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/2307.13173v1
Opinion Mining Using Population-tuned Generative Language Models
Allmin Susaiyah, Abhinay Pandya, Aki Härmä
cs.CL, cs.AI
We present a novel method for mining opinions from text collections using generative language models trained on data collected from different populations. We describe the basic definitions, methodology and a generic algorithm for opinion insight mining. We demonstrate the performance of our method in an experiment wher...
http://arxiv.org/pdf/2307.13173v1
2023-07-24T23:42:32
2023-07-24T23:42:32
{ "title": [ -0.008163452, -0.021606445, -0.016021729, -0.03878784, -0.021987915, 0.0048179626, -0.032073975, -0.00095129013, 0.02458191, 0.033294678, -0.04360962, -0.005756378, -0.020828247, 0.008010864, -0.09338379, -0.008628845, -0.026672363, ...
http://arxiv.org/abs/2307.15715v1
Improving Primary Healthcare Workflow Using Extreme Summarization of Scientific Literature Based on Generative AI
Gregor Stiglic, Leon Kopitar, Lucija Gosak, Primoz Kocbek, Zhe He, Prithwish Chakraborty, Pablo Meyer, Jiang Bian
cs.CL, cs.AI, 68T50, I.2.7
Primary care professionals struggle to keep up to date with the latest scientific literature critical in guiding evidence-based practice related to their daily work. To help solve the above-mentioned problem, we employed generative artificial intelligence techniques based on large-scale language models to summarize abs...
http://arxiv.org/pdf/2307.15715v1
2023-07-24T21:42:27
2023-07-24T21:42:27
{ "title": [ -0.013023376, 0.000060737133, -0.037628174, -0.08630371, -0.00223732, 0.0049858093, -0.0647583, 0.0010271072, 0.010749817, 0.049804688, 0.010398865, 0.0357666, -0.00074625015, -0.026107788, -0.048095703, 0.02331543, 0.0030975342, 0.0...
http://arxiv.org/abs/2307.13128v1
Explaining Math Word Problem Solvers
Abby Newcomb, Jugal Kalita
cs.CL
Automated math word problem solvers based on neural networks have successfully managed to obtain 70-80\% accuracy in solving arithmetic word problems. However, it has been shown that these solvers may rely on superficial patterns to obtain their equations. In order to determine what information math word problem solver...
http://arxiv.org/pdf/2307.13128v1
2023-07-24T21:05:47
2023-07-24T21:05:47
{ "title": [ -0.006122589, -0.0031051636, 0.016143799, 0.015670776, -0.055541992, -0.004348755, -0.010116577, -0.019180298, -0.044830322, 0.021835327, -0.038604736, -0.005130768, -0.021911621, -0.024490356, 0.037017822, -0.041870117, 0.024978638, ...
http://arxiv.org/abs/2307.14361v2
A Hybrid Machine Learning Model for Classifying Gene Mutations in Cancer using LSTM, BiLSTM, CNN, GRU, and GloVe
Sanad Aburass, Osama Dorgham, Jamil Al Shaqsi
q-bio.QM, cs.AI, cs.CL, cs.LG
This study presents an ensemble model combining LSTM, BiLSTM, CNN, GRU, and GloVe to classify gene mutations using Kaggle's Personalized Medicine: Redefining Cancer Treatment dataset. The results were compared against well-known transformers like as BERT, Electra, Roberta, XLNet, Distilbert, and their LSTM ensembles. O...
http://arxiv.org/pdf/2307.14361v2
2023-08-07T17:09:07
2023-07-24T21:01:46
{ "title": [ -0.011436462, 0.019943237, -0.035705566, -0.03262329, 0.005004883, -0.020065308, -0.024368286, -0.0035858154, -0.015106201, 0.036071777, -0.02166748, -0.01360321, 0.028030396, -0.023239136, -0.026504517, -0.019851685, -0.004043579, 0...
http://arxiv.org/abs/2308.00121v3
Getting pwn'd by AI: Penetration Testing with Large Language Models
Andreas Happe, Jürgen Cito
cs.CL, cs.AI, cs.CR, cs.SE
The field of software security testing, more specifically penetration testing, is an activity that requires high levels of expertise and involves many manual testing and analysis steps. This paper explores the potential usage of large-language models, such as GPT3.5, to augment penetration testers with AI sparring part...
http://arxiv.org/pdf/2308.00121v3
2023-08-17T12:26:40
2023-07-24T19:59:22
{ "title": [ -0.04534912, -0.010551453, -0.03149414, -0.028640747, -0.006526947, -0.0042648315, -0.019378662, -0.014335632, -0.0335083, 0.024932861, -0.012054443, -0.002796173, -0.006996155, -0.02658081, -0.024276733, -0.022964478, -0.0043411255, ...
http://arxiv.org/abs/2307.13106v1
How to use LLMs for Text Analysis
Petter Törnberg
cs.CL, cs.AI
This guide introduces Large Language Models (LLM) as a highly versatile text analysis method within the social sciences. As LLMs are easy-to-use, cheap, fast, and applicable on a broad range of text analysis tasks, ranging from text annotation and classification to sentiment analysis and critical discourse analysis, ma...
http://arxiv.org/pdf/2307.13106v1
2023-07-24T19:54:15
2023-07-24T19:54:15
{ "title": [ -0.028915405, 0.011428833, -0.040649414, 0.002445221, -0.03829956, -0.013656616, -0.042999268, -0.01209259, -0.054718018, 0.053985596, -0.045837402, 0.01197052, 0.00088500977, 0.008766174, -0.00013399124, -0.03564453, 0.0017719269, -...
http://arxiv.org/abs/2307.16850v1
Comparative Analysis of Drug-GPT and ChatGPT LLMs for Healthcare Insights: Evaluating Accuracy and Relevance in Patient and HCP Contexts
Giorgos Lysandrou, Roma English Owen, Kirsty Mursec, Grant Le Brun, Elizabeth A. L. Fairley
cs.CL, cs.AI
This study presents a comparative analysis of three Generative Pre-trained Transformer (GPT) solutions in a question and answer (Q&A) setting: Drug-GPT 3, Drug-GPT 4, and ChatGPT, in the context of healthcare applications. The objective is to determine which model delivers the most accurate and relevant information in ...
http://arxiv.org/pdf/2307.16850v1
2023-07-24T19:27:11
2023-07-24T19:27:11
{ "title": [ -0.027999878, -0.028335571, -0.006248474, -0.022567749, -0.053955078, 0.006542206, -0.070007324, -0.011566162, -0.053588867, 0.0041923523, -0.04977417, 0.066101074, -0.04168701, -0.01864624, 0.02734375, -0.030731201, -0.01852417, -0....
http://arxiv.org/abs/2307.13085v1
Making Metadata More FAIR Using Large Language Models
Sowmya S. Sundaram, Mark A. Musen
cs.CL, cs.AI
With the global increase in experimental data artifacts, harnessing them in a unified fashion leads to a major stumbling block - bad metadata. To bridge this gap, this work presents a Natural Language Processing (NLP) informed application, called FAIRMetaText, that compares metadata. Specifically, FAIRMetaText analyzes...
http://arxiv.org/pdf/2307.13085v1
2023-07-24T19:14:38
2023-07-24T19:14:38
{ "title": [ -0.026733398, -0.0112838745, -0.05230713, -0.07922363, 0.010810852, -0.00029420853, -0.062927246, 0.009773254, 0.00554657, 0.042907715, -0.025238037, -0.015335083, 0.019042969, -0.009857178, -0.052825928, 0.010360718, -0.018218994, 0...
http://arxiv.org/abs/2307.15780v2
LLM-Rec: Personalized Recommendation via Prompting Large Language Models
Hanjia Lyu, Song Jiang, Hanqing Zeng, Qifan Wang, Si Zhang, Ren Chen, Chris Leung, Jiajie Tang, Yinglong Xia, Jiebo Luo
cs.CL, cs.AI, cs.IR
We investigate various prompting strategies for enhancing personalized recommendation performance with large language models (LLMs) through input augmentation. Our proposed approach, termed LLM-Rec, encompasses four distinct prompting strategies: (1) basic prompting, (2) recommendation-driven prompting, (3) engagement-...
http://arxiv.org/pdf/2307.15780v2
2023-08-16T17:59:07
2023-07-24T18:47:38
{ "title": [ -0.01928711, -0.017700195, -0.045288086, -0.061950684, -0.00466156, 0.0066070557, -0.040161133, 0.0064964294, -0.013755798, 0.022888184, -0.017456055, 0.019577026, -0.008102417, -0.016036987, -0.022079468, 0.00409317, -0.0055732727, ...
http://arxiv.org/abs/2307.12981v1
3D-LLM: Injecting the 3D World into Large Language Models
Yining Hong, Haoyu Zhen, Peihao Chen, Shuhong Zheng, Yilun Du, Zhenfang Chen, Chuang Gan
cs.CV, cs.AI, cs.CL, cs.LG, cs.RO
Large language models (LLMs) and Vision-Language Models (VLMs) have been proven to excel at multiple tasks, such as commonsense reasoning. Powerful as these models can be, they are not grounded in the 3D physical world, which involves richer concepts such as spatial relationships, affordances, physics, layout, and so o...
http://arxiv.org/pdf/2307.12981v1
2023-07-24T17:59:02
2023-07-24T17:59:02
{ "title": [ -0.039398193, -0.015159607, -0.04119873, -0.06512451, -0.015151978, 0.013931274, -0.064086914, 0.0039787292, -0.021835327, 0.031082153, -0.012321472, 0.021591187, 0.005306244, -0.025100708, -0.016799927, 0.007041931, -0.009887695, 0....
http://arxiv.org/abs/2307.12976v1
Evaluating the Ripple Effects of Knowledge Editing in Language Models
Roi Cohen, Eden Biran, Ori Yoran, Amir Globerson, Mor Geva
cs.CL
Modern language models capture a large body of factual knowledge. However, some facts can be incorrectly induced or become obsolete over time, resulting in factually incorrect generations. This has led to the development of various editing methods that allow updating facts encoded by the model. Evaluation of these meth...
http://arxiv.org/pdf/2307.12976v1
2023-07-24T17:52:46
2023-07-24T17:52:46
{ "title": [ -0.045562744, -0.015144348, -0.04953003, -0.052337646, -0.0029888153, -0.0026073456, -0.049041748, 0.013832092, 0.004798889, 0.038146973, -0.01826477, -0.014793396, 0.027160645, -0.0073165894, -0.06317139, 0.009284973, -0.022140503, ...
http://arxiv.org/abs/2307.12973v1
Leveraging Label Variation in Large Language Models for Zero-Shot Text Classification
Flor Miriam Plaza-del-Arco, Debora Nozza, Dirk Hovy
cs.CL
The zero-shot learning capabilities of large language models (LLMs) make them ideal for text classification without annotation or supervised training. Many studies have shown impressive results across multiple tasks. While tasks, data, and results differ widely, their similarities to human annotation can aid us in tack...
http://arxiv.org/pdf/2307.12973v1
2023-07-24T17:49:31
2023-07-24T17:49:31
{ "title": [ -0.024154663, -0.004722595, -0.045654297, -0.042938232, -0.0041007996, -0.010826111, -0.034729004, 0.007949829, 0.016525269, 0.0287323, -0.01335144, 0.015289307, -0.01789856, -0.005935669, -0.057525635, 0.0006403923, -0.012428284, 0....
http://arxiv.org/abs/2307.12966v1
Aligning Large Language Models with Human: A Survey
Yufei Wang, Wanjun Zhong, Liangyou Li, Fei Mi, Xingshan Zeng, Wenyong Huang, Lifeng Shang, Xin Jiang, Qun Liu
cs.CL
Large Language Models (LLMs) trained on extensive textual corpora have emerged as leading solutions for a broad array of Natural Language Processing (NLP) tasks. Despite their notable performance, these models are prone to certain limitations such as misunderstanding human instructions, generating potentially biased co...
