for OCR Error Correction of Historical Hebrew Texts
Omri Suissa, Maayan Zhitomirsky-Geffet, Avshalom Elmalech
cs.CL, cs.LG
Over the past few decades, large archives of paper-based historical documents, such as books and newspapers, have been digitized using the Optical Character Recognition (OCR) technology. Unfortunately, this broadly used technology is error-prone, especially when an OCRed document was written hundreds of years ago. Neur...
http://arxiv.org/pdf/2307.16213v1
2023-07-30T12:40:31
2023-07-30T12:40:31
{ "title": [ -0.0048179626, 0.014892578, -0.007835388, -0.028121948, 0.010108948, -0.0019493103, -0.010528564, -0.022277832, 0.018173218, 0.0647583, -0.02458191, 0.020980835, 0.05529785, 0.0044822693, -0.056243896, 0.024093628, -0.013313293, 0.02...
http://arxiv.org/abs/2307.16208v1
Around the GLOBE: Numerical Aggregation Question-Answering on Heterogeneous Genealogical Knowledge Graphs with Deep Neural Networks
Omri Suissa, Maayan Zhitomirsky-Geffet, Avshalom Elmalech
cs.CL, cs.AI, cs.IR, cs.LG
One of the key AI tools for textual corpora exploration is natural language question-answering (QA). Unlike keyword-based search engines, QA algorithms receive and process natural language questions and produce precise answers to these questions, rather than long lists of documents that need to be manually scanned by t...
http://arxiv.org/pdf/2307.16208v1
2023-07-30T12:09:00
2023-07-30T12:09:00
{ "title": [ -0.0071258545, 0.012008667, -0.057373047, -0.081604004, 0.022476196, 0.0027160645, -0.046936035, 0.01675415, 0.047546387, 0.030838013, -0.042510986, 0.0043563843, 0.02355957, -0.007106781, -0.041992188, 0.01083374, 0.011993408, 0.000...
http://arxiv.org/abs/2307.16200v3
A Knowledge-enhanced Two-stage Generative Framework for Medical Dialogue Information Extraction
Zefa Hu, Ziyi Ni, Jing Shi, Shuang Xu, Bo Xu
cs.CL
This paper focuses on term-status pair extraction from medical dialogues (MD-TSPE), which is essential in diagnosis dialogue systems and the automatic scribe of electronic medical records (EMRs). In the past few years, works on MD-TSPE have attracted increasing research attention, especially after the remarkable progre...
http://arxiv.org/pdf/2307.16200v3
2023-10-26T04:55:52
2023-07-30T10:51:32
{ "title": [ -0.007194519, -0.012245178, -0.047668457, -0.09698486, 0.008842468, 0.004524231, -0.058807373, 0.024017334, 0.00013184547, 0.03286743, 0.015838623, 0.013160706, -0.02923584, -0.026855469, -0.060455322, 0.002122879, -0.003255844, 0.00...
http://arxiv.org/abs/2307.16199v1
Improving TTS for Shanghainese: Addressing Tone Sandhi via Word Segmentation
Yuanhao Chen
cs.CL
Tone is a crucial component of the prosody of Shanghainese, a Wu Chinese variety spoken primarily in urban Shanghai. Tone sandhi, which applies to all multi-syllabic words in Shanghainese, then, is key to natural-sounding speech. Unfortunately, recent work on Shanghainese TTS (text-to-speech) such as Apple's VoiceOver ...
http://arxiv.org/pdf/2307.16199v1
2023-07-30T10:50:18
2023-07-30T10:50:18
{ "title": [ 0.020080566, -0.010482788, -0.005657196, -0.036010742, -0.03189087, 0.004547119, -0.06378174, -0.01626587, -0.0064735413, 0.0513916, 0.004470825, -0.035369873, 0.041290283, -0.05014038, -0.051757812, -0.003824234, 0.0049362183, -0.00...
http://arxiv.org/abs/2307.16180v1
Do LLMs Possess a Personality? Making the MBTI Test an Amazing Evaluation for Large Language Models
Keyu Pan, Yawen Zeng
cs.CL
The field of large language models (LLMs) has made significant progress, and their knowledge storage capacity is approaching that of human beings. Furthermore, advanced techniques, such as prompt learning and reinforcement learning, are being employed to address ethical concerns and hallucination problems associated wi...
http://arxiv.org/pdf/2307.16180v1
2023-07-30T09:34:35
2023-07-30T09:34:35
{ "title": [ -0.057647705, -0.009178162, -0.0395813, -0.008865356, -0.02658081, -0.016723633, -0.043121338, -0.013290405, -0.025039673, 0.028030396, -0.007881165, -0.0024585724, 0.02128601, -0.037475586, -0.02381897, -0.0051116943, -0.017730713, ...
http://arxiv.org/abs/2307.16139v1
User-Controlled Knowledge Fusion in Large Language Models: Balancing Creativity and Hallucination
Chen Zhang
cs.CL, cs.AI, cs.LG
In modern dialogue systems, the use of Large Language Models (LLMs) has grown exponentially due to their capacity to generate diverse, relevant, and creative responses. Despite their strengths, striking a balance between the LLMs' creativity and their faithfulness to external knowledge remains a key challenge. This pap...
http://arxiv.org/pdf/2307.16139v1
2023-07-30T06:06:35
2023-07-30T06:06:35
{ "title": [ -0.028869629, -0.0039901733, -0.037261963, -0.056488037, -0.00819397, 0.0031089783, -0.053833008, -0.011276245, 0.014419556, 0.04421997, -0.019088745, -0.0012054443, 0.015396118, -0.0072250366, -0.059143066, 0.0029506683, -0.015556335, ...
http://arxiv.org/abs/2307.16125v2
SEED-Bench: Benchmarking Multimodal LLMs with Generative Comprehension
Bohao Li, Rui Wang, Guangzhi Wang, Yuying Ge, Yixiao Ge, Ying Shan
cs.CL, cs.CV
Based on powerful Large Language Models (LLMs), recent generative Multimodal Large Language Models (MLLMs) have gained prominence as a pivotal research area, exhibiting remarkable capability for both comprehension and generation. In this work, we address the evaluation of generative comprehension in MLLMs as a prelimin...
http://arxiv.org/pdf/2307.16125v2
2023-08-02T08:02:35
2023-07-30T04:25:16
{ "title": [ -0.072143555, -0.013587952, -0.036590576, -0.035858154, -0.053649902, 0.013504028, -0.103637695, -0.0053977966, 0.0024757385, 0.025558472, -0.034973145, 0.02810669, -0.013015747, -0.01361084, -0.015037537, -0.0082473755, -0.01789856, ...
http://arxiv.org/abs/2308.02037v1
Proposing a conceptual framework: social media listening for public health behavior
Shu-Feng Tsao, Helen Chen, Samantha Meyer, Zahid A. Butt
cs.CY, cs.CL, cs.SI
Existing communications and behavioral theories have been adopted to address health misinformation. Although various theories and models have been used to investigate the COVID-19 pandemic, there is no framework specially designed for social listening or misinformation studies using social media data and natural langua...
