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/2305.06535v1
KGA: A General Machine Unlearning Framework Based on Knowledge Gap Alignment
Lingzhi Wang, Tong Chen, Wei Yuan, Xingshan Zeng, Kam-Fai Wong, Hongzhi Yin
cs.CL
Recent legislation of the "right to be forgotten" has led to the interest in machine unlearning, where the learned models are endowed with the function to forget information about specific training instances as if they have never existed in the training set. Previous work mainly focuses on computer vision scenarios and...
http://arxiv.org/pdf/2305.06535v1
2023-05-11T02:44:29
2023-05-11T02:44:29
{ "title": [ -0.033935547, -0.011657715, -0.04888916, -0.018569946, -0.021362305, -0.006843567, -0.04159546, -0.025421143, 0.031341553, 0.04107666, -0.037902832, 0.016967773, 0.003955841, -0.008628845, -0.05065918, 0.0050086975, 0.0018796921, 0.0...
http://arxiv.org/abs/2305.06530v1
How Good are Commercial Large Language Models on African Languages?
Jessica Ojo, Kelechi Ogueji
cs.CL, cs.AI, cs.LG
Recent advancements in Natural Language Processing (NLP) has led to the proliferation of large pretrained language models. These models have been shown to yield good performance, using in-context learning, even on unseen tasks and languages. They have also been exposed as commercial APIs as a form of language-model-as-...
http://arxiv.org/pdf/2305.06530v1
2023-05-11T02:29:53
2023-05-11T02:29:53
{ "title": [ -0.042816162, -0.02571106, 0.008636475, -0.014076233, -0.087524414, 0.036315918, -0.0881958, -0.010643005, -0.026275635, 0.020187378, 0.0034866333, -0.014122009, 0.040771484, -0.056488037, 0.03189087, 0.020492554, 0.021331787, 0.0102...
http://arxiv.org/abs/2305.06522v1
Randomized Smoothing with Masked Inference for Adversarially Robust Text Classifications
Han Cheol Moon, Shafiq Joty, Ruochen Zhao, Megh Thakkar, Xu Chi
cs.CL, cs.AI
Large-scale pre-trained language models have shown outstanding performance in a variety of NLP tasks. However, they are also known to be significantly brittle against specifically crafted adversarial examples, leading to increasing interest in probing the adversarial robustness of NLP systems. We introduce RSMI, a nove...
http://arxiv.org/pdf/2305.06522v1
2023-05-11T01:50:16
2023-05-11T01:50:16
{ "title": [ -0.02748108, 0.030090332, -0.05859375, -0.02772522, 0.0317688, -0.0014219284, -0.017089844, -0.028015137, -0.019180298, 0.006843567, -0.0066490173, 0.041015625, -0.016113281, -0.025268555, -0.013160706, -0.018966675, -0.02104187, 0.0...
http://arxiv.org/abs/2305.06485v1
Multimodal Contextualized Plan Prediction for Embodied Task Completion
Mert İnan, Aishwarya Padmakumar, Spandana Gella, Patrick Lange, Dilek Hakkani-Tur
cs.RO, cs.AI, cs.CL, cs.HC
Task planning is an important component of traditional robotics systems enabling robots to compose fine grained skills to perform more complex tasks. Recent work building systems for translating natural language to executable actions for task completion in simulated embodied agents is focused on directly predicting low...
http://arxiv.org/pdf/2305.06485v1
2023-05-10T22:29:12
2023-05-10T22:29:12
{ "title": [ -0.014305115, 0.009048462, -0.03527832, -0.064697266, -0.006362915, -0.0010633469, -0.053619385, -0.0064468384, -0.016677856, 0.048095703, -0.02432251, 0.018203735, 0.033203125, -0.0149002075, -0.0069122314, -0.0019292831, -0.0031585693,...
http://arxiv.org/abs/2305.06472v2
ChatGPT-Like Large-Scale Foundation Models for Prognostics and Health Management: A Survey and Roadmaps
Yan-Fu Li, Huan Wang, Muxia Sun
cs.LG, cs.AI, cs.CL
Prognostics and health management (PHM) technology plays a critical role in industrial production and equipment maintenance by identifying and predicting possible equipment failures and damages, thereby allowing necessary maintenance measures to be taken to enhance equipment service life and reliability while reducing ...
http://arxiv.org/pdf/2305.06472v2
2023-05-12T10:41:35
2023-05-10T21:37:44
{ "title": [ -0.0079574585, -0.007232666, -0.040649414, -0.085998535, 0.007347107, 0.0022010803, -0.06359863, 0.010810852, -0.011100769, 0.019500732, -0.000036895275, 0.024719238, -0.01889038, -0.024734497, -0.021087646, 0.0077705383, -0.00068473816,...
http://arxiv.org/abs/2305.06434v1
Word Grounded Graph Convolutional Network
Zhibin Lu, Qianqian Xie, Benyou Wang, Jian-yun Nie
cs.CL, cs.LG
Graph Convolutional Networks (GCNs) have shown strong performance in learning text representations for various tasks such as text classification, due to its expressive power in modeling graph structure data (e.g., a literature citation network). Most existing GCNs are limited to deal with documents included in a pre-de...
http://arxiv.org/pdf/2305.06434v1
2023-05-10T19:56:55
2023-05-10T19:56:55
{ "title": [ -0.022583008, 0.00080919266, -0.062683105, -0.0501709, -0.018218994, -0.022857666, -0.057678223, 0.013938904, 0.006248474, 0.041290283, -0.03564453, 0.008705139, 0.03201294, -0.034942627, 0.00011354685, 0.014701843, 0.007820129, 0.05...
http://arxiv.org/abs/2305.06429v1
Mispronunciation Detection of Basic Quranic Recitation Rules using Deep Learning
Ahmad Al Harere, Khloud Al Jallad
cs.SD, cs.AI, cs.CL, cs.LG, eess.AS
In Islam, readers must apply a set of pronunciation rules called Tajweed rules to recite the Quran in the same way that the angel Jibrael taught the Prophet, Muhammad. The traditional process of learning the correct application of these rules requires a human who must have a license and great experience to detect mispr...
http://arxiv.org/pdf/2305.06429v1
2023-05-10T19:31:25
2023-05-10T19:31:25
{ "title": [ -0.0050582886, 0.0131073, -0.07159424, -0.0423584, 0.005153656, -0.02079773, -0.017990112, 0.007194519, 0.005794525, 0.04336548, -0.032989502, -0.00059604645, 0.029968262, -0.03768921, -0.026168823, -0.0027484894, -0.0025463104, 0.04...
http://arxiv.org/abs/2305.06424v2
Bot or Human? Detecting ChatGPT Imposters with A Single Question
Hong Wang, Xuan Luo, Weizhi Wang, Xifeng Yan
cs.CL
Large language models like ChatGPT have recently demonstrated impressive capabilities in natural language understanding and generation, enabling various applications including translation, essay writing, and chit-chatting. However, there is a concern that they can be misused for malicious purposes, such as fraud or den...
http://arxiv.org/pdf/2305.06424v2
2023-05-16T17:59:17
2023-05-10T19:09:24
{ "title": [ 0.002216339, -0.0048065186, -0.01953125, -0.021972656, 0.010925293, -0.007003784, 0.013008118, -0.023803711, -0.024368286, 0.020523071, -0.015777588, 0.027770996, -0.01878357, -0.01361084, -0.023208618, -0.0057868958, -0.03060913, 0....
http://arxiv.org/abs/2305.06416v1
A Method to Automate the Discharge Summary Hospital Course for Neurology Patients
Vince C. Hartman, Sanika S. Bapat, Mark G. Weiner, Babak B. Navi, Evan T. Sholle, Thomas R. Campion, Jr
cs.CL, cs.LG
Generation of automated clinical notes have been posited as a strategy to mitigate physician burnout. In particular, an automated narrative summary of a patient's hospital stay could supplement the hospital course section of the discharge summary that inpatient physicians document in electronic health record (EHR) syst...
http://arxiv.org/pdf/2305.06416v1
2023-05-10T18:53:51
2023-05-10T18:53:51
{ "title": [ 0.027908325, -0.00894928, -0.03967285, -0.066589355, -0.0141067505, -0.01272583, -0.040618896, 0.00623703, -0.01914978, 0.04888916, -0.044769287, 0.024978638, 0.017089844, -0.014770508, 0.0038261414, -0.014678955, 0.015579224, -0.008...
http://arxiv.org/abs/2305.07034v1
Quran Recitation Recognition using End-to-End Deep Learning
Ahmad Al Harere, Khloud Al Jallad
eess.AS, cs.AI, cs.CL, cs.LG, cs.SD
The Quran is the holy scripture of Islam, and its recitation is an important aspect of the religion. Recognizing the recitation of the Holy Quran automatically is a challenging task due to its unique rules that are not applied in normal speaking speeches. A lot of research has been done in this domain, but previous wor...
http://arxiv.org/pdf/2305.07034v1
2023-05-10T18:40:01
2023-05-10T18:40:01
{ "title": [ -0.011405945, 0.03555298, -0.05505371, -0.06695557, 0.030639648, -0.019546509, -0.030944824, -0.012374878, 0.020553589, 0.035095215, -0.00036358833, -0.008804321, 0.0029182434, -0.034729004, -0.053344727, 0.014122009, 0.0116119385, 0...
http://arxiv.org/abs/2306.01742v1
Beyond Negativity: Re-Analysis and Follow-Up Experiments on Hope Speech Detection
Neemesh Yadav, Mohammad Aflah Khan, Diksha Sethi, Raghav Sahni
cs.CL, cs.LG
Health experts assert that hope plays a crucial role in enhancing individuals' physical and mental well-being, facilitating their recovery, and promoting restoration. Hope speech refers to comments, posts and other social media messages that offer support, reassurance, suggestions, inspiration, and insight. The detecti...
http://arxiv.org/pdf/2306.01742v1
2023-05-10T18:38:48
2023-05-10T18:38:48
{ "title": [ 0.0046958923, 0.018814087, -0.055999756, -0.03186035, 0.018035889, -0.044006348, -0.022735596, -0.019973755, -0.0012264252, 0.04046631, -0.025802612, -0.02305603, 0.033081055, -0.030303955, -0.051513672, -0.017730713, -0.023330688, 0...
http://arxiv.org/abs/2305.06404v1
LACoS-BLOOM: Low-rank Adaptation with Contrastive objective on 8 bits Siamese-BLOOM
Wen-Yu Hua, Brian Williams, Davood Shamsi
cs.CL, cs.AI
Text embeddings are useful features for several NLP applications, such as sentence similarity, text clustering, and semantic search. In this paper, we present a Low-rank Adaptation with a Contrastive objective on top of 8-bit Siamese-BLOOM, a multilingual large language model optimized to produce semantically meaningfu...
http://arxiv.org/pdf/2305.06404v1
2023-05-10T18:26:42
2023-05-10T18:26:42
{ "title": [ -0.011123657, 0.0015630722, -0.023269653, -0.015945435, -0.024505615, -0.00932312, -0.0039405823, -0.030838013, 0.011924744, 0.06915283, -0.02406311, 0.01586914, 0.029876709, -0.0037593842, -0.053863525, 0.0039634705, 0.017684937, -0...
http://arxiv.org/abs/2305.06355v1
VideoChat: Chat-Centric Video Understanding
KunChang Li, Yinan He, Yi Wang, Yizhuo Li, Wenhai Wang, Ping Luo, Yali Wang, Limin Wang, Yu Qiao
cs.CV, cs.CL
In this study, we initiate an exploration into video understanding by introducing VideoChat, an end-to-end chat-centric video understanding system. It integrates video foundation models and large language models via a learnable neural interface, excelling in spatiotemporal reasoning, event localization, and causal rela...
