id
stringlengths
33
34
title
stringlengths
6
218
authors
stringlengths
5
6.98k
categories
stringlengths
5
188
summary
stringlengths
34
3.26k
link_pdf
stringlengths
33
34
updated
timestamp[s]
published
timestamp[s]
embeddings
dict
http://arxiv.org/abs/2307.02971v1
On the Cultural Gap in Text-to-Image Generation
Bingshuai Liu, Longyue Wang, Chenyang Lyu, Yong Zhang, Jinsong Su, Shuming Shi, Zhaopeng Tu
cs.CV, cs.AI, cs.CL
One challenge in text-to-image (T2I) generation is the inadvertent reflection of culture gaps present in the training data, which signifies the disparity in generated image quality when the cultural elements of the input text are rarely collected in the training set. Although various T2I models have shown impressive bu...
http://arxiv.org/pdf/2307.02971v1
2023-07-06T13:17:55
2023-07-06T13:17:55
{ "title": [ -0.003332138, -0.010978699, 0.012489319, -0.012435913, -0.026779175, -0.0042266846, -0.012435913, -0.041809082, 0.009307861, 0.051116943, -0.034606934, 0.008972168, 0.00198555, -0.009170532, -0.043304443, 0.010353088, 0.0026416779, 0...
http://arxiv.org/abs/2307.04821v1
Amplifying Limitations, Harms and Risks of Large Language Models
Michael O'Neill, Mark Connor
cs.CL, cs.AI, cs.CY
We present this article as a small gesture in an attempt to counter what appears to be exponentially growing hype around Artificial Intelligence (AI) and its capabilities, and the distraction provided by the associated talk of science-fiction scenarios that might arise if AI should become sentient and super-intelligent...
http://arxiv.org/pdf/2307.04821v1
2023-07-06T11:53:45
2023-07-06T11:53:45
{ "title": [ -0.013046265, -0.008255005, -0.042633057, -0.07421875, -0.002866745, -0.0062713623, -0.05456543, 0.0022697449, -0.0138168335, 0.036590576, -0.027633667, 0.011955261, 0.0046043396, -0.015464783, -0.021270752, 0.012878418, 0.0044021606, ...
http://arxiv.org/abs/2307.02912v1
LEA: Improving Sentence Similarity Robustness to Typos Using Lexical Attention Bias
Mario Almagro, Emilio Almazán, Diego Ortego, David Jiménez
cs.CL, cs.AI
Textual noise, such as typos or abbreviations, is a well-known issue that penalizes vanilla Transformers for most downstream tasks. We show that this is also the case for sentence similarity, a fundamental task in multiple domains, e.g. matching, retrieval or paraphrasing. Sentence similarity can be approached using cr...
http://arxiv.org/pdf/2307.02912v1
2023-07-06T10:53:50
2023-07-06T10:53:50
{ "title": [ -0.012031555, -0.005004883, -0.0032463074, -0.0368042, -0.005256653, 0.010253906, -0.04083252, -0.029510498, 0.043945312, 0.039520264, -0.03161621, 0.013244629, 0.050964355, 0.01096344, -0.05908203, 0.020492554, -0.04324341, 0.045623...
http://arxiv.org/abs/2307.02910v1
Agentività e telicità in GilBERTo: implicazioni cognitive
Agnese Lombardi, Alessandro Lenci
cs.CL
The goal of this study is to investigate whether a Transformer-based neural language model infers lexical semantics and use this information for the completion of morphosyntactic patterns. The semantic properties considered are telicity (also combined with definiteness) and agentivity. Both act at the interface between...
http://arxiv.org/pdf/2307.02910v1
2023-07-06T10:52:22
2023-07-06T10:52:22
{ "title": [ 0.004623413, 0.035858154, -0.01751709, -0.026748657, 0.009796143, -0.007534027, -0.045013428, -0.039489746, 0.010284424, 0.03265381, 0.0090408325, -0.015083313, 0.026168823, -0.06506348, -0.01890564, -0.009742737, -0.03491211, 0.0775...
http://arxiv.org/abs/2307.02892v2
The Relationship Between Speech Features Changes When You Get Depressed: Feature Correlations for Improving Speed and Performance of Depression Detection
Fuxiang Tao, Wei Ma, Xuri Ge, Anna Esposito, Alessandro Vinciarelli
cs.CL, cs.SD, eess.AS
This work shows that depression changes the correlation between features extracted from speech. Furthermore, it shows that using such an insight can improve the training speed and performance of depression detectors based on SVMs and LSTMs. The experiments were performed over the Androids Corpus, a publicly available d...
http://arxiv.org/pdf/2307.02892v2
2023-07-07T09:31:28
2023-07-06T09:54:35
{ "title": [ 0.009506226, 0.021057129, -0.031585693, -0.035186768, 0.015335083, -0.025772095, -0.04373169, -0.011260986, -0.0037574768, 0.056518555, -0.011154175, -0.02885437, 0.030014038, -0.030288696, -0.04156494, -0.024261475, 0.0149383545, 0....
http://arxiv.org/abs/2307.02882v1
Contrast Is All You Need
Burak Kilic, Florix Bex, Albert Gatt
cs.CL, cs.AI
In this study, we analyze data-scarce classification scenarios, where available labeled legal data is small and imbalanced, potentially hurting the quality of the results. We focused on two finetuning objectives; SetFit (Sentence Transformer Finetuning), a contrastive learning setup, and a vanilla finetuning setup on a...
