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/2303.09092v1
Investigating Failures to Generalize for Coreference Resolution Models
Ian Porada, Alexandra Olteanu, Kaheer Suleman, Adam Trischler, Jackie Chi Kit Cheung
cs.CL
Coreference resolution models are often evaluated on multiple datasets. Datasets vary, however, in how coreference is realized -- i.e., how the theoretical concept of coreference is operationalized in the dataset -- due to factors such as the choice of corpora and annotation guidelines. We investigate the extent to whi...
http://arxiv.org/pdf/2303.09092v1
2023-03-16T05:32:02
2023-03-16T05:32:02
{ "title": [ 0.0063285828, 0.004371643, -0.009262085, -0.07006836, 0.0011672974, 0.003528595, -0.022949219, -0.01159668, 0.005748749, 0.06774902, -0.027938843, 0.019622803, 0.035003662, 0.012428284, -0.06677246, 0.0141067505, -0.020004272, 0.0533...
http://arxiv.org/abs/2303.09075v3
Enhancing Text Generation with Cooperative Training
Tong Wu, Hao Wang, Zhongshen Zeng, Wei Wang, Hai-Tao Zheng, Jiaxing Zhang
cs.CL
Recently, there has been a surge in the use of generated data to enhance the performance of downstream models, largely due to the advancements in pre-trained language models. However, most prevailing methods trained generative and discriminative models in isolation, which left them unable to adapt to changes in each ot...
http://arxiv.org/pdf/2303.09075v3
2023-09-23T13:24:28
2023-03-16T04:21:19
{ "title": [ -0.025268555, -0.009628296, -0.03164673, -0.046661377, -0.02809143, 0.0077552795, -0.06109619, 0.0015172958, -0.020202637, 0.047088623, -0.0011043549, 0.016708374, -0.004638672, -0.0093307495, -0.041656494, -0.00001013279, -0.033416748, ...
http://arxiv.org/abs/2303.09067v1
Secret-Keeping in Question Answering
Nathaniel W. Rollings, Kent O'Sullivan, Sakshum Kulshrestha
cs.CL
Existing question-answering research focuses on unanswerable questions in the context of always providing an answer when a system can\dots but what about cases where a system {\bf should not} answer a question. This can either be to protect sensitive users or sensitive information. Many models expose sensitive informat...
http://arxiv.org/pdf/2303.09067v1
2023-03-16T03:56:17
2023-03-16T03:56:17
{ "title": [ -0.008613586, -0.002248764, -0.028503418, -0.022628784, -0.03778076, -0.028411865, -0.0006532669, -0.007537842, -0.012329102, 0.04034424, -0.018463135, -0.015670776, 0.008911133, -0.03086853, -0.02255249, -0.0022888184, 0.018661499, ...
http://arxiv.org/abs/2303.09038v3
Translating Radiology Reports into Plain Language using ChatGPT and GPT-4 with Prompt Learning: Promising Results, Limitations, and Potential
Qing Lyu, Josh Tan, Michael E. Zapadka, Janardhana Ponnatapura, Chuang Niu, Kyle J. Myers, Ge Wang, Christopher T. Whitlow
cs.CL, cs.AI, physics.med-ph
The large language model called ChatGPT has drawn extensively attention because of its human-like expression and reasoning abilities. In this study, we investigate the feasibility of using ChatGPT in experiments on using ChatGPT to translate radiology reports into plain language for patients and healthcare providers so...
http://arxiv.org/pdf/2303.09038v3
2023-03-29T03:22:52
2023-03-16T02:21:39
{ "title": [ -0.021881104, -0.0064582825, -0.008796692, -0.013717651, -0.024337769, 0.010871887, -0.031311035, -0.024505615, -0.0010671616, 0.01966858, 0.008384705, 0.002357483, -0.03225708, -0.048736572, -0.030303955, -0.0003874302, -0.0129852295, ...
http://arxiv.org/abs/2303.09031v1
A Picture is Worth a Thousand Words: Language Models Plan from Pixels
Anthony Z. Liu, Lajanugen Logeswaran, Sungryull Sohn, Honglak Lee
cs.CL, cs.AI, cs.LG
Planning is an important capability of artificial agents that perform long-horizon tasks in real-world environments. In this work, we explore the use of pre-trained language models (PLMs) to reason about plan sequences from text instructions in embodied visual environments. Prior PLM based approaches for planning eithe...
http://arxiv.org/pdf/2303.09031v1
2023-03-16T02:02:18
2023-03-16T02:02:18
{ "title": [ -0.025161743, -0.01966858, -0.025238037, -0.027832031, -0.0524292, -0.018737793, -0.03250122, -0.0057525635, -0.015556335, 0.022338867, -0.046691895, 0.01159668, -0.0015325546, -0.025863647, -0.014335632, -0.020843506, 0.01259613, 0....
http://arxiv.org/abs/2303.09014v1
ART: Automatic multi-step reasoning and tool-use for large language models
Bhargavi Paranjape, Scott Lundberg, Sameer Singh, Hannaneh Hajishirzi, Luke Zettlemoyer, Marco Tulio Ribeiro
cs.CL
Large language models (LLMs) can perform complex reasoning in few- and zero-shot settings by generating intermediate chain of thought (CoT) reasoning steps. Further, each reasoning step can rely on external tools to support computation beyond the core LLM capabilities (e.g. search/running code). Prior work on CoT promp...
http://arxiv.org/pdf/2303.09014v1
2023-03-16T01:04:45
2023-03-16T01:04:45
{ "title": [ -0.025421143, -0.030059814, -0.038116455, -0.099243164, 0.0026950836, 0.026535034, -0.07922363, 0.02041626, 0.013305664, 0.021240234, -0.011260986, 0.0015258789, 0.011993408, -0.025131226, -0.029571533, 0.016815186, -0.00957489, 0.02...
http://arxiv.org/abs/2303.08991v5
DeltaScore: Fine-Grained Story Evaluation with Perturbations
Zhuohan Xie, Miao Li, Trevor Cohn, Jey Han Lau
cs.CL
Numerous evaluation metrics have been developed for natural language generation tasks, but their effectiveness in evaluating stories is limited as they are not specifically tailored to assess intricate aspects of storytelling, such as fluency and interestingness. In this paper, we introduce DELTASCORE, a novel methodol...
http://arxiv.org/pdf/2303.08991v5
2023-11-02T06:08:44
2023-03-15T23:45:54
{ "title": [ 0.00010317564, -0.008506775, -0.018051147, -0.045532227, -0.0013399124, 0.0028514862, 0.004184723, -0.022064209, -0.012588501, 0.01335144, -0.039245605, 0.008773804, -0.024337769, -0.010559082, -0.038360596, -0.026351929, -0.021636963, ...
http://arxiv.org/abs/2303.08985v1
Cross-domain Sentiment Classification in Spanish
Lautaro Estienne, Matias Vera, Leonardo Rey Vega
cs.CL
Sentiment Classification is a fundamental task in the field of Natural Language Processing, and has very important academic and commercial applications. It aims to automatically predict the degree of sentiment present in a text that contains opinions and subjectivity at some level, like product and movie reviews, or tw...
http://arxiv.org/pdf/2303.08985v1
2023-03-15T23:11:30
2023-03-15T23:11:30
{ "title": [ -0.0057792664, 0.0021591187, -0.027282715, -0.05480957, 0.018814087, -0.008995056, -0.023529053, 0.000426054, 0.022125244, 0.03149414, -0.020965576, -0.0059280396, 0.0035095215, -0.017578125, -0.055603027, 0.0067634583, -0.029968262, ...
http://arxiv.org/abs/2303.08954v2
PRESTO: A Multilingual Dataset for Parsing Realistic Task-Oriented Dialogs
Rahul Goel, Waleed Ammar, Aditya Gupta, Siddharth Vashishtha, Motoki Sano, Faiz Surani, Max Chang, HyunJeong Choe, David Greene, Kyle He, Rattima Nitisaroj, Anna Trukhina, Shachi Paul, Pararth Shah, Rushin Shah, Zhou Yu
cs.CL
Research interest in task-oriented dialogs has increased as systems such as Google Assistant, Alexa and Siri have become ubiquitous in everyday life. However, the impact of academic research in this area has been limited by the lack of datasets that realistically capture the wide array of user pain points. To enable re...
http://arxiv.org/pdf/2303.08954v2
2023-03-17T02:26:52
2023-03-15T21:51:13
{ "title": [ -0.008056641, -0.010314941, -0.041503906, -0.09124756, 0.0032749176, 0.0011119843, -0.09069824, 0.0118637085, 0.00630188, 0.033203125, 0.008293152, -0.0036010742, -0.0008869171, -0.014732361, -0.052215576, 0.0028152466, -0.009315491, ...
http://arxiv.org/abs/2303.08928v1
Applying unsupervised keyphrase methods on concepts extracted from discharge sheets
Hoda Memarzadeh, Nasser Ghadiri, Matthias Samwald, Maryam Lotfi Shahreza
cs.CL, cs.LG, 68P20, H.1; J.3
Clinical notes containing valuable patient information are written by different health care providers with various scientific levels and writing styles. It might be helpful for clinicians and researchers to understand what information is essential when dealing with extensive electronic medical records. Entities recogni...
http://arxiv.org/pdf/2303.08928v1
2023-03-15T20:55:25
2023-03-15T20:55:25
{ "title": [ -0.027191162, 0.012420654, -0.064819336, -0.0647583, -0.0035324097, -0.008514404, -0.0703125, -0.0013160706, 0.05203247, 0.04714966, -0.019927979, -0.019744873, 0.000790596, -0.017669678, -0.060791016, 0.008399963, -0.030517578, 0.04...
http://arxiv.org/abs/2303.08896v3
SelfCheckGPT: Zero-Resource Black-Box Hallucination Detection for Generative Large Language Models
Potsawee Manakul, Adian Liusie, Mark J. F. Gales
cs.CL
Generative Large Language Models (LLMs) such as GPT-3 are capable of generating highly fluent responses to a wide variety of user prompts. However, LLMs are known to hallucinate facts and make non-factual statements which can undermine trust in their output. Existing fact-checking approaches either require access to th...
http://arxiv.org/pdf/2303.08896v3
2023-10-11T17:43:28
2023-03-15T19:31:21
{ "title": [ 0.024734497, 0.00459671, -0.03186035, -0.07080078, 0.0027618408, -0.011581421, -0.045806885, -0.0066070557, 0.011001587, 0.03540039, -0.010070801, 0.017944336, 0.024795532, -0.02355957, -0.04550171, 0.024871826, -0.014060974, 0.05444...
http://arxiv.org/abs/2303.08894v1
A Formalization of Operads in Coq
Zachary Flores, Angelo Taranto, Eric Bond, Yakir Forman
math.CT, cs.CL, cs.PL
What provides the highest level of assurance for correctness of execution within a programming language? One answer, and our solution in particular, to this problem is to provide a formalization for, if it exists, the denotational semantics of a programming language. Achieving such a formalization provides a gold stand...
http://arxiv.org/pdf/2303.08894v1
2023-03-15T19:29:40
2023-03-15T19:29:40
{ "title": [ 0.015617371, -0.01751709, -0.03427124, -0.10192871, -0.026290894, 0.008598328, -0.07775879, 0.032043457, -0.0317688, 0.029678345, -0.022750854, -0.0057754517, 0.0069885254, -0.047058105, 0.045288086, -0.009941101, 0.021392822, -0.003...
http://arxiv.org/abs/2303.08877v1
ROSE: A Neurocomputational Architecture for Syntax
Elliot Murphy
cs.CL, q-bio.NC
A comprehensive model of natural language processing in the brain must accommodate four components: representations, operations, structures and encoding. It further requires a principled account of how these components mechanistically, and causally, relate to each another. While previous models have isolated regions of...