http://arxiv.org/pdf/2305.06355v1
2023-05-10T17:59:04
2023-05-10T17:59:04
{ "title": [ -0.009025574, -0.009475708, -0.02154541, -0.06921387, 0.004711151, 0.004634857, -0.03543091, -0.0029850006, 0.020523071, 0.043701172, -0.041534424, 0.004398346, 0.02180481, -0.003414154, -0.055114746, 0.00006765127, -0.0026378632, 0....
http://arxiv.org/abs/2305.06349v3
RECKONING: Reasoning through Dynamic Knowledge Encoding
Zeming Chen, Gail Weiss, Eric Mitchell, Asli Celikyilmaz, Antoine Bosselut
cs.CL, cs.AI, cs.LG
Recent studies on transformer-based language models show that they can answer questions by reasoning over knowledge provided as part of the context (i.e., in-context reasoning). However, since the available knowledge is often not filtered for a particular question, in-context reasoning can be sensitive to distractor fa...
http://arxiv.org/pdf/2305.06349v3
2023-11-05T21:20:54
2023-05-10T17:54:51
{ "title": [ -0.0259552, -0.027069092, -0.052703857, -0.06695557, 0.0049972534, -0.0103302, -0.011520386, 0.013786316, 0.009849548, 0.017807007, -0.021316528, -0.02041626, -0.022323608, -0.018859863, -0.051635742, -0.015396118, 0.014923096, 0.021...
http://arxiv.org/abs/2305.06335v3
K-UniMorph: Korean Universal Morphology and its Feature Schema
Eunkyul Leah Jo, Kyuwon Kim, Xihan Wu, KyungTae Lim, Jungyeul Park, Chulwoo Park
cs.CL
We present in this work a new Universal Morphology dataset for Korean. Previously, the Korean language has been underrepresented in the field of morphological paradigms amongst hundreds of diverse world languages. Hence, we propose this Universal Morphological paradigms for the Korean language that preserve its distinc...
http://arxiv.org/pdf/2305.06335v3
2023-05-17T08:29:58
2023-05-10T17:44:01
{ "title": [ -0.062561035, -0.00258255, -0.050811768, -0.059173584, -0.031829834, 0.029815674, -0.10687256, -0.005088806, -0.021850586, 0.039794922, -0.008460999, 0.017929077, 0.031982422, -0.03237915, -0.0062065125, 0.022949219, -0.027389526, 0....
http://arxiv.org/abs/2305.06330v1
Korean Named Entity Recognition Based on Language-Specific Features
Yige Chen, KyungTae Lim, Jungyeul Park
cs.CL
In the paper, we propose a novel way of improving named entity recognition in the Korean language using its language-specific features. While the field of named entity recognition has been studied extensively in recent years, the mechanism of efficiently recognizing named entities in Korean has hardly been explored. Th...
http://arxiv.org/pdf/2305.06330v1
2023-05-10T17:34:52
2023-05-10T17:34:52
{ "title": [ -0.03564453, -0.008483887, -0.03552246, -0.057647705, -0.005153656, 0.0071487427, -0.06506348, 0.0016546249, 0.0033111572, 0.028930664, -0.026733398, 0.010231018, 0.0137786865, -0.0078086853, -0.03225708, 0.007434845, -0.0019016266, ...
http://arxiv.org/abs/2305.06311v2
Automatic Evaluation of Attribution by Large Language Models
Xiang Yue, Boshi Wang, Ziru Chen, Kai Zhang, Yu Su, Huan Sun
cs.CL
A recent focus of large language model (LLM) development, as exemplified by generative search engines, is to incorporate external references to generate and support its claims. However, evaluating the attribution, i.e., verifying whether the generated statement is fully supported by the cited reference, remains an open...
http://arxiv.org/pdf/2305.06311v2
2023-10-07T22:46:33
2023-05-10T16:58:33
{ "title": [ -0.018066406, -0.011116028, -0.02558899, -0.04373169, -0.0016622543, 0.004890442, -0.053771973, -0.033966064, -0.007183075, 0.021224976, -0.010047913, 0.009567261, -0.011558533, -0.019042969, -0.024780273, 0.00088119507, -0.029418945, ...
http://arxiv.org/abs/2305.06300v2
Evaluating Embedding APIs for Information Retrieval
Ehsan Kamalloo, Xinyu Zhang, Odunayo Ogundepo, Nandan Thakur, David Alfonso-Hermelo, Mehdi Rezagholizadeh, Jimmy Lin
cs.IR, cs.CL
The ever-increasing size of language models curtails their widespread availability to the community, thereby galvanizing many companies into offering access to large language models through APIs. One particular type, suitable for dense retrieval, is a semantic embedding service that builds vector representations of inp...
http://arxiv.org/pdf/2305.06300v2
2023-07-06T18:47:02
2023-05-10T16:40:52
{ "title": [ 0.00374794, -0.009750366, -0.027893066, -0.036376953, 0.02394104, -0.011695862, 0.013938904, -0.0057640076, -0.00013709068, 0.033172607, -0.049713135, 0.003479004, -0.012550354, 0.0047721863, -0.052612305, -0.014228821, -0.018112183, ...
http://arxiv.org/abs/2305.06299v2
Summarizing, Simplifying, and Synthesizing Medical Evidence Using GPT-3 (with Varying Success)
Chantal Shaib, Millicent L. Li, Sebastian Joseph, Iain J. Marshall, Junyi Jessy Li, Byron C. Wallace
cs.CL
Large language models, particularly GPT-3, are able to produce high quality summaries of general domain news articles in few- and zero-shot settings. However, it is unclear if such models are similarly capable in more specialized, high-stakes domains such as biomedicine. In this paper, we enlist domain experts (individ...
http://arxiv.org/pdf/2305.06299v2
2023-05-11T15:51:55
2023-05-10T16:40:37
{ "title": [ 0.008132935, 0.024139404, -0.041900635, -0.045806885, -0.0064735413, -0.005039215, -0.050048828, -0.02204895, -0.041992188, 0.053619385, -0.024475098, 0.039916992, 0.01020813, -0.012039185, -0.006164551, -0.0056381226, 0.015686035, 0...
http://arxiv.org/abs/2305.06294v2
CADGE: Context-Aware Dialogue Generation Enhanced with Graph-Structured Knowledge Aggregation
Hongbo Zhang, Chen Tang, Tyler Loakman, Chenghua Lin, Stefan Goetze
cs.CL
Commonsense knowledge is crucial to many natural language processing tasks. Existing works usually incorporate graph knowledge with conventional graph neural networks (GNNs), leading to the text and graph knowledge encoding processes being separated in a serial pipeline. We argue that these separate representation lear...
http://arxiv.org/pdf/2305.06294v2
2023-06-07T05:33:07
2023-05-10T16:31:35
{ "title": [ -0.01600647, -0.018844604, -0.07086182, -0.09094238, 0.0059280396, 0.0024471283, -0.059631348, 0.039154053, 0.0037937164, 0.03857422, -0.0043029785, 0.011932373, 0.002538681, -0.017227173, -0.048950195, 0.0048446655, -0.011383057, 0....
http://arxiv.org/abs/2305.06274v1
Context-Aware Document Simplification
Liam Cripwell, Joël Legrand, Claire Gardent
cs.CL
To date, most work on text simplification has focused on sentence-level inputs. Early attempts at document simplification merely applied these approaches iteratively over the sentences of a document. However, this fails to coherently preserve the discourse structure, leading to suboptimal output quality. Recently, stra...
http://arxiv.org/pdf/2305.06274v1
2023-05-10T16:06:36
2023-05-10T16:06:36
{ "title": [ -0.020935059, -0.008148193, -0.06088257, -0.08856201, 0.014099121, 0.006061554, -0.06628418, 0.01737976, 0.010879517, 0.0736084, -0.010353088, 0.012123108, 0.03286743, -0.012542725, -0.040405273, 0.010917664, 0.019241333, 0.016906738...
http://arxiv.org/abs/2305.06212v1
Privacy-Preserving Prompt Tuning for Large Language Model Services
Yansong Li, Zhixing Tan, Yang Liu
cs.CL, cs.CR
Prompt tuning provides an efficient way for users to customize Large Language Models (LLMs) with their private data in the emerging LLM service scenario. However, the sensitive nature of private data brings the need for privacy preservation in LLM service customization. Based on prompt tuning, we propose Privacy-Preser...
http://arxiv.org/pdf/2305.06212v1
2023-05-10T14:41:51
2023-05-10T14:41:51
{ "title": [ -0.028137207, -0.019119263, -0.045654297, -0.0770874, -0.0016927719, 0.004421234, -0.060394287, 0.029037476, -0.013580322, 0.017303467, -0.0032749176, -0.00566864, 0.0062561035, -0.02583313, -0.019699097, 0.008392334, -0.007751465, 0...
http://arxiv.org/abs/2305.16326v1
Large language models in biomedical natural language processing: benchmarks, baselines, and recommendations
Qingyu Chen, Jingcheng Du, Yan Hu, Vipina Kuttichi Keloth, Xueqing Peng, Kalpana Raja, Rui Zhang, Zhiyong Lu, Hua Xu
cs.CL, cs.AI, cs.IR, cs.LG
Biomedical literature is growing rapidly, making it challenging to curate and extract knowledge manually. Biomedical natural language processing (BioNLP) techniques that can automatically extract information from biomedical literature help alleviate this burden. Recently, large Language Models (LLMs), such as GPT-3 and...
http://arxiv.org/pdf/2305.16326v1
2023-05-10T13:40:06
2023-05-10T13:40:06
{ "title": [ -0.034973145, -0.0105896, -0.034973145, -0.05545044, 0.004760742, 0.013542175, -0.058258057, -0.0067367554, 0.014556885, 0.024276733, -0.0073051453, -0.009414673, 0.009895325, -0.02798462, -0.04434204, 0.018539429, -0.005130768, 0.03...
http://arxiv.org/abs/2305.11068v1
ORKG-Leaderboards: A Systematic Workflow for Mining Leaderboards as a Knowledge Graph
Salomon Kabongo, Jennifer D'Souza, Sören Auer
cs.CL, cs.AI
The purpose of this work is to describe the Orkg-Leaderboard software designed to extract leaderboards defined as Task-Dataset-Metric tuples automatically from large collections of empirical research papers in Artificial Intelligence (AI). The software can support both the main workflows of scholarly publishing, viz. a...
http://arxiv.org/pdf/2305.11068v1
2023-05-10T13:19:18
2023-05-10T13:19:18
{ "title": [ -0.019210815, -0.020385742, -0.03111267, -0.03717041, -0.028579712, -0.0005264282, -0.030578613, 0.0051879883, 0.0005340576, 0.040405273, -0.019836426, 0.037384033, -0.027053833, -0.01399231, -0.039764404, 0.0010118484, -0.0109939575, ...
http://arxiv.org/abs/2305.06099v1
PAI at SemEval-2023 Task 2: A Universal System for Named Entity Recognition with External Entity Information
Long Ma, Kai Lu, Tianbo Che, Hailong Huang, Weiguo Gao, Xuan Li
cs.CL, cs.AI
The MultiCoNER II task aims to detect complex, ambiguous, and fine-grained named entities in low-context situations and noisy scenarios like the presence of spelling mistakes and typos for multiple languages. The task poses significant challenges due to the scarcity of contextual information, the high granularity of th...
http://arxiv.org/pdf/2305.06099v1
2023-05-10T12:40:48
2023-05-10T12:40:48
{ "title": [ -0.019699097, -0.006465912, -0.045318604, -0.07623291, 0.02458191, 0.008430481, -0.04901123, 0.0017061234, 0.0049591064, 0.023513794, -0.032470703, 0.028579712, 0.0079956055, 0.0010766983, -0.018753052, 0.0107803345, -0.0099487305, 0...