http://arxiv.org/pdf/2307.12966v1
2023-07-24T17:44:58
2023-07-24T17:44:58
{ "title": [ -0.038024902, -0.0008535385, -0.045013428, -0.058502197, -0.0026187897, 0.0032024384, -0.077697754, 0.007534027, 0.005279541, 0.042938232, -0.009292603, -0.01259613, 0.023773193, -0.01007843, -0.055419922, 0.01260376, -0.009010315, 0...
http://arxiv.org/abs/2308.03767v1
Enhancing image captioning with depth information using a Transformer-based framework
Aya Mahmoud Ahmed, Mohamed Yousef, Khaled F. Hussain, Yousef Bassyouni Mahdy
cs.CV, cs.AI, cs.CL, eess.IV
Captioning images is a challenging scene-understanding task that connects computer vision and natural language processing. While image captioning models have been successful in producing excellent descriptions, the field has primarily focused on generating a single sentence for 2D images. This paper investigates whethe...
http://arxiv.org/pdf/2308.03767v1
2023-07-24T17:31:51
2023-07-24T17:31:51
{ "title": [ 0.007255554, 0.016067505, -0.031219482, -0.027557373, 0.0067825317, -0.022338867, -0.0132751465, -0.031585693, -0.0013360977, 0.05984497, -0.022338867, 0.016693115, 0.005622864, -0.002193451, -0.062683105, -0.010437012, -0.006252289, ...
http://arxiv.org/abs/2307.12950v2
RLCD: Reinforcement Learning from Contrast Distillation for Language Model Alignment
Kevin Yang, Dan Klein, Asli Celikyilmaz, Nanyun Peng, Yuandong Tian
cs.CL, cs.AI
We propose Reinforcement Learning from Contrast Distillation (RLCD), a method for aligning language models to follow natural language principles without using human feedback. RLCD trains a preference model using simulated preference pairs that contain both a high-quality and low-quality example, generated using contras...
http://arxiv.org/pdf/2307.12950v2
2023-08-18T22:16:51
2023-07-24T17:23:22
{ "title": [ -0.014533997, -0.0055274963, -0.037628174, -0.055389404, -0.002664566, -0.0042648315, -0.061553955, -0.004535675, -0.01574707, 0.03164673, -0.0132751465, 0.0016708374, -0.0047683716, -0.018859863, -0.030151367, -0.015838623, -0.014099121...
http://arxiv.org/abs/2307.12949v1
Boosting Punctuation Restoration with Data Generation and Reinforcement Learning
Viet Dac Lai, Abel Salinas, Hao Tan, Trung Bui, Quan Tran, Seunghyun Yoon, Hanieh Deilamsalehy, Franck Dernoncourt, Thien Huu Nguyen
cs.CL
Punctuation restoration is an important task in automatic speech recognition (ASR) which aim to restore the syntactic structure of generated ASR texts to improve readability. While punctuated texts are abundant from written documents, the discrepancy between written punctuated texts and ASR texts limits the usability o...
http://arxiv.org/pdf/2307.12949v1
2023-07-24T17:22:04
2023-07-24T17:22:04
{ "title": [ -0.008300781, 0.00970459, -0.022109985, -0.03829956, -0.0014019012, -0.003288269, -0.03060913, -0.0066223145, -0.034484863, 0.032440186, -0.01461792, -0.0014362335, 0.00045728683, -0.017120361, -0.021728516, -0.018310547, -0.027954102, ...
http://arxiv.org/abs/2307.13018v1
The potential of LLMs for coding with low-resource and domain-specific programming languages
Artur Tarassow
cs.CL, cs.SE, D.2.3; I.2.7
This paper presents a study on the feasibility of using large language models (LLM) for coding with low-resource and domain-specific programming languages that typically lack the amount of data required for effective LLM processing techniques. This study focuses on the econometric scripting language named hansl of the ...
http://arxiv.org/pdf/2307.13018v1
2023-07-24T17:17:13
2023-07-24T17:17:13
{ "title": [ -0.037200928, -0.021728516, -0.056884766, -0.04623413, -0.038269043, -0.018249512, -0.06222534, 0.025466919, -0.014198303, 0.022827148, -0.0048713684, 0.002937317, -0.009803772, -0.04824829, -0.0034713745, -0.012039185, -0.009407043, ...
http://arxiv.org/abs/2307.12935v1
Rule By Example: Harnessing Logical Rules for Explainable Hate Speech Detection
Christopher Clarke, Matthew Hall, Gaurav Mittal, Ye Yu, Sandra Sajeev, Jason Mars, Mei Chen
cs.CL, cs.AI
Classic approaches to content moderation typically apply a rule-based heuristic approach to flag content. While rules are easily customizable and intuitive for humans to interpret, they are inherently fragile and lack the flexibility or robustness needed to moderate the vast amount of undesirable content found online t...
http://arxiv.org/pdf/2307.12935v1
2023-07-24T16:55:37
2023-07-24T16:55:37
{ "title": [ 0.019424438, -0.0029010773, -0.037872314, -0.04675293, 0.024719238, -0.023971558, 0.007911682, -0.0071258545, -0.030044556, 0.018218994, -0.03314209, 0.035491943, -0.0009493828, -0.008140564, -0.028381348, -0.026046753, -0.023544312, ...
http://arxiv.org/abs/2307.12896v3
Corrections of Zipf's and Heaps' Laws Derived from Hapax Rate Models
Łukasz Dębowski
cs.CL, stat.AP, 62P99, I.2.7
The article introduces corrections to Zipf's and Heaps' laws based on systematic models of the hapax rate. The derivation rests on two assumptions: The first one is the standard urn model which predicts that marginal frequency distributions for shorter texts look as if word tokens were sampled blindly from a given long...
http://arxiv.org/pdf/2307.12896v3
2023-09-28T10:12:07
2023-07-24T15:44:23
{ "title": [ -0.03289795, -0.010848999, -0.017974854, 0.022094727, -0.04421997, -0.013519287, -0.002790451, -0.0065460205, -0.031280518, 0.02355957, -0.007205963, -0.012619019, -0.017059326, -0.042266846, -0.007133484, -0.03302002, -0.006603241, ...
http://arxiv.org/abs/2308.00071v1
Interpretable Stereotype Identification through Reasoning
Jacob-Junqi Tian, Omkar Dige, David Emerson, Faiza Khan Khattak
cs.CL, cs.AI, cs.CY, cs.LG
Given that language models are trained on vast datasets that may contain inherent biases, there is a potential danger of inadvertently perpetuating systemic discrimination. Consequently, it becomes essential to examine and address biases in language models, integrating fairness into their development to ensure these mo...
http://arxiv.org/pdf/2308.00071v1
2023-07-24T15:12:13
2023-07-24T15:12:13
{ "title": [ 0.0067863464, -0.0023441315, -0.016174316, -0.062286377, 0.0074920654, 0.0002477169, -0.025924683, -0.021850586, -0.011528015, 0.006816864, -0.020126343, -0.023376465, -0.0395813, -0.03274536, -0.029006958, -0.03579712, -0.01939392, ...
http://arxiv.org/abs/2307.12856v3
A Real-World WebAgent with Planning, Long Context Understanding, and Program Synthesis
Izzeddin Gur, Hiroki Furuta, Austin Huang, Mustafa Safdari, Yutaka Matsuo, Douglas Eck, Aleksandra Faust
cs.LG, cs.AI, cs.CL
Pre-trained large language models (LLMs) have recently achieved better generalization and sample efficiency in autonomous web automation. However, the performance on real-world websites has still suffered from (1) open domainness, (2) limited context length, and (3) lack of inductive bias on HTML. We introduce WebAgent...
http://arxiv.org/pdf/2307.12856v3
2023-10-03T03:51:14
2023-07-24T14:56:30
{ "title": [ 0.048858643, -0.0146102905, -0.031799316, -0.11077881, -0.006286621, -0.029190063, -0.04937744, 0.044158936, -0.030914307, 0.011528015, -0.050476074, 0.023239136, -0.009223938, -0.0135650635, 0.025482178, -0.048217773, 0.019241333, -...
http://arxiv.org/abs/2307.12835v1
Joint Dropout: Improving Generalizability in Low-Resource Neural Machine Translation through Phrase Pair Variables
Ali Araabi, Vlad Niculae, Christof Monz
cs.CL, 68T50, I.2.7
Despite the tremendous success of Neural Machine Translation (NMT), its performance on low-resource language pairs still remains subpar, partly due to the limited ability to handle previously unseen inputs, i.e., generalization. In this paper, we propose a method called Joint Dropout, that addresses the challenge of lo...
http://arxiv.org/pdf/2307.12835v1
2023-07-24T14:33:49
2023-07-24T14:33:49
{ "title": [ -0.02835083, -0.003604889, -0.000521183, -0.047454834, -0.0074043274, 0.020004272, -0.05206299, -0.016952515, 0.01058197, 0.049224854, -0.0058250427, 0.023239136, 0.0073623657, 0.002248764, -0.054992676, 0.024749756, -0.013824463, 0....
http://arxiv.org/abs/2308.02432v1
Performance of Large Language Models in a Computer Science Degree Program
Tim Krüger, Michael Gref
cs.CY, cs.AI, cs.CL
Large language models such as ChatGPT-3.5 and GPT-4.0 are ubiquitous and dominate the current discourse. Their transformative capabilities have led to a paradigm shift in how we interact with and utilize (text-based) information. Each day, new possibilities to leverage the capabilities of these models emerge. This pape...
http://arxiv.org/pdf/2308.02432v1
2023-07-24T14:17:00
2023-07-24T14:17:00
{ "title": [ -0.045776367, -0.008308411, -0.047729492, -0.04373169, -0.012832642, -0.01171875, -0.03363037, 0.01247406, -0.0035591125, 0.03656006, -0.025054932, -0.020721436, 0.01902771, -0.024597168, -0.037841797, -0.0043296814, -0.007873535, 0....
http://arxiv.org/abs/2307.12803v1
Guidance in Radiology Report Summarization: An Empirical Evaluation and Error Analysis
Jan Trienes, Paul Youssef, Jörg Schlötterer, Christin Seifert
cs.CL
Automatically summarizing radiology reports into a concise impression can reduce the manual burden of clinicians and improve the consistency of reporting. Previous work aimed to enhance content selection and factuality through guided abstractive summarization. However, two key issues persist. First, current methods hea...
http://arxiv.org/pdf/2307.12803v1
2023-07-24T13:54:37
2023-07-24T13:54:37
{ "title": [ -0.0077590942, -0.0017004013, -0.027053833, -0.061950684, -0.01007843, 0.008285522, -0.048065186, -0.012580872, -0.013206482, 0.03704834, -0.01474762, 0.020050049, -0.024124146, -0.0050621033, -0.052947998, 0.0026226044, -0.0038909912, ...
http://arxiv.org/abs/2307.12798v3
RRAML: Reinforced Retrieval Augmented Machine Learning
Andrea Bacciu, Florin Cuconasu, Federico Siciliano, Fabrizio Silvestri, Nicola Tonellotto, Giovanni Trappolini
cs.CL, cs.IR
The emergence of large language models (LLMs) has revolutionized machine learning and related fields, showcasing remarkable abilities in comprehending, generating, and manipulating human language. However, their conventional usage through API-based text prompt submissions imposes certain limitations in terms of context...
http://arxiv.org/pdf/2307.12798v3
2023-07-27T07:20:28
2023-07-24T13:51:19
{ "title": [ -0.00793457, -0.0011253357, -0.026763916, -0.061065674, 0.018234253, 0.00818634, -0.014953613, -0.019836426, 0.009483337, 0.015823364, -0.009140015, 0.009887695, -0.00592041, -0.02684021, -0.014930725, 0.016418457, -0.01689148, 0.028...