http://arxiv.org/pdf/2308.02037v1
2023-07-30T03:03:48
2023-07-30T03:03:48
{ "title": [ 0.03543091, -0.0015192032, -0.027801514, -0.074035645, 0.015426636, -0.017059326, -0.026184082, 0.0034885406, -0.058013916, 0.016433716, -0.0060272217, 0.033416748, -0.061035156, -0.031463623, -0.0006580353, -0.021392822, 0.005958557, ...
http://arxiv.org/abs/2307.16082v2
EnrichEvent: Enriching Social Data with Contextual Information for Emerging Event Extraction
Mohammadali Sefidi Esfahani, Mohammad Akbari
cs.CL, cs.AI
Social platforms have emerged as crucial platforms for disseminating information and discussing real-life social events, which offers an excellent opportunity for researchers to design and implement novel event detection frameworks. However, most existing approaches merely exploit keyword burstiness or network structur...
http://arxiv.org/pdf/2307.16082v2
2023-08-16T09:00:25
2023-07-29T21:37:55
{ "title": [ 0.026367188, 0.011108398, -0.04727173, -0.11480713, 0.048919678, -0.0029087067, -0.036193848, -0.002128601, 0.0059547424, 0.024536133, -0.017303467, 0.036193848, -0.007896423, -0.0141067505, -0.021575928, 0.0154418945, -0.0044021606, ...
http://arxiv.org/abs/2307.16081v1
Roll Up Your Sleeves: Working with a Collaborative and Engaging Task-Oriented Dialogue System
Lingbo Mo, Shijie Chen, Ziru Chen, Xiang Deng, Ashley Lewis, Sunit Singh, Samuel Stevens, Chang-You Tai, Zhen Wang, Xiang Yue, Tianshu Zhang, Yu Su, Huan Sun
cs.CL
We introduce TacoBot, a user-centered task-oriented digital assistant designed to guide users through complex real-world tasks with multiple steps. Covering a wide range of cooking and how-to tasks, we aim to deliver a collaborative and engaging dialogue experience. Equipped with language understanding, dialogue manage...
http://arxiv.org/pdf/2307.16081v1
2023-07-29T21:37:24
2023-07-29T21:37:24
{ "title": [ -0.028457642, -0.0076179504, -0.06536865, -0.04071045, 0.016738892, -0.029846191, -0.0181427, 0.013519287, -0.0129470825, 0.02532959, 0.0050621033, -0.007873535, -0.028213501, -0.033111572, -0.04925537, -0.026489258, -0.008384705, 0....
http://arxiv.org/abs/2307.16071v1
ÌròyìnSpeech: A multi-purpose Yorùbá Speech Corpus
Tolulope Ogunremi, Kola Tubosun, Anuoluwapo Aremu, Iroro Orife, David Ifeoluwa Adelani
cs.CL, cs.SD, eess.AS
We introduce the \`{I}r\`{o}y\`{i}nSpeech corpus -- a new dataset influenced by a desire to increase the amount of high quality, freely available, contemporary Yor\`{u}b\'{a} speech. We release a multi-purpose dataset that can be used for both TTS and ASR tasks. We curated text sentences from the news and creative writ...
http://arxiv.org/pdf/2307.16071v1
2023-07-29T20:42:50
2023-07-29T20:42:50
{ "title": [ -0.016418457, -0.017745972, -0.015823364, -0.060394287, -0.0055618286, 0.005092621, -0.046417236, 0.0034770966, -0.0116119385, 0.046813965, -0.026870728, -0.0012016296, 0.03161621, -0.017669678, -0.044189453, -0.011329651, -0.000702858, ...
http://arxiv.org/abs/2307.16045v1
Automatic Extraction of the Romanian Academic Word List: Data and Methods
Ana-Maria Bucur, Andreea Dincă, Mădălina Chitez, Roxana Rogobete
cs.CL
This paper presents the methodology and data used for the automatic extraction of the Romanian Academic Word List (Ro-AWL). Academic Word Lists are useful in both L2 and L1 teaching contexts. For the Romanian language, no such resource exists so far. Ro-AWL has been generated by combining methods from corpus and comput...
http://arxiv.org/pdf/2307.16045v1
2023-07-29T18:21:38
2023-07-29T18:21:38
{ "title": [ -0.043060303, 0.016220093, -0.050933838, 0.015640259, 0.014831543, -0.00630188, 0.011688232, -0.04550171, -0.02168274, 0.04397583, -0.022567749, -0.0027580261, 0.015083313, -0.009719849, -0.04360962, -0.007762909, -0.022842407, 0.018...
http://arxiv.org/abs/2307.16039v2
Okapi: Instruction-tuned Large Language Models in Multiple Languages with Reinforcement Learning from Human Feedback
Viet Dac Lai, Chien Van Nguyen, Nghia Trung Ngo, Thuat Nguyen, Franck Dernoncourt, Ryan A. Rossi, Thien Huu Nguyen
cs.CL, cs.LG
A key technology for the development of large language models (LLMs) involves instruction tuning that helps align the models' responses with human expectations to realize impressive learning abilities. Two major approaches for instruction tuning characterize supervised fine-tuning (SFT) and reinforcement learning from ...
http://arxiv.org/pdf/2307.16039v2
2023-08-02T00:39:25
2023-07-29T18:01:46
{ "title": [ -0.046966553, -0.026306152, -0.031585693, -0.042236328, -0.01828003, 0.0024776459, -0.055633545, 0.0013751984, -0.033599854, 0.02330017, -0.019607544, 0.020629883, -0.005722046, -0.012550354, -0.011505127, -0.012428284, -0.009056091, ...
http://arxiv.org/abs/2307.16013v1
Marrying Dialogue Systems with Data Visualization: Interactive Data Visualization Generation from Natural Language Conversations
Yuanfeng Song, Xuefang Zhao, Raymond Chi-Wing Wong
cs.AI, cs.CL, cs.DB
Data visualization (DV) has become the prevailing tool in the market due to its effectiveness into illustrating insights in vast amounts of data. To lower the barrier of using DVs, automatic DV tasks, such as natural language question (NLQ) to visualization translation (formally called text-to-vis), have been investiga...
http://arxiv.org/pdf/2307.16013v1
2023-07-29T15:50:07
2023-07-29T15:50:07
{ "title": [ -0.016983032, -0.015213013, -0.04232788, -0.071899414, -0.010536194, 0.007358551, -0.040740967, 0.0047836304, 0.0131073, 0.019943237, -0.035736084, 0.0019817352, -0.04309082, -0.009277344, -0.05960083, -0.034454346, -0.0042304993, 0....
http://arxiv.org/abs/2307.15997v1
RoCar: A Relationship Network-based Evaluation Method to Large Language Models
Ming Wang, Wenfang Wu, Chongyun Gao, Daling Wang, Shi Feng, Yifei Zhang
cs.CL, cs.AI
Large language models (LLMs) have received increasing attention. However, due to the complexity of its capabilities, how to rationally evaluate the capabilities of LLMs is still a task to be solved. We propose the RoCar method, which utilizes the defined basic schemas to randomly construct a task graph and generates na...