http://arxiv.org/pdf/2303.08877v1
2023-03-15T18:44:37
2023-03-15T18:44:37
{ "title": [ -0.0317688, -0.02079773, -0.015113831, -0.058776855, -0.007030487, 0.019577026, -0.041381836, 0.012481689, 0.0015325546, 0.03466797, -0.016403198, -0.015640259, 0.00014090538, -0.0206604, -0.059295654, 0.00063991547, -0.0053367615, 0...
http://arxiv.org/abs/2303.08809v2
Cascading and Direct Approaches to Unsupervised Constituency Parsing on Spoken Sentences
Yuan Tseng, Cheng-I Lai, Hung-yi Lee
cs.CL, eess.AS
Past work on unsupervised parsing is constrained to written form. In this paper, we present the first study on unsupervised spoken constituency parsing given unlabeled spoken sentences and unpaired textual data. The goal is to determine the spoken sentences' hierarchical syntactic structure in the form of constituency ...
http://arxiv.org/pdf/2303.08809v2
2023-05-09T10:36:55
2023-03-15T17:57:22
{ "title": [ -0.0022411346, 0.010826111, -0.01914978, -0.08508301, 0.013595581, 0.014404297, -0.074279785, -0.002035141, -0.0118637085, 0.028808594, -0.027786255, 0.008666992, -0.017547607, -0.0041542053, -0.04840088, -0.014251709, -0.013320923, ...
http://arxiv.org/abs/2303.08806v1
Understanding Post-hoc Explainers: The Case of Anchors
Gianluigi Lopardo, Frederic Precioso, Damien Garreau
stat.ML, cs.AI, cs.CL, cs.LG
In many scenarios, the interpretability of machine learning models is a highly required but difficult task. To explain the individual predictions of such models, local model-agnostic approaches have been proposed. However, the process generating the explanations can be, for a user, as mysterious as the prediction to be...
http://arxiv.org/pdf/2303.08806v1
2023-03-15T17:56:34
2023-03-15T17:56:34
{ "title": [ 0.012542725, -0.0072746277, 0.00005364418, -0.054260254, -0.000014781952, -0.009635925, -0.008964539, -0.036254883, -0.029464722, 0.015312195, -0.0059776306, 0.0021247864, -0.020477295, -0.029785156, -0.014816284, -0.012832642, 0.0053634...
http://arxiv.org/abs/2303.08774v3
GPT-4 Technical Report
OpenAI
cs.CL, cs.AI
We report the development of GPT-4, a large-scale, multimodal model which can accept image and text inputs and produce text outputs. While less capable than humans in many real-world scenarios, GPT-4 exhibits human-level performance on various professional and academic benchmarks, including passing a simulated bar exam...
http://arxiv.org/pdf/2303.08774v3
2023-03-27T17:46:54
2023-03-15T17:15:04
{ "title": [ 0.015541077, 0.007987976, -0.0079574585, -0.019882202, -0.012825012, -0.014289856, -0.035095215, -0.039031982, -0.022003174, 0.044891357, 0.00705719, -0.018463135, 0.018203735, -0.047668457, 0.006374359, -0.007095337, 0.007965088, -0...
http://arxiv.org/abs/2303.08721v1
Artificial Influence: An Analysis Of AI-Driven Persuasion
Matthew Burtell, Thomas Woodside
cs.CY, cs.AI, cs.CL, cs.LG
Persuasion is a key aspect of what it means to be human, and is central to business, politics, and other endeavors. Advancements in artificial intelligence (AI) have produced AI systems that are capable of persuading humans to buy products, watch videos, click on search results, and more. Even systems that are not expl...
http://arxiv.org/pdf/2303.08721v1
2023-03-15T16:05:11
2023-03-15T16:05:11
{ "title": [ 0.00031757355, -0.0035419464, -0.0050849915, -0.026794434, 0.013755798, -0.009147644, 0.031173706, -0.026107788, -0.045562744, 0.01725769, -0.019973755, 0.028289795, -0.025039673, -0.017196655, -0.014549255, -0.015716553, -0.02494812, ...
http://arxiv.org/abs/2303.08697v1
Mirror: A Natural Language Interface for Data Querying, Summarization, and Visualization
Canwen Xu, Julian McAuley, Penghan Wang
cs.DB, cs.AI, cs.CL, cs.SE
We present Mirror, an open-source platform for data exploration and analysis powered by large language models. Mirror offers an intuitive natural language interface for querying databases, and automatically generates executable SQL commands to retrieve relevant data and summarize it in natural language. In addition, us...
http://arxiv.org/pdf/2303.08697v1
2023-03-15T15:31:51
2023-03-15T15:31:51
{ "title": [ -0.024963379, -0.020812988, -0.04537964, -0.05758667, 0.006717682, 0.004585266, -0.02545166, 0.0011520386, 0.019439697, 0.037963867, -0.01360321, -0.012329102, 0.012008667, -0.014228821, -0.05682373, -0.004760742, 0.00092840195, 0.02...
http://arxiv.org/abs/2303.08652v1
Automated Query Generation for Evidence Collection from Web Search Engines
Nestor Prieto-Chavana, Julie Weeds, David Weir
cs.CL, cs.IR
It is widely accepted that so-called facts can be checked by searching for information on the Internet. This process requires a fact-checker to formulate a search query based on the fact and to present it to a search engine. Then, relevant and believable passages need to be identified in the search results before a dec...
http://arxiv.org/pdf/2303.08652v1
2023-03-15T14:32:00
2023-03-15T14:32:00
{ "title": [ -0.0042762756, -0.010047913, -0.045288086, -0.058685303, -0.0042762756, -0.010681152, -0.013755798, 0.007659912, -0.016937256, 0.029724121, -0.0178833, 0.021133423, -0.015930176, -0.017288208, -0.03286743, 0.007270813, -0.0013942719, ...
http://arxiv.org/abs/2303.08606v1
On the Calibration and Uncertainty with Pólya-Gamma Augmentation for Dialog Retrieval Models
Tong Ye, Shijing Si, Jianzong Wang, Ning Cheng, Zhitao Li, Jing Xiao
cs.CL, cs.AI
Deep neural retrieval models have amply demonstrated their power but estimating the reliability of their predictions remains challenging. Most dialog response retrieval models output a single score for a response on how relevant it is to a given question. However, the bad calibration of deep neural network results in v...
http://arxiv.org/pdf/2303.08606v1
2023-03-15T13:26:25
2023-03-15T13:26:25
{ "title": [ 0.0015363693, 0.016998291, -0.038146973, -0.044158936, -0.0067863464, -0.011627197, -0.041381836, -0.018722534, 0.03945923, 0.031433105, -0.0014791489, 0.0034770966, 0.005355835, -0.026245117, -0.06384277, 0.014602661, -0.011604309, ...
http://arxiv.org/abs/2303.08601v1
GCRE-GPT: A Generative Model for Comparative Relation Extraction
Yequan Wang, Hengran Zhang, Aixin Sun, Xuying Meng
cs.CL
Given comparative text, comparative relation extraction aims to extract two targets (\eg two cameras) in comparison and the aspect they are compared for (\eg image quality). The extracted comparative relations form the basis of further opinion analysis.Existing solutions formulate this task as a sequence labeling task,...
http://arxiv.org/pdf/2303.08601v1
2023-03-15T13:15:22
2023-03-15T13:15:22
{ "title": [ -0.0068511963, -0.016448975, -0.040924072, -0.078552246, 0.010688782, 0.0154953, -0.05404663, -0.017578125, 0.029281616, 0.023849487, -0.01701355, 0.016571045, -0.015457153, -0.0047454834, -0.05419922, 0.0058135986, -0.02041626, 0.04...
http://arxiv.org/abs/2303.08599v1
Efficient Uncertainty Estimation with Gaussian Process for Reliable Dialog Response Retrieval
Tong Ye, Zhitao Li, Jianzong Wang, Ning Cheng, Jing Xiao
cs.CL, cs.AI
Deep neural networks have achieved remarkable performance in retrieval-based dialogue systems, but they are shown to be ill calibrated. Though basic calibration methods like Monte Carlo Dropout and Ensemble can calibrate well, these methods are time-consuming in the training or inference stages. To tackle these challen...
http://arxiv.org/pdf/2303.08599v1
2023-03-15T13:12:16
2023-03-15T13:12:16
{ "title": [ -0.004611969, 0.0005197525, -0.04071045, -0.044952393, -0.004951477, -0.025344849, -0.04006958, -0.025665283, 0.019424438, 0.037475586, 0.0008955002, -0.003370285, -0.011489868, -0.021514893, -0.080078125, 0.0029277802, -0.009223938, ...
http://arxiv.org/abs/2303.08559v2
Large Language Model Is Not a Good Few-shot Information Extractor, but a Good Reranker for Hard Samples!
Yubo Ma, Yixin Cao, YongChing Hong, Aixin Sun
cs.CL, cs.AI
Large Language Models (LLMs) have made remarkable strides in various tasks. Whether LLMs are competitive few-shot solvers for information extraction (IE) tasks, however, remains an open problem. In this work, we aim to provide a thorough answer to this question. Through extensive experiments on nine datasets across fou...
http://arxiv.org/pdf/2303.08559v2
2023-10-21T14:44:11
2023-03-15T12:20:13
{ "title": [ -0.028137207, -0.03656006, 0.026885986, -0.025253296, -0.06427002, 0.046936035, -0.072265625, -0.024841309, 0.015029907, 0.014328003, -0.04638672, 0.024765015, -0.030517578, -0.01309967, 0.0009036064, -0.01473999, -0.0047416687, -0.0...
http://arxiv.org/abs/2303.08553v2
The Image of the Process Interpretation of Regular Expressions is Not Closed under Bisimulation Collapse
Clemens Grabmayer
cs.LO, cs.CL, F.4.1
Axiomatization and expressibility problems for Milner's process semantics (1984) of regular expressions modulo bisimilarity have turned out to be difficult for the full class of expressions with deadlock 0 and empty step~1. We report on a phenomenon that arises from the added presence of 1 when 0 is available, and that...
http://arxiv.org/pdf/2303.08553v2
2023-11-12T12:52:31
2023-03-15T12:13:30
{ "title": [ 0.03842163, 0.0034656525, 0.010505676, -0.03656006, -0.027175903, 0.011917114, -0.031173706, -0.015556335, -0.010528564, 0.023086548, -0.0552063, 0.0050811768, -0.013786316, -0.03427124, 0.0004749298, -0.035369873, 0.0031757355, 0.01...
http://arxiv.org/abs/2303.08518v3
UPRISE: Universal Prompt Retrieval for Improving Zero-Shot Evaluation
Daixuan Cheng, Shaohan Huang, Junyu Bi, Yuefeng Zhan, Jianfeng Liu, Yujing Wang, Hao Sun, Furu Wei, Denvy Deng, Qi Zhang
cs.CL
Large Language Models (LLMs) are popular for their impressive abilities, but the need for model-specific fine-tuning or task-specific prompt engineering can hinder their generalization. We propose UPRISE (Universal Prompt Retrieval for Improving zero-Shot Evaluation), which tunes a lightweight and versatile retriever t...
http://arxiv.org/pdf/2303.08518v3
2023-10-11T05:40:41
2023-03-15T10:53:49
{ "title": [ 0.03842163, -0.016479492, -0.019577026, -0.046936035, -0.010261536, -0.025848389, 0.0056877136, -0.01473999, 0.030883789, 0.026855469, -0.005870819, -0.006286621, -0.020141602, -0.03942871, -0.04940796, 0.009124756, -0.003452301, 0.0...