http://arxiv.org/pdf/2307.02882v1
2023-07-06T09:36:54
2023-07-06T09:36:54
{ "title": [ 0.06707764, -0.0022850037, 0.009262085, -0.07989502, -0.012199402, -0.007865906, -0.042022705, -0.0029735565, 0.0289917, 0.06311035, -0.02053833, -0.0309906, 0.0262146, -0.019714355, -0.0435791, -0.013122559, 0.02394104, -0.009506226...
http://arxiv.org/abs/2307.02863v4
ValiTex -- a unified validation framework for computational text-based measures of social science constructs
Lukas Birkenmaier, Claudia Wagner, Clemens Lechner
cs.CL
Guidance on how to validate computational text-based measures of social science constructs is fragmented. While scholars generally acknowledge the importance of validating their text-based measures, they often lack common terminology and a unified framework to do so. This paper introduces ValiTex, a new validation fram...
http://arxiv.org/pdf/2307.02863v4
2023-10-04T15:34:00
2023-07-06T09:03:10
{ "title": [ -0.018035889, 0.006198883, -0.026687622, -0.03866577, 0.009803772, -0.0030384064, -0.041046143, -0.04373169, -0.008872986, 0.024032593, -0.03555298, 0.039245605, -0.026397705, -0.014541626, -0.040252686, -0.010429382, -0.024368286, 0...
http://arxiv.org/abs/2307.02849v1
NatLogAttack: A Framework for Attacking Natural Language Inference Models with Natural Logic
Zi'ou Zheng, Xiaodan Zhu
cs.CL
Reasoning has been a central topic in artificial intelligence from the beginning. The recent progress made on distributed representation and neural networks continues to improve the state-of-the-art performance of natural language inference. However, it remains an open question whether the models perform real reasoning...
http://arxiv.org/pdf/2307.02849v1
2023-07-06T08:32:14
2023-07-06T08:32:14
{ "title": [ -0.019760132, -0.023925781, -0.05331421, -0.05795288, -0.0014438629, -0.0037498474, -0.03213501, 0.0027618408, -0.024978638, 0.008796692, -0.013931274, 0.0046043396, -0.008468628, -0.029556274, -0.017486572, -0.024383545, -0.033111572, ...
http://arxiv.org/abs/2307.02839v2
Enhancing LLM with Evolutionary Fine Tuning for News Summary Generation
Le Xiao, Xiaolin Chen
cs.CL, cs.AI
News summary generation is an important task in the field of intelligence analysis, which can provide accurate and comprehensive information to help people better understand and respond to complex real-world events. However, traditional news summary generation methods face some challenges, which are limited by the mode...
http://arxiv.org/pdf/2307.02839v2
2023-07-08T07:50:44
2023-07-06T08:13:53
{ "title": [ -0.07110596, -0.017242432, -0.043884277, -0.054656982, -0.043304443, 0.022827148, -0.0703125, 0.02357483, -0.020553589, 0.059387207, -0.008514404, 0.0574646, -0.032958984, 0.00025224686, -0.054595947, 0.017166138, -0.018508911, 0.010...
http://arxiv.org/abs/2307.04532v1
Read, Look or Listen? What's Needed for Solving a Multimodal Dataset
Netta Madvil, Yonatan Bitton, Roy Schwartz
cs.CV, cs.AI, cs.CL, eess.AS
The prevalence of large-scale multimodal datasets presents unique challenges in assessing dataset quality. We propose a two-step method to analyze multimodal datasets, which leverages a small seed of human annotation to map each multimodal instance to the modalities required to process it. Our method sheds light on the...
http://arxiv.org/pdf/2307.04532v1
2023-07-06T08:02:45
2023-07-06T08:02:45
{ "title": [ -0.039489746, -0.008323669, -0.072753906, -0.049835205, -0.0032043457, -0.04019165, -0.055114746, 0.01335144, -0.010566711, 0.009094238, -0.033294678, 0.0031585693, -0.02545166, -0.018157959, -0.031799316, -0.024230957, 0.009017944, ...
http://arxiv.org/abs/2307.02830v1
Generative Zero-Shot Prompt Learning for Cross-Domain Slot Filling with Inverse Prompting
Xuefeng Li, Liwen Wang, Guanting Dong, Keqing He, Jinzheng Zhao, Hao Lei, Jiachi Liu, Weiran Xu
cs.CL
Zero-shot cross-domain slot filling aims to transfer knowledge from the labeled source domain to the unlabeled target domain. Existing models either encode slot descriptions and examples or design handcrafted question templates using heuristic rules, suffering from poor generalization capability or robustness. In this ...
http://arxiv.org/pdf/2307.02830v1
2023-07-06T07:53:46
2023-07-06T07:53:46
{ "title": [ 0.01084137, -0.0031719208, -0.022827148, -0.05126953, -0.003780365, -0.0062789917, -0.03074646, -0.022109985, -0.00044250488, 0.015327454, -0.044006348, 0.010894775, -0.01020813, -0.027999878, -0.013870239, -0.01285553, -0.01574707, ...