http://arxiv.org/abs/2305.16324v1
Talking with Machines: A Comprehensive Survey of Emergent Dialogue Systems
William Tholke
cs.CL
From the earliest experiments in the 20th century to the utilization of large language models and transformers, dialogue systems research has continued to evolve, playing crucial roles in numerous fields. This paper offers a comprehensive review of these systems, tracing their historical development and examining their...
http://arxiv.org/pdf/2305.16324v1
2023-05-10T12:24:03
2023-05-10T12:24:03
{ "title": [ -0.027862549, -0.010307312, -0.03326416, -0.036834717, 0.014144897, -0.0029792786, -0.0078048706, -0.003435135, 0.010345459, 0.030731201, -0.015899658, 0.0047836304, -0.014755249, -0.010978699, -0.05984497, 0.0009293556, -0.016921997, ...
http://arxiv.org/abs/2305.06087v1
A Glimpse in ChatGPT Capabilities and its impact for AI research
Frank Joublin, Antonello Ceravola, Joerg Deigmoeller, Michael Gienger, Mathias Franzius, Julian Eggert
cs.AI, cs.CL, cs.HC, cs.LG, cs.RO, I.2; I.7
Large language models (LLMs) have recently become a popular topic in the field of Artificial Intelligence (AI) research, with companies such as Google, Amazon, Facebook, Amazon, Tesla, and Apple (GAFA) investing heavily in their development. These models are trained on massive amounts of data and can be used for a wide...
http://arxiv.org/pdf/2305.06087v1
2023-05-10T12:10:51
2023-05-10T12:10:51
{ "title": [ -0.015014648, -0.017623901, -0.03086853, -0.034179688, -0.0115737915, -0.017150879, -0.0066947937, -0.0042877197, -0.00724411, 0.030929565, -0.02267456, 0.020736694, 0.0035953522, -0.021270752, -0.015357971, 0.014823914, -0.0017633438, ...
http://arxiv.org/abs/2305.06074v1
iLab at SemEval-2023 Task 11 Le-Wi-Di: Modelling Disagreement or Modelling Perspectives?
Nikolas Vitsakis, Amit Parekh, Tanvi Dinkar, Gavin Abercrombie, Ioannis Konstas, Verena Rieser
cs.CL, cs.LG
There are two competing approaches for modelling annotator disagreement: distributional soft-labelling approaches (which aim to capture the level of disagreement) or modelling perspectives of individual annotators or groups thereof. We adapt a multi-task architecture -- which has previously shown success in modelling p...
http://arxiv.org/pdf/2305.06074v1
2023-05-10T11:55:17
2023-05-10T11:55:17
{ "title": [ -0.01725769, -0.039154053, -0.021194458, -0.050842285, -0.037078857, 0.0025558472, -0.06561279, -0.004901886, -0.0054244995, 0.0061073303, 0.0041885376, -0.021316528, -0.01625061, -0.06524658, -0.015060425, -0.006881714, -0.016052246, ...
http://arxiv.org/abs/2305.11070v1
Enriching language models with graph-based context information to better understand textual data
Albert Roethel, Maria Ganzha, Anna Wróblewska
cs.CL, cs.AI, cs.LG, cs.NE
A considerable number of texts encountered daily are somehow connected with each other. For example, Wikipedia articles refer to other articles via hyperlinks, scientific papers relate to others via citations or (co)authors, while tweets relate via users that follow each other or reshare content. Hence, a graph-like st...
http://arxiv.org/pdf/2305.11070v1
2023-05-10T10:57:21
2023-05-10T10:57:21
{ "title": [ -0.029022217, -0.014915466, -0.047454834, -0.08319092, -0.0019683838, 0.0038795471, -0.062347412, 0.022613525, 0.008895874, 0.0335083, -0.027282715, 0.009986877, 0.0099487305, -0.0012331009, -0.04296875, 0.00541687, -0.0140686035, 0....
http://arxiv.org/abs/2306.01741v1
GPT Models Meet Robotic Applications: Co-Speech Gesturing Chat System
Naoki Wake, Atsushi Kanehira, Kazuhiro Sasabuchi, Jun Takamatsu, Katsushi Ikeuchi
cs.RO, cs.CL
This technical paper introduces a chatting robot system that utilizes recent advancements in large-scale language models (LLMs) such as GPT-3 and ChatGPT. The system is integrated with a co-speech gesture generation system, which selects appropriate gestures based on the conceptual meaning of speech. Our motivation is ...
http://arxiv.org/pdf/2306.01741v1
2023-05-10T10:14:16
2023-05-10T10:14:16
{ "title": [ -0.002948761, 0.0017280579, -0.020645142, -0.016464233, -0.029800415, -0.016174316, -0.04269409, -0.024520874, -0.019058228, 0.054595947, -0.05102539, -0.016540527, 0.06262207, -0.018753052, -0.019622803, -0.032836914, -0.008354187, ...
http://arxiv.org/abs/2305.11061v1
SPSQL: Step-by-step Parsing Based Framework for Text-to-SQL Generation
Ran Shen, Gang Sun, Hao Shen, Yiling Li, Liangfeng Jin, Han Jiang
cs.CL, cs.AI, cs.DB
Converting text into the structured query language (Text2SQL) is a research hotspot in the field of natural language processing (NLP), which has broad application prospects. In the era of big data, the use of databases has penetrated all walks of life, in which the collected data is large in scale, diverse in variety, ...
http://arxiv.org/pdf/2305.11061v1
2023-05-10T10:01:36
2023-05-10T10:01:36
{ "title": [ -0.023757935, -0.006965637, -0.05758667, -0.08605957, 0.02330017, 0.015777588, -0.052947998, 0.015464783, 0.008468628, 0.039611816, -0.022567749, 0.0063056946, -0.0021095276, -0.01626587, -0.044525146, -0.0068130493, 0.0010089874, 0....
http://arxiv.org/abs/2305.05994v1
ANALOGYKB: Unlocking Analogical Reasoning of Language Models with A Million-scale Knowledge Base
Siyu Yuan, Jiangjie Chen, Changzhi Sun, Jiaqing Liang, Yanghua Xiao, Deqing Yang
cs.CL, cs.AI
Analogical reasoning is a fundamental cognitive ability of humans. However, current language models (LMs) still struggle to achieve human-like performance in analogical reasoning tasks due to a lack of resources for model training. In this work, we address this gap by proposing ANALOGYKB, a million-scale analogy knowle...
http://arxiv.org/pdf/2305.05994v1
2023-05-10T09:03:01
2023-05-10T09:03:01
{ "title": [ -0.023468018, -0.024047852, -0.015686035, -0.066589355, -0.011497498, 0.024017334, -0.06088257, -0.004447937, 0.02796936, 0.028808594, -0.021499634, 0.010986328, 0.030151367, -0.012199402, -0.033599854, 0.027755737, -0.0037136078, 0....
http://arxiv.org/abs/2305.11064v1
Bits of Grass: Does GPT already know how to write like Whitman?
Piotr Sawicki, Marek Grzes, Fabricio Goes, Dan Brown, Max Peeperkorn, Aisha Khatun
cs.CL
This study examines the ability of GPT-3.5, GPT-3.5-turbo (ChatGPT) and GPT-4 models to generate poems in the style of specific authors using zero-shot and many-shot prompts (which use the maximum context length of 8192 tokens). We assess the performance of models that are not fine-tuned for generating poetry in the st...
http://arxiv.org/pdf/2305.11064v1
2023-05-10T09:02:34
2023-05-10T09:02:34
{ "title": [ -0.018722534, 0.0009613037, -0.012016296, 0.01687622, -0.030471802, 0.0017328262, -0.003255844, -0.023254395, -0.04144287, 0.028961182, -0.04849243, 0.02015686, 0.024261475, -0.02331543, 0.002216339, -0.041656494, -0.008346558, 0.023...
http://arxiv.org/abs/2305.05982v1
Generating medically-accurate summaries of patient-provider dialogue: A multi-stage approach using large language models
Varun Nair, Elliot Schumacher, Anitha Kannan
cs.CL, cs.AI, cs.LG
A medical provider's summary of a patient visit serves several critical purposes, including clinical decision-making, facilitating hand-offs between providers, and as a reference for the patient. An effective summary is required to be coherent and accurately capture all the medically relevant information in the dialogu...
http://arxiv.org/pdf/2305.05982v1
2023-05-10T08:48:53
2023-05-10T08:48:53
{ "title": [ -0.007270813, -0.0024909973, -0.030792236, -0.088256836, -0.0061683655, 0.012931824, -0.084472656, 0.0023097992, -0.0129470825, 0.039123535, 0.011131287, 0.013031006, -0.0058898926, -0.017349243, -0.051361084, 0.004508972, -0.0055732727,...
http://arxiv.org/abs/2305.05976v2
Say What You Mean! Large Language Models Speak Too Positively about Negative Commonsense Knowledge
Jiangjie Chen, Wei Shi, Ziquan Fu, Sijie Cheng, Lei Li, Yanghua Xiao
cs.CL
Large language models (LLMs) have been widely studied for their ability to store and utilize positive knowledge. However, negative knowledge, such as "lions don't live in the ocean", is also ubiquitous in the world but rarely mentioned explicitly in the text. What do LLMs know about negative knowledge? This work examin...
http://arxiv.org/pdf/2305.05976v2
2023-05-13T13:34:04
2023-05-10T08:35:50
{ "title": [ -0.013702393, -0.025161743, -0.019577026, -0.04156494, -0.026977539, -0.018157959, -0.026794434, -0.0079422, 0.016174316, 0.013877869, -0.03390503, -0.009063721, -0.009178162, -0.018051147, -0.055236816, -0.015914917, -0.0057029724, ...
http://arxiv.org/abs/2305.05973v1
Privacy-Preserving Recommender Systems with Synthetic Query Generation using Differentially Private Large Language Models
Aldo Gael Carranza, Rezsa Farahani, Natalia Ponomareva, Alex Kurakin, Matthew Jagielski, Milad Nasr
cs.CL, cs.CR, cs.IR
We propose a novel approach for developing privacy-preserving large-scale recommender systems using differentially private (DP) large language models (LLMs) which overcomes certain challenges and limitations in DP training these complex systems. Our method is particularly well suited for the emerging area of LLM-based ...
http://arxiv.org/pdf/2305.05973v1
2023-05-10T08:30:31
2023-05-10T08:30:31
{ "title": [ -0.0085372925, -0.008621216, -0.04473877, -0.057373047, 0.010314941, 0.0047302246, -0.025772095, 0.012367249, -0.034851074, 0.019805908, -0.024246216, -0.0005660057, 0.0011386871, -0.016052246, -0.013069153, -0.01939392, -0.013938904, ...
http://arxiv.org/abs/2305.05968v1
Investigating Forgetting in Pre-Trained Representations Through Continual Learning
Yun Luo, Zhen Yang, Xuefeng Bai, Fandong Meng, Jie Zhou, Yue Zhang
cs.CL
Representation forgetting refers to the drift of contextualized representations during continual training. Intuitively, the representation forgetting can influence the general knowledge stored in pre-trained language models (LMs), but the concrete effect is still unclear. In this paper, we study the effect of represent...
http://arxiv.org/pdf/2305.05968v1
2023-05-10T08:27:59
2023-05-10T08:27:59
{ "title": [ 0.021118164, 0.014907837, -0.034332275, -0.07434082, 0.017730713, -0.026535034, -0.005680084, -0.004714966, 0.0061683655, 0.03729248, -0.0062408447, 0.0020198822, -0.011756897, -0.024459839, -0.05154419, 0.0058555603, 0.0067596436, 0...