http://arxiv.org/pdf/2307.15997v1
2023-07-29T14:47:07
2023-07-29T14:47:07
{ "title": [ -0.043151855, -0.014289856, -0.046722412, -0.059814453, 0.002374649, 0.00995636, -0.06542969, 0.0056114197, 0.010978699, 0.008911133, -0.0025253296, -0.0035934448, -0.0075912476, -0.027023315, -0.044830322, 0.001739502, -0.03366089, ...
http://arxiv.org/abs/2307.15992v1
Towards Codable Text Watermarking for Large Language Models
Lean Wang, Wenkai Yang, Deli Chen, Hao Zhou, Yankai Lin, Fandong Meng, Jie Zhou, Xu Sun
cs.CL
As large language models (LLMs) generate texts with increasing fluency and realism, there is a growing need to identify the source of texts to prevent the abuse of LLMs. Text watermarking techniques have proven reliable in distinguishing whether a text is generated by LLMs by injecting hidden patterns into the generate...
http://arxiv.org/pdf/2307.15992v1
2023-07-29T14:11:15
2023-07-29T14:11:15
{ "title": [ -0.028778076, 0.0073242188, -0.02319336, -0.044708252, 0.013130188, 0.006122589, -0.040252686, -0.010276794, 0.00035452843, 0.042541504, -0.0037784576, 0.007133484, 0.01399231, -0.011497498, -0.045684814, 0.015357971, -0.023208618, 0...
http://arxiv.org/abs/2307.15936v2
A Theory for Emergence of Complex Skills in Language Models
Sanjeev Arora, Anirudh Goyal
cs.LG, cs.AI, cs.CL, stat.ML
A major driver of AI products today is the fact that new skills emerge in language models when their parameter set and training corpora are scaled up. This phenomenon is poorly understood, and a mechanistic explanation via mathematical analysis of gradient-based training seems difficult. The current paper takes a diffe...
http://arxiv.org/pdf/2307.15936v2
2023-11-06T00:36:24
2023-07-29T09:22:54
{ "title": [ -0.017059326, -0.0031929016, -0.0317688, -0.040771484, 0.001121521, -0.00970459, -0.022354126, 0.006881714, -0.01638794, 0.03781128, -0.0018625259, 0.01134491, 0.0003437996, -0.017364502, -0.050445557, 0.013160706, -0.021362305, 0.04...
http://arxiv.org/abs/2307.15933v1
GeneMask: Fast Pretraining of Gene Sequences to Enable Few-Shot Learning
Soumyadeep Roy, Jonas Wallat, Sowmya S Sundaram, Wolfgang Nejdl, Niloy Ganguly
cs.CL
Large-scale language models such as DNABert and LOGO aim to learn optimal gene representations and are trained on the entire Human Reference Genome. However, standard tokenization schemes involve a simple sliding window of tokens like k-mers that do not leverage any gene-based semantics and thus may lead to (trivial) m...
http://arxiv.org/pdf/2307.15933v1
2023-07-29T09:17:16
2023-07-29T09:17:16
{ "title": [ -0.01953125, -0.011657715, -0.055877686, -0.051513672, 0.009094238, -0.022033691, -0.030227661, 0.00605011, 0.00083065033, 0.013908386, -0.021209717, 0.011039734, -0.016433716, -0.038330078, -0.02998352, 0.006362915, 0.021636963, 0.0...
http://arxiv.org/abs/2308.01203v1
Analysing the Resourcefulness of the Paragraph for Precedence Retrieval
Bhoomeendra Singh Sisodiya, Narendra Babu Unnam, P. Krishna Reddy, Apala Das, K. V. K. Santhy, V. Balakista Reddy
cs.IR, cs.CL
Developing methods for extracting relevant legal information to aid legal practitioners is an active research area. In this regard, research efforts are being made by leveraging different kinds of information, such as meta-data, citations, keywords, sentences, paragraphs, etc. Similar to any text document, legal docume...
http://arxiv.org/pdf/2308.01203v1
2023-07-29T08:55:38
2023-07-29T08:55:38
{ "title": [ 0.011550903, 0.0004954338, -0.028305054, -0.06304932, -0.0020751953, -0.022109985, -0.017990112, -0.0055503845, -0.013954163, 0.052581787, -0.00472641, 0.0118255615, 0.000647068, -0.004299164, -0.060180664, -0.00217247, -0.006565094, ...
http://arxiv.org/abs/2307.15920v1
ATESA-BÆRT: A Heterogeneous Ensemble Learning Model for Aspect-Based Sentiment Analysis
Elena-Simona Apostol, Alin-Georgian Pisică, Ciprian-Octavian Truică
cs.CL, cs.AI
The increasing volume of online reviews has made possible the development of sentiment analysis models for determining the opinion of customers regarding different products and services. Until now, sentiment analysis has proven to be an effective tool for determining the overall polarity of reviews. To improve the gran...
http://arxiv.org/pdf/2307.15920v1
2023-07-29T07:35:19
2023-07-29T07:35:19
{ "title": [ 0.0072135925, -0.002866745, -0.02319336, -0.06439209, 0.011894226, 0.00712204, -0.032562256, -0.0040359497, 0.037017822, 0.05218506, -0.036895752, -0.0012378693, 0.00024843216, 0.005832672, -0.07574463, 0.0009975433, -0.013320923, 0....
http://arxiv.org/abs/2307.15833v1
Dialogue Shaping: Empowering Agents through NPC Interaction
Wei Zhou, Xiangyu Peng, Mark Riedl
cs.CL
One major challenge in reinforcement learning (RL) is the large amount of steps for the RL agent needs to converge in the training process and learn the optimal policy, especially in text-based game environments where the action space is extensive. However, non-player characters (NPCs) sometimes hold some key informati...
http://arxiv.org/pdf/2307.15833v1
2023-07-28T22:44:54
2023-07-28T22:44:54
{ "title": [ 0.010246277, -0.013557434, -0.03451538, -0.068237305, 0.010429382, -0.003566742, -0.008384705, 0.006576538, -0.018829346, 0.010803223, -0.0041618347, 0.0022716522, -0.022216797, -0.03253174, -0.021820068, -0.01675415, -0.02949524, 0....
http://arxiv.org/abs/2307.15818v1
RT-2: Vision-Language-Action Models Transfer Web Knowledge to Robotic Control
Anthony Brohan, Noah Brown, Justice Carbajal, Yevgen Chebotar, Xi Chen, Krzysztof Choromanski, Tianli Ding, Danny Driess, Avinava Dubey, Chelsea Finn, Pete Florence, Chuyuan Fu, Montse Gonzalez Arenas, Keerthana Gopalakrishnan, Kehang Han, Karol Hausman, Alexander Herzog, Jasmine Hsu, Brian Ichter, Alex Irpan, Nikhil J...
cs.RO, cs.CL, cs.CV, cs.LG
We study how vision-language models trained on Internet-scale data can be incorporated directly into end-to-end robotic control to boost generalization and enable emergent semantic reasoning. Our goal is to enable a single end-to-end trained model to both learn to map robot observations to actions and enjoy the benefit...