http://arxiv.org/abs/2307.02796v2
VerifAI: Verified Generative AI
Nan Tang, Chenyu Yang, Ju Fan, Lei Cao, Yuyu Luo, Alon Halevy
cs.DB, cs.CL, cs.LG
Generative AI has made significant strides, yet concerns about the accuracy and reliability of its outputs continue to grow. Such inaccuracies can have serious consequences such as inaccurate decision-making, the spread of false information, privacy violations, legal liabilities, and more. Although efforts to address t...
http://arxiv.org/pdf/2307.02796v2
2023-10-11T03:16:23
2023-07-06T06:11:51
{ "title": [ -0.017044067, -0.009521484, -0.06262207, -0.05783081, 0.0061149597, -0.012023926, -0.03152466, 0.0016431808, -0.007789612, 0.01197052, -0.0090408325, 0.014457703, -0.009513855, -0.03390503, -0.018005371, 0.005844116, -0.013908386, 0....
http://arxiv.org/abs/2307.02792v2
What Should Data Science Education Do with Large Language Models?
Xinming Tu, James Zou, Weijie J. Su, Linjun Zhang
cs.CY, cs.AI, cs.CL
The rapid advances of large language models (LLMs), such as ChatGPT, are revolutionizing data science and statistics. These state-of-the-art tools can streamline complex processes. As a result, it reshapes the role of data scientists. We argue that LLMs are transforming the responsibilities of data scientists, shifting...
http://arxiv.org/pdf/2307.02792v2
2023-07-07T17:56:39
2023-07-06T06:07:29
{ "title": [ -0.014404297, -0.017166138, -0.039855957, -0.025863647, -0.044189453, -0.016555786, -0.022232056, 0.011482239, -0.0041236877, 0.012756348, -0.029144287, -0.011360168, -0.0068740845, -0.04019165, 0.011894226, -0.023635864, 0.01940918, ...
http://arxiv.org/abs/2307.05355v1
UniCoRN: Unified Cognitive Signal ReconstructioN bridging cognitive signals and human language
Nuwa Xi, Sendong Zhao, Haochun Wang, Chi Liu, Bing Qin, Ting Liu
eess.SP, cs.CL
Decoding text stimuli from cognitive signals (e.g. fMRI) enhances our understanding of the human language system, paving the way for building versatile Brain-Computer Interface. However, existing studies largely focus on decoding individual word-level fMRI volumes from a restricted vocabulary, which is far too idealize...
http://arxiv.org/pdf/2307.05355v1
2023-07-06T05:26:49
2023-07-06T05:26:49
{ "title": [ -0.032592773, -0.011657715, -0.026870728, -0.018600464, -0.014060974, -0.008781433, -0.030548096, -0.03314209, 0.00422287, 0.03488159, -0.0119018555, -0.00566864, 0.039001465, -0.029083252, -0.032287598, 0.021850586, -0.0017948151, 0...
http://arxiv.org/abs/2307.02768v1
Training Models to Generate, Recognize, and Reframe Unhelpful Thoughts
Mounica Maddela, Megan Ung, Jing Xu, Andrea Madotto, Heather Foran, Y-Lan Boureau
cs.CL
Many cognitive approaches to well-being, such as recognizing and reframing unhelpful thoughts, have received considerable empirical support over the past decades, yet still lack truly widespread adoption in self-help format. A barrier to that adoption is a lack of adequately specific and diverse dedicated practice mate...
http://arxiv.org/pdf/2307.02768v1
2023-07-06T04:40:52
2023-07-06T04:40:52
{ "title": [ 0.052490234, 0.0128479, -0.0028495789, -0.03692627, -0.02557373, -0.02116394, -0.04046631, -0.04058838, 0.0012722015, 0.028717041, -0.018051147, -0.031173706, 0.04498291, -0.04547119, -0.01374054, -0.017837524, 0.008911133, 0.0312347...
http://arxiv.org/abs/2307.03679v1
Undecimated Wavelet Transform for Word Embedded Semantic Marginal Autoencoder in Security improvement and Denoising different Languages
Shreyanth S
cs.CL, cs.CR, cs.IR, cs.LG
By combining the undecimated wavelet transform within a Word Embedded Semantic Marginal Autoencoder (WESMA), this research study provides a novel strategy for improving security measures and denoising multiple languages. The incorporation of these strategies is intended to address the issues of robustness, privacy, and...
http://arxiv.org/pdf/2307.03679v1
2023-07-06T04:10:40
2023-07-06T04:10:40
{ "title": [ 0.013755798, 0.02067566, -0.032562256, -0.03253174, -0.00015878677, -0.018249512, -0.03616333, -0.025283813, -0.031707764, 0.036224365, -0.031280518, 0.0023288727, 0.058776855, -0.030410767, 0.022262573, -0.006225586, 0.010169983, 0....
http://arxiv.org/abs/2307.02763v1
Your spouse needs professional help: Determining the Contextual Appropriateness of Messages through Modeling Social Relationships
David Jurgens, Agrima Seth, Jackson Sargent, Athena Aghighi, Michael Geraci
cs.CL, cs.CY
Understanding interpersonal communication requires, in part, understanding the social context and norms in which a message is said. However, current methods for identifying offensive content in such communication largely operate independent of context, with only a few approaches considering community norms or prior con...