http://arxiv.org/abs/2307.12759v1
Code-Switched Urdu ASR for Noisy Telephonic Environment using Data Centric Approach with Hybrid HMM and CNN-TDNN
Muhammad Danyal Khan, Raheem Ali, Arshad Aziz
cs.CL, cs.SD, eess.AS
Call Centers have huge amount of audio data which can be used for achieving valuable business insights and transcription of phone calls is manually tedious task. An effective Automated Speech Recognition system can accurately transcribe these calls for easy search through call history for specific context and content a...
http://arxiv.org/pdf/2307.12759v1
2023-07-24T13:04:21
2023-07-24T13:04:21
{ "title": [ -0.013496399, -0.008453369, -0.034240723, -0.02633667, -0.0073547363, -0.016204834, -0.013641357, 0.021102905, 0.007205963, 0.039367676, -0.042144775, -0.009681702, 0.041412354, -0.028839111, -0.01486969, -0.008857727, 0.007965088, 0...
http://arxiv.org/abs/2307.12659v1
A Model for Every User and Budget: Label-Free and Personalized Mixed-Precision Quantization
Edward Fish, Umberto Michieli, Mete Ozay
cs.SD, cs.CL, eess.AS
Recent advancement in Automatic Speech Recognition (ASR) has produced large AI models, which become impractical for deployment in mobile devices. Model quantization is effective to produce compressed general-purpose models, however such models may only be deployed to a restricted sub-domain of interest. We show that AS...
http://arxiv.org/pdf/2307.12659v1
2023-07-24T10:03:28
2023-07-24T10:03:28
{ "title": [ 0.0077934265, 0.0090789795, -0.023239136, -0.058166504, -0.01687622, -0.028274536, -0.054595947, -0.015823364, -0.043273926, 0.057678223, 0.0107040405, 0.019180298, 0.006000519, -0.03744507, -0.010917664, 0.027557373, 0.030288696, 0....
http://arxiv.org/abs/2307.12639v1
Fake News Detection Through Graph-based Neural Networks: A Survey
Shuzhi Gong, Richard O. Sinnott, Jianzhong Qi, Cecile Paris
cs.SI, cs.CL, cs.GR, cs.LG
The popularity of online social networks has enabled rapid dissemination of information. People now can share and consume information much more rapidly than ever before. However, low-quality and/or accidentally/deliberately fake information can also spread rapidly. This can lead to considerable and negative impacts on ...
http://arxiv.org/pdf/2307.12639v1
2023-07-24T09:30:30
2023-07-24T09:30:30
{ "title": [ -0.021575928, 0.010391235, -0.026733398, -0.013267517, 0.0066871643, -0.0076141357, -0.006225586, -0.024520874, -0.052581787, 0.027572632, -0.0012693405, 0.016311646, -0.021743774, -0.021438599, -0.014915466, -0.018829346, -0.04055786, ...
http://arxiv.org/abs/2307.12573v1
Tachikuma: Understading Complex Interactions with Multi-Character and Novel Objects by Large Language Models
Yuanzhi Liang, Linchao Zhu, Yi Yang
cs.CL
Recent advancements in natural language and Large Language Models (LLMs) have enabled AI agents to simulate human-like interactions within virtual worlds. However, these interactions still face limitations in complexity and flexibility, particularly in scenarios involving multiple characters and novel objects. Pre-defi...
http://arxiv.org/pdf/2307.12573v1
2023-07-24T07:40:59
2023-07-24T07:40:59
{ "title": [ -0.028869629, -0.026763916, -0.02029419, -0.06750488, -0.018875122, 0.005470276, -0.079833984, -0.009902954, 0.013710022, 0.037017822, -0.00052022934, 0.009506226, 0.027770996, -0.023040771, -0.03543091, 0.02079773, -0.01739502, 0.04...
http://arxiv.org/abs/2307.12564v1
Towards Generalising Neural Topical Representations
Xiaohao Yang, He Zhao, Dinh Phung, Lan Du
cs.CL, cs.LG
Topic models have evolved from conventional Bayesian probabilistic models to Neural Topic Models (NTMs) over the last two decays. Although NTMs have achieved promising performance when trained and tested on a specific corpus, their generalisation ability across corpora is rarely studied. In practice, we often expect th...
http://arxiv.org/pdf/2307.12564v1
2023-07-24T07:17:33
2023-07-24T07:17:33
{ "title": [ 0.015945435, 0.025421143, -0.015716553, -0.059906006, 0.017166138, 0.0064849854, -0.038208008, -0.019729614, 0.024887085, 0.043273926, 0.0030117035, -0.013755798, -0.008407593, -0.018249512, -0.049835205, 0.023864746, -0.009033203, 0...
http://arxiv.org/abs/2307.12520v1
Lost In Translation: Generating Adversarial Examples Robust to Round-Trip Translation
Neel Bhandari, Pin-Yu Chen
cs.CL, cs.LG
Language Models today provide a high accuracy across a large number of downstream tasks. However, they remain susceptible to adversarial attacks, particularly against those where the adversarial examples maintain considerable similarity to the original text. Given the multilingual nature of text, the effectiveness of a...
http://arxiv.org/pdf/2307.12520v1
2023-07-24T04:29:43
2023-07-24T04:29:43
{ "title": [ -0.018096924, -0.0064964294, 0.024002075, -0.02671814, -0.00019335747, 0.02444458, -0.021820068, -0.045684814, -0.022369385, 0.020401001, 0.0109939575, 0.028671265, -0.032684326, -0.029586792, -0.027069092, -0.002576828, 0.005443573, ...
http://arxiv.org/abs/2307.12507v2
Gradient-Based Word Substitution for Obstinate Adversarial Examples Generation in Language Models
Yimu Wang, Peng Shi, Hongyang Zhang
cs.CL, cs.AI, cs.CR, cs.CY
In this paper, we study the problem of generating obstinate (over-stability) adversarial examples by word substitution in NLP, where input text is meaningfully changed but the model's prediction does not, even though it should. Previous word substitution approaches have predominantly focused on manually designed antony...
http://arxiv.org/pdf/2307.12507v2
2023-08-18T03:07:02
2023-07-24T03:44:17
{ "title": [ -0.008369446, -0.004245758, 0.011779785, -0.027526855, -0.030776978, 0.016586304, -0.047698975, -0.049743652, -0.018417358, 0.019042969, -0.008972168, 0.02079773, 0.0056381226, -0.017501831, -0.02482605, -0.011405945, -0.020706177, 0...
http://arxiv.org/abs/2307.12498v1
Robust Automatic Speech Recognition via WavAugment Guided Phoneme Adversarial Training
Gege Qi, Yuefeng Chen, Xiaofeng Mao, Xiaojun Jia, Ranjie Duan, Rong Zhang, Hui Xue
cs.SD, cs.CL, eess.AS
Developing a practically-robust automatic speech recognition (ASR) is challenging since the model should not only maintain the original performance on clean samples, but also achieve consistent efficacy under small volume perturbations and large domain shifts. To address this problem, we propose a novel WavAugment Guid...
http://arxiv.org/pdf/2307.12498v1
2023-07-24T03:07:40
2023-07-24T03:07:40
{ "title": [ 0.006778717, -0.0023002625, -0.040252686, -0.056549072, 0.0077590942, -0.0050735474, -0.023284912, 0.008430481, -0.042022705, 0.05166626, -0.0029563904, 0.038848877, -0.014640808, -0.019302368, -0.03717041, -0.012825012, -0.004497528, ...
http://arxiv.org/abs/2307.12425v1
On the Effectiveness of Offline RL for Dialogue Response Generation
Paloma Sodhi, Felix Wu, Ethan R. Elenberg, Kilian Q. Weinberger, Ryan McDonald
cs.CL
A common training technique for language models is teacher forcing (TF). TF attempts to match human language exactly, even though identical meanings can be expressed in different ways. This motivates use of sequence-level objectives for dialogue response generation. In this paper, we study the efficacy of various offli...
http://arxiv.org/pdf/2307.12425v1
2023-07-23T20:43:21
2023-07-23T20:43:21
{ "title": [ -0.019500732, -0.027114868, -0.013923645, -0.059631348, -0.021636963, 0.014884949, -0.052093506, 0.012420654, -0.00970459, 0.016113281, 0.03704834, -0.024414062, -0.020385742, -0.049346924, -0.040771484, -0.0062065125, -0.016647339, ...
http://arxiv.org/abs/2307.12418v1
Testing Hateful Speeches against Policies
Jiangrui Zheng, Xueqing Liu, Girish Budhrani, Wei Yang, Ravishka Rathnasuriya
cs.SE, cs.CL
In the recent years, many software systems have adopted AI techniques, especially deep learning techniques. Due to their black-box nature, AI-based systems brought challenges to traceability, because AI system behaviors are based on models and data, whereas the requirements or policies are rules in the form of natural ...
http://arxiv.org/pdf/2307.12418v1
2023-07-23T20:08:38
2023-07-23T20:08:38
{ "title": [ 0.05596924, -0.016052246, 0.010093689, -0.0113220215, -0.033996582, -0.018859863, 0.012130737, 0.0055274963, -0.031402588, 0.04623413, -0.012886047, 0.012664795, 0.008361816, -0.023040771, 0.0104522705, -0.016815186, 0.007858276, 0.0...
http://arxiv.org/abs/2308.00107v1
Validation of a Zero-Shot Learning Natural Language Processing Tool for Data Abstraction from Unstructured Healthcare Data
Basil Kaufmann, Dallin Busby, Chandan Krushna Das, Neeraja Tillu, Mani Menon, Ashutosh K. Tewari, Michael A. Gorin
cs.CL, cs.AI, 68T50, I.2.7
Objectives: To describe the development and validation of a zero-shot learning natural language processing (NLP) tool for abstracting data from unstructured text contained within PDF documents, such as those found within electronic health records. Materials and Methods: A data abstraction tool based on the GPT-3.5 mode...
http://arxiv.org/pdf/2308.00107v1
2023-07-23T17:52:28
2023-07-23T17:52:28
{ "title": [ -0.00033521652, -0.0053596497, -0.058807373, -0.09729004, 0.013702393, -0.0066833496, -0.0524292, 0.013824463, -0.010116577, 0.02558899, -0.019836426, -0.0027885437, -0.016311646, -0.029510498, -0.0395813, -0.022338867, -0.008132935, ...
http://arxiv.org/abs/2307.12382v1
CommonsenseVIS: Visualizing and Understanding Commonsense Reasoning Capabilities of Natural Language Models
Xingbo Wang, Renfei Huang, Zhihua Jin, Tianqing Fang, Huamin Qu
cs.CL, cs.HC
Recently, large pretrained language models have achieved compelling performance on commonsense benchmarks. Nevertheless, it is unclear what commonsense knowledge the models learn and whether they solely exploit spurious patterns. Feature attributions are popular explainability techniques that identify important input c...
http://arxiv.org/pdf/2307.12382v1
2023-07-23T17:16:13
2023-07-23T17:16:13
{ "title": [ 0.009513855, -0.02999878, -0.009819031, -0.057250977, -0.011375427, 0.010505676, -0.019821167, -0.018936157, 0.02734375, 0.011726379, -0.04373169, -0.0031433105, -0.006931305, -0.019088745, -0.04147339, -0.0020618439, -0.013137817, 0...
http://arxiv.org/abs/2307.12375v3
In-Context Learning Learns Label Relationships but Is Not Conventional Learning
Jannik Kossen, Yarin Gal, Tom Rainforth
cs.CL, cs.AI, cs.LG
The predictions of Large Language Models (LLMs) on downstream tasks often improve significantly when including examples of the input--label relationship in the context. However, there is currently no consensus about how this in-context learning (ICL) ability of LLMs works. For example, while Xie et al. (2021) liken ICL...
http://arxiv.org/pdf/2307.12375v3
2023-10-03T08:19:44
2023-07-23T16:54:41
{ "title": [ 0.034729004, 0.0047912598, -0.040405273, -0.026168823, -0.010681152, -0.045898438, 0.008255005, -0.0069770813, -0.021881104, 0.015144348, -0.034606934, 0.008628845, -0.01309967, -0.035461426, 0.037139893, -0.028030396, 0.0036239624, ...