http://arxiv.org/pdf/2307.15818v1
2023-07-28T21:18:02
2023-07-28T21:18:02
{ "title": [ -0.026245117, -0.016555786, -0.048583984, -0.035369873, 0.003419876, -0.022598267, -0.018249512, -0.008956909, -0.009544373, 0.017349243, -0.022262573, 0.0022068024, -0.0007777214, -0.022857666, -0.021484375, -0.013267517, 0.030197144, ...
http://arxiv.org/abs/2307.15776v2
Select and Augment: Enhanced Dense Retrieval Knowledge Graph Augmentation
Micheal Abaho, Yousef H. Alfaifi
cs.CL, cs.AI
Injecting textual information into knowledge graph (KG) entity representations has been a worthwhile expedition in terms of improving performance in KG oriented tasks within the NLP community. External knowledge often adopted to enhance KG embeddings ranges from semantically rich lexical dependency parsed features to a...
http://arxiv.org/pdf/2307.15776v2
2023-10-25T10:35:03
2023-07-28T19:33:18
{ "title": [ 0.0030498505, -0.0060806274, -0.060333252, -0.0715332, 0.0072784424, -0.0068244934, -0.027420044, 0.010421753, 0.03466797, 0.031402588, -0.042144775, 0.03375244, 0.012207031, -0.001364708, -0.039886475, 0.015914917, 0.009925842, 0.01...
http://arxiv.org/abs/2307.15771v1
The Hydra Effect: Emergent Self-repair in Language Model Computations
Thomas McGrath, Matthew Rahtz, Janos Kramar, Vladimir Mikulik, Shane Legg
cs.LG, cs.AI, cs.CL
We investigate the internal structure of language model computations using causal analysis and demonstrate two motifs: (1) a form of adaptive computation where ablations of one attention layer of a language model cause another layer to compensate (which we term the Hydra effect) and (2) a counterbalancing function of l...
http://arxiv.org/pdf/2307.15771v1
2023-07-28T19:13:26
2023-07-28T19:13:26
{ "title": [ -0.025619507, -0.011360168, -0.029067993, -0.07165527, -0.013839722, 0.008552551, -0.074401855, 0.024215698, 0.004058838, 0.047698975, -0.013145447, 0.014144897, 0.024505615, -0.0031223297, -0.060760498, 0.02027893, -0.012466431, 0.0...
http://arxiv.org/abs/2307.15770v2
CHATREPORT: Democratizing Sustainability Disclosure Analysis through LLM-based Tools
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, cs.AI
In the face of climate change, are companies really taking substantial steps toward more sustainable operations? A comprehensive answer lies in the dense, information-rich landscape of corporate sustainability reports. However, the sheer volume and complexity of these reports make human analysis very costly. Therefore,...
http://arxiv.org/pdf/2307.15770v2
2023-10-11T16:49:29
2023-07-28T18:58:16
{ "title": [ -0.05618286, -0.017623901, -0.07537842, -0.07458496, -0.019302368, 0.016189575, -0.06677246, 0.033966064, -0.00762558, 0.023086548, -0.048461914, 0.026657104, -0.036102295, -0.0064468384, -0.041381836, -0.03466797, -0.038757324, 0.05...
http://arxiv.org/abs/2307.15759v1
Lessons in Reproducibility: Insights from NLP Studies in Materials Science
Xiangyun Lei, Edward Kim, Viktoriia Baibakova, Shijing Sun
physics.chem-ph, cs.AI, cs.CL
Natural Language Processing (NLP), a cornerstone field within artificial intelligence, has been increasingly utilized in the field of materials science literature. Our study conducts a reproducibility analysis of two pioneering works within this domain: "Machine-learned and codified synthesis parameters of oxide materi...
http://arxiv.org/pdf/2307.15759v1
2023-07-28T18:36:42
2023-07-28T18:36:42
{ "title": [ -0.062683105, -0.015777588, -0.054779053, -0.025436401, -0.02708435, -0.0013618469, -0.062927246, 0.00907135, -0.00019836426, 0.011016846, -0.014717102, 0.014053345, -0.014511108, -0.026412964, -0.018051147, 0.019363403, -0.04168701, ...
http://arxiv.org/abs/2307.15752v1
Resume Evaluation through Latent Dirichlet Allocation and Natural Language Processing for Effective Candidate Selection
Vidhita Jagwani, Smit Meghani, Krishna Pai, Sudhir Dhage
cs.CL, cs.LG
In this paper, we propose a method for resume rating using Latent Dirichlet Allocation (LDA) and entity detection with SpaCy. The proposed method first extracts relevant entities such as education, experience, and skills from the resume using SpaCy's Named Entity Recognition (NER). The LDA model then uses these entitie...
http://arxiv.org/pdf/2307.15752v1
2023-07-28T18:11:17
2023-07-28T18:11:17
{ "title": [ 0.048828125, -0.01689148, 0.0021533966, -0.049560547, -0.001001358, -0.017913818, 0.032409668, -0.023742676, -0.0039405823, 0.020889282, -0.04220581, 0.017593384, 0.0005722046, -0.015426636, -0.025604248, -0.03567505, -0.024536133, 0...
http://arxiv.org/abs/2307.15745v2
Context-VQA: Towards Context-Aware and Purposeful Visual Question Answering
Nandita Naik, Christopher Potts, Elisa Kreiss
cs.CL, cs.CV
Visual question answering (VQA) has the potential to make the Internet more accessible in an interactive way, allowing people who cannot see images to ask questions about them. However, multiple studies have shown that people who are blind or have low-vision prefer image explanations that incorporate the context in whi...
http://arxiv.org/pdf/2307.15745v2
2023-08-30T15:58:56
2023-07-28T18:01:08
{ "title": [ 0.007789612, 0.001250267, -0.028518677, -0.049560547, 0.00680542, -0.0057296753, -0.0024719238, -0.0046463013, 0.032470703, 0.03656006, -0.027801514, -0.011886597, 0.005722046, -0.0035629272, -0.07043457, -0.006385803, 0.016525269, 0...
http://arxiv.org/abs/2307.15703v1
Uncertainty in Natural Language Generation: From Theory to Applications
Joris Baan, Nico Daheim, Evgenia Ilia, Dennis Ulmer, Haau-Sing Li, Raquel Fernández, Barbara Plank, Rico Sennrich, Chrysoula Zerva, Wilker Aziz
cs.CL, cs.AI, cs.LG
Recent advances of powerful Language Models have allowed Natural Language Generation (NLG) to emerge as an important technology that can not only perform traditional tasks like summarisation or translation, but also serve as a natural language interface to a variety of applications. As such, it is crucial that NLG syst...
http://arxiv.org/pdf/2307.15703v1
2023-07-28T17:51:21
2023-07-28T17:51:21
{ "title": [ -0.05328369, -0.012809753, -0.05419922, -0.050445557, 0.0010194778, 0.008598328, -0.060943604, 0.000016629696, 0.0010509491, 0.016708374, -0.015449524, 0.009155273, -0.016235352, -0.024139404, -0.04650879, -0.00016355515, -0.019134521, ...