http://arxiv.org/pdf/2307.02763v1
2023-07-06T04:06:05
2023-07-06T04:06:05
{ "title": [ 0.023391724, 0.04537964, -0.004611969, 0.0061454773, 0.010246277, -0.02720642, -0.0054397583, -0.058135986, -0.05014038, 0.037261963, -0.00035381317, -0.015655518, 0.027511597, -0.05029297, 0.013023376, -0.023284912, 0.020019531, 0.0...
http://arxiv.org/abs/2307.02762v1
PRD: Peer Rank and Discussion Improve Large Language Model based Evaluations
Ruosen Li, Teerth Patel, Xinya Du
cs.CL, cs.AI
Nowadays, the quality of responses generated by different modern large language models (LLMs) are hard to evaluate and compare automatically. Recent studies suggest and predominantly use LLMs as a reference-free metric for open-ended question answering. More specifically, they use the recognized "strongest" LLM as the ...
http://arxiv.org/pdf/2307.02762v1
2023-07-06T04:05:44
2023-07-06T04:05:44
{ "title": [ -0.0093688965, -0.013824463, -0.025558472, -0.029281616, -0.025665283, -0.015281677, -0.035186768, -0.02079773, -0.018493652, 0.013504028, 0.005344391, -0.012359619, 0.0023517609, -0.05038452, -0.009437561, -0.005012512, -0.026748657, ...
http://arxiv.org/abs/2307.02758v2
Exploring Linguistic Style Matching in Online Communities: The Role of Social Context and Conversation Dynamics
Aparna Ananthasubramaniam, Hong Chen, Jason Yan, Kenan Alkiek, Jiaxin Pei, Agrima Seth, Lavinia Dunagan, Minje Choi, Benjamin Litterer, David Jurgens
cs.CL
Linguistic style matching (LSM) in conversations can be reflective of several aspects of social influence such as power or persuasion. However, how LSM relates to the outcomes of online communication on platforms such as Reddit is an unknown question. In this study, we analyze a large corpus of two-party conversation t...
http://arxiv.org/pdf/2307.02758v2
2023-08-26T16:56:27
2023-07-06T03:43:45
{ "title": [ -0.015670776, 0.006767273, -0.050933838, -0.019897461, 0.01966858, -0.021118164, -0.01159668, -0.0035934448, -0.025909424, 0.043518066, 0.006187439, 0.021850586, 0.029174805, -0.020629883, -0.011138916, 0.022354126, -0.0048828125, 0....
http://arxiv.org/abs/2307.02740v1
Dense Retrieval Adaptation using Target Domain Description
Helia Hashemi, Yong Zhuang, Sachith Sri Ram Kothur, Srivas Prasad, Edgar Meij, W. Bruce Croft
cs.IR, cs.CL
In information retrieval (IR), domain adaptation is the process of adapting a retrieval model to a new domain whose data distribution is different from the source domain. Existing methods in this area focus on unsupervised domain adaptation where they have access to the target document collection or supervised (often f...
http://arxiv.org/pdf/2307.02740v1
2023-07-06T02:59:47
2023-07-06T02:59:47
{ "title": [ -0.0021686554, 0.0069351196, -0.039978027, -0.036956787, 0.018615723, -0.018081665, 0.0009689331, -0.02104187, 0.031204224, 0.04034424, 0.015159607, -0.010169983, 0.024658203, -0.026184082, -0.036315918, 0.033447266, 0.007675171, -0....
http://arxiv.org/abs/2307.02738v3
RecallM: An Adaptable Memory Mechanism with Temporal Understanding for Large Language Models
Brandon Kynoch, Hugo Latapie, Dwane van der Sluis
cs.AI, cs.CL, cs.SC
Large Language Models (LLMs) have made extraordinary progress in the field of Artificial Intelligence and have demonstrated remarkable capabilities across a large variety of tasks and domains. However, as we venture closer to creating Artificial General Intelligence (AGI) systems, we recognize the need to supplement LL...
http://arxiv.org/pdf/2307.02738v3
2023-10-03T01:16:33
2023-07-06T02:51:54
{ "title": [ -0.021209717, -0.022109985, -0.047058105, -0.10235596, -0.0021209717, 0.0012989044, -0.062927246, 0.033203125, 0.028808594, 0.025665283, -0.013435364, -0.012771606, 0.009849548, -0.009010315, -0.06604004, 0.015335083, -0.0065994263, ...
http://arxiv.org/abs/2307.02729v2
Text Alignment Is An Efficient Unified Model for Massive NLP Tasks
Yuheng Zha, Yichi Yang, Ruichen Li, Zhiting Hu
cs.CL
Large language models (LLMs), typically designed as a function of next-word prediction, have excelled across extensive NLP tasks. Despite the generality, next-word prediction is often not an efficient formulation for many of the tasks, demanding an extreme scale of model parameters (10s or 100s of billions) and sometim...
http://arxiv.org/pdf/2307.02729v2
2023-11-02T03:49:19
2023-07-06T02:28:31
{ "title": [ -0.0031414032, -0.015281677, -0.055847168, -0.04864502, -0.04449463, -0.020599365, -0.058135986, 0.021652222, 0.009750366, 0.07354736, -0.016296387, 0.009475708, 0.079711914, -0.021911621, -0.0047912598, 0.029266357, -0.012161255, 0....