http://arxiv.org/abs/2305.06934v1
Humans are Still Better than ChatGPT: Case of the IEEEXtreme Competition
Anis Koubaa, Basit Qureshi, Adel Ammar, Zahid Khan, Wadii Boulila, Lahouari Ghouti
cs.SE, cs.AI, cs.CL, cs.CY, cs.LG, cs.PL
Since the release of ChatGPT, numerous studies have highlighted the remarkable performance of ChatGPT, which often rivals or even surpasses human capabilities in various tasks and domains. However, this paper presents a contrasting perspective by demonstrating an instance where human performance excels in typical tasks...
http://arxiv.org/pdf/2305.06934v1
2023-05-10T08:16:46
2023-05-10T08:16:46
{ "title": [ 0.01751709, -0.01727295, -0.026824951, -0.00086545944, -0.0093307495, -0.048065186, 0.010696411, -0.03665161, -0.024414062, 0.041137695, -0.011810303, -0.0036582947, -0.0047073364, -0.0045547485, -0.060455322, -0.0076446533, -0.019195557...
http://arxiv.org/abs/2305.05964v1
Interpretable Multimodal Misinformation Detection with Logic Reasoning
Hui Liu, Wenya Wang, Haoliang Li
cs.MM, cs.AI, cs.CL
Multimodal misinformation on online social platforms is becoming a critical concern due to increasing credibility and easier dissemination brought by multimedia content, compared to traditional text-only information. While existing multimodal detection approaches have achieved high performance, the lack of interpretabi...
http://arxiv.org/pdf/2305.05964v1
2023-05-10T08:16:36
2023-05-10T08:16:36
{ "title": [ -0.0026359558, 0.02293396, -0.051208496, -0.06402588, 0.020645142, -0.007408142, -0.043640137, -0.008071899, -0.03552246, 0.02204895, 0.0017910004, 0.020355225, -0.0209198, -0.02015686, -0.015945435, -0.018997192, -0.036621094, 0.054...
http://arxiv.org/abs/2305.05948v1
Multi-Path Transformer is Better: A Case Study on Neural Machine Translation
Ye Lin, Shuhan Zhou, Yanyang Li, Anxiang Ma, Tong Xiao, Jingbo Zhu
cs.CL, cs.AI
For years the model performance in machine learning obeyed a power-law relationship with the model size. For the consideration of parameter efficiency, recent studies focus on increasing model depth rather than width to achieve better performance. In this paper, we study how model width affects the Transformer model th...
http://arxiv.org/pdf/2305.05948v1
2023-05-10T07:39:57
2023-05-10T07:39:57
{ "title": [ -0.01838684, -0.005264282, -0.023590088, -0.041259766, -0.008049011, -0.010093689, -0.025238037, 0.00053071976, -0.028884888, 0.035186768, -0.003042221, -0.009628296, 0.019119263, -0.03604126, -0.01890564, 0.005672455, -0.001288414, ...
http://arxiv.org/abs/2305.05945v1
Adapter-TST: A Parameter Efficient Method for Multiple-Attribute Text Style Transfer
Zhiqiang Hu, Roy Ka-Wei Lee, Nancy F. Chen
cs.CL
Adapting a large language model for multiple-attribute text style transfer via fine-tuning can be challenging due to the significant amount of computational resources and labeled data required for the specific task. In this paper, we address this challenge by introducing AdapterTST, a framework that freezes the pre-tra...
http://arxiv.org/pdf/2305.05945v1
2023-05-10T07:33:36
2023-05-10T07:33:36
{ "title": [ -0.018585205, -0.010269165, -0.03967285, -0.051513672, 0.010749817, 0.001203537, -0.028778076, -0.01687622, 0.008621216, 0.053955078, 0.010940552, 0.029037476, 0.040100098, -0.013961792, -0.028915405, 0.033081055, -0.0014314651, 0.01...
http://arxiv.org/abs/2305.05940v3
Multilingual LLMs are Better Cross-lingual In-context Learners with Alignment
Eshaan Tanwar, Subhabrata Dutta, Manish Borthakur, Tanmoy Chakraborty
cs.CL
In-context learning (ICL) unfolds as large language models become capable of inferring test labels conditioned on a few labeled samples without any gradient update. ICL-enabled large language models provide a promising step forward toward bypassing recurrent annotation costs in a low-resource setting. Yet, only a handf...
http://arxiv.org/pdf/2305.05940v3
2023-06-24T07:43:35
2023-05-10T07:24:36
{ "title": [ -0.013183594, -0.0018672943, -0.053497314, -0.040802002, -0.038482666, -0.015670776, -0.071899414, -0.0043525696, -0.036224365, 0.052520752, -0.013191223, 0.016464233, 0.019882202, -0.041656494, 0.034820557, -0.0038604736, -0.007129669, ...
http://arxiv.org/abs/2305.05936v1
Multi-hop Commonsense Knowledge Injection Framework for Zero-Shot Commonsense Question Answering
Xin Guan, Biwei Cao, Qingqing Gao, Zheng Yin, Bo Liu, Jiuxin Cao
cs.CL
Commonsense question answering (QA) research requires machines to answer questions based on commonsense knowledge. However, this research requires expensive labor costs to annotate data as the basis of research, and models that rely on fine-tuning paradigms only apply to specific tasks, rather than learn a general comm...
http://arxiv.org/pdf/2305.05936v1
2023-05-10T07:13:47
2023-05-10T07:13:47
{ "title": [ 0.017822266, -0.036743164, -0.026306152, -0.047607422, -0.03265381, -0.019821167, 0.0048980713, 0.027297974, 0.02772522, 0.027954102, -0.03817749, 0.011756897, -0.008049011, -0.019317627, -0.037506104, -0.00472641, -0.0013122559, 0.0...
http://arxiv.org/abs/2305.05928v1
WikiSQE: A Large-Scale Dataset for Sentence Quality Estimation in Wikipedia
Kenichiro Ando, Satoshi Sekine, Mamoru Komachi
cs.CL
Wikipedia can be edited by anyone and thus contains various quality sentences. Therefore, Wikipedia includes some poor-quality edits, which are often marked up by other editors. While editors' reviews enhance the credibility of Wikipedia, it is hard to check all edited text. Assisting in this process is very important,...
http://arxiv.org/pdf/2305.05928v1
2023-05-10T06:45:13
2023-05-10T06:45:13
{ "title": [ -0.012619019, -0.00082063675, -0.031021118, -0.04168701, 0.018188477, -0.008148193, -0.02204895, -0.024139404, 0.009681702, 0.022445679, -0.016937256, 0.00894928, 0.012893677, -0.0030021667, -0.040405273, 0.014823914, -0.019302368, 0...
http://arxiv.org/abs/2305.05921v2
Decker: Double Check with Heterogeneous Knowledge for Commonsense Fact Verification
Anni Zou, Zhuosheng Zhang, Hai Zhao
cs.CL
Commonsense fact verification, as a challenging branch of commonsense question-answering (QA), aims to verify through facts whether a given commonsense claim is correct or not. Answering commonsense questions necessitates a combination of knowledge from various levels. However, existing studies primarily rest on graspi...
http://arxiv.org/pdf/2305.05921v2
2023-05-27T08:49:05
2023-05-10T06:28:16
{ "title": [ 0.0016288757, -0.018341064, -0.04598999, -0.081604004, 0.012268066, -0.009094238, -0.025680542, 0.0012960434, -0.009719849, 0.04458618, -0.025283813, 0.021652222, 0.014320374, -0.0037345886, -0.035736084, -0.0026855469, -0.027740479, ...
http://arxiv.org/abs/2305.05874v1
Address Matching Based On Hierarchical Information
Chengxian Zhang, Jintao Tang, Ting Wang, Shasha Li
cs.CL
There is evidence that address matching plays a crucial role in many areas such as express delivery, online shopping and so on. Address has a hierarchical structure, in contrast to unstructured texts, which can contribute valuable information for address matching. Based on this idea, this paper proposes a novel method ...
http://arxiv.org/pdf/2305.05874v1
2023-05-10T03:45:22
2023-05-10T03:45:22
{ "title": [ 0.010574341, 0.01159668, -0.02998352, -0.06750488, -0.013954163, -0.011795044, -0.029403687, -0.010864258, -0.007019043, 0.052612305, -0.03375244, -0.00031733513, 0.030273438, -0.023269653, -0.0098724365, 0.009162903, 0.010253906, 0....
http://arxiv.org/abs/2305.05862v2
Are ChatGPT and GPT-4 General-Purpose Solvers for Financial Text Analytics? A Study on Several Typical Tasks
Xianzhi Li, Samuel Chan, Xiaodan Zhu, Yulong Pei, Zhiqiang Ma, Xiaomo Liu, Sameena Shah
cs.CL, cs.AI
The most recent large language models(LLMs) such as ChatGPT and GPT-4 have shown exceptional capabilities of generalist models, achieving state-of-the-art performance on a wide range of NLP tasks with little or no adaptation. How effective are such models in the financial domain? Understanding this basic question would...
http://arxiv.org/pdf/2305.05862v2
2023-10-10T18:54:43
2023-05-10T03:13:54
{ "title": [ -0.012626648, 0.0016803741, -0.013397217, -0.011505127, -0.021484375, -0.0069274902, -0.027496338, -0.024475098, -0.0011920929, 0.030197144, -0.026855469, 0.015426636, 0.021209717, -0.016311646, -0.019256592, -0.011650085, -0.011100769, ...
http://arxiv.org/abs/2305.05858v1
Vārta: A Large-Scale Headline-Generation Dataset for Indic Languages
Rahul Aralikatte, Ziling Cheng, Sumanth Doddapaneni, Jackie Chi Kit Cheung
cs.CL
We present V\=arta, a large-scale multilingual dataset for headline generation in Indic languages. This dataset includes 41.8 million news articles in 14 different Indic languages (and English), which come from a variety of high-quality sources. To the best of our knowledge, this is the largest collection of curated ar...
http://arxiv.org/pdf/2305.05858v1
2023-05-10T03:07:17
2023-05-10T03:07:17
{ "title": [ -0.036865234, -0.009666443, -0.048706055, -0.061157227, -0.0061187744, 0.0026416779, -0.06903076, 0.016494751, -0.0017156601, 0.017974854, -0.009300232, 0.033203125, -0.026397705, -0.012893677, -0.042510986, 0.010612488, -0.01675415, ...
http://arxiv.org/abs/2305.05834v1
Unsupervised Dense Retrieval Training with Web Anchors
Yiqing Xie, Xiao Liu, Chenyan Xiong
cs.IR, cs.CL
In this work, we present an unsupervised retrieval method with contrastive learning on web anchors. The anchor text describes the content that is referenced from the linked page. This shows similarities to search queries that aim to retrieve pertinent information from relevant documents. Based on their commonalities, w...
http://arxiv.org/pdf/2305.05834v1
2023-05-10T01:46:17
2023-05-10T01:46:17
{ "title": [ 0.0049362183, 0.021636963, -0.03225708, -0.031799316, 0.034210205, -0.027679443, 0.009841919, -0.026382446, 0.01574707, 0.034301758, -0.016113281, -0.009635925, 0.024353027, -0.021972656, -0.023376465, 0.0037727356, 0.00579834, 0.008...
http://arxiv.org/abs/2305.05821v1
Context-dependent communication under environmental constraints
Krzysztof Główka, Julian Zubek, Joanna Rączaszek-Leonardi
cs.AI, cs.CL
There is significant evidence that real-world communication cannot be reduced to sending signals with context-independent meaning. In this work, based on a variant of the classical Lewis (1969) signaling model, we explore the conditions for the emergence of context-dependent communication in a situated scenario. In par...
http://arxiv.org/pdf/2305.05821v1
2023-05-10T00:33:08
2023-05-10T00:33:08
{ "title": [ -0.02935791, 0.008811951, -0.05557251, -0.05593872, 0.031219482, 0.0046691895, -0.024276733, -0.00004863739, 0.005695343, 0.020584106, -0.011993408, -0.016174316, -0.017745972, -0.023773193, -0.04864502, -0.0074806213, -0.008880615, ...