http://arxiv.org/abs/2307.12371v1
Evaluating Emotional Nuances in Dialogue Summarization
Yongxin Zhou, Fabien Ringeval, François Portet
cs.CL
Automatic dialogue summarization is a well-established task that aims to identify the most important content from human conversations to create a short textual summary. Despite recent progress in the field, we show that most of the research has focused on summarizing the factual information, leaving aside the affective...
http://arxiv.org/pdf/2307.12371v1
2023-07-23T16:46:01
2023-07-23T16:46:01
{ "title": [ -0.0058898926, -0.003452301, -0.017181396, -0.055419922, -0.014350891, 0.00033211708, -0.04736328, -0.006629944, 0.019104004, 0.03955078, 0.00037145615, 0.0011281967, -0.0045318604, -0.015838623, -0.072387695, 0.0065689087, -0.010803223,...
http://arxiv.org/abs/2307.12369v1
Early Prediction of Alzheimers Disease Leveraging Symptom Occurrences from Longitudinal Electronic Health Records of US Military Veterans
Rumeng Li, Xun Wang, Dan Berlowitz, Brian Silver, Wen Hu, Heather Keating, Raelene Goodwin, Weisong Liu, Honghuang Lin, Hong Yu
cs.LG, cs.AI, cs.CL
Early prediction of Alzheimer's disease (AD) is crucial for timely intervention and treatment. This study aims to use machine learning approaches to analyze longitudinal electronic health records (EHRs) of patients with AD and identify signs and symptoms that can predict AD onset earlier. We used a case-control design ...
http://arxiv.org/pdf/2307.12369v1
2023-07-23T16:38:16
2023-07-23T16:38:16
{ "title": [ 0.03768921, 0.01914978, -0.025527954, -0.076293945, 0.042816162, -0.022369385, -0.009025574, -0.026916504, -0.020690918, 0.008277893, -0.023498535, -0.008506775, -0.030960083, -0.042633057, -0.020996094, -0.031677246, 0.008621216, 0....
http://arxiv.org/abs/2307.12332v1
X-CapsNet For Fake News Detection
Mohammad Hadi Goldani, Reza Safabakhsh, Saeedeh Momtazi
cs.CL, cs.CY
News consumption has significantly increased with the growing popularity and use of web-based forums and social media. This sets the stage for misinforming and confusing people. To help reduce the impact of misinformation on users' potential health-related decisions and other intents, it is desired to have machine lear...
http://arxiv.org/pdf/2307.12332v1
2023-07-23T13:58:00
2023-07-23T13:58:00
{ "title": [ -0.017044067, 0.006954193, -0.036071777, -0.025009155, 0.005302429, -0.00207901, -0.017303467, -0.026184082, -0.04043579, 0.03475952, -0.02041626, 0.017700195, 0.0058288574, -0.01725769, -0.028549194, -0.008018494, -0.03286743, 0.083...
http://arxiv.org/abs/2307.12282v1
Milimili. Collecting Parallel Data via Crowdsourcing
Alexander Antonov
cs.CL
We present a methodology for gathering a parallel corpus through crowdsourcing, which is more cost-effective than hiring professional translators, albeit at the expense of quality. Additionally, we have made available experimental parallel data collected for Chechen-Russian and Fula-English language pairs.
http://arxiv.org/pdf/2307.12282v1
2023-07-23T10:23:00
2023-07-23T10:23:00
{ "title": [ -0.01448822, 0.010223389, -0.072143555, -0.025543213, -0.01977539, -0.018035889, -0.0115737915, 0.015617371, -0.05984497, 0.052856445, -0.024993896, 0.028259277, 0.012191772, -0.020721436, -0.0023765564, -0.024765015, 0.0029678345, 0...
http://arxiv.org/abs/2307.12267v4
Towards Automatic Boundary Detection for Human-AI Collaborative Hybrid Essay in Education
Zijie Zeng, Lele Sha, Yuheng Li, Kaixun Yang, Dragan Gašević, Guanliang Chen
cs.CL, cs.AI
The recent large language models (LLMs), e.g., ChatGPT, have been able to generate human-like and fluent responses when provided with specific instructions. While admitting the convenience brought by technological advancement, educators also have concerns that students might leverage LLMs to complete their writing assi...
http://arxiv.org/pdf/2307.12267v4
2023-08-16T08:44:08
2023-07-23T08:47:51
{ "title": [ -0.000600338, 0.021728516, 0.00390625, 0.011917114, 0.0043296814, -0.006752014, 0.022720337, -0.052764893, -0.030929565, 0.020553589, -0.04272461, 0.013214111, -0.031173706, -0.001868248, -0.047546387, -0.051757812, -0.033050537, 0.0...
http://arxiv.org/abs/2307.12266v1
Transformer-based Joint Source Channel Coding for Textual Semantic Communication
Shicong Liu, Zhen Gao, Gaojie Chen, Yu Su, Lu Peng
cs.CL, eess.SP
The Space-Air-Ground-Sea integrated network calls for more robust and secure transmission techniques against jamming. In this paper, we propose a textual semantic transmission framework for robust transmission, which utilizes the advanced natural language processing techniques to model and encode sentences. Specificall...
http://arxiv.org/pdf/2307.12266v1
2023-07-23T08:42:05
2023-07-23T08:42:05
{ "title": [ -0.06112671, -0.007801056, -0.044006348, -0.03326416, -0.00089883804, 0.0042877197, -0.043823242, 0.00005811453, -0.0014123917, 0.059692383, 0.0068893433, 0.013587952, 0.06793213, -0.030441284, -0.0079193115, 0.03717041, -0.007926941, ...
http://arxiv.org/abs/2307.12262v1
A meta learning scheme for fast accent domain expansion in Mandarin speech recognition
Ziwei Zhu, Changhao Shan, Bihong Zhang, Jian Yu
cs.SD, cs.CL, cs.HC, eess.AS
Spoken languages show significant variation across mandarin and accent. Despite the high performance of mandarin automatic speech recognition (ASR), accent ASR is still a challenge task. In this paper, we introduce meta-learning techniques for fast accent domain expansion in mandarin speech recognition, which expands t...
http://arxiv.org/pdf/2307.12262v1
2023-07-23T08:23:26
2023-07-23T08:23:26
{ "title": [ -0.0029449463, 0.012161255, -0.02192688, -0.019454956, 0.00055503845, -0.010009766, -0.019241333, -0.018463135, -0.0090789795, 0.045074463, -0.0042495728, 0.00422287, 0.014144897, -0.042022705, -0.0390625, 0.0090408325, -0.00983429, ...
http://arxiv.org/abs/2308.02503v1
MyVoice: Arabic Speech Resource Collaboration Platform
Yousseif Elshahawy, Yassine El Kheir, Shammur Absar Chowdhury, Ahmed Ali
eess.AS, cs.CL, cs.SD
We introduce MyVoice, a crowdsourcing platform designed to collect Arabic speech to enhance dialectal speech technologies. This platform offers an opportunity to design large dialectal speech datasets; and makes them publicly available. MyVoice allows contributors to select city/country-level fine-grained dialect and r...
http://arxiv.org/pdf/2308.02503v1
2023-07-23T07:13:30
2023-07-23T07:13:30
{ "title": [ -0.013641357, -0.037750244, -0.04840088, -0.09210205, -0.023208618, 0.0039863586, -0.05014038, 0.039886475, -0.019897461, 0.037750244, 0.021209717, 0.00016570091, -0.0019006729, -0.06549072, -0.027755737, -0.011749268, 0.008956909, 0...
http://arxiv.org/abs/2307.12231v1
Exploring the Integration of Speech Separation and Recognition with Self-Supervised Learning Representation
Yoshiki Masuyama, Xuankai Chang, Wangyou Zhang, Samuele Cornell, Zhong-Qiu Wang, Nobutaka Ono, Yanmin Qian, Shinji Watanabe
cs.SD, cs.CL, eess.AS
Neural speech separation has made remarkable progress and its integration with automatic speech recognition (ASR) is an important direction towards realizing multi-speaker ASR. This work provides an insightful investigation of speech separation in reverberant and noisy-reverberant scenarios as an ASR front-end. In deta...
http://arxiv.org/pdf/2307.12231v1
2023-07-23T05:39:39
2023-07-23T05:39:39
{ "title": [ 0.0075263977, 0.026535034, -0.07537842, -0.07348633, 0.022064209, -0.023147583, -0.02684021, 0.014305115, -0.044830322, 0.040863037, -0.014137268, 0.023666382, -0.013786316, -0.022033691, -0.0178833, -0.0057144165, 0.002412796, 0.053...
http://arxiv.org/abs/2307.12221v1
FATRER: Full-Attention Topic Regularizer for Accurate and Robust Conversational Emotion Recognition
Yuzhao Mao, Di Lu, Xiaojie Wang, Yang Zhang
cs.CL, cs.AI
This paper concentrates on the understanding of interlocutors' emotions evoked in conversational utterances. Previous studies in this literature mainly focus on more accurate emotional predictions, while ignoring model robustness when the local context is corrupted by adversarial attacks. To maintain robustness while e...
http://arxiv.org/pdf/2307.12221v1
2023-07-23T04:01:24
2023-07-23T04:01:24
{ "title": [ -0.004047394, 0.009490967, -0.037261963, -0.0574646, 0.009361267, 0.00076150894, -0.051757812, -0.010765076, 0.048339844, 0.020431519, -0.010192871, 0.0053596497, 0.021118164, -0.025863647, -0.044036865, 0.014663696, -0.005680084, 0....
http://arxiv.org/abs/2307.12166v2
The Imitation Game: Detecting Human and AI-Generated Texts in the Era of ChatGPT and BARD
Kadhim Hayawi, Sakib Shahriar, Sujith Samuel Mathew
cs.CL, cs.AI
The potential of artificial intelligence (AI)-based large language models (LLMs) holds considerable promise in revolutionizing education, research, and practice. However, distinguishing between human-written and AI-generated text has become a significant task. This paper presents a comparative study, introducing a nove...
http://arxiv.org/pdf/2307.12166v2
2023-11-12T01:26:48
2023-07-22T21:00:14
{ "title": [ -0.013587952, 0.001455307, -0.01953125, -0.0087890625, 0.00016140938, -0.010398865, 0.017913818, -0.02406311, -0.027648926, 0.021255493, -0.018539429, 0.03390503, -0.01423645, -0.009346008, -0.023651123, -0.007797241, -0.0335083, 0.0...
http://arxiv.org/abs/2307.12155v1
Identifying Misinformation on YouTube through Transcript Contextual Analysis with Transformer Models
Christos Christodoulou, Nikos Salamanos, Pantelitsa Leonidou, Michail Papadakis, Michael Sirivianos
cs.CL
Misinformation on YouTube is a significant concern, necessitating robust detection strategies. In this paper, we introduce a novel methodology for video classification, focusing on the veracity of the content. We convert the conventional video classification task into a text classification task by leveraging the textua...
http://arxiv.org/pdf/2307.12155v1
2023-07-22T19:59:16
2023-07-22T19:59:16
{ "title": [ 0.027252197, 0.04135132, -0.05203247, -0.064208984, 0.044433594, -0.020629883, -0.016235352, -0.015625, -0.04168701, 0.054901123, -0.023788452, 0.00894165, 0.028076172, -0.012863159, -0.03274536, -0.020141602, -0.045013428, 0.0816650...
http://arxiv.org/abs/2307.12134v1
Modality Confidence Aware Training for Robust End-to-End Spoken Language Understanding
Suyoun Kim, Akshat Shrivastava, Duc Le, Ju Lin, Ozlem Kalinli, Michael L. Seltzer
cs.CL, cs.SD, eess.AS
End-to-end (E2E) spoken language understanding (SLU) systems that generate a semantic parse from speech have become more promising recently. This approach uses a single model that utilizes audio and text representations from pre-trained speech recognition models (ASR), and outperforms traditional pipeline SLU systems i...