http://arxiv.org/abs/2307.02720v1
On-Device Constrained Self-Supervised Speech Representation Learning for Keyword Spotting via Knowledge Distillation
Gene-Ping Yang, Yue Gu, Qingming Tang, Dongsu Du, Yuzong Liu
cs.CL, cs.SD, eess.AS
Large self-supervised models are effective feature extractors, but their application is challenging under on-device budget constraints and biased dataset collection, especially in keyword spotting. To address this, we proposed a knowledge distillation-based self-supervised speech representation learning (S3RL) architec...
http://arxiv.org/pdf/2307.02720v1
2023-07-06T02:03:31
2023-07-06T02:03:31
{ "title": [ -0.0046157837, 0.012924194, -0.05001831, -0.052856445, -0.010253906, -0.009971619, -0.039398193, 0.013725281, -0.0045776367, 0.039978027, -0.06188965, 0.005630493, 0.013305664, -0.00434494, -0.025817871, -0.033843994, -0.0032291412, ...
http://arxiv.org/abs/2307.02716v1
CFSum: A Coarse-to-Fine Contribution Network for Multimodal Summarization
Min Xiao, Junnan Zhu, Haitao Lin, Yu Zhou, Chengqing Zong
cs.CL, cs.CV
Multimodal summarization usually suffers from the problem that the contribution of the visual modality is unclear. Existing multimodal summarization approaches focus on designing the fusion methods of different modalities, while ignoring the adaptive conditions under which visual modalities are useful. Therefore, we pr...
http://arxiv.org/pdf/2307.02716v1
2023-07-06T01:46:00
2023-07-06T01:46:00
{ "title": [ -0.009223938, 0.0066452026, -0.050231934, -0.096191406, -0.00051164627, 0.008346558, -0.08294678, 0.012168884, 0.0037765503, 0.049713135, -0.011734009, 0.021850586, 0.0031223297, 0.0034828186, -0.053466797, 0.011833191, -0.013221741, ...
http://arxiv.org/abs/2307.02697v3
Strahler Number of Natural Language Sentences in Comparison with Random Trees
Kumiko Tanaka-Ishii, Akira Tanaka
cs.CL, physics.data-an
The Strahler number was originally proposed to characterize the complexity of river bifurcation and has found various applications. This article proposes computation of the Strahler number's upper and lower limits for natural language sentence tree structures. Through empirical measurements across grammatically annotat...
http://arxiv.org/pdf/2307.02697v3
2023-11-09T02:54:04
2023-07-06T00:06:14
{ "title": [ -0.035461426, 0.015853882, -0.066223145, 0.013832092, -0.031341553, -0.0473938, -0.018951416, -0.016586304, -0.03552246, 0.010070801, -0.009361267, 0.018447876, -0.011154175, -0.04751587, 0.035858154, -0.02067566, -0.023544312, 0.070...
http://arxiv.org/abs/2307.02690v1
Scaling In-Context Demonstrations with Structured Attention
Tianle Cai, Kaixuan Huang, Jason D. Lee, Mengdi Wang
cs.CL, cs.AI, cs.LG
The recent surge of large language models (LLMs) highlights their ability to perform in-context learning, i.e., "learning" to perform a task from a few demonstrations in the context without any parameter updates. However, their capabilities of in-context learning are limited by the model architecture: 1) the use of dem...
http://arxiv.org/pdf/2307.02690v1
2023-07-05T23:26:01
2023-07-05T23:26:01
{ "title": [ 0.028381348, 0.021438599, -0.038024902, -0.0770874, 0.01626587, -0.022994995, -0.021240234, -0.00843811, 0.032684326, 0.02168274, -0.05267334, -0.0050849915, -0.01524353, -0.014205933, -0.030258179, -0.021194458, 0.023742676, -0.0043...
http://arxiv.org/abs/2307.02689v1
Learning Symbolic Rules over Abstract Meaning Representations for Textual Reinforcement Learning
Subhajit Chaudhury, Sarathkrishna Swaminathan, Daiki Kimura, Prithviraj Sen, Keerthiram Murugesan, Rosario Uceda-Sosa, Michiaki Tatsubori, Achille Fokoue, Pavan Kapanipathi, Asim Munawar, Alexander Gray
cs.CL
Text-based reinforcement learning agents have predominantly been neural network-based models with embeddings-based representation, learning uninterpretable policies that often do not generalize well to unseen games. On the other hand, neuro-symbolic methods, specifically those that leverage an intermediate formal repre...
http://arxiv.org/pdf/2307.02689v1
2023-07-05T23:21:05
2023-07-05T23:21:05
{ "title": [ -0.014160156, -0.012359619, -0.011550903, -0.010055542, -0.015235901, -0.0048828125, -0.011756897, -0.022277832, -0.019561768, 0.03125, -0.025268555, 0.037200928, -0.039031982, -0.0005402565, -0.03012085, -0.01713562, -0.013885498, 0...
http://arxiv.org/abs/2307.03691v2
Comparing Apples to Apples: Generating Aspect-Aware Comparative Sentences from User Reviews
Jessica Echterhoff, An Yan, Julian McAuley
cs.CL, cs.AI, cs.IR
It is time-consuming to find the best product among many similar alternatives. Comparative sentences can help to contrast one item from others in a way that highlights important features of an item that stand out. Given reviews of one or multiple items and relevant item features, we generate comparative review sentence...