http://arxiv.org/abs/2305.05759v1
Ranking & Reweighting Improves Group Distributional Robustness
Yachuan Liu, Bohan Zhang, Qiaozhu Mei, Paramveer Dhillon
cs.LG, cs.AI, cs.CL, stat.ML
Recent work has shown that standard training via empirical risk minimization (ERM) can produce models that achieve high accuracy on average but low accuracy on underrepresented groups due to the prevalence of spurious features. A predominant approach to tackle this group robustness problem minimizes the worst group err...
http://arxiv.org/pdf/2305.05759v1
2023-05-09T20:37:16
2023-05-09T20:37:16
{ "title": [ 0.0059661865, 0.021469116, -0.03881836, -0.031463623, 0.0072517395, -0.010025024, -0.052612305, -0.04949951, 0.007446289, 0.026367188, -0.018737793, 0.017700195, -0.0051994324, -0.017425537, -0.035186768, 0.0077781677, -0.012428284, ...
http://arxiv.org/abs/2305.05754v1
When and What to Ask Through World States and Text Instructions: IGLU NLP Challenge Solution
Zhengxiang Shi, Jerome Ramos, To Eun Kim, Xi Wang, Hossein A. Rahmani, Aldo Lipani
cs.CL, cs.AI, cs.LG
In collaborative tasks, effective communication is crucial for achieving joint goals. One such task is collaborative building where builders must communicate with each other to construct desired structures in a simulated environment such as Minecraft. We aim to develop an intelligent builder agent to build structures b...
http://arxiv.org/pdf/2305.05754v1
2023-05-09T20:23:17
2023-05-09T20:23:17
{ "title": [ -0.026473999, -0.005504608, -0.03967285, -0.0032615662, -0.008911133, -0.00093889236, -0.015266418, -0.00447464, -0.041748047, 0.016479492, -0.049987793, 0.019256592, 0.0027809143, -0.016937256, 0.011894226, -0.04248047, 0.006832123, ...
http://arxiv.org/abs/2305.05748v1
Multilevel Sentence Embeddings for Personality Prediction
Paolo Tirotta, Akira Yuasa, Masashi Morita
cs.CL
Representing text into a multidimensional space can be done with sentence embedding models such as Sentence-BERT (SBERT). However, training these models when the data has a complex multilevel structure requires individually trained class-specific models, which increases time and computing costs. We propose a two step a...
http://arxiv.org/pdf/2305.05748v1
2023-05-09T20:02:18
2023-05-09T20:02:18
{ "title": [ 0.0069999695, -0.0074272156, -0.022003174, -0.062286377, 0.0066452026, 0.0044174194, -0.033111572, -0.0008497238, 0.013496399, 0.031951904, 0.0075149536, 0.013839722, 0.030715942, -0.02178955, -0.030197144, 0.031311035, -0.024337769, ...
http://arxiv.org/abs/2305.05711v2
CodeIE: Large Code Generation Models are Better Few-Shot Information Extractors
Peng Li, Tianxiang Sun, Qiong Tang, Hang Yan, Yuanbin Wu, Xuanjing Huang, Xipeng Qiu
cs.CL, cs.AI
Large language models (LLMs) pre-trained on massive corpora have demonstrated impressive few-shot learning ability on many NLP tasks. A common practice is to recast the task into a text-to-text format such that generative LLMs of natural language (NL-LLMs) like GPT-3 can be prompted to solve it. However, it is nontrivi...
http://arxiv.org/pdf/2305.05711v2
2023-05-11T01:27:43
2023-05-09T18:40:31
{ "title": [ -0.034179688, -0.037902832, -0.053222656, -0.031585693, -0.043823242, -0.0211792, -0.03741455, 0.0077056885, -0.0023612976, 0.021148682, -0.02746582, 0.025436401, -0.013267517, -0.028930664, -0.0047454834, -0.0041618347, 0.010345459, ...
http://arxiv.org/abs/2305.06159v1
A Review of Vision-Language Models and their Performance on the Hateful Memes Challenge
Bryan Zhao, Andrew Zhang, Blake Watson, Gillian Kearney, Isaac Dale
cs.CL, cs.AI
Moderation of social media content is currently a highly manual task, yet there is too much content posted daily to do so effectively. With the advent of a number of multimodal models, there is the potential to reduce the amount of manual labor for this task. In this work, we aim to explore different models and determi...
http://arxiv.org/pdf/2305.06159v1
2023-05-09T17:55:29
2023-05-09T17:55:29
{ "title": [ 0.023330688, -0.009140015, -0.020355225, -0.021224976, -0.018615723, -0.040740967, -0.01638794, -0.028778076, 0.049591064, 0.04623413, -0.020248413, -0.010864258, 0.044677734, -0.017990112, -0.06439209, 0.019851685, 0.003627777, 0.02...
http://arxiv.org/abs/2305.05658v2
TidyBot: Personalized Robot Assistance with Large Language Models
Jimmy Wu, Rika Antonova, Adam Kan, Marion Lepert, Andy Zeng, Shuran Song, Jeannette Bohg, Szymon Rusinkiewicz, Thomas Funkhouser
cs.RO, cs.AI, cs.CL, cs.CV, cs.LG
For a robot to personalize physical assistance effectively, it must learn user preferences that can be generally reapplied to future scenarios. In this work, we investigate personalization of household cleanup with robots that can tidy up rooms by picking up objects and putting them away. A key challenge is determining...
http://arxiv.org/pdf/2305.05658v2
2023-10-11T17:59:44
2023-05-09T17:52:59
{ "title": [ -0.020370483, -0.016998291, -0.03277588, -0.050598145, -0.002193451, 0.0013742447, -0.040740967, -0.0066070557, -0.012794495, 0.044311523, -0.004776001, -0.0063056946, 0.0184021, -0.017089844, -0.031921387, 0.008392334, 0.0048217773, ...
http://arxiv.org/abs/2305.05644v1
Towards Building the Federated GPT: Federated Instruction Tuning
Jianyi Zhang, Saeed Vahidian, Martin Kuo, Chunyuan Li, Ruiyi Zhang, Guoyin Wang, Yiran Chen
cs.CL, cs.DC, cs.SY, eess.SY
While ``instruction-tuned" generative large language models (LLMs) have demonstrated an impressive ability to generalize to new tasks, the training phases heavily rely on large amounts of diverse and high-quality instruction data (such as ChatGPT and GPT-4). Unfortunately, acquiring high-quality data, especially when i...
http://arxiv.org/pdf/2305.05644v1
2023-05-09T17:42:34
2023-05-09T17:42:34
{ "title": [ -0.052337646, 0.0062561035, -0.05392456, -0.08905029, 0.015617371, 0.01361084, -0.04724121, 0.030517578, -0.020324707, 0.03616333, -0.04852295, 0.030334473, -0.019836426, 0.0029392242, -0.03829956, -0.0051193237, 0.0023002625, 0.0190...
http://arxiv.org/abs/2305.05640v3
Representation Learning for Person or Entity-centric Knowledge Graphs: An Application in Healthcare
Christos Theodoropoulos, Natasha Mulligan, Thaddeus Stappenbeck, Joao Bettencourt-Silva
cs.AI, cs.CL, cs.LG
Knowledge graphs (KGs) are a popular way to organise information based on ontologies or schemas and have been used across a variety of scenarios from search to recommendation. Despite advances in KGs, representing knowledge remains a non-trivial task across industries and it is especially challenging in the biomedical ...
http://arxiv.org/pdf/2305.05640v3
2023-10-09T11:58:55
2023-05-09T17:39:45
{ "title": [ -0.0077056885, -0.00491333, -0.06640625, -0.08062744, 0.017028809, -0.013458252, -0.05239868, 0.017028809, -0.0041046143, 0.04522705, -0.024612427, 0.022613525, 0.015060425, -0.006259918, -0.03665161, 0.0062561035, 0.0051116943, 0.04...
http://arxiv.org/abs/2305.05627v1
An Exploration of Encoder-Decoder Approaches to Multi-Label Classification for Legal and Biomedical Text
Yova Kementchedjhieva, Ilias Chalkidis
cs.CL
Standard methods for multi-label text classification largely rely on encoder-only pre-trained language models, whereas encoder-decoder models have proven more effective in other classification tasks. In this study, we compare four methods for multi-label classification, two based on an encoder only, and two based on an...
http://arxiv.org/pdf/2305.05627v1
2023-05-09T17:13:53
2023-05-09T17:13:53
{ "title": [ -0.026382446, -0.014076233, -0.044189453, -0.033416748, -0.014839172, -0.014419556, -0.04309082, 0.0029659271, 0.0074005127, 0.024810791, -0.017471313, 0.019210815, 0.0018348694, -0.012359619, -0.03277588, 0.0014257431, -0.02128601, ...
http://arxiv.org/abs/2305.06176v2
Fine-tuning Language Models with Generative Adversarial Feedback
Zhang Ze Yu, Lau Jia Jaw, Wong Qin Jiang, Zhang Hui, Bryan Kian Hsiang Low
cs.CL, cs.AI, cs.LG
Reinforcement Learning with Human Feedback (RLHF) has been demonstrated to significantly enhance the performance of large language models (LLMs) by aligning their outputs with desired human values through instruction tuning. However, RLHF is constrained by the expertise and productivity limitations of human evaluators....
http://arxiv.org/pdf/2305.06176v2
2023-10-08T00:40:42
2023-05-09T17:06:06
{ "title": [ -0.025604248, -0.008880615, -0.010284424, -0.049957275, -0.006904602, 0.017486572, -0.053894043, -0.010025024, -0.0036354065, 0.036834717, -0.0069847107, 0.018432617, -0.0009098053, -0.008422852, -0.049438477, 0.01146698, -0.0014266968, ...
http://arxiv.org/abs/2305.05609v1
The Case Records of ChatGPT: Language Models and Complex Clinical Questions
Timothy Poterucha, Pierre Elias, Christopher M. Haggerty
cs.CL, stat.AP
Background: Artificial intelligence language models have shown promise in various applications, including assisting with clinical decision-making as demonstrated by strong performance of large language models on medical licensure exams. However, their ability to solve complex, open-ended cases, which may be representat...
http://arxiv.org/pdf/2305.05609v1
2023-05-09T16:58:32
2023-05-09T16:58:32
{ "title": [ -0.013900757, -0.02659607, -0.006149292, -0.03451538, -0.017242432, 0.019424438, -0.042297363, -0.012382507, -0.010360718, 0.014015198, -0.009857178, 0.011154175, -0.015701294, -0.02973938, -0.017974854, 0.019454956, -0.022750854, 0....
http://arxiv.org/abs/2305.05588v2
StrAE: Autoencoding for Pre-Trained Embeddings using Explicit Structure
Mattia Opper, Victor Prokhorov, N. Siddharth
cs.CL, cs.LG
This work presents StrAE: a Structured Autoencoder framework that through strict adherence to explicit structure, and use of a novel contrastive objective over tree-structured representations, enables effective learning of multi-level representations. Through comparison over different forms of structure, we verify that...
http://arxiv.org/pdf/2305.05588v2
2023-10-25T15:30:59
2023-05-09T16:20:48
{ "title": [ -0.003250122, -0.004940033, -0.039245605, -0.07141113, 0.012611389, -0.0025348663, -0.033569336, -0.017791748, 0.0065307617, 0.01878357, -0.0067481995, 0.013221741, -0.006362915, -0.03277588, -0.006793976, 0.017822266, 0.0052337646, ...
http://arxiv.org/abs/2305.05576v1
Large Language Models Humanize Technology
Pratyush Kumar
cs.CY, cs.CL
Large Language Models (LLMs) have made rapid progress in recent months and weeks, garnering significant public attention. This has sparked concerns about aligning these models with human values, their impact on labor markets, and the potential need for regulation in further research and development. However, the discou...