http://arxiv.org/pdf/2307.12134v1
2023-07-22T17:47:31
2023-07-22T17:47:31
{ "title": [ -0.009178162, 0.008735657, -0.04763794, -0.07904053, 0.005168915, -0.0045394897, -0.066345215, 0.005554199, -0.0023155212, 0.034698486, -0.014640808, -0.00491333, 0.022445679, -0.031311035, -0.04208374, 0.0104599, -0.01977539, 0.0109...
http://arxiv.org/abs/2307.12131v1
Explainable Topic-Enhanced Argument Mining from Heterogeneous Sources
Jiasheng Si, Yingjie Zhu, Xingyu Shi, Deyu Zhou, Yulan He
cs.CL
Given a controversial target such as ``nuclear energy'', argument mining aims to identify the argumentative text from heterogeneous sources. Current approaches focus on exploring better ways of integrating the target-associated semantic information with the argumentative text. Despite their empirical successes, two iss...
http://arxiv.org/pdf/2307.12131v1
2023-07-22T17:26:55
2023-07-22T17:26:55
{ "title": [ 0.02947998, 0.0043525696, -0.015686035, -0.09326172, 0.028213501, 0.015174866, -0.02394104, -0.004863739, -0.017608643, 0.0022830963, -0.021591187, 0.036193848, -0.0496521, -0.004070282, -0.015106201, -0.025360107, -0.024307251, 0.03...
http://arxiv.org/abs/2307.12114v1
A Zero-shot and Few-shot Study of Instruction-Finetuned Large Language Models Applied to Clinical and Biomedical Tasks
Yanis Labrak, Mickael Rouvier, Richard Dufour
cs.CL, cs.AI, cs.LG
We evaluate four state-of-the-art instruction-tuned large language models (LLMs) -- ChatGPT, Flan-T5 UL2, Tk-Instruct, and Alpaca -- on a set of 13 real-world clinical and biomedical natural language processing (NLP) tasks in English, such as named-entity recognition (NER), question-answering (QA), relation extraction ...
http://arxiv.org/pdf/2307.12114v1
2023-07-22T15:58:17
2023-07-22T15:58:17
{ "title": [ -0.015655518, -0.038513184, -0.031585693, -0.049621582, -0.026809692, -0.005908966, -0.06512451, 0.0022525787, -0.00056791306, 0.040100098, 0.002603531, 0.0009813309, 0.0063667297, -0.030288696, -0.03451538, 0.008514404, -0.0016078949, ...
http://arxiv.org/abs/2307.12045v1
Revisiting Distillation for Continual Learning on Visual Question Localized-Answering in Robotic Surgery
Long Bai, Mobarakol Islam, Hongliang Ren
cs.CV, cs.CL, cs.RO
The visual-question localized-answering (VQLA) system can serve as a knowledgeable assistant in surgical education. Except for providing text-based answers, the VQLA system can highlight the interested region for better surgical scene understanding. However, deep neural networks (DNNs) suffer from catastrophic forgetti...
http://arxiv.org/pdf/2307.12045v1
2023-07-22T10:35:25
2023-07-22T10:35:25
{ "title": [ 0.011909485, 0.0029315948, -0.031188965, -0.051727295, 0.01158905, -0.017196655, -0.029220581, -0.003047943, 0.041107178, 0.008377075, -0.06933594, -0.016616821, -0.022644043, -0.017196655, -0.05126953, -0.042785645, 0.037628174, -0....
http://arxiv.org/abs/2307.12996v1
Extracting Molecular Properties from Natural Language with Multimodal Contrastive Learning
Romain Lacombe, Andrew Gaut, Jeff He, David Lüdeke, Kateryna Pistunova
cs.LG, cs.AI, cs.CL, cs.IR, q-bio.QM
Deep learning in computational biochemistry has traditionally focused on molecular graphs neural representations; however, recent advances in language models highlight how much scientific knowledge is encoded in text. To bridge these two modalities, we investigate how molecular property information can be transferred f...
http://arxiv.org/pdf/2307.12996v1
2023-07-22T10:32:58
2023-07-22T10:32:58
{ "title": [ -0.044921875, -0.01940918, -0.05722046, -0.049438477, -0.01751709, -0.0016107559, -0.047058105, 0.020355225, 0.018814087, 0.014472961, -0.011917114, 0.019638062, -0.0178833, -0.019042969, -0.036468506, 0.009506226, -0.02470398, 0.051...
http://arxiv.org/abs/2308.00065v1
FinPT: Financial Risk Prediction with Profile Tuning on Pretrained Foundation Models
Yuwei Yin, Yazheng Yang, Jian Yang, Qi Liu
q-fin.RM, cs.CE, cs.CL, cs.LG, q-fin.ST
Financial risk prediction plays a crucial role in the financial sector. Machine learning methods have been widely applied for automatically detecting potential risks and thus saving the cost of labor. However, the development in this field is lagging behind in recent years by the following two facts: 1) the algorithms ...
http://arxiv.org/pdf/2308.00065v1
2023-07-22T09:27:05
2023-07-22T09:27:05
{ "title": [ -0.02003479, -0.0010032654, -0.0395813, -0.04660034, 0.011047363, -0.015060425, -0.02003479, -0.025009155, -0.030319214, 0.017822266, -0.016815186, 0.023376465, -0.025512695, -0.019104004, -0.022216797, -0.0109939575, -0.011077881, 0...
http://arxiv.org/abs/2307.11991v2
Psy-LLM: Scaling up Global Mental Health Psychological Services with AI-based Large Language Models
Tin Lai, Yukun Shi, Zicong Du, Jiajie Wu, Ken Fu, Yichao Dou, Ziqi Wang
cs.CL, cs.AI
The demand for psychological counselling has grown significantly in recent years, particularly with the global outbreak of COVID-19, which has heightened the need for timely and professional mental health support. Online psychological counselling has emerged as the predominant mode of providing services in response to ...
http://arxiv.org/pdf/2307.11991v2
2023-09-01T04:52:38
2023-07-22T06:21:41
{ "title": [ -0.035949707, -0.009010315, -0.035064697, -0.050628662, -0.007965088, 0.001288414, -0.059509277, -0.008529663, -0.011077881, 0.027542114, -0.011482239, 0.012214661, 0.01826477, -0.04260254, 0.0025157928, 0.006504059, 0.0011615753, 0....
http://arxiv.org/abs/2307.11984v1
Learning Vision-and-Language Navigation from YouTube Videos
Kunyang Lin, Peihao Chen, Diwei Huang, Thomas H. Li, Mingkui Tan, Chuang Gan
cs.CV, cs.CL
Vision-and-language navigation (VLN) requires an embodied agent to navigate in realistic 3D environments using natural language instructions. Existing VLN methods suffer from training on small-scale environments or unreasonable path-instruction datasets, limiting the generalization to unseen environments. There are mas...
http://arxiv.org/pdf/2307.11984v1
2023-07-22T05:26:50
2023-07-22T05:26:50
{ "title": [ -0.0047416687, -0.005443573, -0.027175903, -0.039794922, 0.014091492, -0.027893066, 0.0025177002, -0.0044136047, 0.0034294128, 0.022583008, -0.028701782, 0.0025691986, -0.017913818, -0.0072021484, -0.05368042, -0.014282227, -0.0027275085...
http://arxiv.org/abs/2307.11922v1
Selective Perception: Optimizing State Descriptions with Reinforcement Learning for Language Model Actors
Kolby Nottingham, Yasaman Razeghi, Kyungmin Kim, JB Lanier, Pierre Baldi, Roy Fox, Sameer Singh
cs.LG, cs.AI, cs.CL
Large language models (LLMs) are being applied as actors for sequential decision making tasks in domains such as robotics and games, utilizing their general world knowledge and planning abilities. However, previous work does little to explore what environment state information is provided to LLM actors via language. Ex...
http://arxiv.org/pdf/2307.11922v1
2023-07-21T22:02:50
2023-07-21T22:02:50
{ "title": [ -0.02319336, -0.005908966, -0.0052719116, -0.050476074, -0.009414673, 0.0018548965, -0.041259766, -0.013450623, -0.014465332, 0.01184845, -0.031799316, 0.012138367, -0.027359009, -0.007171631, -0.028549194, -0.025405884, -0.016418457, ...
http://arxiv.org/abs/2307.11865v2
CARTIER: Cartographic lAnguage Reasoning Targeted at Instruction Execution for Robots
Dmitriy Rivkin, Nikhil Kakodkar, Francois Hogan, Bobak H. Baghi, Gregory Dudek
cs.RO, cs.CL
This work explores the capacity of large language models (LLMs) to address problems at the intersection of spatial planning and natural language interfaces for navigation.Our focus is on following relatively complex instructions that are more akin to natural conversation than traditional explicit procedural directives ...
http://arxiv.org/pdf/2307.11865v2
2023-10-06T15:41:54
2023-07-21T19:09:37
{ "title": [ -0.053344727, -0.047790527, -0.014762878, -0.022445679, -0.04196167, 0.005870819, -0.046813965, 0.003358841, 0.00066041946, 0.017105103, -0.029571533, 0.0026855469, -0.012138367, -0.015388489, -0.025405884, -0.016601562, 0.0101623535, ...
http://arxiv.org/abs/2307.11864v1
The Looming Threat of Fake and LLM-generated LinkedIn Profiles: Challenges and Opportunities for Detection and Prevention
Navid Ayoobi, Sadat Shahriar, Arjun Mukherjee
cs.SI, cs.CL, cs.CR, cs.LG
In this paper, we present a novel method for detecting fake and Large Language Model (LLM)-generated profiles in the LinkedIn Online Social Network immediately upon registration and before establishing connections. Early fake profile identification is crucial to maintaining the platform's integrity since it prevents im...
http://arxiv.org/pdf/2307.11864v1
2023-07-21T19:09:24
2023-07-21T19:09:24
{ "title": [ 0.017166138, 0.009986877, -0.030792236, -0.02116394, 0.008773804, -0.013420105, -0.005924225, -0.015136719, -0.099121094, 0.012023926, 0.016296387, 0.05328369, -0.021072388, -0.049072266, 0.035186768, -0.015899658, -0.023544312, 0.04...
http://arxiv.org/abs/2307.11848v1
MythQA: Query-Based Large-Scale Check-Worthy Claim Detection through Multi-Answer Open-Domain Question Answering
Yang Bai, Anthony Colas, Daisy Zhe Wang
cs.CL, cs.IR
Check-worthy claim detection aims at providing plausible misinformation to downstream fact-checking systems or human experts to check. This is a crucial step toward accelerating the fact-checking process. Many efforts have been put into how to identify check-worthy claims from a small scale of pre-collected claims, but...
http://arxiv.org/pdf/2307.11848v1
2023-07-21T18:35:24
2023-07-21T18:35:24
{ "title": [ 0.030563354, 0.028701782, -0.040527344, -0.056671143, 0.046051025, -0.01689148, -0.003112793, -0.027175903, -0.036712646, 0.005531311, 0.003993988, 0.02116394, 0.011009216, -0.03591919, 0.0038452148, -0.006504059, -0.02734375, 0.0453...
http://arxiv.org/abs/2307.11845v1
Multimodal Document Analytics for Banking Process Automation
Christopher Gerling, Stefan Lessmann
cs.CL, cs.AI, cs.CV, q-fin.CP
In response to growing FinTech competition and the need for improved operational efficiency, this research focuses on understanding the potential of advanced document analytics, particularly using multimodal models, in banking processes. We perform a comprehensive analysis of the diverse banking document landscape, hig...
http://arxiv.org/pdf/2307.11845v1
2023-07-21T18:29:04
2023-07-21T18:29:04
{ "title": [ -0.036865234, -0.007007599, -0.066589355, -0.05303955, -0.017166138, -0.014343262, -0.0541687, 0.018371582, 0.006916046, 0.05166626, -0.05154419, -0.004371643, 0.053771973, -0.0061912537, -0.029708862, -0.012237549, 0.0048332214, 0.0...