http://arxiv.org/abs/2303.08448v1
A Cross-institutional Evaluation on Breast Cancer Phenotyping NLP Algorithms on Electronic Health Records
Sicheng Zhou, Nan Wang, Liwei Wang, Ju Sun, Anne Blaes, Hongfang Liu, Rui Zhang
cs.CL, cs.IR, cs.LG
Objective: The generalizability of clinical large language models is usually ignored during the model development process. This study evaluated the generalizability of BERT-based clinical NLP models across different clinical settings through a breast cancer phenotype extraction task. Materials and Methods: Two clinic...
http://arxiv.org/pdf/2303.08448v1
2023-03-15T08:44:07
2023-03-15T08:44:07
{ "title": [ 0.004432678, -0.010688782, -0.03137207, -0.026184082, -0.0054626465, -0.020462036, 0.010299683, -0.009025574, -0.0030231476, 0.021316528, -0.04949951, 0.0048942566, 0.010803223, -0.020462036, -0.004989624, -0.011444092, -0.01802063, ...
http://arxiv.org/abs/2303.08389v1
PR-MCS: Perturbation Robust Metric for MultiLingual Image Captioning
Yongil Kim, Yerin Hwang, Hyeongu Yun, Seunghyun Yoon, Trung Bui, Kyomin Jung
cs.CL
Vulnerability to lexical perturbation is a critical weakness of automatic evaluation metrics for image captioning. This paper proposes Perturbation Robust Multi-Lingual CLIPScore(PR-MCS), which exhibits robustness to such perturbations, as a novel reference-free image captioning metric applicable to multiple languages....
http://arxiv.org/pdf/2303.08389v1
2023-03-15T06:37:26
2023-03-15T06:37:26
{ "title": [ -0.0071411133, 0.008720398, 0.0047187805, -0.009887695, -0.0037384033, -0.008598328, -0.0069618225, -0.045288086, 0.023529053, 0.03540039, -0.038726807, 0.010223389, 0.019866943, -0.0079422, -0.031066895, -0.005428314, -0.002325058, ...
http://arxiv.org/abs/2303.13364v1
Reevaluating Data Partitioning for Emotion Detection in EmoWOZ
Moeen Mostafavi, Michael D. Porter
cs.CL, cs.LG
This paper focuses on the EmoWoz dataset, an extension of MultiWOZ that provides emotion labels for the dialogues. MultiWOZ was partitioned initially for another purpose, resulting in a distributional shift when considering the new purpose of emotion recognition. The emotion tags in EmoWoz are highly imbalanced and une...
http://arxiv.org/pdf/2303.13364v1
2023-03-15T03:06:13
2023-03-15T03:06:13
{ "title": [ -0.021331787, -0.01525116, -0.0340271, -0.019302368, -0.0025939941, -0.01876831, 0.0073432922, -0.004234314, 0.008407593, 0.048553467, -0.042144775, -0.019210815, 0.032348633, -0.0005054474, -0.06768799, -0.005142212, 0.0034160614, 0...
http://arxiv.org/abs/2303.08335v2
FactReranker: Fact-guided Reranker for Faithful Radiology Report Summarization
Qianqian Xie, Jiayu Zhou, Yifan Peng, Fei Wang
cs.CL, cs.AI
Automatic radiology report summarization is a crucial clinical task, whose key challenge is to maintain factual accuracy between produced summaries and ground truth radiology findings. Existing research adopts reinforcement learning to directly optimize factual consistency metrics such as CheXBert or RadGraph score. Ho...
http://arxiv.org/pdf/2303.08335v2
2023-03-16T03:09:09
2023-03-15T02:51:57
{ "title": [ 0.0027179718, -0.0036621094, -0.020065308, -0.078186035, -0.015274048, 0.01727295, -0.061431885, -0.009544373, -0.0019025803, 0.037384033, 0.0051193237, 0.0046691895, 0.0029029846, -0.031585693, -0.030532837, 0.008285522, -0.0027770996, ...
http://arxiv.org/abs/2303.08329v1
Cross-speaker Emotion Transfer by Manipulating Speech Style Latents
Suhee Jo, Younggun Lee, Yookyung Shin, Yeongtae Hwang, Taesu Kim
cs.SD, cs.CL, eess.AS
In recent years, emotional text-to-speech has shown considerable progress. However, it requires a large amount of labeled data, which is not easily accessible. Even if it is possible to acquire an emotional speech dataset, there is still a limitation in controlling emotion intensity. In this work, we propose a novel me...
http://arxiv.org/pdf/2303.08329v1
2023-03-15T02:34:03
2023-03-15T02:34:03
{ "title": [ 0.0023555756, -0.002374649, -0.03277588, -0.046569824, 0.0032215118, -0.019729614, -0.0101623535, -0.0039253235, 0.0054130554, 0.046173096, 0.0026130676, 0.012786865, 0.043945312, -0.033721924, -0.015625, 0.028533936, 0.001159668, 0....
http://arxiv.org/abs/2303.11163v1
Finding Similar Exercises in Retrieval Manner
Tongwen Huang, Xihua Li, Chao Yi, Xuemin Zhao, Yunbo Cao
cs.IR, cs.AI, cs.CL, cs.CY
When students make a mistake in an exercise, they can consolidate it by ``similar exercises'' which have the same concepts, purposes and methods. Commonly, for a certain subject and study stage, the size of the exercise bank is in the range of millions to even tens of millions, how to find similar exercises for a given...
http://arxiv.org/pdf/2303.11163v1
2023-03-15T01:40:32
2023-03-15T01:40:32
{ "title": [ 0.0496521, 0.039001465, -0.0041618347, 0.003704071, -0.002008438, -0.039886475, 0.00894928, -0.06958008, -0.0010490417, 0.047698975, -0.02029419, -0.020996094, 0.0127334595, -0.021347046, -0.010284424, -0.008918762, 0.017410278, 0.00...
http://arxiv.org/abs/2303.08302v3
ZeroQuant-V2: Exploring Post-training Quantization in LLMs from Comprehensive Study to Low Rank Compensation
Zhewei Yao, Xiaoxia Wu, Cheng Li, Stephen Youn, Yuxiong He
cs.LG, cs.AI, cs.CL
Post-training quantization (PTQ) has emerged as a promising technique for mitigating memory consumption and computational costs in large language models (LLMs). However, a systematic examination of various quantization schemes, model families, and quantization bit precision has been absent from the literature. In this ...
http://arxiv.org/pdf/2303.08302v3
2023-05-26T00:17:06
2023-03-15T01:27:15
{ "title": [ -0.03866577, 0.0024738312, -0.045166016, -0.0034751892, -0.033081055, -0.03237915, -0.037750244, -0.024780273, -0.017623901, 0.05114746, -0.0047416687, 0.018615723, 0.016464233, -0.03842163, -0.005115509, 0.01625061, -0.008041382, 0....
http://arxiv.org/abs/2303.08290v2
Rediscovery of CNN's Versatility for Text-based Encoding of Raw Electronic Health Records
Eunbyeol Cho, Min Jae Lee, Kyunghoon Hur, Jiyoun Kim, Jinsung Yoon, Edward Choi
cs.LG, cs.CL
Making the most use of abundant information in electronic health records (EHR) is rapidly becoming an important topic in the medical domain. Recent work presented a promising framework that embeds entire features in raw EHR data regardless of its form and medical code standards. The framework, however, only focuses on ...
http://arxiv.org/pdf/2303.08290v2
2023-05-10T09:11:10
2023-03-15T00:37:18
{ "title": [ -0.011932373, 0.007987976, -0.035064697, -0.023513794, -0.023025513, -0.011810303, -0.032165527, -0.018310547, -0.014854431, 0.05480957, -0.04815674, 0.017974854, 0.03427124, -0.013824463, -0.009780884, -0.014770508, 0.015670776, 0.0...
http://arxiv.org/abs/2303.13648v1
ChatGPT or Grammarly? Evaluating ChatGPT on Grammatical Error Correction Benchmark
Haoran Wu, Wenxuan Wang, Yuxuan Wan, Wenxiang Jiao, Michael Lyu
cs.CL
ChatGPT is a cutting-edge artificial intelligence language model developed by OpenAI, which has attracted a lot of attention due to its surprisingly strong ability in answering follow-up questions. In this report, we aim to evaluate ChatGPT on the Grammatical Error Correction(GEC) task, and compare it with commercial G...
http://arxiv.org/pdf/2303.13648v1
2023-03-15T00:35:50
2023-03-15T00:35:50
{ "title": [ -0.0011987686, -0.006942749, -0.015174866, -0.030227661, -0.0073661804, -0.00932312, -0.022415161, -0.007671356, -0.01737976, 0.038116455, -0.02154541, 0.028503418, -0.0003271103, -0.0054359436, -0.03704834, 0.00566864, -0.046844482, ...
http://arxiv.org/abs/2303.08288v1
Attention-likelihood relationship in transformers
Valeria Ruscio, Valentino Maiorca, Fabrizio Silvestri
cs.CL, cs.LG
We analyze how large language models (LLMs) represent out-of-context words, investigating their reliance on the given context to capture their semantics. Our likelihood-guided text perturbations reveal a correlation between token likelihood and attention values in transformer-based language models. Extensive experiment...
http://arxiv.org/pdf/2303.08288v1
2023-03-15T00:23:49
2023-03-15T00:23:49
{ "title": [ 0.02432251, 0.022918701, -0.02027893, -0.048339844, 0.01864624, -0.015686035, -0.0019111633, -0.02407837, 0.0038261414, 0.023971558, -0.054901123, -0.0077285767, 0.010772705, -0.018127441, -0.02243042, -0.0036239624, -0.009506226, 0....
http://arxiv.org/abs/2303.08268v3
Chat with the Environment: Interactive Multimodal Perception Using Large Language Models
Xufeng Zhao, Mengdi Li, Cornelius Weber, Muhammad Burhan Hafez, Stefan Wermter
cs.RO, cs.AI, cs.CL, cs.LG, cs.SD, eess.AS
Programming robot behavior in a complex world faces challenges on multiple levels, from dextrous low-level skills to high-level planning and reasoning. Recent pre-trained Large Language Models (LLMs) have shown remarkable reasoning ability in few-shot robotic planning. However, it remains challenging to ground LLMs in ...
http://arxiv.org/pdf/2303.08268v3
2023-10-11T16:17:20
2023-03-14T23:01:27
{ "title": [ -0.01687622, -0.0008773804, -0.017547607, -0.047729492, -0.0075149536, -0.009277344, -0.03842163, -0.016693115, 0.024307251, 0.027236938, -0.035888672, -0.01625061, -0.007774353, -0.009819031, -0.054229736, -0.015380859, -0.004547119, ...
http://arxiv.org/abs/2303.08262v1
Clinical Concept and Relation Extraction Using Prompt-based Machine Reading Comprehension
Cheng Peng, Xi Yang, Zehao Yu, Jiang Bian, William R. Hogan, Yonghui Wu
cs.CL
Objective: To develop a natural language processing system that solves both clinical concept extraction and relation extraction in a unified prompt-based machine reading comprehension (MRC) architecture with good generalizability for cross-institution applications. Methods: We formulate both clinical concept extracti...
http://arxiv.org/pdf/2303.08262v1
2023-03-14T22:37:31
2023-03-14T22:37:31
{ "title": [ -0.018676758, -0.015357971, -0.033996582, -0.0703125, -0.012672424, 0.014442444, -0.06756592, -0.0052261353, 0.026519775, 0.020202637, -0.01763916, -0.015037537, 0.0018663406, -0.023422241, -0.056762695, 0.00027537346, -0.015182495, ...
http://arxiv.org/abs/2303.08259v1
Contextualized Medication Information Extraction Using Transformer-based Deep Learning Architectures
Aokun Chen, Zehao Yu, Xi Yang, Yi Guo, Jiang Bian, Yonghui Wu
cs.CL, cs.AI
Objective: To develop a natural language processing (NLP) system to extract medications and contextual information that help understand drug changes. This project is part of the 2022 n2c2 challenge. Materials and methods: We developed NLP systems for medication mention extraction, event classification (indicating med...