http://arxiv.org/pdf/2305.05576v1
2023-05-09T16:05:36
2023-05-09T16:05:36
{ "title": [ 0.010848999, 0.026794434, -0.029388428, -0.03945923, -0.029968262, -0.026367188, -0.045654297, -0.003293991, -0.042755127, 0.064086914, -0.038635254, -0.016296387, 0.036834717, -0.0134887695, 0.011543274, -0.013908386, 0.0072898865, ...
http://arxiv.org/abs/2305.05496v1
Exploiting Pseudo Image Captions for Multimodal Summarization
Chaoya Jiang, Rui Xie, Wei Ye, Jinan Sun, Shikun Zhang
cs.CL
Cross-modal contrastive learning in vision language pretraining (VLP) faces the challenge of (partial) false negatives. In this paper, we study this problem from the perspective of Mutual Information (MI) optimization. It is common sense that InfoNCE loss used in contrastive learning will maximize the lower bound of MI...
http://arxiv.org/pdf/2305.05496v1
2023-05-09T14:47:25
2023-05-09T14:47:25
{ "title": [ 0.008468628, 0.005268097, -0.017150879, -0.060424805, -0.00680542, -0.006214142, -0.03427124, -0.022766113, 0.0077209473, 0.06573486, 0.008476257, 0.010345459, 0.004699707, -0.009651184, -0.059783936, 0.014419556, -0.0026283264, 0.01...
http://arxiv.org/abs/2305.05486v1
MAUPQA: Massive Automatically-created Polish Question Answering Dataset
Piotr Rybak
cs.CL
Recently, open-domain question answering systems have begun to rely heavily on annotated datasets to train neural passage retrievers. However, manually annotating such datasets is both difficult and time-consuming, which limits their availability for less popular languages. In this work, we experiment with several meth...
http://arxiv.org/pdf/2305.05486v1
2023-05-09T14:36:04
2023-05-09T14:36:04
{ "title": [ -0.018737793, -0.027023315, -0.03579712, -0.034118652, -0.010116577, -0.007663727, -0.015792847, 0.01071167, 0.04888916, 0.028320312, -0.02305603, -0.009361267, 0.02394104, -0.018112183, -0.05633545, 0.024032593, -0.009010315, 0.0292...
http://arxiv.org/abs/2305.05480v3
Effects of sub-word segmentation on performance of transformer language models
Jue Hou, Anisia Katinskaia, Anh-Duc Vu, Roman Yangarber
cs.CL, cs.AI, cs.LG
Language modeling is a fundamental task in natural language processing, which has been thoroughly explored with various architectures and hyperparameters. However, few studies focus on the effect of sub-word segmentation on the performance of language models (LMs). In this paper, we compare GPT and BERT models trained ...
http://arxiv.org/pdf/2305.05480v3
2023-10-26T20:36:36
2023-05-09T14:30:29
{ "title": [ -0.02458191, 0.0020809174, -0.03186035, -0.03668213, -0.0043029785, -0.011047363, -0.05166626, -0.009864807, -0.025283813, 0.04345703, -0.028686523, 0.0064964294, 0.046173096, -0.015365601, -0.027770996, 0.00983429, -0.0076179504, 0....
http://arxiv.org/abs/2305.05474v1
Going beyond research datasets: Novel intent discovery in the industry setting
Aleksandra Chrabrowa, Tsimur Hadeliya, Dariusz Kajtoch, Robert Mroczkowski, Piotr Rybak
cs.CL, cs.LG
Novel intent discovery automates the process of grouping similar messages (questions) to identify previously unknown intents. However, current research focuses on publicly available datasets which have only the question field and significantly differ from real-life datasets. This paper proposes methods to improve the i...
http://arxiv.org/pdf/2305.05474v1
2023-05-09T14:21:29
2023-05-09T14:21:29
{ "title": [ -0.021347046, -0.006340027, -0.070617676, -0.045532227, 0.025482178, -0.025009155, -0.023849487, -0.0037403107, -0.02041626, 0.006072998, -0.023101807, 0.005542755, -0.04458618, -0.025405884, -0.03250122, -0.028656006, -0.01184845, 0...
http://arxiv.org/abs/2305.05471v3
Beyond Good Intentions: Reporting the Research Landscape of NLP for Social Good
Fernando Gonzalez, Zhijing Jin, Bernhard Schölkopf, Tom Hope, Mrinmaya Sachan, Rada Mihalcea
cs.CL
With the recent advances in natural language processing (NLP), a vast number of applications have emerged across various use cases. Among the plethora of NLP applications, many academic researchers are motivated to do work that has a positive social impact, in line with the recent initiatives of NLP for Social Good (NL...
http://arxiv.org/pdf/2305.05471v3
2023-10-21T13:41:57
2023-05-09T14:16:25
{ "title": [ -0.02218628, -0.0024337769, -0.049713135, -0.058746338, 0.012741089, -0.01222229, -0.030471802, -0.0069122314, -0.02458191, 0.0064086914, -0.04034424, 0.024017334, -0.027008057, -0.005077362, -0.023834229, -0.013267517, -0.029830933, ...
http://arxiv.org/abs/2305.05461v1
What is the best recipe for character-level encoder-only modelling?
Kris Cao
cs.CL
This paper aims to benchmark recent progress in language understanding models that output contextualised representations at the character level. Many such modelling architectures and methods to train those architectures have been proposed, but it is currently unclear what the relative contributions of the architecture ...
http://arxiv.org/pdf/2305.05461v1
2023-05-09T14:00:15
2023-05-09T14:00:15
{ "title": [ -0.01826477, -0.028335571, 0.016235352, 0.0060577393, -0.07678223, 0.0016517639, -0.039489746, -0.0043144226, 0.009353638, 0.0446167, 0.013000488, -0.008216858, 0.045135498, -0.043121338, 0.021591187, 0.035827637, 0.00806427, -0.0252...
http://arxiv.org/abs/2305.05432v1
WikiWeb2M: A Page-Level Multimodal Wikipedia Dataset
Andrea Burns, Krishna Srinivasan, Joshua Ainslie, Geoff Brown, Bryan A. Plummer, Kate Saenko, Jianmo Ni, Mandy Guo
cs.CL, cs.CV
Webpages have been a rich resource for language and vision-language tasks. Yet only pieces of webpages are kept: image-caption pairs, long text articles, or raw HTML, never all in one place. Webpage tasks have resultingly received little attention and structured image-text data underused. To study multimodal webpage un...
http://arxiv.org/pdf/2305.05432v1
2023-05-09T13:20:59
2023-05-09T13:20:59
{ "title": [ -0.032348633, 0.011772156, -0.06518555, -0.05935669, 0.018203735, -0.015571594, -0.036254883, -0.004096985, 0.013977051, 0.033569336, -0.013389587, 0.013015747, 0.007408142, -0.019424438, -0.024963379, 0.02041626, 0.0015058517, 0.033...
http://arxiv.org/abs/2305.05420v1
Estimating related words computationally using language model from the Mahabharata -- an Indian epic
Vrunda Gadesha, Keyur D Joshi, Shefali Naik
cs.CL
'Mahabharata' is the most popular among many Indian pieces of literature referred to in many domains for completely different purposes. This text itself is having various dimension and aspects which is useful for the human being in their personal life and professional life. This Indian Epic is originally written in the...
http://arxiv.org/pdf/2305.05420v1
2023-05-09T13:13:26
2023-05-09T13:13:26
{ "title": [ -0.041778564, 0.010604858, -0.037628174, -0.030853271, -0.02734375, -0.012634277, -0.06390381, -0.010765076, -0.016738892, 0.040496826, -0.029953003, 0.010215759, 0.025863647, -0.017807007, -0.015411377, -0.0038375854, -0.0030345917, ...
http://arxiv.org/abs/2305.06166v2
ChatGPT as a Text Simplification Tool to Remove Bias
Charmaine Barker, Dimitar Kazakov
cs.CL
The presence of specific linguistic signals particular to a certain sub-group of people can be picked up by language models during training. If the model begins to associate specific language with a distinct group, any decisions made based upon this language would hold a strong correlation to a decision based upon thei...
http://arxiv.org/pdf/2305.06166v2
2023-06-01T09:45:30
2023-05-09T13:10:23
{ "title": [ 0.006427765, 0.00036907196, -0.044158936, -0.041778564, -0.0035209656, -0.016326904, -0.016342163, -0.019348145, -0.035369873, 0.055023193, -0.0035209656, 0.029418945, 0.0028705597, -0.028076172, -0.022064209, 0.014762878, -0.013282776, ...
http://arxiv.org/abs/2305.05410v2
Large Language Models Need Holistically Thought in Medical Conversational QA
Yixuan Weng, Bin Li, Fei Xia, Minjun Zhu, Bin Sun, Shizhu He, Kang Liu, Jun Zhao
cs.CL
The medical conversational question answering (CQA) system aims at providing a series of professional medical services to improve the efficiency of medical care. Despite the success of large language models (LLMs) in complex reasoning tasks in various fields, such as mathematics, logic, and commonsense QA, they still n...
http://arxiv.org/pdf/2305.05410v2
2023-05-10T13:40:11
2023-05-09T12:57:28
{ "title": [ -0.0019083023, -0.02456665, -0.033355713, -0.037841797, -0.05722046, -0.028198242, -0.030197144, 0.01600647, -0.01096344, 0.02671814, -0.023666382, -0.035736084, 0.020721436, -0.050872803, 0.005809784, -0.02268982, 0.01676941, 0.0254...
http://arxiv.org/abs/2305.05403v1
Completeness, Recall, and Negation in Open-World Knowledge Bases: A Survey
Simon Razniewski, Hiba Arnaout, Shrestha Ghosh, Fabian Suchanek
cs.AI, cs.CL, cs.DB, cs.DL
General-purpose knowledge bases (KBs) are a cornerstone of knowledge-centric AI. Many of them are constructed pragmatically from Web sources, and are thus far from complete. This poses challenges for the consumption as well as the curation of their content. While several surveys target the problem of completing incompl...
http://arxiv.org/pdf/2305.05403v1
2023-05-09T12:50:16
2023-05-09T12:50:16
{ "title": [ -0.012748718, 0.010498047, -0.030929565, -0.091796875, 0.01600647, 0.025131226, -0.052246094, -0.006717682, 0.027038574, 0.020339966, -0.06896973, 0.0057525635, -0.0014600754, 0.009124756, -0.04522705, -0.018798828, -0.003168106, -0....
http://arxiv.org/abs/2305.05402v2
Consistent Text Categorization using Data Augmentation in e-Commerce
Guy Horowitz, Stav Yanovsky Daye, Noa Avigdor-Elgrabli, Ariel Raviv
cs.LG, cs.CL, cs.IR
The categorization of massive e-Commerce data is a crucial, well-studied task, which is prevalent in industrial settings. In this work, we aim to improve an existing product categorization model that is already in use by a major web company, serving multiple applications. At its core, the product categorization model i...
http://arxiv.org/pdf/2305.05402v2
2023-05-30T08:47:25
2023-05-09T12:47:28
{ "title": [ 0.00008624792, 0.01222229, -0.054901123, -0.04876709, 0.024658203, -0.024993896, -0.008384705, -0.006111145, -0.00217247, 0.032073975, -0.037750244, 0.031677246, 0.0043525696, -0.0013580322, -0.035888672, -0.009399414, -0.020095825, ...
http://arxiv.org/abs/2305.05393v1
CaseEncoder: A Knowledge-enhanced Pre-trained Model for Legal Case Encoding
Yixiao Ma, Yueyue Wu, Weihang Su, Qingyao Ai, Yiqun Liu
cs.IR, cs.CL
Legal case retrieval is a critical process for modern legal information systems. While recent studies have utilized pre-trained language models (PLMs) based on the general domain self-supervised pre-training paradigm to build models for legal case retrieval, there are limitations in using general domain PLMs as backbon...