http://arxiv.org/pdf/2307.03691v2
2023-07-23T17:05:06
2023-07-05T23:19:18
{ "title": [ -0.008811951, -0.0093688965, -0.0057754517, -0.02961731, -0.015701294, 0.0039253235, -0.026519775, -0.020721436, 0.008651733, 0.05029297, -0.0050849915, 0.02394104, 0.0042686462, 0.001496315, -0.07879639, 0.008995056, -0.022155762, 0...
http://arxiv.org/abs/2307.02682v2
Zero-Shot Dense Video Captioning by Jointly Optimizing Text and Moment
Yongrae Jo, Seongyun Lee, Aiden SJ Lee, Hyunji Lee, Hanseok Oh, Minjoon Seo
cs.CV, cs.CL
Dense video captioning, a task of localizing meaningful moments and generating relevant captions for videos, often requires a large, expensive corpus of annotated video segments paired with text. In an effort to minimize the annotation cost, we propose ZeroTA, a novel method for dense video captioning in a zero-shot ma...
http://arxiv.org/pdf/2307.02682v2
2023-07-11T04:10:49
2023-07-05T23:01:26
{ "title": [ -0.011558533, -0.002281189, -0.020812988, -0.03817749, -0.009490967, -0.008140564, -0.011772156, -0.008651733, 0.016311646, 0.0680542, -0.0256958, 0.017684937, 0.012519836, 0.014480591, -0.064819336, 0.010795593, -0.013298035, 0.0444...
http://arxiv.org/abs/2307.02640v1
Unsupervised Sentiment Analysis of Plastic Surgery Social Media Posts
Alexandrea K. Ramnarine
cs.CL, 68T50
The massive collection of user posts across social media platforms is primarily untapped for artificial intelligence (AI) use cases based on the sheer volume and velocity of textual data. Natural language processing (NLP) is a subfield of AI that leverages bodies of documents, known as corpora, to train computers in hu...
http://arxiv.org/pdf/2307.02640v1
2023-07-05T20:16:20
2023-07-05T20:16:20
{ "title": [ -0.0025806427, 0.0075416565, -0.03677368, -0.038208008, 0.019012451, -0.008659363, -0.008361816, -0.008369446, -0.03062439, 0.04058838, -0.02558899, 0.021713257, -0.0020103455, -0.014350891, -0.022140503, -0.0033359528, 0.017211914, ...
http://arxiv.org/abs/2307.02628v1
SkipDecode: Autoregressive Skip Decoding with Batching and Caching for Efficient LLM Inference
Luciano Del Corro, Allie Del Giorno, Sahaj Agarwal, Bin Yu, Ahmed Awadallah, Subhabrata Mukherjee
cs.CL
Autoregressive large language models (LLMs) have made remarkable progress in various natural language generation tasks. However, they incur high computation cost and latency resulting from the autoregressive token-by-token generation. To address this issue, several approaches have been proposed to reduce computational ...
http://arxiv.org/pdf/2307.02628v1
2023-07-05T19:59:09
2023-07-05T19:59:09
{ "title": [ -0.06896973, -0.030380249, -0.074645996, -0.08630371, -0.03366089, 0.010871887, -0.095458984, 0.0395813, -0.004928589, 0.031097412, -0.049346924, 0.029876709, 0.022445679, -0.014427185, -0.004283905, -0.0033473969, 0.0014276505, 0.03...
http://arxiv.org/abs/2307.02615v1
Human Inspired Progressive Alignment and Comparative Learning for Grounded Word Acquisition
Yuwei Bao, Barrett Martin Lattimer, Joyce Chai
cs.CL, cs.AI, cs.LG
Human language acquisition is an efficient, supervised, and continual process. In this work, we took inspiration from how human babies acquire their first language, and developed a computational process for word acquisition through comparative learning. Motivated by cognitive findings, we generated a small dataset that...
http://arxiv.org/pdf/2307.02615v1
2023-07-05T19:38:04
2023-07-05T19:38:04
{ "title": [ -0.019836426, 0.012542725, -0.055114746, -0.031463623, -0.012794495, -0.029815674, -0.01499176, 0.007144928, -0.014007568, 0.047180176, -0.031311035, 0.031341553, -0.0038528442, 0.0040664673, -0.046203613, -0.007637024, -0.0053215027, ...
http://arxiv.org/abs/2307.02599v2
Evade ChatGPT Detectors via A Single Space
Shuyang Cai, Wanyun Cui
cs.CL, cs.AI
ChatGPT brings revolutionary social value but also raises concerns about the misuse of AI-generated text. Consequently, an important question is how to detect whether texts are generated by ChatGPT or by human. Existing detectors are built upon the assumption that there are distributional gaps between human-generated a...
http://arxiv.org/pdf/2307.02599v2
2023-10-13T17:01:11
2023-07-05T18:48:28
{ "title": [ 0.044311523, -0.018371582, 0.030929565, 0.015472412, -0.051361084, -0.015640259, 0.0060272217, -0.05126953, 0.0079193115, 0.019577026, -0.007446289, -0.00006133318, -0.016281128, -0.042053223, 0.0031757355, 0.009773254, 0.0024147034, ...