http://arxiv.org/abs/2307.11729v2
OUTFOX: LLM-generated Essay Detection through In-context Learning with Adversarially Generated Examples
Ryuto Koike, Masahiro Kaneko, Naoaki Okazaki
cs.CL
Large Language Models (LLMs) have achieved human-level fluency in text generation, making it difficult to distinguish between human-written and LLM-generated texts. This poses a growing risk of misuse of LLMs and demands the development of detectors to identify LLM-generated texts. However, existing detectors lack robu...
http://arxiv.org/pdf/2307.11729v2
2023-09-04T10:20:30
2023-07-21T17:40:47
{ "title": [ 0.008087158, 0.022949219, -0.005783081, 0.014961243, 0.0036411285, -0.0096206665, 0.011268616, -0.059906006, -0.025985718, 0.008094788, -0.023864746, 0.056427002, -0.053771973, -0.012672424, -0.01033783, -0.041778564, -0.033569336, 0...
http://arxiv.org/abs/2308.03762v2
GPT-4 Can't Reason
Konstantine Arkoudas
cs.CL
GPT-4 was released in March 2023 to wide acclaim, marking a very substantial improvement across the board over GPT-3.5 (OpenAI's previously best model, which had powered the initial release of ChatGPT). However, despite the genuinely impressive improvement, there are good reasons to be highly skeptical of GPT-4's abili...
http://arxiv.org/pdf/2308.03762v2
2023-08-10T14:24:37
2023-07-21T17:04:25
{ "title": [ 0.002231598, 0.0020942688, -0.013168335, -0.012260437, -0.01826477, 0.006034851, -0.015670776, -0.024475098, -0.008361816, 0.024856567, -0.019256592, -0.01940918, -0.043029785, -0.035186768, -0.026824951, -0.024978638, 0.0044555664, ...
http://arxiv.org/abs/2307.11661v2
Enhancing CLIP with GPT-4: Harnessing Visual Descriptions as Prompts
Mayug Maniparambil, Chris Vorster, Derek Molloy, Noel Murphy, Kevin McGuinness, Noel E. O'Connor
cs.CV, cs.AI, cs.CL, cs.LG
Contrastive pretrained large Vision-Language Models (VLMs) like CLIP have revolutionized visual representation learning by providing good performance on downstream datasets. VLMs are 0-shot adapted to a downstream dataset by designing prompts that are relevant to the dataset. Such prompt engineering makes use of domain...
http://arxiv.org/pdf/2307.11661v2
2023-08-08T13:44:12
2023-07-21T15:49:59
{ "title": [ 0.010925293, 0.0051612854, -0.026321411, -0.032806396, -0.0012235641, -0.0063705444, -0.026916504, -0.047424316, -0.0057373047, 0.0036621094, -0.0209198, 0.00111866, -0.020980835, -0.026870728, -0.029815674, -0.007980347, -0.004508972, ...
http://arxiv.org/abs/2307.11636v1
OxfordTVG-HIC: Can Machine Make Humorous Captions from Images?
Runjia Li, Shuyang Sun, Mohamed Elhoseiny, Philip Torr
cs.CV, cs.CL
This paper presents OxfordTVG-HIC (Humorous Image Captions), a large-scale dataset for humour generation and understanding. Humour is an abstract, subjective, and context-dependent cognitive construct involving several cognitive factors, making it a challenging task to generate and interpret. Hence, humour generation a...
http://arxiv.org/pdf/2307.11636v1
2023-07-21T14:58:44
2023-07-21T14:58:44
{ "title": [ 0.033233643, -0.030319214, -0.0041389465, 0.009605408, -0.04269409, -0.019104004, 0.032714844, -0.017593384, 0.020065308, 0.04727173, -0.051330566, 0.0036506653, 0.0014162064, -0.01612854, -0.046142578, -0.034057617, -0.0059127808, 0...
http://arxiv.org/abs/2307.11610v2
CausE: Towards Causal Knowledge Graph Embedding
Yichi Zhang, Wen Zhang
cs.CL, cs.AI
Knowledge graph embedding (KGE) focuses on representing the entities and relations of a knowledge graph (KG) into the continuous vector spaces, which can be employed to predict the missing triples to achieve knowledge graph completion (KGC). However, KGE models often only briefly learn structural correlations of triple...
http://arxiv.org/pdf/2307.11610v2
2023-07-24T01:35:47
2023-07-21T14:25:39
{ "title": [ -0.01889038, -0.012458801, -0.057617188, -0.10119629, 0.008354187, 0.0009803772, -0.06365967, 0.023895264, 0.0046844482, -0.007713318, -0.01838684, 0.010528564, -0.005554199, -0.028244019, 0.001244545, -0.009475708, -0.009986877, 0.0...
http://arxiv.org/abs/2307.11584v1
A Change of Heart: Improving Speech Emotion Recognition through Speech-to-Text Modality Conversion
Zeinab Sadat Taghavi, Ali Satvaty, Hossein Sameti
cs.SD, cs.CL, cs.LG, eess.AS
Speech Emotion Recognition (SER) is a challenging task. In this paper, we introduce a modality conversion concept aimed at enhancing emotion recognition performance on the MELD dataset. We assess our approach through two experiments: first, a method named Modality-Conversion that employs automatic speech recognition (A...
http://arxiv.org/pdf/2307.11584v1
2023-07-21T13:48:11
2023-07-21T13:48:11
{ "title": [ -0.0079956055, 0.0012207031, -0.03086853, -0.036254883, 0.0067596436, -0.016036987, -0.025802612, 0.00048065186, 0.017440796, 0.05239868, -0.03314209, 0.0012378693, 0.039123535, -0.016174316, -0.06585693, 0.011047363, -0.0058898926, ...
http://arxiv.org/abs/2307.11558v1
Advancing Visual Grounding with Scene Knowledge: Benchmark and Method
Zhihong Chen, Ruifei Zhang, Yibing Song, Xiang Wan, Guanbin Li
cs.CV, cs.CL
Visual grounding (VG) aims to establish fine-grained alignment between vision and language. Ideally, it can be a testbed for vision-and-language models to evaluate their understanding of the images and texts and their reasoning abilities over their joint space. However, most existing VG datasets are constructed using s...
http://arxiv.org/pdf/2307.11558v1
2023-07-21T13:06:02
2023-07-21T13:06:02
{ "title": [ 0.010169983, 0.0066070557, -0.046905518, -0.05630493, 0.027008057, -0.027893066, -0.011138916, -0.012435913, 0.026046753, 0.01576233, -0.03567505, 0.0049743652, -0.014678955, -0.016525269, -0.053527832, -0.0047683716, 0.012794495, 0....
http://arxiv.org/abs/2307.11545v1
Bridging Vision and Language Encoders: Parameter-Efficient Tuning for Referring Image Segmentation
Zunnan Xu, Zhihong Chen, Yong Zhang, Yibing Song, Xiang Wan, Guanbin Li
cs.CV, cs.CL
Parameter Efficient Tuning (PET) has gained attention for reducing the number of parameters while maintaining performance and providing better hardware resource savings, but few studies investigate dense prediction tasks and interaction between modalities. In this paper, we do an investigation of efficient tuning probl...
http://arxiv.org/pdf/2307.11545v1
2023-07-21T12:46:15
2023-07-21T12:46:15
{ "title": [ -0.022262573, -0.011222839, 0.012840271, -0.031219482, -0.031555176, 0.003332138, -0.028213501, -0.029632568, -0.004627228, 0.03363037, -0.017868042, 0.015472412, -0.0016498566, -0.0074539185, -0.013626099, 0.0014925003, 0.013748169, ...
http://arxiv.org/abs/2307.11516v1
IndigoVX: Where Human Intelligence Meets AI for Optimal Decision Making
Kais Dukes
cs.AI, cs.CL
This paper defines a new approach for augmenting human intelligence with AI for optimal goal solving. Our proposed AI, Indigo, is an acronym for Informed Numerical Decision-making through Iterative Goal-Oriented optimization. When combined with a human collaborator, we term the joint system IndigoVX, for Virtual eXpert...
http://arxiv.org/pdf/2307.11516v1
2023-07-21T11:54:53
2023-07-21T11:54:53
{ "title": [ -0.013053894, -0.015304565, -0.02986145, -0.021194458, -0.010421753, -0.023254395, 0.0018901825, -0.03540039, -0.054534912, 0.027816772, -0.049926758, 0.010032654, -0.018829346, -0.015335083, -0.023086548, -0.04232788, -0.0053710938, ...
http://arxiv.org/abs/2307.11457v1
Incorporating Human Translator Style into English-Turkish Literary Machine Translation
Zeynep Yirmibeşoğlu, Olgun Dursun, Harun Dallı, Mehmet Şahin, Ena Hodzik, Sabri Gürses, Tunga Güngör
cs.CL, cs.AI
Although machine translation systems are mostly designed to serve in the general domain, there is a growing tendency to adapt these systems to other domains like literary translation. In this paper, we focus on English-Turkish literary translation and develop machine translation models that take into account the stylis...
http://arxiv.org/pdf/2307.11457v1
2023-07-21T09:39:50
2023-07-21T09:39:50
{ "title": [ -0.035308838, 0.0030765533, -0.024597168, 0.01838684, -0.019882202, -0.02230835, 0.019805908, -0.018356323, -0.02583313, 0.07336426, 0.0035858154, -0.003162384, 0.06304932, -0.03125, -0.037902832, 0.021835327, -0.011680603, 0.0192413...
http://arxiv.org/abs/2307.11450v1
Topic Identification For Spontaneous Speech: Enriching Audio Features With Embedded Linguistic Information
Dejan Porjazovski, Tamás Grósz, Mikko Kurimo
eess.AS, cs.CL
Traditional topic identification solutions from audio rely on an automatic speech recognition system (ASR) to produce transcripts used as input to a text-based model. These approaches work well in high-resource scenarios, where there are sufficient data to train both components of the pipeline. However, in low-resource...
http://arxiv.org/pdf/2307.11450v1
2023-07-21T09:30:46
2023-07-21T09:30:46
{ "title": [ -0.010070801, 0.0068092346, -0.044158936, -0.037872314, 0.007637024, -0.013145447, -0.032562256, -0.025772095, -0.015449524, 0.023513794, -0.013534546, 0.03466797, -0.005367279, -0.018875122, -0.0043945312, -0.003862381, -0.016342163, ...
http://arxiv.org/abs/2307.11795v1
Prompting Large Language Models with Speech Recognition Abilities
Yassir Fathullah, Chunyang Wu, Egor Lakomkin, Junteng Jia, Yuan Shangguan, Ke Li, Jinxi Guo, Wenhan Xiong, Jay Mahadeokar, Ozlem Kalinli, Christian Fuegen, Mike Seltzer
eess.AS, cs.AI, cs.CL, cs.LG
Large language models have proven themselves highly flexible, able to solve a wide range of generative tasks, such as abstractive summarization and open-ended question answering. In this paper we extend the capabilities of LLMs by directly attaching a small audio encoder allowing it to perform speech recognition. By di...
http://arxiv.org/pdf/2307.11795v1
2023-07-21T08:39:15
2023-07-21T08:39:15
{ "title": [ -0.015380859, 0.00032496452, -0.03616333, -0.07678223, -0.001909256, 0.0008831024, -0.073791504, 0.0069465637, 0.007587433, 0.010536194, -0.044555664, -0.016418457, -0.014945984, -0.011703491, -0.040985107, -0.0121536255, 0.0053634644, ...