http://arxiv.org/pdf/2305.05393v1
2023-05-09T12:40:19
2023-05-09T12:40:19
{ "title": [ -0.013786316, -0.019012451, -0.04660034, -0.061462402, -0.0018014908, 0.00029993057, -0.046875, 0.007003784, 0.025772095, 0.040100098, -0.036132812, -0.0025787354, 0.03881836, -0.0062789917, -0.04171753, 0.009552002, -0.013244629, 0....
http://arxiv.org/abs/2305.05390v1
COKE: A Cognitive Knowledge Graph for Machine Theory of Mind
Jincenzi Wu, Zhuang Chen, Jiawen Deng, Sahand Sabour, Minlie Huang
cs.CL
Theory of mind (ToM) refers to humans' ability to understand and infer the desires, beliefs, and intentions of others. The acquisition of ToM plays a key role in humans' social cognition and interpersonal relations. Though indispensable for social intelligence, ToM is still lacking for modern AI and NLP systems since t...
http://arxiv.org/pdf/2305.05390v1
2023-05-09T12:36:58
2023-05-09T12:36:58
{ "title": [ -0.04058838, -0.022018433, -0.028930664, -0.028518677, 0.0027313232, 0.000269413, -0.008224487, -0.0031280518, -0.0076026917, 0.02017212, -0.02571106, 0.006931305, -0.00015544891, -0.011405945, -0.048034668, -0.007762909, -0.008171082, ...
http://arxiv.org/abs/2305.05378v1
PLM-GNN: A Webpage Classification Method based on Joint Pre-trained Language Model and Graph Neural Network
Qiwei Lang, Jingbo Zhou, Haoyi Wang, Shiqi Lyu, Rui Zhang
cs.CL
The number of web pages is growing at an exponential rate, accumulating massive amounts of data on the web. It is one of the key processes to classify webpages in web information mining. Some classical methods are based on manually building features of web pages and training classifiers based on machine learning or dee...
http://arxiv.org/pdf/2305.05378v1
2023-05-09T12:19:10
2023-05-09T12:19:10
{ "title": [ -0.003414154, -0.004512787, -0.033111572, -0.032592773, -0.00018656254, -0.017166138, -0.017913818, -0.015037537, -0.00182724, 0.00028538704, 0.012710571, 0.0087509155, -0.025222778, -0.045928955, -0.0022506714, -0.002998352, -0.02052307...
http://arxiv.org/abs/2305.05364v1
Large Language Model Programs
Imanol Schlag, Sainbayar Sukhbaatar, Asli Celikyilmaz, Wen-tau Yih, Jason Weston, Jürgen Schmidhuber, Xian Li
cs.LG, cs.AI, cs.CL
In recent years, large pre-trained language models (LLMs) have demonstrated the ability to follow instructions and perform novel tasks from a few examples. The possibility to parameterise an LLM through such in-context examples widens their capability at a much lower cost than finetuning. We extend this line of reasoni...
http://arxiv.org/pdf/2305.05364v1
2023-05-09T11:55:36
2023-05-09T11:55:36
{ "title": [ -0.027938843, -0.012321472, -0.037597656, -0.07879639, -0.01725769, 0.007144928, -0.08081055, 0.0143966675, 0.0046463013, 0.043945312, -0.015083313, -0.02067566, 0.012489319, -0.047424316, -0.01727295, 0.02166748, 0.0020656586, 0.022...
http://arxiv.org/abs/2305.05352v4
A Taxonomy of Foundation Model based Systems for Responsible-AI-by-Design
Qinghua Lu, Liming Zhu, Xiwei Xu, Zhenchang Xing, Jon Whittle
cs.SE, cs.AI, cs.CL
The recent release of large language model (LLM) based chatbots, such as ChatGPT, has attracted significant attention on foundation models. It is widely believed that foundation models will serve as the fundamental building blocks for future AI systems. As foundation models are in their early stages, the design of foun...
http://arxiv.org/pdf/2305.05352v4
2023-06-25T23:14:13
2023-05-09T11:37:16
{ "title": [ -0.026321411, 0.00082588196, -0.042022705, -0.045074463, 0.012336731, -0.016952515, -0.0068969727, -0.009719849, -0.010734558, 0.011238098, -0.0423584, 0.0099105835, -0.022338867, -0.02053833, -0.017028809, -0.011367798, -0.01713562, ...
http://arxiv.org/abs/2305.06154v1
Alleviating Over-smoothing for Unsupervised Sentence Representation
Nuo Chen, Linjun Shou, Ming Gong, Jian Pei, Bowen Cao, Jianhui Chang, Daxin Jiang, Jia Li
cs.CL
Currently, learning better unsupervised sentence representations is the pursuit of many natural language processing communities. Lots of approaches based on pre-trained language models (PLMs) and contrastive learning have achieved promising results on this task. Experimentally, we observe that the over-smoothing proble...
http://arxiv.org/pdf/2305.06154v1
2023-05-09T11:00:02
2023-05-09T11:00:02
{ "title": [ 0.00843811, 0.010093689, -0.029678345, -0.04598999, 0.02168274, -0.009765625, -0.026489258, -0.020599365, 0.021224976, 0.024002075, -0.022827148, -0.0027008057, 0.020843506, -0.0057296753, -0.026428223, 0.009468079, -0.0154953, 0.055...
http://arxiv.org/abs/2305.05335v1
Rudolf Christoph Eucken at SemEval-2023 Task 4: An Ensemble Approach for Identifying Human Values from Arguments
Sougata Saha, Rohini Srihari
cs.CL
The subtle human values we acquire through life experiences govern our thoughts and gets reflected in our speech. It plays an integral part in capturing the essence of our individuality and making it imperative to identify such values in computational systems that mimic human actions. Computational argumentation is a f...
http://arxiv.org/pdf/2305.05335v1
2023-05-09T10:54:34
2023-05-09T10:54:34
{ "title": [ 0.049194336, 0.030853271, -0.029541016, 0.004951477, 0.014587402, -0.040130615, 0.018127441, -0.02468872, -0.01940918, 0.034484863, 0.02645874, -0.017745972, 0.00070142746, -0.049682617, 0.0011606216, -0.01828003, -0.020996094, 0.051...
http://arxiv.org/abs/2305.05334v1
ArgU: A Controllable Factual Argument Generator
Sougata Saha, Rohini Srihari
cs.CL
Effective argumentation is essential towards a purposeful conversation with a satisfactory outcome. For example, persuading someone to reconsider smoking might involve empathetic, well founded arguments based on facts and expert opinions about its ill-effects and the consequences on one's family. However, the automatic...
http://arxiv.org/pdf/2305.05334v1
2023-05-09T10:49:45
2023-05-09T10:49:45
{ "title": [ 0.018447876, -0.018310547, -0.03744507, -0.07244873, 0.006580353, -0.0062217712, -0.009353638, -0.00409317, -0.034423828, 0.034240723, 0.025802612, 0.0035915375, -0.018066406, -0.029296875, -0.036865234, 0.0004515648, -0.014350891, 0...
http://arxiv.org/abs/2305.05331v1
Explainable Recommender with Geometric Information Bottleneck
Hanqi Yan, Lin Gui, Menghan Wang, Kun Zhang, Yulan He
cs.IR, cs.CL
Explainable recommender systems can explain their recommendation decisions, enhancing user trust in the systems. Most explainable recommender systems either rely on human-annotated rationales to train models for explanation generation or leverage the attention mechanism to extract important text spans from reviews as e...
http://arxiv.org/pdf/2305.05331v1
2023-05-09T10:38:36
2023-05-09T10:38:36
{ "title": [ -0.0046310425, -0.00097608566, -0.031051636, -0.037872314, -0.009178162, -0.007091522, 0.001616478, 0.0018405914, -0.015975952, 0.021316528, -0.01965332, 0.021621704, -0.010482788, -0.006336212, -0.012336731, -0.0057258606, -0.0066184998...
http://arxiv.org/abs/2305.05325v1
Detection of depression on social networks using transformers and ensembles
Ilija Tavchioski, Marko Robnik-Šikonja, Senja Pollak
cs.CL, cs.AI
As the impact of technology on our lives is increasing, we witness increased use of social media that became an essential tool not only for communication but also for sharing information with community about our thoughts and feelings. This can be observed also for people with mental health disorders such as depression ...
http://arxiv.org/pdf/2305.05325v1
2023-05-09T10:21:14
2023-05-09T10:21:14
{ "title": [ 0.04321289, 0.00970459, 0.0016365051, -0.04486084, 0.02859497, -0.001074791, -0.015380859, -0.03378296, 0.009017944, 0.02281189, -0.039093018, 0.009887695, 0.019714355, -0.018371582, -0.035614014, 0.001996994, -0.010787964, 0.0339355...
http://arxiv.org/abs/2305.05311v1
Structured Sentiment Analysis as Transition-based Dependency Parsing
Daniel Fernández-González
cs.CL, cs.AI, 68T50, I.2.7
Structured sentiment analysis (SSA) aims to automatically extract people's opinions from a text in natural language and adequately represent that information in a graph structure. One of the most accurate methods for performing SSA was recently proposed and consists of approaching it as a dependency parsing task. Altho...
http://arxiv.org/pdf/2305.05311v1
2023-05-09T10:03:34
2023-05-09T10:03:34
{ "title": [ -0.015129089, -0.0019741058, -0.03817749, -0.09466553, 0.012001038, 0.018859863, -0.067871094, 0.0146102905, -0.0033016205, 0.02633667, -0.023803711, -0.00982666, -0.03378296, -0.0009012222, -0.07019043, -0.027648926, -0.011726379, 0...
http://arxiv.org/abs/2305.05302v1
The Perfect Victim: Computational Analysis of Judicial Attitudes towards Victims of Sexual Violence
Eliya Habba, Renana Keydar, Dan Bareket, Gabriel Stanovsky
cs.CL
We develop computational models to analyze court statements in order to assess judicial attitudes toward victims of sexual violence in the Israeli court system. The study examines the resonance of "rape myths" in the criminal justice system's response to sex crimes, in particular in judicial assessment of victim's cred...
http://arxiv.org/pdf/2305.05302v1
2023-05-09T09:45:44
2023-05-09T09:45:44
{ "title": [ 0.010902405, 0.010635376, -0.024932861, -0.02267456, 0.0021800995, -0.030899048, 0.029464722, -0.0128479, 0.021728516, 0.028808594, -0.045196533, 0.019973755, 0.016983032, -0.0079574585, -0.049102783, -0.0028190613, -0.01902771, 0.00...
http://arxiv.org/abs/2305.06156v2
The Vault: A Comprehensive Multilingual Dataset for Advancing Code Understanding and Generation
Dung Nguyen Manh, Nam Le Hai, Anh T. V. Dau, Anh Minh Nguyen, Khanh Nghiem, Jin Guo, Nghi D. Q. Bui
cs.CL, cs.AI, cs.PL, cs.SE
We present The Vault, a dataset of high-quality code-text pairs in multiple programming languages for training large language models to understand and generate code. We present methods for thoroughly extracting samples that use both rule-based and deep learning-based methods to ensure that they contain high-quality pai...
http://arxiv.org/pdf/2305.06156v2
2023-10-30T11:05:38
2023-05-09T09:35:03
{ "title": [ -0.037109375, -0.03930664, -0.0390625, -0.08087158, -0.0013160706, 0.008110046, -0.07196045, 0.030548096, 0.01966858, 0.023071289, -0.0058059692, -0.0016012192, 0.009216309, -0.02758789, -0.02645874, 0.010017395, -0.0009551048, 0.018...