http://arxiv.org/pdf/2303.08259v1
2023-03-14T22:22:28
2023-03-14T22:22:28
{ "title": [ 0.0016498566, 0.014816284, -0.051513672, -0.070739746, 0.019973755, -0.01096344, -0.0501709, -0.0152282715, -0.017486572, 0.034851074, -0.026168823, 0.031402588, -0.008453369, -0.010398865, -0.025421143, 0.0031661987, -0.0070533752, ...
http://arxiv.org/abs/2303.08233v2
NL4Opt Competition: Formulating Optimization Problems Based on Their Natural Language Descriptions
Rindranirina Ramamonjison, Timothy T. Yu, Raymond Li, Haley Li, Giuseppe Carenini, Bissan Ghaddar, Shiqi He, Mahdi Mostajabdaveh, Amin Banitalebi-Dehkordi, Zirui Zhou, Yong Zhang
cs.CL, cs.AI
The Natural Language for Optimization (NL4Opt) Competition was created to investigate methods of extracting the meaning and formulation of an optimization problem based on its text description. Specifically, the goal of the competition is to increase the accessibility and usability of optimization solvers by allowing n...
http://arxiv.org/pdf/2303.08233v2
2023-03-27T01:10:12
2023-03-14T20:59:04
{ "title": [ -0.02406311, -0.02305603, -0.010765076, -0.04119873, -0.009742737, 0.008010864, -0.028884888, -0.014808655, -0.01449585, 0.01600647, -0.0048713684, 0.021347046, -0.022155762, -0.021362305, -0.007789612, 0.0019016266, -0.008720398, 0....
http://arxiv.org/abs/2303.08179v2
MEDBERT.de: A Comprehensive German BERT Model for the Medical Domain
Keno K. Bressem, Jens-Michalis Papaioannou, Paul Grundmann, Florian Borchert, Lisa C. Adams, Leonhard Liu, Felix Busch, Lina Xu, Jan P. Loyen, Stefan M. Niehues, Moritz Augustin, Lennart Grosser, Marcus R. Makowski, Hugo JWL. Aerts, Alexander Löser
cs.CL, cs.AI
This paper presents medBERTde, a pre-trained German BERT model specifically designed for the German medical domain. The model has been trained on a large corpus of 4.7 Million German medical documents and has been shown to achieve new state-of-the-art performance on eight different medical benchmarks covering a wide ra...
http://arxiv.org/pdf/2303.08179v2
2023-03-24T23:05:31
2023-03-14T18:58:08
{ "title": [ -0.010139465, -0.016189575, -0.020462036, -0.07470703, 0.0028438568, 0.015617371, -0.05621338, -0.021499634, -0.008224487, 0.042816162, 0.038635254, -0.015457153, 0.038085938, -0.054992676, -0.037109375, 0.04232788, -0.009628296, 0.0...
http://arxiv.org/abs/2303.08127v3
CB2: Collaborative Natural Language Interaction Research Platform
Jacob Sharf, Mustafa Omer Gul, Yoav Artzi
cs.LG, cs.AI, cs.CL, cs.MA
CB2 is a multi-agent platform to study collaborative natural language interaction in a grounded task-oriented scenario. It includes a 3D game environment, a backend server designed to serve trained models to human agents, and various tools and processes to enable scalable studies. We deploy CB2 at https://cb2.ai as a s...
http://arxiv.org/pdf/2303.08127v3
2023-05-30T00:45:39
2023-03-14T17:57:06
{ "title": [ -0.05206299, -0.016143799, -0.05230713, -0.04147339, -0.003921509, 0.0032863617, -0.04559326, 0.0014839172, -0.011291504, 0.03778076, 0.006259918, -0.0029773712, -0.004585266, -0.028915405, -0.047943115, 0.005580902, -0.0236969, 0.04...
http://arxiv.org/abs/2303.08117v2
Do Transformers Parse while Predicting the Masked Word?
Haoyu Zhao, Abhishek Panigrahi, Rong Ge, Sanjeev Arora
cs.CL, cs.LG
Pre-trained language models have been shown to encode linguistic structures, e.g. dependency and constituency parse trees, in their embeddings while being trained on unsupervised loss functions like masked language modeling. Some doubts have been raised whether the models actually are doing parsing or only some computa...
http://arxiv.org/pdf/2303.08117v2
2023-10-16T03:27:53
2023-03-14T17:49:50
{ "title": [ -0.0065193176, 0.01309967, -0.010017395, -0.02104187, -0.023254395, 0.01977539, -0.040863037, -0.017654419, -0.050354004, 0.024749756, -0.051574707, 0.00017464161, 0.011741638, -0.027709961, 0.03640747, -0.036376953, -0.014190674, 0....
http://arxiv.org/abs/2303.08114v1
Simfluence: Modeling the Influence of Individual Training Examples by Simulating Training Runs
Kelvin Guu, Albert Webson, Ellie Pavlick, Lucas Dixon, Ian Tenney, Tolga Bolukbasi
cs.LG, cs.CL
Training data attribution (TDA) methods offer to trace a model's prediction on any given example back to specific influential training examples. Existing approaches do so by assigning a scalar influence score to each training example, under a simplifying assumption that influence is additive. But in reality, we observe...
http://arxiv.org/pdf/2303.08114v1
2023-03-14T17:47:25
2023-03-14T17:47:25
{ "title": [ 0.011756897, -0.023406982, -0.0124053955, -0.05041504, -0.021713257, -0.008087158, -0.030685425, -0.028823853, -0.043914795, 0.008361816, 0.004875183, 0.025787354, -0.036468506, -0.050201416, -0.003124237, -0.010238647, -0.009223938, ...
http://arxiv.org/abs/2303.09455v1
Learning Cross-lingual Visual Speech Representations
Andreas Zinonos, Alexandros Haliassos, Pingchuan Ma, Stavros Petridis, Maja Pantic
cs.CL, cs.CV, cs.LG, cs.SD, eess.AS
Cross-lingual self-supervised learning has been a growing research topic in the last few years. However, current works only explored the use of audio signals to create representations. In this work, we study cross-lingual self-supervised visual representation learning. We use the recently-proposed Raw Audio-Visual Spee...
http://arxiv.org/pdf/2303.09455v1
2023-03-14T17:05:08
2023-03-14T17:05:08
{ "title": [ -0.0002565384, 0.0027885437, -0.023422241, -0.07366943, -0.002937317, 0.0034713745, -0.05606079, 0.0037155151, 0.010307312, 0.039001465, -0.034698486, 0.01763916, 0.00032138824, -0.0067977905, -0.044525146, 0.0020580292, 0.0076141357, ...
http://arxiv.org/abs/2303.08044v1
Happy-GLL: modular, reusable and complete top-down parsers for parameterized nonterminals
L. Thomas van Binsbergen, Damian Frolich
cs.CL, cs.PL, cs.SE
Parser generators and parser combinator libraries are the most popular tools for producing parsers. Parser combinators use the host language to provide reusable components in the form of higher-order functions with parsers as parameters. Very few parser generators support this kind of reuse through abstraction and even...
http://arxiv.org/pdf/2303.08044v1
2023-03-14T16:23:23
2023-03-14T16:23:23
{ "title": [ -0.012634277, -0.01537323, -0.044525146, -0.12426758, -0.00010550022, 0.032562256, -0.107421875, 0.023834229, -0.036865234, 0.026229858, 0.0061912537, 0.016098022, 0.025344849, -0.024963379, -0.0057640076, 0.015602112, -0.027023315, ...
http://arxiv.org/abs/2303.08038v1
Progress Note Understanding -- Assessment and Plan Reasoning: Overview of the 2022 N2C2 Track 3 Shared Task
Yanjun Gao, Dmitriy Dligach, Timothy Miller, Matthew M Churpek, Ozlem Uzuner, Majid Afshar
cs.AI, cs.CL
Daily progress notes are common types in the electronic health record (EHR) where healthcare providers document the patient's daily progress and treatment plans. The EHR is designed to document all the care provided to patients, but it also enables note bloat with extraneous information that distracts from the diagnose...
http://arxiv.org/pdf/2303.08038v1
2023-03-14T16:17:55
2023-03-14T16:17:55
{ "title": [ -0.015289307, -0.007751465, -0.026397705, -0.052459717, -0.03552246, 0.015296936, -0.08441162, 0.005962372, -0.054138184, 0.02217102, -0.0206604, 0.0074691772, 0.00083732605, -0.035980225, 0.039367676, -0.025268555, 0.024795532, 0.00...
http://arxiv.org/abs/2303.08032v2
Verifying the Robustness of Automatic Credibility Assessment
Piotr Przybyła, Alexander Shvets, Horacio Saggion
cs.CL, cs.LG
Text classification methods have been widely investigated as a way to detect content of low credibility: fake news, social media bots, propaganda, etc. Quite accurate models (likely based on deep neural networks) help in moderating public electronic platforms and often cause content creators to face rejection of their ...
http://arxiv.org/pdf/2303.08032v2
2023-08-11T09:59:07
2023-03-14T16:11:47
{ "title": [ -0.00054359436, 0.025421143, -0.0335083, -0.038726807, 0.024307251, -0.0070266724, -0.01890564, -0.050109863, -0.02331543, 0.038909912, 0.007396698, -0.006542206, -0.020904541, -0.027267456, -0.055725098, -0.0019664764, -0.033081055, ...
http://arxiv.org/abs/2303.08021v2
OptBA: Optimizing Hyperparameters with the Bees Algorithm for Improved Medical Text Classification
Mai A. Shaaban, Mariam Kashkash, Maryam Alghfeli, Adham Ibrahim
cs.CL, cs.AI
One of the challenges that artificial intelligence engineers face, specifically in the field of deep learning is obtaining the optimal model hyperparameters. The search for optimal hyperparameters usually hinders the progress of solutions to real-world problems such as healthcare. To overcome this hurdle, the proposed ...
http://arxiv.org/pdf/2303.08021v2
2023-10-08T05:30:33
2023-03-14T16:04:13
{ "title": [ -0.02949524, 0.005290985, -0.029220581, -0.05206299, 0.006462097, 0.025344849, -0.042175293, -0.026397705, -0.0010147095, 0.03149414, -0.025405884, 0.022216797, 0.013793945, -0.014007568, -0.029006958, 0.00409317, -0.0096588135, 0.03...
http://arxiv.org/abs/2303.07992v3
Can ChatGPT Replace Traditional KBQA Models? An In-depth Analysis of the Question Answering Performance of the GPT LLM Family
Yiming Tan, Dehai Min, Yu Li, Wenbo Li, Nan Hu, Yongrui Chen, Guilin Qi
cs.CL
ChatGPT is a powerful large language model (LLM) that covers knowledge resources such as Wikipedia and supports natural language question answering using its own knowledge. Therefore, there is growing interest in exploring whether ChatGPT can replace traditional knowledge-based question answering (KBQA) models. Althoug...
http://arxiv.org/pdf/2303.07992v3
2023-09-20T05:25:22
2023-03-14T15:46:28
{ "title": [ -0.06390381, -0.011512756, -0.026885986, 0.010726929, -0.06188965, -0.0015993118, -0.06262207, -0.021942139, 0.00705719, 0.028274536, -0.03652954, -0.004714966, 0.0062713623, -0.021118164, -0.022918701, -0.0075798035, -0.016036987, 0...
http://arxiv.org/abs/2303.07991v1
Finding the Needle in a Haystack: Unsupervised Rationale Extraction from Long Text Classifiers
Kamil Bujel, Andrew Caines, Helen Yannakoudakis, Marek Rei
cs.CL, cs.LG
Long-sequence transformers are designed to improve the representation of longer texts by language models and their performance on downstream document-level tasks. However, not much is understood about the quality of token-level predictions in long-form models. We investigate the performance of such architectures in the...