http://arxiv.org/abs/2307.15644v2
Scaling Data Generation in Vision-and-Language Navigation
Zun Wang, Jialu Li, Yicong Hong, Yi Wang, Qi Wu, Mohit Bansal, Stephen Gould, Hao Tan, Yu Qiao
cs.CV, cs.AI, cs.CL, cs.LG
Recent research in language-guided visual navigation has demonstrated a significant demand for the diversity of traversable environments and the quantity of supervision for training generalizable agents. To tackle the common data scarcity issue in existing vision-and-language navigation datasets, we propose an effectiv...
http://arxiv.org/pdf/2307.15644v2
2023-08-09T23:51:22
2023-07-28T16:03:28
{ "title": [ -0.019561768, -0.020523071, -0.032592773, -0.03955078, -0.012542725, -0.017471313, -0.024505615, -0.0035419464, 0.015007019, 0.024398804, -0.04660034, -0.0113220215, -0.037078857, -0.014915466, -0.057434082, -0.024978638, 0.021911621, ...
http://arxiv.org/abs/2307.15593v2
Robust Distortion-free Watermarks for Language Models
Rohith Kuditipudi, John Thickstun, Tatsunori Hashimoto, Percy Liang
cs.LG, cs.CL, cs.CR
We propose a methodology for planting watermarks in text from an autoregressive language model that are robust to perturbations without changing the distribution over text up to a certain maximum generation budget. We generate watermarked text by mapping a sequence of random numbers -- which we compute using a randomiz...
http://arxiv.org/pdf/2307.15593v2
2023-09-27T15:42:24
2023-07-28T14:52:08
{ "title": [ -0.016296387, 0.013832092, -0.033203125, -0.060180664, 0.01713562, 0.0013141632, -0.054901123, -0.02293396, 0.0030460358, 0.04663086, -0.0069770813, 0.010665894, 0.012870789, -0.009536743, -0.047332764, 0.022491455, -0.012016296, 0.0...
http://arxiv.org/abs/2307.15582v1
When to generate hedges in peer-tutoring interactions
Alafate Abulimiti, Chloé Clavel, Justine Cassell
cs.CL
This paper explores the application of machine learning techniques to predict where hedging occurs in peer-tutoring interactions. The study uses a naturalistic face-to-face dataset annotated for natural language turns, conversational strategies, tutoring strategies, and nonverbal behaviours. These elements are processe...
http://arxiv.org/pdf/2307.15582v1
2023-07-28T14:29:19
2023-07-28T14:29:19
{ "title": [ 0.041107178, 0.005252838, -0.008522034, 0.002773285, -0.028518677, -0.030059814, 0.019927979, -0.043182373, -0.051849365, 0.04711914, 0.0010976791, 0.028076172, -0.014831543, -0.024597168, -0.01828003, -0.01939392, -0.0073623657, 0.0...
http://arxiv.org/abs/2307.15555v1
All-for-One and One-For-All: Deep learning-based feature fusion for Synthetic Speech Detection
Daniele Mari, Davide Salvi, Paolo Bestagini, Simone Milani
cs.SD, cs.CL, cs.CR, eess.AS
Recent advances in deep learning and computer vision have made the synthesis and counterfeiting of multimedia content more accessible than ever, leading to possible threats and dangers from malicious users. In the audio field, we are witnessing the growth of speech deepfake generation techniques, which solicit the deve...
http://arxiv.org/pdf/2307.15555v1
2023-07-28T13:50:25
2023-07-28T13:50:25
{ "title": [ -0.017196655, 0.01637268, -0.04925537, -0.06048584, 0.020599365, -0.008460999, -0.025527954, 0.00349617, -0.019256592, 0.033355713, -0.023544312, 0.000056803226, 0.015319824, -0.027709961, -0.014968872, -0.0184021, -0.027755737, 0.05...
http://arxiv.org/abs/2307.15554v1
'What are you referring to?' Evaluating the Ability of Multi-Modal Dialogue Models to Process Clarificational Exchanges
Javier Chiyah-Garcia, Alessandro Suglia, Arash Eshghi, Helen Hastie
cs.CL
Referential ambiguities arise in dialogue when a referring expression does not uniquely identify the intended referent for the addressee. Addressees usually detect such ambiguities immediately and work with the speaker to repair it using meta-communicative, Clarificational Exchanges (CE): a Clarification Request (CR) a...
http://arxiv.org/pdf/2307.15554v1
2023-07-28T13:44:33
2023-07-28T13:44:33
{ "title": [ -0.015686035, 0.004432678, -0.040740967, -0.020477295, 0.00029468536, -0.024841309, -0.05307007, -0.034210205, -0.0044555664, 0.028778076, 0.006504059, 0.003250122, -0.009292603, -0.02268982, -0.06161499, -0.0076942444, -0.026184082, ...
http://arxiv.org/abs/2307.15543v1
Oracle Computability and Turing Reducibility in the Calculus of Inductive Constructions
Yannick Forster, Dominik Kirst, Niklas Mück
cs.LO, cs.CL, math.LO
We develop synthetic notions of oracle computability and Turing reducibility in the Calculus of Inductive Constructions (CIC), the constructive type theory underlying the Coq proof assistant. As usual in synthetic approaches, we employ a definition of oracle computations based on meta-level functions rather than object...
http://arxiv.org/pdf/2307.15543v1
2023-07-28T13:16:46
2023-07-28T13:16:46
{ "title": [ -0.015037537, -0.007785797, -0.039398193, -0.05041504, 0.0046081543, -0.017120361, -0.0132751465, 0.0012235641, -0.031311035, -0.0014324188, -0.052001953, 0.026412964, -0.0637207, -0.01675415, -0.00705719, -0.030639648, -0.01159668, ...
http://arxiv.org/abs/2307.15508v1
The Road to Quality is Paved with Good Revisions: A Detailed Evaluation Methodology for Revision Policies in Incremental Sequence Labelling
Brielen Madureira, Patrick Kahardipraja, David Schlangen
cs.CL
Incremental dialogue model components produce a sequence of output prefixes based on incoming input. Mistakes can occur due to local ambiguities or to wrong hypotheses, making the ability to revise past outputs a desirable property that can be governed by a policy. In this work, we formalise and characterise edits and ...
http://arxiv.org/pdf/2307.15508v1
2023-07-28T12:08:15
2023-07-28T12:08:15
{ "title": [ -0.014831543, -0.002708435, -0.052246094, -0.099121094, 0.0018930435, -0.014892578, -0.07543945, 0.024887085, -0.016143799, 0.0074501038, -0.046051025, 0.0019330978, -0.021240234, -0.023895264, -0.021743774, -0.040222168, -0.015899658, ...
http://arxiv.org/abs/2307.15504v1
Exploring Format Consistency for Instruction Tuning
Shihao Liang, Kunlun Zhu, Runchu Tian, Yujia Qin, Huadong Wang, Xin Cong, Zhiyuan Liu, Xiaojiang Liu, Maosong Sun
cs.CL, cs.AI
Instruction tuning has emerged as a promising approach to enhancing large language models in following human instructions. It is shown that increasing the diversity and number of instructions in the training data can consistently enhance generalization performance, which facilitates a recent endeavor to collect various...