http://arxiv.org/abs/2307.02591v2
ODD: A Benchmark Dataset for the NLP-based Opioid Related Aberrant Behavior Detection
Sunjae Kwon, Xun Wang, Weisong Liu, Emily Druhl, Minhee L. Sung, Joel I. Reisman, Wenjun Li, Robert D. Kerns, William Becker, Hong Yu
cs.CL, cs.AI
Opioid related aberrant behaviors (ORAB) present novel risk factors for opioid overdose. Previously, ORAB have been mainly assessed by survey results and by monitoring drug administrations. Such methods however, cannot scale up and do not cover the entire spectrum of aberrant behaviors. On the other hand, ORAB are wide...
http://arxiv.org/pdf/2307.02591v2
2023-07-24T00:47:23
2023-07-05T18:41:29
{ "title": [ -0.0008096695, -0.00043058395, -0.024337769, -0.062408447, 0.021087646, -0.005695343, -0.035369873, 0.002128601, -0.01838684, 0.007434845, -0.04724121, 0.042175293, -0.037628174, -0.0067100525, -0.018432617, -0.020645142, -0.023208618, ...
http://arxiv.org/abs/2308.02044v1
Artificial Intelligence in archival and historical scholarship workflow: HTS and ChatGPT
Salvatore Spina
cs.DL, cs.AI, cs.CL
This article examines the impact of Artificial Intelligence on the archival heritage digitization processes, specifically regarding the manuscripts' automatic transcription, their correction, and normalization. It highlights how digitality has compelled scholars to redefine Archive and History field and has facilitated...
http://arxiv.org/pdf/2308.02044v1
2023-07-05T18:32:28
2023-07-05T18:32:28
{ "title": [ -0.00381279, 0.013847351, -0.036315918, -0.006877899, 0.0011577606, -0.028762817, 0.026672363, -0.009994507, 0.031463623, 0.049041748, -0.026931763, 0.0023918152, 0.053527832, -0.032226562, -0.021896362, 0.02557373, 0.0069770813, -0....
http://arxiv.org/abs/2307.10188v1
Several categories of Large Language Models (LLMs): A Short Survey
Saurabh Pahune, Manoj Chandrasekharan
cs.CL, cs.LG
Large Language Models(LLMs)have become effective tools for natural language processing and have been used in many different fields. This essay offers a succinct summary of various LLM subcategories. The survey emphasizes recent developments and efforts made for various LLM kinds, including task-based financial LLMs, mu...
http://arxiv.org/pdf/2307.10188v1
2023-07-05T18:18:23
2023-07-05T18:18:23
{ "title": [ -0.04699707, -0.01574707, -0.050598145, -0.078063965, -0.0034503937, 0.011077881, -0.0814209, 0.018707275, -0.0064964294, 0.025772095, -0.009353638, 0.004650116, 0.008033752, -0.030166626, -0.028305054, 0.018341064, -0.019561768, 0.0...
http://arxiv.org/abs/2307.02570v1
Named Entity Inclusion in Abstractive Text Summarization
Sergey Berezin, Tatiana Batura
cs.CL, cs.AI, cs.LG, cs.SI
We address the named entity omission - the drawback of many current abstractive text summarizers. We suggest a custom pretraining objective to enhance the model's attention on the named entities in a text. At first, the named entity recognition model RoBERTa is trained to determine named entities in the text. After tha...
http://arxiv.org/pdf/2307.02570v1
2023-07-05T18:12:31
2023-07-05T18:12:31
{ "title": [ 0.007698059, 0.01675415, -0.024169922, -0.07946777, 0.011985779, 0.0074768066, -0.040527344, -0.017486572, 0.020950317, 0.049743652, -0.019622803, 0.04257202, 0.013290405, 0.011482239, -0.053344727, 0.020904541, -0.012123108, 0.04592...
http://arxiv.org/abs/2307.02486v2
LongNet: Scaling Transformers to 1,000,000,000 Tokens
Jiayu Ding, Shuming Ma, Li Dong, Xingxing Zhang, Shaohan Huang, Wenhui Wang, Nanning Zheng, Furu Wei
cs.CL, cs.LG
Scaling sequence length has become a critical demand in the era of large language models. However, existing methods struggle with either computational complexity or model expressivity, rendering the maximum sequence length restricted. To address this issue, we introduce LongNet, a Transformer variant that can scale seq...
http://arxiv.org/pdf/2307.02486v2
2023-07-19T12:25:35
2023-07-05T17:59:38
{ "title": [ 0.023605347, -0.011199951, -0.03677368, -0.019958496, -0.02619934, -0.010673523, -0.0016355515, 0.0015583038, -0.053649902, 0.020812988, -0.05053711, -0.010681152, 0.007007599, -0.033477783, 0.0146102905, -0.066223145, -0.0016403198, ...
http://arxiv.org/abs/2307.02485v1
Building Cooperative Embodied Agents Modularly with Large Language Models
Hongxin Zhang, Weihua Du, Jiaming Shan, Qinhong Zhou, Yilun Du, Joshua B. Tenenbaum, Tianmin Shu, Chuang Gan
cs.AI, cs.CL, cs.CV
Large Language Models (LLMs) have demonstrated impressive planning abilities in single-agent embodied tasks across various domains. However, their capacity for planning and communication in multi-agent cooperation remains unclear, even though these are crucial skills for intelligent embodied agents. In this paper, we p...