http://arxiv.org/abs/2307.11394v2
MeetEval: A Toolkit for Computation of Word Error Rates for Meeting Transcription Systems
Thilo von Neumann, Christoph Boeddeker, Marc Delcroix, Reinhold Haeb-Umbach
cs.CL, eess.AS
MeetEval is an open-source toolkit to evaluate all kinds of meeting transcription systems. It provides a unified interface for the computation of commonly used Word Error Rates (WERs), specifically cpWER, ORC WER and MIMO WER along other WER definitions. We extend the cpWER computation by a temporal constraint to ensur...
http://arxiv.org/pdf/2307.11394v2
2023-10-02T05:25:18
2023-07-21T07:22:18
{ "title": [ -0.046081543, -0.0008773804, -0.06762695, -0.07244873, 0.0019607544, 0.0055503845, -0.07312012, 0.005470276, 0.031982422, 0.050994873, -0.04940796, 0.008079529, 0.036376953, -0.00019872189, -0.06536865, -0.007904053, -0.019180298, 0....
http://arxiv.org/abs/2307.11380v1
Is ChatGPT Involved in Texts? Measure the Polish Ratio to Detect ChatGPT-Generated Text
Lingyi Yang, Feng Jiang, Haizhou Li
cs.CL
The remarkable capabilities of large-scale language models, such as ChatGPT, in text generation have incited awe and spurred researchers to devise detectors to mitigate potential risks, including misinformation, phishing, and academic dishonesty. Despite this, most previous studies, including HC3, have been predominant...
http://arxiv.org/pdf/2307.11380v1
2023-07-21T06:38:37
2023-07-21T06:38:37
{ "title": [ -0.05029297, -0.0049438477, -0.015045166, 0.046051025, -0.059173584, -0.0135650635, -0.020385742, -0.023132324, -0.045532227, 0.055786133, -0.019729614, 0.02178955, 0.012023926, -0.01789856, -0.008415222, -0.0027637482, -0.027908325, ...
http://arxiv.org/abs/2307.11346v1
CohortGPT: An Enhanced GPT for Participant Recruitment in Clinical Study
Zihan Guan, Zihao Wu, Zhengliang Liu, Dufan Wu, Hui Ren, Quanzheng Li, Xiang Li, Ninghao Liu
cs.CL, cs.AI
Participant recruitment based on unstructured medical texts such as clinical notes and radiology reports has been a challenging yet important task for the cohort establishment in clinical research. Recently, Large Language Models (LLMs) such as ChatGPT have achieved tremendous success in various downstream tasks thanks...
http://arxiv.org/pdf/2307.11346v1
2023-07-21T04:43:00
2023-07-21T04:43:00
{ "title": [ 0.0087890625, 0.008934021, -0.058685303, -0.07647705, 0.03704834, -0.016738892, -0.032562256, -0.019943237, -0.029678345, 0.011306763, -0.017929077, 0.02609253, -0.035095215, -0.03439331, -0.006275177, -0.0051612854, -0.015335083, 0....
http://arxiv.org/abs/2307.11344v1
DEFTri: A Few-Shot Label Fused Contextual Representation Learning For Product Defect Triage in e-Commerce
Ipsita Mohanty
cs.SE, cs.CL
Defect Triage is a time-sensitive and critical process in a large-scale agile software development lifecycle for e-commerce. Inefficiencies arising from human and process dependencies in this domain have motivated research in automated approaches using machine learning to accurately assign defects to qualified teams. T...
http://arxiv.org/pdf/2307.11344v1
2023-07-21T04:22:43
2023-07-21T04:22:43
{ "title": [ 0.036468506, -0.010627747, -0.032714844, -0.068847656, 0.008918762, -0.016799927, -0.02067566, 0.0036849976, 0.0052719116, 0.0357666, -0.05618286, 0.009361267, 0.043182373, -0.018325806, -0.0033893585, -0.024291992, -0.009277344, 0.0...
http://arxiv.org/abs/2307.11316v1
Making Pre-trained Language Models both Task-solvers and Self-calibrators
Yangyi Chen, Xingyao Wang, Heng Ji
cs.CL, cs.LG
Pre-trained language models (PLMs) serve as backbones for various real-world systems. For high-stake applications, it's equally essential to have reasonable confidence estimations in predictions. While the vanilla confidence scores of PLMs can already be effectively utilized, PLMs consistently become overconfident in t...
http://arxiv.org/pdf/2307.11316v1
2023-07-21T02:51:41
2023-07-21T02:51:41
{ "title": [ -0.000059127808, -0.0021629333, -0.016311646, -0.04046631, -0.04385376, -0.008796692, -0.052215576, -0.015319824, -0.029525757, 0.024932861, -0.028137207, 0.011260986, -0.013328552, -0.035186768, 0.011138916, -0.009185791, 0.019683838, ...
http://arxiv.org/abs/2307.11315v2
GIST: Generating Image-Specific Text for Fine-grained Object Classification
Kathleen M. Lewis, Emily Mu, Adrian V. Dalca, John Guttag
cs.CV, cs.CL
Recent vision-language models outperform vision-only models on many image classification tasks. However, because of the absence of paired text/image descriptions, it remains difficult to fine-tune these models for fine-grained image classification. In this work, we propose a method, GIST, for generating image-specific ...
http://arxiv.org/pdf/2307.11315v2
2023-08-04T19:36:31
2023-07-21T02:47:18
{ "title": [ -0.016281128, -0.00088071823, -0.028823853, -0.030136108, 0.002445221, -0.007896423, -0.014289856, -0.022018433, 0.019454956, 0.028259277, -0.0385437, 0.025665283, -0.01626587, 0.0060310364, -0.056518555, -0.005329132, 0.004776001, 0...
http://arxiv.org/abs/2308.04524v1
Who should I Collaborate with? A Comparative Study of Academia and Industry Research Collaboration in NLP
Hussain Sadiq Abuwala, Bohan Zhang, Mushi Wang
cs.DL, cs.CL, cs.IR, cs.LG
The goal of our research was to investigate the effects of collaboration between academia and industry on Natural Language Processing (NLP). To do this, we created a pipeline to extract affiliations and citations from NLP papers and divided them into three categories: academia, industry, and hybrid (collaborations betw...
http://arxiv.org/pdf/2308.04524v1
2023-07-21T01:26:29
2023-07-21T01:26:29
{ "title": [ -0.042755127, -0.013542175, -0.058410645, -0.018432617, -0.012123108, -0.025634766, -0.028442383, -0.0021457672, -0.044555664, 0.002691269, -0.035858154, 0.023422241, -0.051696777, -0.0072288513, -0.025436401, -0.048309326, -0.038909912,...
http://arxiv.org/abs/2307.11278v1
Generator-Retriever-Generator: A Novel Approach to Open-domain Question Answering
Abdelrahman Abdallah, Adam Jatowt
cs.CL
Open-domain question answering (QA) tasks usually require the retrieval of relevant information from a large corpus to generate accurate answers. We propose a novel approach called Generator-Retriever-Generator (GRG) that combines document retrieval techniques with a large language model (LLM), by first prompting the m...
http://arxiv.org/pdf/2307.11278v1
2023-07-21T00:34:38
2023-07-21T00:34:38
{ "title": [ 0.0047950745, 0.01033783, -0.023925781, -0.0769043, 0.023254395, 0.014381409, -0.031799316, -0.005340576, 0.0035057068, 0.02859497, -0.006439209, 0.01838684, 0.00016748905, -0.010543823, -0.046691895, 0.008956909, -0.0054130554, 0.01...
http://arxiv.org/abs/2307.11254v2
An In-Depth Evaluation of Federated Learning on Biomedical Natural Language Processing
Le Peng, Gaoxiang Luo, sicheng zhou, jiandong chen, Rui Zhang, Ziyue Xu, Ju Sun
cs.CL
Language models (LMs) such as BERT and GPT have revolutionized natural language processing (NLP). However, the medical field faces challenges in training LMs due to limited data access and privacy constraints imposed by regulations like the Health Insurance Portability and Accountability Act (HIPPA) and the General Dat...
http://arxiv.org/pdf/2307.11254v2
2023-11-11T23:50:59
2023-07-20T22:10:04
{ "title": [ -0.011627197, 0.0025749207, -0.05569458, -0.081970215, 0.022354126, -0.0024814606, -0.036468506, 0.014961243, -0.024520874, 0.041900635, -0.033416748, 0.003692627, -0.009674072, -0.007835388, -0.04663086, -0.021240234, -0.011726379, ...
http://arxiv.org/abs/2307.11224v3
Jina Embeddings: A Novel Set of High-Performance Sentence Embedding Models
Michael Günther, Louis Milliken, Jonathan Geuter, Georgios Mastrapas, Bo Wang, Han Xiao
cs.CL, cs.AI, cs.IR, cs.LG, 68T50, H.3.1; H.3.3; I.2.7; I.5.4
Jina Embeddings constitutes a set of high-performance sentence embedding models adept at translating textual inputs into numerical representations, capturing the semantics of the text. These models excel in applications like dense retrieval and semantic textual similarity. This paper details the development of Jina Emb...
http://arxiv.org/pdf/2307.11224v3
2023-10-20T14:09:36
2023-07-20T20:37:24
{ "title": [ -0.0061950684, 0.0021648407, -0.040893555, -0.05987549, 0.022384644, -0.00004285574, -0.042510986, -0.009986877, 0.02305603, 0.018707275, 0.0010032654, 0.013046265, 0.002319336, -0.018081665, -0.039398193, 0.021133423, -0.025146484, ...
http://arxiv.org/abs/2307.11788v3
Applying QNLP to sentiment analysis in finance
Jonas Stein, Ivo Christ, Nicolas Kraus, Maximilian Balthasar Mansky, Robert Müller, Claudia Linnhoff-Popien
cs.CL, cs.AI, quant-ph
As an application domain where the slightest qualitative improvements can yield immense value, finance is a promising candidate for early quantum advantage. Focusing on the rapidly advancing field of Quantum Natural Language Processing (QNLP), we explore the practical applicability of the two central approaches DisCoCa...
http://arxiv.org/pdf/2307.11788v3
2023-09-11T07:21:10
2023-07-20T18:30:35
{ "title": [ 0.00793457, 0.00094509125, -0.024032593, 0.00051259995, -0.0076828003, -0.02607727, 0.0031414032, -0.043884277, 0.0017213821, 0.0317688, -0.039611816, -0.023757935, -0.009094238, -0.0152282715, -0.07519531, -0.02557373, -0.024475098, ...
http://arxiv.org/abs/2307.11170v1
UMLS-KGI-BERT: Data-Centric Knowledge Integration in Transformers for Biomedical Entity Recognition
Aidan Mannion, Thierry Chevalier, Didier Schwab, Lorraine Geouriot
cs.CL
Pre-trained transformer language models (LMs) have in recent years become the dominant paradigm in applied NLP. These models have achieved state-of-the-art performance on tasks such as information extraction, question answering, sentiment analysis, document classification and many others. In the biomedical domain, sign...
http://arxiv.org/pdf/2307.11170v1
2023-07-20T18:08:34
2023-07-20T18:08:34
{ "title": [ -0.017837524, -0.018112183, -0.049224854, -0.09991455, 0.0034694672, 0.017807007, -0.06616211, 0.02053833, 0.0015544891, 0.03878784, -0.03982544, 0.0044517517, 0.04031372, -0.00680542, -0.03744507, -0.0030059814, -0.0004711151, 0.049...
http://arxiv.org/abs/2307.11088v3
L-Eval: Instituting Standardized Evaluation for Long Context Language Models
Chenxin An, Shansan Gong, Ming Zhong, Xingjian Zhao, Mukai Li, Jun Zhang, Lingpeng Kong, Xipeng Qiu
cs.CL
Recently, there has been growing interest in extending the context length of large language models (LLMs), aiming to effectively process long inputs of one turn or conversations with more extensive histories. While proprietary models such as GPT-4 and Claude can largely preserve the reasoning ability in an extended con...