http://arxiv.org/pdf/2307.15504v1
2023-07-28T12:00:13
2023-07-28T12:00:13
{ "title": [ -0.06384277, -0.0059661865, -0.041778564, -0.04724121, -0.010292053, 0.0050697327, -0.0362854, 0.04095459, -0.01499176, 0.04711914, -0.02178955, 0.0067863464, -0.010543823, -0.015541077, -0.039093018, -0.003894806, 0.0039863586, 0.01...
http://arxiv.org/abs/2307.15494v1
ETHER: Aligning Emergent Communication for Hindsight Experience Replay
Kevin Denamganaï, Daniel Hernandez, Ozan Vardal, Sondess Missaoui, James Alfred Walker
cs.CL, cs.AI, cs.HC
Natural language instruction following is paramount to enable collaboration between artificial agents and human beings. Natural language-conditioned reinforcement learning (RL) agents have shown how natural languages' properties, such as compositionality, can provide a strong inductive bias to learn complex policies. P...
http://arxiv.org/pdf/2307.15494v1
2023-07-28T11:42:31
2023-07-28T11:42:31
{ "title": [ -0.0368042, -0.012802124, -0.028015137, -0.039764404, -0.026123047, 0.007698059, -0.035705566, -0.012321472, 0.03050232, 0.044036865, -0.014709473, -0.0049209595, 0.03012085, -0.021652222, -0.05886841, 0.01713562, 0.013664246, 0.0093...
http://arxiv.org/abs/2307.15493v1
The timing bottleneck: Why timing and overlap are mission-critical for conversational user interfaces, speech recognition and dialogue systems
Andreas Liesenfeld, Alianda Lopez, Mark Dingemanse
cs.CL
Speech recognition systems are a key intermediary in voice-driven human-computer interaction. Although speech recognition works well for pristine monologic audio, real-life use cases in open-ended interactive settings still present many challenges. We argue that timing is mission-critical for dialogue systems, and eval...
http://arxiv.org/pdf/2307.15493v1
2023-07-28T11:38:05
2023-07-28T11:38:05
{ "title": [ -0.005619049, -0.004875183, -0.054656982, -0.039001465, -0.0027751923, -0.035827637, -0.036193848, -0.0038261414, -0.027679443, 0.058288574, -0.022628784, 0.009559631, 0.009185791, -0.03213501, -0.028701782, -0.029830933, -0.018310547, ...
http://arxiv.org/abs/2307.15484v2
Minimally-Supervised Speech Synthesis with Conditional Diffusion Model and Language Model: A Comparative Study of Semantic Coding
Chunyu Qiang, Hao Li, Hao Ni, He Qu, Ruibo Fu, Tao Wang, Longbiao Wang, Jianwu Dang
cs.SD, cs.AI, cs.CL, eess.AS
Recently, there has been a growing interest in text-to-speech (TTS) methods that can be trained with minimal supervision by combining two types of discrete speech representations and using two sequence-to-sequence tasks to decouple TTS. However, existing methods suffer from three problems: the high dimensionality and w...
http://arxiv.org/pdf/2307.15484v2
2023-09-01T12:16:20
2023-07-28T11:20:23
{ "title": [ -0.029052734, -0.0033893585, -0.061279297, -0.053710938, -0.016174316, 0.0055770874, -0.06774902, 0.026428223, -0.01651001, 0.03717041, -0.033447266, 0.0051574707, 0.019119263, -0.02609253, -0.01914978, -0.014701843, -0.015029907, 0....
http://arxiv.org/abs/2307.16806v1
Testing the Depth of ChatGPT's Comprehension via Cross-Modal Tasks Based on ASCII-Art: GPT3.5's Abilities in Regard to Recognizing and Generating ASCII-Art Are Not Totally Lacking
David Bayani
cs.CL, cs.AI, cs.LG
Over the eight months since its release, ChatGPT and its underlying model, GPT3.5, have garnered massive attention, due to their potent mix of capability and accessibility. While a niche-industry of papers have emerged examining the scope of capabilities these models possess, the information fed to and extracted from t...
http://arxiv.org/pdf/2307.16806v1
2023-07-28T10:45:14
2023-07-28T10:45:14
{ "title": [ -0.0030021667, 0.03741455, 0.008636475, 0.026306152, -0.03036499, 0.020980835, -0.017868042, -0.053344727, 0.03540039, 0.039794922, -0.050598145, -0.013442993, 0.042114258, 0.0053520203, -0.01889038, -0.0073280334, 0.0146865845, 0.05...
http://arxiv.org/abs/2307.15460v1
Cross-Modal Concept Learning and Inference for Vision-Language Models
Yi Zhang, Ce Zhang, Yushun Tang, Zhihai He
cs.CV, cs.CL
Large-scale pre-trained Vision-Language Models (VLMs), such as CLIP, establish the correlation between texts and images, achieving remarkable success on various downstream tasks with fine-tuning. In existing fine-tuning methods, the class-specific text description is matched against the whole image. We recognize that t...
http://arxiv.org/pdf/2307.15460v1
2023-07-28T10:26:28
2023-07-28T10:26:28
{ "title": [ -0.01940918, -0.005252838, -0.03765869, -0.04824829, -0.0029850006, -0.021453857, -0.046722412, -0.012893677, 0.038238525, 0.020095825, -0.049316406, 0.0065231323, 0.010528564, -0.0062942505, -0.035858154, -0.004989624, 0.0056877136, ...
http://arxiv.org/abs/2307.15455v2
Trie-NLG: Trie Context Augmentation to Improve Personalized Query Auto-Completion for Short and Unseen Prefixes
Kaushal Kumar Maurya, Maunendra Sankar Desarkar, Manish Gupta, Puneet Agrawal
cs.CL
Query auto-completion (QAC) aims to suggest plausible completions for a given query prefix. Traditionally, QAC systems have leveraged tries curated from historical query logs to suggest most popular completions. In this context, there are two specific scenarios that are difficult to handle for any QAC system: short pre...
http://arxiv.org/pdf/2307.15455v2
2023-10-23T05:16:56
2023-07-28T10:17:30
{ "title": [ 0.03137207, -0.023635864, -0.044952393, -0.04827881, -0.0053634644, -0.021972656, 0.0032958984, 0.01939392, -0.012237549, 0.0725708, -0.018829346, 0.043701172, 0.042510986, -0.0072135925, -0.002401352, 0.023986816, -0.017059326, 0.02...
http://arxiv.org/abs/2307.15453v2
From Probabilistic Programming to Complexity-based Programming
Giovanni Sileno, Jean-Louis Dessalles
cs.AI, cs.CL
The paper presents the main characteristics and a preliminary implementation of a novel computational framework named CompLog. Inspired by probabilistic programming systems like ProbLog, CompLog builds upon the inferential mechanisms proposed by Simplicity Theory, relying on the computation of two Kolmogorov complexiti...