http://arxiv.org/pdf/2303.07991v1
2023-03-14T15:45:35
2023-03-14T15:45:35
{ "title": [ 0.0014724731, 0.003484726, -0.018417358, -0.040527344, -0.00831604, -0.021697998, -0.012588501, -0.027877808, 0.013435364, 0.015823364, -0.025161743, 0.012931824, -0.021255493, -0.013916016, -0.036834717, -0.026977539, -0.030700684, ...
http://arxiv.org/abs/2303.07971v1
A Theory of Emergent In-Context Learning as Implicit Structure Induction
Michael Hahn, Navin Goyal
cs.CL, cs.LG
Scaling large language models (LLMs) leads to an emergent capacity to learn in-context from example demonstrations. Despite progress, theoretical understanding of this phenomenon remains limited. We argue that in-context learning relies on recombination of compositional operations found in natural language data. We der...
http://arxiv.org/pdf/2303.07971v1
2023-03-14T15:24:05
2023-03-14T15:24:05
{ "title": [ 0.0064811707, -0.020217896, -0.025253296, -0.04925537, -0.0014696121, -0.017318726, 0.0035305023, -0.0044670105, -0.0012178421, 0.008079529, -0.039520264, 0.014656067, -0.04095459, -0.020248413, -0.03466797, -0.012832642, -0.013916016, ...
http://arxiv.org/abs/2303.07924v1
Improving Accented Speech Recognition with Multi-Domain Training
Lucas Maison, Yannick Estève
cs.LG, cs.CL, cs.SD, eess.AS
Thanks to the rise of self-supervised learning, automatic speech recognition (ASR) systems now achieve near-human performance on a wide variety of datasets. However, they still lack generalization capability and are not robust to domain shifts like accent variations. In this work, we use speech audio representing four ...
http://arxiv.org/pdf/2303.07924v1
2023-03-14T14:10:16
2023-03-14T14:10:16
{ "title": [ 0.009140015, 0.00642395, -0.029159546, -0.047729492, 0.0067863464, -0.014015198, -0.023223877, 0.0070495605, -0.026107788, 0.0345459, -0.0055999756, 0.0048599243, 0.00092458725, -0.03427124, -0.031463623, -0.009002686, -0.015716553, ...
http://arxiv.org/abs/2303.07914v2
Adapting Offline Speech Translation Models for Streaming with Future-Aware Distillation and Inference
Biao Fu, Minpeng Liao, Kai Fan, Zhongqiang Huang, Boxing Chen, Yidong Chen, Xiaodong Shi
cs.CL
A popular approach to streaming speech translation is to employ a single offline model with a wait-k policy to support different latency requirements, which is simpler than training multiple online models with different latency constraints. However, there is a mismatch problem in using a model trained with complete utt...
http://arxiv.org/pdf/2303.07914v2
2023-10-26T11:27:24
2023-03-14T13:56:36
{ "title": [ -0.002357483, -0.0049324036, -0.033355713, -0.10205078, 0.0007843971, 0.007911682, -0.077941895, 0.03555298, 0.025787354, 0.05038452, -0.017089844, -0.038482666, 0.0413208, -0.024246216, -0.061523438, -0.0013008118, 0.0056266785, 0.0...
http://arxiv.org/abs/2303.08572v1
Distinguishing Cause from Effect on Categorical Data: The Uniform Channel Model
Mário A. T. Figueiredo, Catarina A. Oliveira
cs.LG, cs.CL, cs.IT, math.IT, 62D20
Distinguishing cause from effect using observations of a pair of random variables is a core problem in causal discovery. Most approaches proposed for this task, namely additive noise models (ANM), are only adequate for quantitative data. We propose a criterion to address the cause-effect problem with categorical variab...
http://arxiv.org/pdf/2303.08572v1
2023-03-14T13:54:11
2023-03-14T13:54:11
{ "title": [ -0.0059890747, -0.0027675629, -0.04724121, -0.09057617, 0.015068054, -0.00969696, -0.03933716, -0.0044288635, -0.013458252, 0.016815186, 0.00667572, -0.010986328, -0.008773804, -0.059326172, -0.012237549, -0.006198883, -0.007522583, ...
http://arxiv.org/abs/2303.12804v1
Features matching using natural language processing
Muhammad Danial Khilji
cs.DB, cs.CL, cs.LG
The feature matching is a basic step in matching different datasets. This article proposes shows a new hybrid model of a pretrained Natural Language Processing (NLP) based model called BERT used in parallel with a statistical model based on Jaccard similarity to measure the similarity between list of features from two ...
http://arxiv.org/pdf/2303.12804v1
2023-03-14T13:31:19
2023-03-14T13:31:19
{ "title": [ -0.036132812, -0.014190674, -0.02709961, -0.04107666, 0.014259338, 0.0073051453, -0.025939941, -0.028457642, 0.0132369995, 0.024429321, -0.020751953, -0.024505615, 0.0025100708, -0.02748108, -0.04763794, -0.0018529892, 0.0048103333, ...
http://arxiv.org/abs/2303.07895v1
The Learnability of In-Context Learning
Noam Wies, Yoav Levine, Amnon Shashua
cs.CL
In-context learning is a surprising and important phenomenon that emerged when modern language models were scaled to billions of learned parameters. Without modifying a large language model's weights, it can be tuned to perform various downstream natural language tasks simply by including concatenated training examples...
http://arxiv.org/pdf/2303.07895v1
2023-03-14T13:28:39
2023-03-14T13:28:39
{ "title": [ 0.028320312, 0.006084442, -0.015159607, -0.038909912, 0.017288208, -0.023239136, 0.011924744, -0.030380249, -0.003900528, 0.012741089, -0.042816162, -0.0017690659, -0.041992188, -0.020629883, -0.031082153, -0.02746582, 0.0071525574, ...
http://arxiv.org/abs/2303.07865v2
Geolocation Predicting of Tweets Using BERT-Based Models
Kateryna Lutsai, Christoph H. Lampert
cs.CL, cs.AI, 68T50, I.2.7
This research is aimed to solve the tweet/user geolocation prediction task and provide a flexible methodology for the geotagging of textual big data. The suggested approach implements neural networks for natural language processing (NLP) to estimate the location as coordinate pairs (longitude, latitude) and two-dimensi...
http://arxiv.org/pdf/2303.07865v2
2023-05-08T15:55:10
2023-03-14T12:56:47
{ "title": [ 0.0231781, -0.0055389404, 0.008178711, -0.04550171, -0.0045204163, 0.0048065186, -0.01663208, -0.029785156, -0.0031719208, 0.058685303, -0.017593384, 0.009971619, 0.027770996, -0.015853882, -0.047821045, 0.01966858, -0.012863159, 0.0...
http://arxiv.org/abs/2303.07833v1
X-ReCoSa: Multi-Scale Context Aggregation For Multi-Turn Dialogue Generation
Danqin Wu
cs.CL
In multi-turn dialogue generation, responses are not only related to the topic and background of the context but also related to words and phrases in the sentences of the context. However, currently widely used hierarchical dialog models solely rely on context representations from the utterance-level encoder, ignoring ...
http://arxiv.org/pdf/2303.07833v1
2023-03-14T12:15:52
2023-03-14T12:15:52
{ "title": [ -0.00774765, -0.0058937073, -0.06347656, -0.09197998, 0.015899658, 0.0013656616, -0.044433594, 0.034942627, 0.0050315857, 0.03652954, -0.021453857, 0.0069465637, -0.005130768, -0.01676941, -0.041046143, -0.012870789, -0.016967773, 0....
http://arxiv.org/abs/2303.09446v1
Controlling High-Dimensional Data With Sparse Input
Dan Andrei Iliescu, Devang Savita Ram Mohan, Tian Huey Teh, Zack Hodari
eess.AS, cs.AI, cs.CL, cs.LG
We address the problem of human-in-the-loop control for generating highly-structured data. This task is challenging because existing generative models lack an efficient interface through which users can modify the output. Users have the option to either manually explore a non-interpretable latent space, or to laborious...
http://arxiv.org/pdf/2303.09446v1
2023-03-14T09:47:23
2023-03-14T09:47:23
{ "title": [ -0.027801514, -0.011657715, -0.039489746, -0.03817749, -0.0007338524, -0.008956909, -0.012428284, -0.0057296753, 0.006259918, 0.020614624, -0.020050049, -0.0038928986, -0.022994995, -0.015350342, -0.038146973, -0.0023422241, 0.017410278,...
http://arxiv.org/abs/2303.07740v1
Efficient Image-Text Retrieval via Keyword-Guided Pre-Screening
Min Cao, Yang Bai, Jingyao Wang, Ziqiang Cao, Liqiang Nie, Min Zhang
cs.CV, cs.CL
Under the flourishing development in performance, current image-text retrieval methods suffer from $N$-related time complexity, which hinders their application in practice. Targeting at efficiency improvement, this paper presents a simple and effective keyword-guided pre-screening framework for the image-text retrieval...
http://arxiv.org/pdf/2303.07740v1
2023-03-14T09:36:42
2023-03-14T09:36:42
{ "title": [ 0.007171631, 0.009552002, 0.0007109642, -0.044067383, 0.010292053, 0.0010738373, 0.009407043, -0.038726807, 0.027267456, 0.043518066, -0.042785645, 0.032928467, -0.012901306, 0.015174866, -0.058410645, -0.00063562393, 0.0018901825, 0...
http://arxiv.org/abs/2303.07726v1
Good Neighbors Are All You Need for Chinese Grapheme-to-Phoneme Conversion
Jungjun Kim, Changjin Han, Gyuhyeon Nam, Gyeongsu Chae
cs.CL, cs.LG, eess.AS
Most Chinese Grapheme-to-Phoneme (G2P) systems employ a three-stage framework that first transforms input sequences into character embeddings, obtains linguistic information using language models, and then predicts the phonemes based on global context about the entire input sequence. However, linguistic knowledge alone...
http://arxiv.org/pdf/2303.07726v1
2023-03-14T09:15:51
2023-03-14T09:15:51
{ "title": [ 0.022201538, -0.029403687, 0.0041046143, -0.033203125, -0.042297363, -0.015266418, -0.018081665, -0.00052547455, -0.01625061, 0.055480957, -0.003332138, 0.000033557415, 0.0317688, -0.041778564, 0.026947021, 0.0026493073, -0.0056991577, ...
http://arxiv.org/abs/2303.07711v1
Improving Prosody for Cross-Speaker Style Transfer by Semi-Supervised Style Extractor and Hierarchical Modeling in Speech Synthesis
Chunyu Qiang, Peng Yang, Hao Che, Ying Zhang, Xiaorui Wang, Zhongyuan Wang
cs.SD, cs.AI, cs.CL, eess.AS
Cross-speaker style transfer in speech synthesis aims at transferring a style from source speaker to synthesized speech of a target speaker's timbre. In most previous methods, the synthesized fine-grained prosody features often represent the source speaker's average style, similar to the one-to-many problem(i.e., multi...
http://arxiv.org/pdf/2303.07711v1
2023-03-14T08:52:58
2023-03-14T08:52:58
{ "title": [ -0.023498535, 0.020111084, -0.053863525, -0.061706543, -0.0031414032, 0.0027160645, -0.070739746, 0.0103302, -0.013877869, 0.04953003, 0.026351929, 0.017227173, 0.0692749, -0.05279541, 0.010002136, 0.03390503, -0.018310547, 0.0107650...
http://arxiv.org/abs/2303.07689v1
Dual-Attention Model for Aspect-Level Sentiment Classification
Mengfei Ye
cs.CL, cs.AI
I propose a novel dual-attention model(DAM) for aspect-level sentiment classification. Many methods have been proposed, such as support vector machines for artificial design features, long short-term memory networks based on attention mechanisms, and graph neural networks based on dependency parsing. While these method...