http://arxiv.org/abs/2305.05295v2
Boosting Zero-shot Cross-lingual Retrieval by Training on Artificially Code-Switched Data
Robert Litschko, Ekaterina Artemova, Barbara Plank
cs.CL, cs.IR
Transferring information retrieval (IR) models from a high-resource language (typically English) to other languages in a zero-shot fashion has become a widely adopted approach. In this work, we show that the effectiveness of zero-shot rankers diminishes when queries and documents are present in different languages. Mot...
http://arxiv.org/pdf/2305.05295v2
2023-05-26T13:16:42
2023-05-09T09:32:19
{ "title": [ 0.023971558, -0.014266968, -0.015350342, -0.026412964, -0.0013742447, -0.018936157, 0.014091492, -0.011474609, 0.028915405, 0.044952393, -0.025650024, 0.007259369, 0.013511658, -0.006412506, -0.04244995, 0.0067749023, -0.012397766, 0...
http://arxiv.org/abs/2305.05290v1
Dialogue Planning via Brownian Bridge Stochastic Process for Goal-directed Proactive Dialogue
Jian Wang, Dongding Lin, Wenjie Li
cs.CL, cs.LG
Goal-directed dialogue systems aim to proactively reach a pre-determined target through multi-turn conversations. The key to achieving this task lies in planning dialogue paths that smoothly and coherently direct conversations towards the target. However, this is a challenging and under-explored task. In this work, we ...
http://arxiv.org/pdf/2305.05290v1
2023-05-09T09:28:23
2023-05-09T09:28:23
{ "title": [ -0.01651001, 0.0178833, -0.05105591, -0.068603516, 0.011917114, -0.00060129166, -0.034973145, 0.018707275, -0.021713257, 0.006542206, 0.03552246, -0.010032654, -0.04034424, -0.05645752, -0.013877869, -0.016967773, -0.010887146, 0.038...
http://arxiv.org/abs/2305.05280v2
VCSUM: A Versatile Chinese Meeting Summarization Dataset
Han Wu, Mingjie Zhan, Haochen Tan, Zhaohui Hou, Ding Liang, Linqi Song
cs.CL, cs.AI
Compared to news and chat summarization, the development of meeting summarization is hugely decelerated by the limited data. To this end, we introduce a versatile Chinese meeting summarization dataset, dubbed VCSum, consisting of 239 real-life meetings, with a total duration of over 230 hours. We claim our dataset is v...
http://arxiv.org/pdf/2305.05280v2
2023-05-15T09:30:39
2023-05-09T09:07:15
{ "title": [ -0.016067505, 0.014701843, -0.06695557, -0.094055176, 0.020599365, 0.0010871887, -0.07757568, 0.0009174347, 0.026901245, 0.04611206, -0.010978699, 0.00054836273, 0.020599365, -0.00440979, -0.062072754, 0.015792847, -0.002040863, 0.03...
http://arxiv.org/abs/2305.05271v1
Robust Acoustic and Semantic Contextual Biasing in Neural Transducers for Speech Recognition
Xuandi Fu, Kanthashree Mysore Sathyendra, Ankur Gandhe, Jing Liu, Grant P. Strimel, Ross McGowan, Athanasios Mouchtaris
cs.CL, cs.LG, cs.SD, eess.AS
Attention-based contextual biasing approaches have shown significant improvements in the recognition of generic and/or personal rare-words in End-to-End Automatic Speech Recognition (E2E ASR) systems like neural transducers. These approaches employ cross-attention to bias the model towards specific contextual entities ...
http://arxiv.org/pdf/2305.05271v1
2023-05-09T08:51:44
2023-05-09T08:51:44
{ "title": [ 0.015563965, 0.029922485, -0.04525757, -0.03842163, 0.02494812, -0.025405884, -0.005821228, -0.02180481, -0.016433716, 0.0715332, -0.02758789, -0.0031814575, 0.027389526, -0.013381958, -0.04711914, -0.0016393661, -0.023132324, 0.0579...
http://arxiv.org/abs/2305.05253v1
Attack Named Entity Recognition by Entity Boundary Interference
Yifei Yang, Hongqiu Wu, Hai Zhao
cs.CL
Named Entity Recognition (NER) is a cornerstone NLP task while its robustness has been given little attention. This paper rethinks the principles of NER attacks derived from sentence classification, as they can easily violate the label consistency between the original and adversarial NER examples. This is due to the fi...
http://arxiv.org/pdf/2305.05253v1
2023-05-09T08:21:11
2023-05-09T08:21:11
{ "title": [ 0.014701843, 0.008605957, -0.047088623, -0.087890625, 0.030456543, -0.012016296, -0.02355957, 0.0027542114, -0.047943115, 0.038208008, -0.00749588, 0.024383545, 0.023895264, -0.0317688, 0.0041389465, 0.0039024353, -0.024154663, 0.046...
http://arxiv.org/abs/2305.05252v5
Distilling Script Knowledge from Large Language Models for Constrained Language Planning
Siyu Yuan, Jiangjie Chen, Ziquan Fu, Xuyang Ge, Soham Shah, Charles Robert Jankowski, Yanghua Xiao, Deqing Yang
cs.CL, cs.AI
In everyday life, humans often plan their actions by following step-by-step instructions in the form of goal-oriented scripts. Previous work has exploited language models (LMs) to plan for abstract goals of stereotypical activities (e.g., "make a cake"), but leaves more specific goals with multi-facet constraints under...
http://arxiv.org/pdf/2305.05252v5
2023-05-26T06:17:17
2023-05-09T08:19:32
{ "title": [ -0.049865723, -0.028305054, -0.041870117, -0.06781006, -0.023117065, 0.017791748, -0.07495117, 0.03378296, -0.010322571, 0.011856079, -0.04360962, 0.0129852295, -0.00409317, -0.015586853, -0.011749268, -0.021636963, 0.010246277, 0.02...
http://arxiv.org/abs/2305.06161v1
StarCoder: may the source be with you!
Raymond Li, Loubna Ben Allal, Yangtian Zi, Niklas Muennighoff, Denis Kocetkov, Chenghao Mou, Marc Marone, Christopher Akiki, Jia Li, Jenny Chim, Qian Liu, Evgenii Zheltonozhskii, Terry Yue Zhuo, Thomas Wang, Olivier Dehaene, Mishig Davaadorj, Joel Lamy-Poirier, João Monteiro, Oleh Shliazhko, Nicolas Gontier, Nicholas M...
cs.CL, cs.AI, cs.PL, cs.SE
The BigCode community, an open-scientific collaboration working on the responsible development of Large Language Models for Code (Code LLMs), introduces StarCoder and StarCoderBase: 15.5B parameter models with 8K context length, infilling capabilities and fast large-batch inference enabled by multi-query attention. Sta...
http://arxiv.org/pdf/2305.06161v1
2023-05-09T08:16:42
2023-05-09T08:16:42
{ "title": [ 0.024047852, -0.012184143, -0.021057129, -0.06750488, -0.025894165, 0.013198853, -0.06173706, 0.0057868958, 0.007835388, 0.038513184, 0.0053977966, -0.017623901, 0.077819824, -0.053863525, 0.013092041, 0.015808105, 0.016616821, 0.005...
http://arxiv.org/abs/2305.05226v2
Multi-Teacher Knowledge Distillation For Text Image Machine Translation
Cong Ma, Yaping Zhang, Mei Tu, Yang Zhao, Yu Zhou, Chengqing Zong
cs.CL
Text image machine translation (TIMT) has been widely used in various real-world applications, which translates source language texts in images into another target language sentence. Existing methods on TIMT are mainly divided into two categories: the recognition-then-translation pipeline model and the end-to-end model...
http://arxiv.org/pdf/2305.05226v2
2023-05-10T02:31:05
2023-05-09T07:41:17
{ "title": [ -0.023605347, -0.004878998, -0.024780273, -0.04537964, -0.018508911, -0.010131836, -0.047912598, -0.0018262863, 0.029663086, 0.05734253, -0.047821045, 0.0007715225, 0.03475952, -0.004497528, -0.043884277, -0.0069122314, 0.0129470825, ...
http://arxiv.org/abs/2305.05214v1
Utilizing Lexical Similarity to Enable Zero-Shot Machine Translation for Extremely Low-resource Languages
Kaushal Kumar Maurya, Rahul Kejriwal, Maunendra Sankar Desarkar, Anoop Kunchukuttan
cs.CL
We address the task of machine translation from an extremely low-resource language (LRL) to English using cross-lingual transfer from a closely related high-resource language (HRL). For many of these languages, no parallel corpora are available, even monolingual corpora are limited and representations in pre-trained se...
http://arxiv.org/pdf/2305.05214v1
2023-05-09T07:23:01
2023-05-09T07:23:01
{ "title": [ -0.032592773, -0.027801514, -0.011497498, -0.05331421, -0.028366089, 0.012458801, -0.043518066, 0.018081665, 0.009170532, 0.057922363, -0.016708374, 0.01939392, 0.03744507, -0.015792847, -0.050476074, 0.029312134, -0.0014533997, 0.04...
http://arxiv.org/abs/2305.07445v1
QVoice: Arabic Speech Pronunciation Learning Application
Yassine El Kheir, Fouad Khnaisser, Shammur Absar Chowdhury, Hamdy Mubarak, Shazia Afzal, Ahmed Ali
eess.AS, cs.CL, cs.SD
This paper introduces a novel Arabic pronunciation learning application QVoice, powered with end-to-end mispronunciation detection and feedback generator module. The application is designed to support non-native Arabic speakers in enhancing their pronunciation skills, while also helping native speakers mitigate any pot...
http://arxiv.org/pdf/2305.07445v1
2023-05-09T07:21:46
2023-05-09T07:21:46
{ "title": [ 0.0021343231, -0.03668213, -0.029647827, -0.034454346, -0.032104492, -0.011924744, 0.0034217834, 0.03201294, -0.03793335, 0.04928589, -0.0054244995, -0.0013694763, 0.026229858, -0.056671143, -0.009887695, -0.021255493, -0.0004911423, ...
http://arxiv.org/abs/2305.05201v1
Exploration of Language Dependency for Japanese Self-Supervised Speech Representation Models
Takanori Ashihara, Takafumi Moriya, Kohei Matsuura, Tomohiro Tanaka
cs.CL, cs.SD, eess.AS
Self-supervised learning (SSL) has been dramatically successful not only in monolingual but also in cross-lingual settings. However, since the two settings have been studied individually in general, there has been little research focusing on how effective a cross-lingual model is in comparison with a monolingual model....
http://arxiv.org/pdf/2305.05201v1
2023-05-09T06:28:10
2023-05-09T06:28:10
{ "title": [ -0.009613037, -0.0035800934, -0.043945312, -0.095825195, -0.010971069, 0.004348755, -0.0970459, 0.021255493, -0.019866943, 0.031204224, -0.003250122, -0.014968872, 0.007232666, -0.03366089, -0.048858643, -0.0038642883, 0.0023288727, ...
http://arxiv.org/abs/2305.05191v1
COLA: Contextualized Commonsense Causal Reasoning from the Causal Inference Perspective
Zhaowei Wang, Quyet V. Do, Hongming Zhang, Jiayao Zhang, Weiqi Wang, Tianqing Fang, Yangqiu Song, Ginny Y. Wong, Simon See
cs.CL, cs.AI
Detecting commonsense causal relations (causation) between events has long been an essential yet challenging task. Given that events are complicated, an event may have different causes under various contexts. Thus, exploiting context plays an essential role in detecting causal relations. Meanwhile, previous works about...
http://arxiv.org/pdf/2305.05191v1
2023-05-09T05:56:58
2023-05-09T05:56:58
{ "title": [ -0.010932922, -0.021102905, -0.035339355, -0.080078125, 0.0069389343, 0.0062789917, -0.05014038, -0.0026741028, -0.0045433044, 0.0038757324, -0.011909485, 0.018234253, -0.016189575, -0.038269043, 0.00055646896, 0.006954193, -0.016815186,...