http://arxiv.org/pdf/2307.15453v2
2023-09-11T14:06:27
2023-07-28T10:11:01
{ "title": [ -0.04916382, 0.00060510635, -0.031143188, -0.07788086, 0.014320374, 0.020324707, -0.057006836, 0.0059547424, -0.007129669, 0.0042533875, 0.0047073364, 0.018493652, -0.047180176, -0.042297363, -0.02268982, 0.0029087067, -0.012817383, ...
http://arxiv.org/abs/2307.15432v1
CFN-ESA: A Cross-Modal Fusion Network with Emotion-Shift Awareness for Dialogue Emotion Recognition
Jiang Li, Yingjian Liu, Xiaoping Wang, Zhigang Zeng
cs.CL
Multimodal Emotion Recognition in Conversation (ERC) has garnered growing attention from research communities in various fields. In this paper, we propose a cross-modal fusion network with emotion-shift awareness (CFN-ESA) for ERC. Extant approaches employ each modality equally without distinguishing the amount of emot...
http://arxiv.org/pdf/2307.15432v1
2023-07-28T09:29:42
2023-07-28T09:29:42
{ "title": [ -0.00844574, 0.0066375732, -0.04397583, -0.02645874, -0.0042495728, -0.020889282, -0.030563354, -0.018051147, 0.035064697, 0.02154541, -0.012054443, -0.017532349, 0.015914917, -0.03729248, -0.050598145, 0.0018033981, -0.021194458, 0....
http://arxiv.org/abs/2307.15425v1
A Critical Review of Large Language Models: Sensitivity, Bias, and the Path Toward Specialized AI
Arash Hajikhani, Carolyn Cole
cs.CL, cs.AI
This paper examines the comparative effectiveness of a specialized compiled language model and a general-purpose model like OpenAI's GPT-3.5 in detecting SDGs within text data. It presents a critical review of Large Language Models (LLMs), addressing challenges related to bias and sensitivity. The necessity of speciali...
http://arxiv.org/pdf/2307.15425v1
2023-07-28T09:20:22
2023-07-28T09:20:22
{ "title": [ -0.037200928, -0.016799927, -0.025802612, -0.05178833, -0.008598328, 0.0053977966, -0.050567627, -0.0046577454, -0.0015010834, 0.031021118, -0.009597778, -0.0027637482, 0.007019043, -0.018005371, -0.04812622, 0.0146865845, -0.014930725, ...
http://arxiv.org/abs/2307.15413v1
Improving Social Media Popularity Prediction with Multiple Post Dependencies
Zhizhen Zhang, Xiaohui Xie, Mengyu Yang, Ye Tian, Yong Jiang, Yong Cui
cs.MM, cs.AI, cs.CL
Social Media Popularity Prediction has drawn a lot of attention because of its profound impact on many different applications, such as recommendation systems and multimedia advertising. Despite recent efforts to leverage the content of social media posts to improve prediction accuracy, many existing models fail to full...
http://arxiv.org/pdf/2307.15413v1
2023-07-28T09:06:50
2023-07-28T09:06:50
{ "title": [ 0.014984131, 0.013084412, -0.035003662, -0.08325195, 0.03375244, -0.0006260872, -0.023513794, 0.001291275, -0.0051231384, 0.03652954, -0.023544312, 0.0368042, -0.0287323, 0.001909256, -0.040283203, 0.004459381, -0.025009155, 0.049835...
http://arxiv.org/abs/2307.15411v2
Investigating the Learning Behaviour of In-context Learning: A Comparison with Supervised Learning
Xindi Wang, Yufei Wang, Can Xu, Xiubo Geng, Bowen Zhang, Chongyang Tao, Frank Rudzicz, Robert E. Mercer, Daxin Jiang
cs.CL
Large language models (LLMs) have shown remarkable capacity for in-context learning (ICL), where learning a new task from just a few training examples is done without being explicitly pre-trained. However, despite the success of LLMs, there has been little understanding of how ICL learns the knowledge from the given pr...
http://arxiv.org/pdf/2307.15411v2
2023-08-01T16:04:09
2023-07-28T09:03:19
{ "title": [ -0.0032997131, 0.0061302185, -0.04925537, -0.023498535, 0.03375244, -0.03942871, 0.025756836, -0.023071289, -0.015098572, 0.016021729, -0.036071777, 0.016784668, -0.036010742, -0.012916565, -0.033081055, -0.020065308, -0.01687622, 0....
http://arxiv.org/abs/2307.15410v1
Towards a Fully Unsupervised Framework for Intent Induction in Customer Support Dialogues
Rita Costa, Bruno Martins, Sérgio Viana, Luisa Coheur
cs.CL, I.2; I.7
State of the art models in intent induction require annotated datasets. However, annotating dialogues is time-consuming, laborious and expensive. In this work, we propose a completely unsupervised framework for intent induction within a dialogue. In addition, we show how pre-processing the dialogue corpora can improve ...
http://arxiv.org/pdf/2307.15410v1
2023-07-28T09:03:14
2023-07-28T09:03:14
{ "title": [ -0.00032877922, -0.0037078857, -0.042419434, -0.05419922, 0.008308411, -0.015380859, -0.033233643, 0.0071372986, 0.0231781, 0.013938904, -0.042785645, -0.0043945312, -0.002500534, -0.016967773, -0.046081543, -0.014945984, -0.012489319, ...
http://arxiv.org/abs/2307.15376v1
Multilingual Tourist Assistance using ChatGPT: Comparing Capabilities in Hindi, Telugu, and Kannada
Sanjana Kolar, Rohit Kumar
cs.CL
This research investigates the effectiveness of ChatGPT, an AI language model by OpenAI, in translating English into Hindi, Telugu, and Kannada languages, aimed at assisting tourists in India's linguistically diverse environment. To measure the translation quality, a test set of 50 questions from diverse fields such as...
http://arxiv.org/pdf/2307.15376v1
2023-07-28T07:52:26
2023-07-28T07:52:26
{ "title": [ 0.0075149536, -0.01737976, 0.001329422, -0.008682251, -0.010307312, 0.0011100769, -0.010536194, -0.016677856, -0.014045715, 0.036346436, -0.016311646, 0.003955841, -0.006225586, -0.026641846, -0.019058228, -0.0115356445, -0.022781372, ...
http://arxiv.org/abs/2307.15343v2
Med-HALT: Medical Domain Hallucination Test for Large Language Models
Ankit Pal, Logesh Kumar Umapathi, Malaikannan Sankarasubbu
cs.CL, cs.AI, cs.LG, stat.ML
This research paper focuses on the challenges posed by hallucinations in large language models (LLMs), particularly in the context of the medical domain. Hallucination, wherein these models generate plausible yet unverified or incorrect information, can have serious consequences in healthcare applications. We propose a...
http://arxiv.org/pdf/2307.15343v2
2023-10-14T17:04:12
2023-07-28T06:43:04
{ "title": [ -0.010757446, 0.0028591156, -0.036132812, -0.049316406, -0.0009217262, -0.011138916, -0.03933716, -0.014511108, 0.011138916, 0.03466797, -0.0038928986, -0.0054969788, 0.03302002, -0.023452759, -0.04547119, 0.01878357, -0.016281128, 0...