http://arxiv.org/pdf/2307.11088v3
2023-10-04T10:04:25
2023-07-20T17:59:41
{ "title": [ -0.01826477, -0.0061416626, -0.04156494, -0.08288574, 0.018936157, 0.0060577393, -0.060791016, -0.010101318, 0.0015201569, 0.03237915, -0.0076789856, -0.0028476715, 0.0119018555, -0.030548096, -0.039886475, 0.01739502, -0.025619507, ...
http://arxiv.org/abs/2307.11031v1
Embroid: Unsupervised Prediction Smoothing Can Improve Few-Shot Classification
Neel Guha, Mayee F. Chen, Kush Bhatia, Azalia Mirhoseini, Frederic Sala, Christopher Ré
cs.LG, cs.CL
Recent work has shown that language models' (LMs) prompt-based learning capabilities make them well suited for automating data labeling in domains where manual annotation is expensive. The challenge is that while writing an initial prompt is cheap, improving a prompt is costly -- practitioners often require significant...
http://arxiv.org/pdf/2307.11031v1
2023-07-20T17:07:28
2023-07-20T17:07:28
{ "title": [ 0.014030457, -0.0071640015, -0.051971436, -0.049468994, 0.012001038, -0.030685425, 0.01158905, -0.002313614, 0.0007443428, -0.011932373, -0.028793335, 0.00340271, -0.015556335, -0.03692627, -0.0009918213, -0.029800415, 0.0077209473, ...
http://arxiv.org/abs/2307.10811v2
"It Felt Like Having a Second Mind": Investigating Human-AI Co-creativity in Prewriting with Large Language Models
Qian Wan, Siying Hu, Yu Zhang, Piaohong Wang, Bo Wen, Zhicong Lu
cs.HC, cs.AI, cs.CL, H.5.m; K.4.0
Prewriting is the process of discovering and developing ideas before a first draft, which requires divergent thinking and often implies unstructured strategies such as diagramming, outlining, free-writing, etc. Although large language models (LLMs) have been demonstrated to be useful for a variety of tasks including cr...
http://arxiv.org/pdf/2307.10811v2
2023-08-31T14:13:31
2023-07-20T16:55:25
{ "title": [ -0.04940796, -0.0141067505, -0.030731201, -0.011016846, -0.033203125, -0.009460449, -0.031982422, -0.018936157, -0.01789856, 0.034606934, -0.028961182, 0.018005371, 0.0057144165, -0.0109939575, -0.035369873, -0.022460938, -0.026519775, ...
http://arxiv.org/abs/2307.11019v2
Investigating the Factual Knowledge Boundary of Large Language Models with Retrieval Augmentation
Ruiyang Ren, Yuhao Wang, Yingqi Qu, Wayne Xin Zhao, Jing Liu, Hao Tian, Hua Wu, Ji-Rong Wen, Haifeng Wang
cs.CL, cs.IR
Knowledge-intensive tasks (e.g., open-domain question answering (QA)) require a substantial amount of factual knowledge and often rely on external information for assistance. Recently, large language models (LLMs) (e.g., ChatGPT), have demonstrated impressive prowess in solving a wide range of tasks with world knowledg...
http://arxiv.org/pdf/2307.11019v2
2023-07-23T16:52:59
2023-07-20T16:46:10
{ "title": [ 0.0039482117, -0.005657196, -0.044769287, -0.060394287, -0.0037002563, -0.013145447, -0.034820557, 0.0069770813, 0.027542114, 0.03201294, -0.031066895, 0.008453369, 0.016616821, -0.00484848, -0.046875, 0.017318726, -0.006134033, 0.01...
http://arxiv.org/abs/2307.11005v1
Integrating Pretrained ASR and LM to Perform Sequence Generation for Spoken Language Understanding
Siddhant Arora, Hayato Futami, Yosuke Kashiwagi, Emiru Tsunoo, Brian Yan, Shinji Watanabe
cs.CL, cs.SD, eess.AS
There has been an increased interest in the integration of pretrained speech recognition (ASR) and language models (LM) into the SLU framework. However, prior methods often struggle with a vocabulary mismatch between pretrained models, and LM cannot be directly utilized as they diverge from its NLU formulation. In this...
http://arxiv.org/pdf/2307.11005v1
2023-07-20T16:34:40
2023-07-20T16:34:40
{ "title": [ -0.005924225, -0.011230469, -0.04043579, -0.09124756, -0.024932861, 0.003145218, -0.081604004, 0.034942627, -0.03086853, 0.021133423, -0.020721436, -0.0012845993, 0.0027675629, -0.027954102, -0.020751953, -0.012802124, 0.012130737, 0...
http://arxiv.org/abs/2307.11787v2
LLM Cognitive Judgements Differ From Human
Sotiris Lamprinidis
cs.CL, cs.AI, cs.LG
Large Language Models (LLMs) have lately been on the spotlight of researchers, businesses, and consumers alike. While the linguistic capabilities of such models have been studied extensively, there is growing interest in investigating them as cognitive subjects. In the present work I examine GPT-3 and ChatGPT capabilit...
http://arxiv.org/pdf/2307.11787v2
2023-08-16T14:03:03
2023-07-20T16:22:36
{ "title": [ -0.033843994, -0.0035991669, 0.00041031837, 0.055725098, -0.08679199, -0.020553589, -0.026748657, -0.018371582, -0.021331787, 0.009140015, -0.03677368, -0.016281128, -0.058380127, -0.013053894, 0.0054779053, -0.035461426, 0.0075302124, ...
http://arxiv.org/abs/2307.10982v2
MASR: Multi-label Aware Speech Representation
Anjali Raj, Shikhar Bharadwaj, Sriram Ganapathy, Min Ma, Shikhar Vashishth
cs.SD, cs.CL, cs.LG, eess.AS
In the recent years, speech representation learning is constructed primarily as a self-supervised learning (SSL) task, using the raw audio signal alone, while ignoring the side-information that is often available for a given speech recording. In this paper, we propose MASR, a Multi-label Aware Speech Representation lea...
http://arxiv.org/pdf/2307.10982v2
2023-09-25T12:49:00
2023-07-20T16:09:57
{ "title": [ -0.0047035217, -0.0015878677, -0.048919678, -0.07269287, 0.011657715, -0.003042221, -0.04324341, 0.017211914, -0.006134033, 0.04473877, 0.0050697327, -0.001958847, 0.012397766, -0.028030396, -0.034301758, 0.017166138, -0.019958496, 0...
http://arxiv.org/abs/2307.10932v2
Identical and Fraternal Twins: Fine-Grained Semantic Contrastive Learning of Sentence Representations
Qingfa Xiao, Shuangyin Li, Lei Chen
cs.CL, cs.AI
The enhancement of unsupervised learning of sentence representations has been significantly achieved by the utility of contrastive learning. This approach clusters the augmented positive instance with the anchor instance to create a desired embedding space. However, relying solely on the contrastive objective can resul...
http://arxiv.org/pdf/2307.10932v2
2023-09-14T06:09:34
2023-07-20T15:02:42
{ "title": [ -0.0014829636, 0.0015983582, -0.039276123, -0.054351807, 0.00023257732, -0.027938843, -0.012962341, 0.006385803, 0.00390625, 0.04849243, -0.010826111, 0.034301758, -0.00831604, -0.0062065125, -0.05908203, 0.010139465, -0.0034999847, ...
http://arxiv.org/abs/2307.10930v2
MediaGPT : A Large Language Model For Chinese Media
Zhonghao Wang, Zijia Lu, Bo Jin, Haiying Deng
cs.CL, cs.AI
Large language models (LLMs) have shown remarkable capabilities in generating high-quality text and making predictions based on large amounts of data, including the media domain. However, in practical applications, the differences between the media's use cases and the general-purpose applications of LLMs have become in...
http://arxiv.org/pdf/2307.10930v2
2023-07-26T14:21:47
2023-07-20T14:59:02
{ "title": [ -0.028808594, -0.009613037, -0.026000977, -0.029953003, 0.0021190643, -0.003692627, -0.02658081, -0.0093307495, 0.0027713776, 0.043060303, 0.0051574707, -0.0018100739, 0.020080566, -0.026977539, -0.037628174, 0.022628784, -0.018341064, ...
http://arxiv.org/abs/2307.10928v2
FLASK: Fine-grained Language Model Evaluation based on Alignment Skill Sets
Seonghyeon Ye, Doyoung Kim, Sungdong Kim, Hyeonbin Hwang, Seungone Kim, Yongrae Jo, James Thorne, Juho Kim, Minjoon Seo
cs.CL, cs.AI
Evaluation of Large Language Models (LLMs) is challenging because instruction-following necessitates alignment with human values and the required set of skills varies depending on the instruction. However, previous studies have mainly focused on coarse-grained evaluation (i.e. overall preference-based evaluation), whic...
http://arxiv.org/pdf/2307.10928v2
2023-10-04T04:11:16
2023-07-20T14:56:35
{ "title": [ -0.031585693, -0.0031375885, -0.0592041, -0.0541687, 0.004886627, -0.0022144318, -0.062469482, 0.0013484955, 0.010597229, 0.030578613, -0.0029945374, -0.007659912, 0.0074043274, -0.02017212, -0.063964844, 0.006336212, -0.02104187, 0....
http://arxiv.org/abs/2308.02035v1
What Twitter Data Tell Us about the Future?
Alina Landowska, Marek Robak, Maciej Skorski
cs.CY, cs.CL, cs.LG, cs.SI
Anticipation is a fundamental human cognitive ability that involves thinking about and living towards the future. While language markers reflect anticipatory thinking, research on anticipation from the perspective of natural language processing is limited. This study aims to investigate the futures projected by futuris...
http://arxiv.org/pdf/2308.02035v1
2023-07-20T14:02:47
2023-07-20T14:02:47
{ "title": [ 0.048065186, 0.0006971359, -0.02822876, -0.070007324, 0.00472641, -0.01209259, -0.00033426285, 0.036193848, -0.02003479, 0.034332275, -0.045898438, -0.0021457672, -0.012649536, -0.010116577, -0.009735107, -0.04336548, 0.016143799, 0....
http://arxiv.org/abs/2307.10867v1
FigCaps-HF: A Figure-to-Caption Generative Framework and Benchmark with Human Feedback
Ashish Singh, Prateek Agarwal, Zixuan Huang, Arpita Singh, Tong Yu, Sungchul Kim, Victor Bursztyn, Nikos Vlassis, Ryan A. Rossi
cs.CL, cs.CV, cs.LG
Captions are crucial for understanding scientific visualizations and documents. Existing captioning methods for scientific figures rely on figure-caption pairs extracted from documents for training, many of which fall short with respect to metrics like helpfulness, explainability, and visual-descriptiveness [15] leadin...
http://arxiv.org/pdf/2307.10867v1
2023-07-20T13:40:22
2023-07-20T13:40:22
{ "title": [ -0.005970001, -0.007534027, -0.027709961, -0.049438477, -0.007827759, -0.009666443, -0.037902832, -0.029190063, 0.025619507, 0.043395996, -0.011116028, 0.02619934, -0.00058841705, -0.012817383, -0.04272461, 0.011299133, -0.00070667267, ...
http://arxiv.org/abs/2307.10864v2
Divide & Bind Your Attention for Improved Generative Semantic Nursing
Yumeng Li, Margret Keuper, Dan Zhang, Anna Khoreva
cs.CV, cs.AI, cs.CL, cs.LG
Emerging large-scale text-to-image generative models, e.g., Stable Diffusion (SD), have exhibited overwhelming results with high fidelity. Despite the magnificent progress, current state-of-the-art models still struggle to generate images fully adhering to the input prompt. Prior work, Attend & Excite, has introduced t...
http://arxiv.org/pdf/2307.10864v2
2023-10-09T07:20:00
2023-07-20T13:33:28
{ "title": [ -0.0118637085, -0.019378662, -0.05307007, -0.06689453, -0.009246826, -0.0104522705, -0.03326416, 0.019943237, -0.03692627, 0.03390503, -0.025863647, 0.019714355, 0.003982544, -0.023910522, -0.003578186, -0.013702393, 0.010658264, 0.0...