http://arxiv.org/pdf/2303.07689v1
2023-03-14T08:04:38
2023-03-14T08:04:38
{ "title": [ 0.009063721, 0.0063591003, -0.01637268, -0.063964844, 0.008171082, 0.00166893, -0.029418945, -0.0009880066, 0.037078857, 0.0592041, -0.017364502, -0.023925781, 0.02180481, 0.0010604858, -0.08203125, 0.013412476, -0.00932312, 0.033905...
http://arxiv.org/abs/2303.07687v1
Dynamic Alignment Mask CTC: Improved Mask-CTC with Aligned Cross Entropy
Xulong Zhang, Haobin Tang, Jianzong Wang, Ning Cheng, Jian Luo, Jing Xiao
cs.SD, cs.CL, eess.AS
Because of predicting all the target tokens in parallel, the non-autoregressive models greatly improve the decoding efficiency of speech recognition compared with traditional autoregressive models. In this work, we present dynamic alignment Mask CTC, introducing two methods: (1) Aligned Cross Entropy (AXE), finding the...
http://arxiv.org/pdf/2303.07687v1
2023-03-14T08:01:21
2023-03-14T08:01:21
{ "title": [ -0.0569458, 0.0041236877, -0.08917236, -0.015670776, -0.021331787, -0.011871338, -0.054382324, -0.0017528534, 0.021759033, 0.06542969, -0.006526947, 0.0059890747, 0.068481445, -0.035369873, -0.015075684, 0.037475586, 0.01737976, 0.01...
http://arxiv.org/abs/2303.07682v1
QI-TTS: Questioning Intonation Control for Emotional Speech Synthesis
Haobin Tang, Xulong Zhang, Jianzong Wang, Ning Cheng, Jing Xiao
cs.SD, cs.CL, eess.AS
Recent expressive text to speech (TTS) models focus on synthesizing emotional speech, but some fine-grained styles such as intonation are neglected. In this paper, we propose QI-TTS which aims to better transfer and control intonation to further deliver the speaker's questioning intention while transferring emotion fro...
http://arxiv.org/pdf/2303.07682v1
2023-03-14T07:53:19
2023-03-14T07:53:19
{ "title": [ -0.014404297, -0.013160706, -0.003522873, -0.027938843, -0.021743774, 0.00068998337, -0.03062439, -0.009376526, -0.009429932, 0.043273926, -0.03390503, 0.0007662773, 0.044555664, -0.033569336, -0.034118652, 0.0025615692, -0.0061416626, ...
http://arxiv.org/abs/2303.07678v2
Query2doc: Query Expansion with Large Language Models
Liang Wang, Nan Yang, Furu Wei
cs.IR, cs.CL
This paper introduces a simple yet effective query expansion approach, denoted as query2doc, to improve both sparse and dense retrieval systems. The proposed method first generates pseudo-documents by few-shot prompting large language models (LLMs), and then expands the query with generated pseudo-documents. LLMs are t...
http://arxiv.org/pdf/2303.07678v2
2023-10-11T08:34:42
2023-03-14T07:27:30
{ "title": [ -0.02003479, -0.018875122, -0.05218506, -0.078125, 0.016204834, 0.005470276, -0.042785645, 0.008399963, 0.014976501, 0.03050232, -0.022964478, 0.0066337585, 0.0096588135, -0.011680603, -0.04019165, 0.017303467, -0.0018472672, 0.02174...
http://arxiv.org/abs/2303.07665v1
RenewNAT: Renewing Potential Translation for Non-Autoregressive Transformer
Pei Guo, Yisheng Xiao, Juntao Li, Min Zhang
cs.CL
Non-autoregressive neural machine translation (NAT) models are proposed to accelerate the inference process while maintaining relatively high performance. However, existing NAT models are difficult to achieve the desired efficiency-quality trade-off. For one thing, fully NAT models with efficient inference perform infe...
http://arxiv.org/pdf/2303.07665v1
2023-03-14T07:10:03
2023-03-14T07:10:03
{ "title": [ -0.019592285, 0.010894775, -0.023956299, -0.044647217, -0.004825592, -0.0030021667, -0.04421997, -0.03955078, -0.019638062, 0.04208374, -0.007896423, 0.0010576248, 0.02848816, -0.03225708, -0.030471802, 0.008010864, 0.0035762787, 0.0...
http://arxiv.org/abs/2303.07650v1
Cross-lingual Alzheimer's Disease detection based on paralinguistic and pre-trained features
Xuchu Chen, Yu Pu, Jinpeng Li, Wei-Qiang Zhang
cs.CL, cs.SD, eess.AS
We present our submission to the ICASSP-SPGC-2023 ADReSS-M Challenge Task, which aims to investigate which acoustic features can be generalized and transferred across languages for Alzheimer's Disease (AD) prediction. The challenge consists of two tasks: one is to classify the speech of AD patients and healthy individu...
http://arxiv.org/pdf/2303.07650v1
2023-03-14T06:34:18
2023-03-14T06:34:18
{ "title": [ 0.017837524, 0.0014648438, -0.022872925, -0.066223145, 0.0129470825, -0.008987427, -0.040374756, -0.013252258, -0.013504028, 0.03186035, -0.011169434, -0.012329102, 0.004459381, -0.028427124, -0.02670288, -0.013114929, -0.00027871132, ...
http://arxiv.org/abs/2303.07624v1
I3D: Transformer architectures with input-dependent dynamic depth for speech recognition
Yifan Peng, Jaesong Lee, Shinji Watanabe
cs.CL, cs.SD, eess.AS
Transformer-based end-to-end speech recognition has achieved great success. However, the large footprint and computational overhead make it difficult to deploy these models in some real-world applications. Model compression techniques can reduce the model size and speed up inference, but the compressed model has a fixe...
http://arxiv.org/pdf/2303.07624v1
2023-03-14T04:47:00
2023-03-14T04:47:00
{ "title": [ -0.025146484, 0.0043182373, -0.07019043, -0.06100464, 0.01524353, -0.0047798157, -0.039001465, 0.018859863, -0.0256958, 0.046661377, -0.0289917, -0.0011835098, -0.0025119781, -0.006969452, -0.05783081, -0.019241333, -0.008575439, 0.0...
http://arxiv.org/abs/2303.07616v1
The Life Cycle of Knowledge in Big Language Models: A Survey
Boxi Cao, Hongyu Lin, Xianpei Han, Le Sun
cs.CL
Knowledge plays a critical role in artificial intelligence. Recently, the extensive success of pre-trained language models (PLMs) has raised significant attention about how knowledge can be acquired, maintained, updated and used by language models. Despite the enormous amount of related studies, there still lacks a uni...
http://arxiv.org/pdf/2303.07616v1
2023-03-14T03:49:22
2023-03-14T03:49:22
{ "title": [ -0.049804688, -0.03604126, -0.05908203, -0.09802246, -0.012489319, 0.020584106, -0.07702637, 0.048828125, 0.0048332214, 0.03375244, -0.040771484, 0.012107849, -0.0073547363, -0.0011024475, -0.050323486, -0.0036773682, 0.0013990402, 0...
http://arxiv.org/abs/2303.07610v1
Exploring ChatGPT's Ability to Rank Content: A Preliminary Study on Consistency with Human Preferences
Yunjie Ji, Yan Gong, Yiping Peng, Chao Ni, Peiyan Sun, Dongyu Pan, Baochang Ma, Xiangang Li
cs.CL
As a natural language assistant, ChatGPT is capable of performing various tasks, including but not limited to article generation, code completion, and data analysis. Furthermore, ChatGPT has consistently demonstrated a remarkable level of accuracy and reliability in terms of content evaluation, exhibiting the capabilit...
http://arxiv.org/pdf/2303.07610v1
2023-03-14T03:13:02
2023-03-14T03:13:02
{ "title": [ 0.011833191, 0.006664276, -0.015151978, 0.0059509277, 0.010978699, -0.017684937, 0.017425537, -0.04171753, -0.01914978, 0.019699097, -0.0134887695, 0.020095825, -0.025100708, -0.010559082, -0.0256958, 0.0017881393, -0.043792725, 0.03...
http://arxiv.org/abs/2303.07585v1
Input-length-shortening and text generation via attention values
Neşet Özkan Tan, Alex Yuxuan Peng, Joshua Bensemann, Qiming Bao, Tim Hartill, Mark Gahegan, Michael Witbrock
cs.CL
Identifying words that impact a task's performance more than others is a challenge in natural language processing. Transformers models have recently addressed this issue by incorporating an attention mechanism that assigns greater attention (i.e., relevance) scores to some words than others. Because of the attention me...
http://arxiv.org/pdf/2303.07585v1
2023-03-14T02:11:24
2023-03-14T02:11:24
{ "title": [ -0.022735596, -0.007865906, -0.03427124, -0.02607727, -0.0211792, -0.008430481, -0.04611206, -0.018615723, -0.00014662743, 0.048553467, -0.00090026855, 0.024383545, -0.0018892288, -0.0095825195, -0.04925537, 0.01007843, -0.027526855, ...
http://arxiv.org/abs/2303.07576v1
Diffusion Models in NLP: A Survey
Yuansong Zhu, Yu Zhao
cs.CL, cs.AI
Diffusion models have become a powerful family of deep generative models, with record-breaking performance in many applications. This paper first gives an overview and derivation of the basic theory of diffusion models, then reviews the research results of diffusion models in the field of natural language processing, f...
http://arxiv.org/pdf/2303.07576v1
2023-03-14T01:53:49
2023-03-14T01:53:49
{ "title": [ -0.015060425, 0.0020713806, -0.034332275, -0.033721924, 0.0078048706, 0.0041007996, -0.009643555, -0.0071640015, 0.0110321045, 0.008964539, -0.024612427, 0.010360718, -0.015098572, -0.013511658, -0.042877197, 0.0053710938, -0.033416748, ...
http://arxiv.org/abs/2303.07522v2
Audio Visual Language Maps for Robot Navigation
Chenguang Huang, Oier Mees, Andy Zeng, Wolfram Burgard
cs.RO, cs.AI, cs.CL, cs.CV, cs.LG
While interacting in the world is a multi-sensory experience, many robots continue to predominantly rely on visual perception to map and navigate in their environments. In this work, we propose Audio-Visual-Language Maps (AVLMaps), a unified 3D spatial map representation for storing cross-modal information from audio, ...
http://arxiv.org/pdf/2303.07522v2
2023-03-27T15:10:51
2023-03-13T23:17:51
{ "title": [ -0.017227173, -0.027557373, 0.0033473969, -0.015304565, -0.037506104, -0.011558533, -0.013587952, -0.014808655, -0.012634277, 0.049468994, -0.035705566, -0.010246277, -0.008132935, -0.014640808, -0.029006958, -0.016235352, 0.024429321, ...
http://arxiv.org/abs/2303.07519v3
Architext: Language-Driven Generative Architecture Design
Theodoros Galanos, Antonios Liapis, Georgios N. Yannakakis
cs.CL, cs.LG
Architectural design is a highly complex practice that involves a wide diversity of disciplines, technologies, proprietary design software, expertise, and an almost infinite number of constraints, across a vast array of design tasks. Enabling intuitive, accessible, and scalable design processes is an important step tow...
http://arxiv.org/pdf/2303.07519v3
2023-05-03T09:29:05
2023-03-13T23:11:05
{ "title": [ -0.03744507, -0.028945923, -0.045684814, -0.047729492, -0.004119873, 0.0024356842, -0.045776367, -0.006980896, 0.0034122467, 0.037506104, -0.015579224, 0.005718231, -0.011116028, -0.021408081, -0.035369873, -0.001657486, -0.0020961761, ...