http://arxiv.org/abs/2307.15341v1
Teach Me How to Improve My Argumentation Skills: A Survey on Feedback in Argumentation
Camélia Guerraoui, Paul Reisert, Naoya Inoue, Farjana Sultana Mim, Shoichi Naito, Jungmin Choi, Irfan Robbani, Wenzhi Wang, Kentaro Inui
cs.CL
The use of argumentation in education has been shown to improve critical thinking skills for end-users such as students, and computational models for argumentation have been developed to assist in this process. Although these models are useful for evaluating the quality of an argument, they oftentimes cannot explain wh...
http://arxiv.org/pdf/2307.15341v1
2023-07-28T06:33:09
2023-07-28T06:33:09
{ "title": [ 0.025482178, -0.011222839, -0.03262329, -0.03363037, -0.00121212, -0.03074646, 0.0073623657, -0.01259613, -0.050079346, 0.0087890625, -0.010292053, 0.005039215, -0.006046295, -0.03363037, -0.0011072159, -0.03262329, 0.0035324097, 0.0...
http://arxiv.org/abs/2307.15337v2
Skeleton-of-Thought: Large Language Models Can Do Parallel Decoding
Xuefei Ning, Zinan Lin, Zixuan Zhou, Zifu Wang, Huazhong Yang, Yu Wang
cs.CL, cs.AI
This work aims at decreasing the end-to-end generation latency of large language models (LLMs). One of the major causes of the high generation latency is the sequential decoding approach adopted by almost all state-of-the-art LLMs. In this work, motivated by the thinking and writing process of humans, we propose Skelet...
http://arxiv.org/pdf/2307.15337v2
2023-10-08T03:59:29
2023-07-28T06:31:34
{ "title": [ -0.041992188, -0.029342651, -0.03643799, -0.03741455, -0.087402344, -0.007446289, -0.07513428, 0.013328552, -0.026382446, 0.06088257, 0.0051460266, -0.021621704, 0.049957275, -0.046051025, -0.022766113, -0.0032577515, 0.000046551228, ...
http://arxiv.org/abs/2307.15335v1
BARTPhoBEiT: Pre-trained Sequence-to-Sequence and Image Transformers Models for Vietnamese Visual Question Answering
Khiem Vinh Tran, Kiet Van Nguyen, Ngan Luu Thuy Nguyen
cs.CL, cs.CV
Visual Question Answering (VQA) is an intricate and demanding task that integrates natural language processing (NLP) and computer vision (CV), capturing the interest of researchers. The English language, renowned for its wealth of resources, has witnessed notable advancements in both datasets and models designed for VQ...
http://arxiv.org/pdf/2307.15335v1
2023-07-28T06:23:32
2023-07-28T06:23:32
{ "title": [ 0.007522583, 0.0023994446, 0.008003235, -0.047973633, -0.010017395, 0.010322571, -0.024032593, -0.015289307, 0.031921387, 0.034088135, -0.04345703, 0.005344391, -0.0049858093, -0.0048980713, -0.06915283, -0.010902405, 0.012367249, 0....
http://arxiv.org/abs/2307.15331v1
Tutorials on Stance Detection using Pre-trained Language Models: Fine-tuning BERT and Prompting Large Language Models
Yun-Shiuan Chuang
cs.CL, cs.AI
This paper presents two self-contained tutorials on stance detection in Twitter data using BERT fine-tuning and prompting large language models (LLMs). The first tutorial explains BERT architecture and tokenization, guiding users through training, tuning, and evaluating standard and domain-specific BERT models with Hug...
http://arxiv.org/pdf/2307.15331v1
2023-07-28T06:15:27
2023-07-28T06:15:27
{ "title": [ -0.0029182434, -0.008384705, 0.0013809204, -0.06439209, -0.037109375, 0.005317688, -0.08258057, -0.011192322, -0.011993408, 0.03982544, -0.033935547, -0.029434204, 0.031021118, -0.03173828, -0.023727417, -0.019119263, 0.016937256, 0....
http://arxiv.org/abs/2308.01420v1
SAP-sLDA: An Interpretable Interface for Exploring Unstructured Text
Charumathi Badrinath, Weiwei Pan, Finale Doshi-Velez
cs.CL, cs.LG
A common way to explore text corpora is through low-dimensional projections of the documents, where one hopes that thematically similar documents will be clustered together in the projected space. However, popular algorithms for dimensionality reduction of text corpora, like Latent Dirichlet Allocation (LDA), often pro...
http://arxiv.org/pdf/2308.01420v1
2023-07-28T05:43:39
2023-07-28T05:43:39
{ "title": [ -0.05105591, -0.0037174225, -0.042388916, -0.044433594, -0.014389038, 0.013320923, -0.06793213, -0.0259552, -0.013206482, 0.040130615, -0.009963989, 0.007167816, -0.0013456345, -0.008361816, -0.0574646, -0.010063171, -0.009246826, 0....
http://arxiv.org/abs/2307.15311v1
TrafficSafetyGPT: Tuning a Pre-trained Large Language Model to a Domain-Specific Expert in Transportation Safety
Ou Zheng, Mohamed Abdel-Aty, Dongdong Wang, Chenzhu Wang, Shengxuan Ding
cs.CL
Large Language Models (LLMs) have shown remarkable effectiveness in various general-domain natural language processing (NLP) tasks. However, their performance in transportation safety domain tasks has been suboptimal, primarily attributed to the requirement for specialized transportation safety expertise in generating ...
http://arxiv.org/pdf/2307.15311v1
2023-07-28T05:17:11
2023-07-28T05:17:11
{ "title": [ 0.0054779053, -0.00907135, -0.04510498, -0.067993164, 0.0017290115, -0.02986145, -0.031799316, -0.0021953583, -0.032043457, 0.020996094, -0.030685425, 0.019561768, -0.007610321, -0.020629883, 0.01713562, -0.023117065, 0.014678955, -0...
http://arxiv.org/abs/2307.15293v1
WC-SBERT: Zero-Shot Text Classification via SBERT with Self-Training for Wikipedia Categories
Te-Yu Chi, Yu-Meng Tang, Chia-Wen Lu, Qiu-Xia Zhang, Jyh-Shing Roger Jang
cs.CL, cs.AI
Our research focuses on solving the zero-shot text classification problem in NLP, with a particular emphasis on innovative self-training strategies. To achieve this objective, we propose a novel self-training strategy that uses labels rather than text for training, significantly reducing the model's training time. Spec...
http://arxiv.org/pdf/2307.15293v1
2023-07-28T04:17:41
2023-07-28T04:17:41
{ "title": [ 0.014930725, 0.022628784, -0.031555176, -0.06335449, 0.017700195, -0.014328003, -0.013130188, -0.026733398, 0.008140564, 0.03265381, -0.011116028, 0.006652832, 0.0068855286, -0.023132324, -0.03378296, 0.0062026978, -0.010177612, 0.06...