http://arxiv.org/abs/2303.07457v1
AMOM: Adaptive Masking over Masking for Conditional Masked Language Model
Yisheng Xiao, Ruiyang Xu, Lijun Wu, Juntao Li, Tao Qin, Yan-Tie Liu, Min Zhang
cs.CL, cs.AI
Transformer-based autoregressive (AR) methods have achieved appealing performance for varied sequence-to-sequence generation tasks, e.g., neural machine translation, summarization, and code generation, but suffer from low inference efficiency. To speed up the inference stage, many non-autoregressive (NAR) strategies ha...
http://arxiv.org/pdf/2303.07457v1
2023-03-13T20:34:56
2023-03-13T20:34:56
{ "title": [ -0.052612305, 0.00089645386, -0.077697754, -0.06689453, 0.00932312, 0.004512787, -0.08074951, 0.01689148, -0.00087308884, 0.02229309, -0.015304565, -0.0028514862, 0.029525757, -0.030334473, -0.020828247, 0.014060974, -0.020645142, 0....
http://arxiv.org/abs/2303.07320v1
Model-tuning Via Prompts Makes NLP Models Adversarially Robust
Mrigank Raman, Pratyush Maini, J. Zico Kolter, Zachary C. Lipton, Danish Pruthi
cs.CL, cs.LG
In recent years, NLP practitioners have converged on the following practice: (i) import an off-the-shelf pretrained (masked) language model; (ii) append a multilayer perceptron atop the CLS token's hidden representation (with randomly initialized weights); and (iii) fine-tune the entire model on a downstream task (MLP)...
http://arxiv.org/pdf/2303.07320v1
2023-03-13T17:41:57
2023-03-13T17:41:57
{ "title": [ 0.0039520264, -0.0038433075, -0.013412476, -0.030517578, -0.03842163, -0.002910614, -0.061767578, -0.03894043, -0.023895264, 0.0047073364, -0.039215088, 0.0025787354, -0.00049877167, -0.029037476, -0.004131317, -0.028121948, -0.01084137,...
http://arxiv.org/abs/2303.07295v1
Meet in the Middle: A New Pre-training Paradigm
Anh Nguyen, Nikos Karampatziakis, Weizhu Chen
cs.CL, cs.LG
Most language models (LMs) are trained and applied in an autoregressive left-to-right fashion, assuming that the next token only depends on the preceding ones. However, this assumption ignores the potential benefits of using the full sequence information during training, and the possibility of having context from both ...
http://arxiv.org/pdf/2303.07295v1
2023-03-13T17:17:11
2023-03-13T17:17:11
{ "title": [ 0.01763916, -0.011688232, -0.027389526, -0.079956055, 0.0072898865, -0.015808105, -0.017562866, -0.00497818, -0.05822754, 0.014373779, 0.0066566467, 0.010795593, -0.045288086, -0.054656982, -0.004016876, -0.013839722, -0.0065345764, ...
http://arxiv.org/abs/2303.07292v1
Transformer-based approaches to Sentiment Detection
Olumide Ebenezer Ojo, Hoang Thang Ta, Alexander Gelbukh, Hiram Calvo, Olaronke Oluwayemisi Adebanji, Grigori Sidorov
cs.CL
The use of transfer learning methods is largely responsible for the present breakthrough in Natural Learning Processing (NLP) tasks across multiple domains. In order to solve the problem of sentiment detection, we examined the performance of four different types of well-known state-of-the-art transformer models for tex...
http://arxiv.org/pdf/2303.07292v1
2023-03-13T17:12:03
2023-03-13T17:12:03
{ "title": [ 0.004055023, 0.0071907043, -0.02178955, -0.029907227, 0.022659302, -0.012214661, -0.007217407, -0.031433105, 0.016860962, 0.03540039, -0.044067383, -0.0068244934, 0.007270813, -0.0059051514, -0.056518555, -0.00023972988, -0.013656616, ...
http://arxiv.org/abs/2303.07274v4
Breaking Common Sense: WHOOPS! A Vision-and-Language Benchmark of Synthetic and Compositional Images
Nitzan Bitton-Guetta, Yonatan Bitton, Jack Hessel, Ludwig Schmidt, Yuval Elovici, Gabriel Stanovsky, Roy Schwartz
cs.CV, cs.AI, cs.CL
Weird, unusual, and uncanny images pique the curiosity of observers because they challenge commonsense. For example, an image released during the 2022 world cup depicts the famous soccer stars Lionel Messi and Cristiano Ronaldo playing chess, which playfully violates our expectation that their competition should occur ...
http://arxiv.org/pdf/2303.07274v4
2023-08-12T22:37:31
2023-03-13T16:49:43
{ "title": [ -0.010513306, 0.0044403076, -0.015602112, -0.036590576, -0.0012407303, -0.018859863, -0.033996582, -0.039855957, 0.021865845, 0.028457642, -0.02178955, 0.012382507, -0.007320404, -0.011947632, -0.04043579, 0.011917114, 0.011276245, 0...
http://arxiv.org/abs/2303.07247v2
Are Models Trained on Indian Legal Data Fair?
Sahil Girhepuje, Anmol Goel, Gokul S Krishnan, Shreya Goyal, Satyendra Pandey, Ponnurangam Kumaraguru, Balaraman Ravindran
cs.CL, cs.CY
Recent advances and applications of language technology and artificial intelligence have enabled much success across multiple domains like law, medical and mental health. AI-based Language Models, like Judgement Prediction, have recently been proposed for the legal sector. However, these models are strife with encoded ...
http://arxiv.org/pdf/2303.07247v2
2023-03-14T17:40:21
2023-03-13T16:20:33
{ "title": [ -0.017532349, -0.011123657, -0.047210693, -0.015380859, -0.032989502, -0.0064849854, -0.040496826, 0.009811401, -0.023269653, 0.00003552437, -0.05758667, -0.00730896, -0.036010742, -0.024856567, 0.022735596, -0.04788208, -0.0155181885, ...
http://arxiv.org/abs/2303.07240v1
PMC-CLIP: Contrastive Language-Image Pre-training using Biomedical Documents
Weixiong Lin, Ziheng Zhao, Xiaoman Zhang, Chaoyi Wu, Ya Zhang, Yanfeng Wang, Weidi Xie
cs.CV, cs.CL, cs.LG, cs.MM
Foundation models trained on large-scale dataset gain a recent surge in CV and NLP. In contrast, development in biomedical domain lags far behind due to data scarcity. To address this issue, we build and release PMC-OA, a biomedical dataset with 1.6M image-caption pairs collected from PubMedCentral's OpenAccess subset,...
http://arxiv.org/pdf/2303.07240v1
2023-03-13T16:13:16
2023-03-13T16:13:16
{ "title": [ 0.0029354095, -0.014289856, -0.0033683777, -0.0413208, -0.021621704, -0.0063209534, -0.02909851, -0.037200928, 0.012023926, 0.021469116, -0.008415222, 0.014038086, 0.00058078766, -0.030807495, -0.020889282, 0.00012385845, 0.009063721, ...
http://arxiv.org/abs/2303.07226v1
Scaling Vision-Language Models with Sparse Mixture of Experts
Sheng Shen, Zhewei Yao, Chunyuan Li, Trevor Darrell, Kurt Keutzer, Yuxiong He
cs.CV, cs.CL
The field of natural language processing (NLP) has made significant strides in recent years, particularly in the development of large-scale vision-language models (VLMs). These models aim to bridge the gap between text and visual information, enabling a more comprehensive understanding of multimedia data. However, as t...
http://arxiv.org/pdf/2303.07226v1
2023-03-13T16:00:31
2023-03-13T16:00:31
{ "title": [ -0.044006348, -0.0033950806, -0.05569458, -0.038513184, -0.01424408, -0.014511108, -0.04498291, 0.017700195, 0.022628784, 0.029403687, -0.017440796, -0.011161804, 0.0062446594, -0.013595581, -0.04852295, 0.00038862228, 0.029052734, 0...
http://arxiv.org/abs/2303.07196v1
A Comprehensive Empirical Evaluation of Existing Word Embedding Approaches
Obaidullah Zaland, Muhammad Abulaish, Mohd. Fazil
cs.CL, cs.NE
Vector-based word representations help countless Natural Language Processing (NLP) tasks capture both semantic and syntactic regularities of the language. In this paper, we present the characteristics of existing word embedding approaches and analyze them with regards to many classification tasks. We categorize the met...
http://arxiv.org/pdf/2303.07196v1
2023-03-13T15:34:19
2023-03-13T15:34:19
{ "title": [ -0.008338928, -0.007320404, -0.026168823, -0.03616333, 0.013053894, -0.0031089783, -0.024658203, -0.013122559, 0.023284912, 0.021377563, -0.018920898, 0.017745972, -0.0008368492, -0.0107803345, -0.033721924, 0.02178955, -0.0063476562, ...
http://arxiv.org/abs/2303.07146v1
NeuroQL: A Neuro-Symbolic Language and Dataset for Inter-Subjective Reasoning
Nick Papoulias
cs.PL, cs.AI, cs.CL, cs.SE
We present a new AI task and baseline solution for Inter-Subjective Reasoning. We define inter-subjective information, to be a mixture of objective and subjective information possibly shared by different parties. Examples may include commodities and their objective properties as reported by IR (Information Retrieval) s...
http://arxiv.org/pdf/2303.07146v1
2023-03-13T14:16:59
2023-03-13T14:16:59
{ "title": [ 0.00059986115, -0.01272583, -0.017623901, -0.040985107, 0.007671356, -0.0015363693, -0.015731812, -0.026229858, 0.008834839, 0.021484375, -0.013542175, -0.032989502, 0.008094788, -0.02609253, -0.04650879, -0.011161804, -0.008613586, ...
http://arxiv.org/abs/2303.07142v3
Large Language Models in the Workplace: A Case Study on Prompt Engineering for Job Type Classification
Benjamin Clavié, Alexandru Ciceu, Frederick Naylor, Guillaume Soulié, Thomas Brightwell
cs.CL
This case study investigates the task of job classification in a real-world setting, where the goal is to determine whether an English-language job posting is appropriate for a graduate or entry-level position. We explore multiple approaches to text classification, including supervised approaches such as traditional mo...
http://arxiv.org/pdf/2303.07142v3
2023-04-18T09:58:53
2023-03-13T14:09:53
{ "title": [ -0.020370483, -0.016937256, -0.045715332, -0.033233643, 0.0010957718, -0.024108887, -0.026321411, 0.0074539185, -0.013687134, 0.016555786, -0.0003836155, -0.017089844, 0.006511688, -0.04248047, -0.02067566, -0.008163452, -0.0102005005, ...
http://arxiv.org/abs/2303.07069v1
Generating multiple-choice questions for medical question answering with distractors and cue-masking
Damien Sileo, Kanimozhi Uma, Marie-Francine Moens
cs.CL, H.4; H.5; I.2
Medical multiple-choice question answering (MCQA) is particularly difficult. Questions may describe patient symptoms and ask for the correct diagnosis, which requires domain knowledge and complex reasoning. Standard language modeling pretraining alone is not sufficient to achieve the best results. \citet{jin2020disease...
http://arxiv.org/pdf/2303.07069v1
2023-03-13T12:45:01
2023-03-13T12:45:01
{ "title": [ -0.0074157715, 0.018798828, -0.028823853, -0.033843994, 0.0064430237, 0.0022640228, -0.009880066, -0.026412964, 0.010986328, 0.015274048, -0.014717102, 0.0079193115, -0.005153656, -0.022354126, -0.033843994, -0.012313843, -0.010353088, ...
http://arxiv.org/abs/2303.07024v1
Addressing Biases in the Texts using an End-to-End Pipeline Approach
Shaina Raza, Syed Raza Bashir, Sneha, Urooj Qamar
cs.CL, cs.AI
The concept of fairness is gaining popularity in academia and industry. Social media is especially vulnerable to media biases and toxic language and comments. We propose a fair ML pipeline that takes a text as input and determines whether it contains biases and toxic content. Then, based on pre-trained word embeddings,...
http://arxiv.org/pdf/2303.07024v1
2023-03-13T11:41:28
2023-03-13T11:41:28
{ "title": [ 0.05596924, -0.00217247, 0.0058898926, -0.045135498, -0.014953613, -0.013763428, -0.003036499, -0.007785797, -0.037384033, 0.037322998, -0.013038635, 0.008117676, 0.01309967, -0.041137695, 0.014595032, -0.0309906, -0.019317627, 0.051...
http://arxiv.org/abs/2304.02481v2
Rediscovering Hashed Random Projections for Efficient Quantization of Contextualized Sentence Embeddings
Ulf A. Hamster, Ji-Ung Lee, Alexander Geyken, Iryna Gurevych
cs.CL
Training and inference on edge devices often requires an efficient setup due to computational limitations. While pre-computing data representations and caching them on a server can mitigate extensive edge device computation, this leads to two challenges. First, the amount of storage required on the server that scales l...
http://arxiv.org/pdf/2304.02481v2
2023-05-16T05:39:46
2023-03-13T10:53:00
{ "title": [ 0.0057258606, 0.0067214966, -0.01637268, -0.040283203, 0.0034561157, -0.005706787, -0.023406982, -0.026245117, 0.00051164627, 0.051757812, -0.005138397, 0.0033092499, 0.0071983337, -0.00756073, -0.047851562, 0.0038833618, -0.00019001961,...
http://arxiv.org/abs/2303.06944v1
A Human Subject Study of Named Entity Recognition (NER) in Conversational Music Recommendation Queries
Elena V. Epure, Romain Hennequin
cs.CL
We conducted a human subject study of named entity recognition on a noisy corpus of conversational music recommendation queries, with many irregular and novel named entities. We evaluated the human NER linguistic behaviour in these challenging conditions and compared it with the most common NER systems nowadays, fine-t...
http://arxiv.org/pdf/2303.06944v1
2023-03-13T09:22:48
2023-03-13T09:22:48
{ "title": [ 0.01939392, -0.00409317, -0.05026245, -0.04498291, 0.01675415, -0.03466797, -0.0020313263, 0.025894165, -0.016326904, 0.03869629, -0.04434204, 0.03363037, 0.02835083, -0.0115737915, -0.022857666, -0.002784729, -0.022018433, 0.0293273...
http://arxiv.org/abs/2303.11438v1
Minimizing Fuzzy Interpretations in Fuzzy Description Logics by Using Crisp Bisimulations
Linh Anh Nguyen
cs.AI, cs.CL
The problem of minimizing finite fuzzy interpretations in fuzzy description logics (FDLs) is worth studying. For example, the structure of a fuzzy/weighted social network can be treated as a fuzzy interpretation in FDLs, where actors are individuals and actions are roles. Minimizing the structure of a fuzzy/weighted so...
http://arxiv.org/pdf/2303.11438v1
2023-03-13T09:04:17
2023-03-13T09:04:17
{ "title": [ -0.023040771, 0.018478394, -0.03463745, -0.07562256, 0.024978638, 0.03010559, -0.0552063, -0.0073928833, -0.004371643, 0.032562256, -0.004295349, -0.004722595, -0.002368927, -0.021865845, -0.042510986, 0.0018520355, -0.009803772, 0.0...
http://arxiv.org/abs/2303.06904v1
Contextually-rich human affect perception using multimodal scene information
Digbalay Bose, Rajat Hebbar, Krishna Somandepalli, Shrikanth Narayanan
cs.CV, cs.AI, cs.CL
The process of human affect understanding involves the ability to infer person specific emotional states from various sources including images, speech, and language. Affect perception from images has predominantly focused on expressions extracted from salient face crops. However, emotions perceived by humans rely on mu...
http://arxiv.org/pdf/2303.06904v1
2023-03-13T07:46:41
2023-03-13T07:46:41
{ "title": [ 0.013252258, -0.0010986328, -0.0362854, -0.06137085, 0.0049209595, -0.027938843, -0.018295288, -0.018081665, 0.022750854, 0.026184082, -0.025863647, -0.018432617, -0.0014696121, -0.019088745, -0.05291748, -0.012588501, -0.0033054352, ...
http://arxiv.org/abs/2303.07354v1
MetaTroll: Few-shot Detection of State-Sponsored Trolls with Transformer Adapters
Lin Tian, Xiuzhen Zhang, Jey Han Lau
cs.CL
State-sponsored trolls are the main actors of influence campaigns on social media and automatic troll detection is important to combat misinformation at scale. Existing troll detection models are developed based on training data for known campaigns (e.g.\ the influence campaign by Russia's Internet Research Agency on t...
http://arxiv.org/pdf/2303.07354v1
2023-03-13T06:39:38
2023-03-13T06:39:38
{ "title": [ 0.025360107, 0.0033073425, 0.0023403168, -0.03048706, 0.007091522, -0.007144928, -0.005970001, -0.04699707, -0.018707275, 0.011383057, -0.03137207, 0.0051498413, 0.0069618225, -0.022460938, -0.019210815, -0.018493652, -0.023895264, 0...
http://arxiv.org/abs/2303.06878v1
The System Description of dun_oscar team for The ICPR MSR Challenge
Binbin Du, Rui Deng, Yingxin Zhang
cs.CL, cs.AI
This paper introduces the system submitted by dun_oscar team for the ICPR MSR Challenge. Three subsystems for task1-task3 are descripted respectively. In task1, we develop a visual system which includes a OCR model, a text tracker, and a NLP classifier for distinguishing subtitles and non-subtitles. In task2, we employ...
http://arxiv.org/pdf/2303.06878v1
2023-03-13T05:53:42
2023-03-13T05:53:42
{ "title": [ -0.00655365, -0.0043258667, -0.035308838, -0.035095215, 0.0017948151, -0.0149002075, -0.028015137, -0.0034236908, -0.054351807, 0.05239868, -0.007850647, 0.012283325, -0.010635376, -0.037109375, -0.0055732727, -0.035614014, -0.0002803802...
http://arxiv.org/abs/2303.06854v1
Robust Contrastive Language-Image Pretraining against Adversarial Attacks
Wenhan Yang, Baharan Mirzasoleiman
cs.CV, cs.CL, cs.CR, cs.LG
Contrastive vision-language representation learning has achieved state-of-the-art performance for zero-shot classification, by learning from millions of image-caption pairs crawled from the internet. However, the massive data that powers large multimodal models such as CLIP, makes them extremely vulnerable to various t...
http://arxiv.org/pdf/2303.06854v1
2023-03-13T04:49:46
2023-03-13T04:49:46
{ "title": [ 0.035491943, 0.01209259, -0.003698349, -0.038208008, 0.021240234, -0.021133423, 0.00598526, -0.039215088, -0.01474762, 0.029342651, 0.0061912537, 0.04168701, -0.029342651, -0.022094727, -0.02633667, -0.0073280334, 0.0021762848, 0.033...
http://arxiv.org/abs/2303.06841v3
Learning Transductions and Alignments with RNN Seq2seq Models
Zhengxiang Wang
cs.CL
The paper studies the capabilities of Recurrent-Neural-Network sequence to sequence (RNN seq2seq) models in learning four transduction tasks: identity, reversal, total reduplication, and quadratic copying. These transductions are traditionally well studied under finite state transducers and attributed with increasing c...
http://arxiv.org/pdf/2303.06841v3
2023-06-25T15:06:10
2023-03-13T04:15:33
{ "title": [ -0.051696777, 0.002948761, -0.06298828, -0.056549072, -0.0054397583, -0.0038890839, -0.07775879, 0.0008506775, -0.01701355, 0.030288696, -0.014572144, -0.009506226, -0.014923096, -0.032562256, -0.035339355, -0.0019216537, -0.00016140938,...
http://arxiv.org/abs/2303.06806v1
Neural Diarization with Non-autoregressive Intermediate Attractors
Yusuke Fujita, Tatsuya Komatsu, Robin Scheibler, Yusuke Kida, Tetsuji Ogawa
eess.AS, cs.CL, cs.SD
End-to-end neural diarization (EEND) with encoder-decoder-based attractors (EDA) is a promising method to handle the whole speaker diarization problem simultaneously with a single neural network. While the EEND model can produce all frame-level speaker labels simultaneously, it disregards output label dependency. In th...
http://arxiv.org/pdf/2303.06806v1
2023-03-13T01:28:55
2023-03-13T01:28:55
{ "title": [ 0.032958984, 0.040008545, -0.0035896301, -0.024887085, 0.014190674, -0.022216797, 0.024032593, -0.05279541, -0.018966675, 0.061431885, -0.036254883, 0.019210815, 0.0069847107, 0.0030269623, -0.053375244, -0.00819397, 0.015411377, 0.0...
http://arxiv.org/abs/2303.06791v2
Beyond Single Items: Exploring User Preferences in Item Sets with the Conversational Playlist Curation Dataset
Arun Tejasvi Chaganty, Megan Leszczynski, Shu Zhang, Ravi Ganti, Krisztian Balog, Filip Radlinski
cs.IR, cs.CL
Users in consumption domains, like music, are often able to more efficiently provide preferences over a set of items (e.g. a playlist or radio) than over single items (e.g. songs). Unfortunately, this is an underexplored area of research, with most existing recommendation systems limited to understanding preferences ov...
http://arxiv.org/pdf/2303.06791v2
2023-05-05T20:31:45
2023-03-13T00:39:04
{ "title": [ 0.009162903, -0.0009493828, -0.053619385, -0.023864746, 0.0028209686, -0.040740967, -0.006416321, -0.002998352, 0.016921997, 0.028701782, -0.037231445, 0.007293701, -0.02192688, -0.009002686, -0.053588867, -0.00969696, -0.022872925, ...
http://arxiv.org/abs/2303.17513v1
Improving the Diproche CNL through autoformalization via GPT-3
Merlin Carl
cs.CL, cs.LO
The Diproche system is an automated proof checker for texts written in a controlled fragment of German, designed for didactical applications in classes introducing students to proofs for the first time. The first version of the system used a controlled natural language for which a Prolog formalization routine was writt...
http://arxiv.org/pdf/2303.17513v1
2023-03-12T20:11:25
2023-03-12T20:11:25
{ "title": [ -0.0014247894, 0.02859497, -0.068725586, -0.06021118, 0.0073165894, -0.0029182434, -0.06945801, -0.005428314, -0.021102905, 0.051757812, 0.018569946, -0.022445679, 0.029174805, -0.048736572, -0.0003399849, -0.004261017, -0.002243042, ...
http://arxiv.org/abs/2303.06675v1
LUKE-Graph: A Transformer-based Approach with Gated Relational Graph Attention for Cloze-style Reading Comprehension
Shima Foolad, Kourosh Kiani
cs.CL, cs.LG
Incorporating prior knowledge can improve existing pre-training models in cloze-style machine reading and has become a new trend in recent studies. Notably, most of the existing models have integrated external knowledge graphs (KG) and transformer-based models, such as BERT into a unified data structure. However, selec...
http://arxiv.org/pdf/2303.06675v1
2023-03-12T14:31:44
2023-03-12T14:31:44
{ "title": [ -0.046875, -0.0008072853, -0.044891357, -0.046081543, -0.035186768, 0.009399414, -0.07318115, 0.0019931793, -0.0060310364, 0.054229736, 0.004886627, 0.0065994263, 0.02180481, -0.008804321, -0.046081543, 0.002445221, -0.0158844, 0.038...