http://arxiv.org/pdf/2307.02485v1
2023-07-05T17:59:27
2023-07-05T17:59:27
{ "title": [ -0.040893555, -0.008308411, -0.022064209, -0.041015625, -0.004699707, 0.010421753, -0.043395996, -0.012664795, -0.0072746277, 0.038360596, -0.021362305, 0.0073547363, 0.009178162, -0.004211426, -0.04296875, -0.0027122498, -0.019561768, ...
http://arxiv.org/abs/2307.02477v2
Reasoning or Reciting? Exploring the Capabilities and Limitations of Language Models Through Counterfactual Tasks
Zhaofeng Wu, Linlu Qiu, Alexis Ross, Ekin Akyürek, Boyuan Chen, Bailin Wang, Najoung Kim, Jacob Andreas, Yoon Kim
cs.CL, cs.AI
The impressive performance of recent language models across a wide range of tasks suggests that they possess a degree of abstract reasoning skills. Are these skills general and transferable, or specialized to specific tasks seen during pretraining? To disentangle these effects, we propose an evaluation framework based ...
http://arxiv.org/pdf/2307.02477v2
2023-08-01T16:50:23
2023-07-05T17:50:42
{ "title": [ -0.010215759, -0.020629883, -0.019851685, -0.029953003, -0.026977539, -0.012260437, -0.05734253, -0.021759033, 0.025360107, 0.024536133, -0.014160156, -0.022659302, 0.0043411255, -0.022994995, -0.070007324, -0.0011463165, -0.010070801, ...
http://arxiv.org/abs/2307.02472v2
Deductive Additivity for Planning of Natural Language Proofs
Zayne Sprague, Kaj Bostrom, Swarat Chaudhuri, Greg Durrett
cs.CL, cs.AI
Current natural language systems designed for multi-step claim validation typically operate in two phases: retrieve a set of relevant premise statements using heuristics (planning), then generate novel conclusions from those statements using a large language model (deduction). The planning step often requires expensive...
http://arxiv.org/pdf/2307.02472v2
2023-07-06T02:16:33
2023-07-05T17:45:48
{ "title": [ -0.00055742264, -0.0029182434, -0.022872925, -0.10253906, 0.0025520325, 0.01701355, -0.061340332, -0.0014390945, -0.016296387, -0.009185791, -0.043945312, -0.0010328293, -0.010864258, -0.020217896, 0.013328552, -0.024154663, -0.018554688...
http://arxiv.org/abs/2307.02469v2
What Matters in Training a GPT4-Style Language Model with Multimodal Inputs?
Yan Zeng, Hanbo Zhang, Jiani Zheng, Jiangnan Xia, Guoqiang Wei, Yang Wei, Yuchen Zhang, Tao Kong
cs.CV, cs.CL
Recent advancements in Large Language Models (LLMs) such as GPT4 have displayed exceptional multi-modal capabilities in following open-ended instructions given images. However, the performance of these models heavily relies on design choices such as network structures, training data, and training strategies, and these ...
http://arxiv.org/pdf/2307.02469v2
2023-07-30T13:20:39
2023-07-05T17:44:28
{ "title": [ -0.02130127, 0.005935669, -0.033325195, -0.00026583672, -0.04208374, -0.02909851, -0.037200928, -0.011787415, -0.029510498, 0.010536194, 0.017700195, -0.00038599968, -0.01576233, -0.04675293, 0.0043144226, -0.008590698, -0.03152466, ...
http://arxiv.org/abs/2308.01404v2
Hoodwinked: Deception and Cooperation in a Text-Based Game for Language Models
Aidan O'Gara
cs.CL, cs.CY, cs.LG
Are current language models capable of deception and lie detection? We study this question by introducing a text-based game called $\textit{Hoodwinked}$, inspired by Mafia and Among Us. Players are locked in a house and must find a key to escape, but one player is tasked with killing the others. Each time a murder is c...
http://arxiv.org/pdf/2308.01404v2
2023-08-04T00:57:06
2023-07-05T17:22:09
{ "title": [ -0.05105591, 0.0013608932, -0.029083252, 0.0029850006, -0.015792847, -0.0099105835, -0.012237549, -0.024215698, -0.026184082, 0.03543091, -0.003730774, 0.019744873, -0.005970001, -0.014663696, -0.03390503, -0.008361816, -0.041168213, ...
http://arxiv.org/abs/2307.02443v1
An Exploratory Literature Study on Sharing and Energy Use of Language Models for Source Code
Max Hort, Anastasiia Grishina, Leon Moonen
cs.SE, cs.AI, cs.CL, cs.LG, cs.NE
Large language models trained on source code can support a variety of software development tasks, such as code recommendation and program repair. Large amounts of data for training such models benefit the models' performance. However, the size of the data and models results in long training times and high energy consum...
http://arxiv.org/pdf/2307.02443v1
2023-07-05T17:13:00
2023-07-05T17:13:00
{ "title": [ -0.032684326, -0.019638062, -0.046020508, -0.050476074, -0.0068244934, -0.008049011, -0.031204224, 0.024932861, -0.005672455, 0.03994751, -0.01802063, -0.0044937134, 0.024917603, -0.032226562, -0.02772522, 0.0025348663, -0.0076141357, ...
http://arxiv.org/abs/2307.12057v2
External Reasoning: Towards Multi-Large-Language-Models Interchangeable Assistance with Human Feedback
Akide Liu
cs.CL, cs.AI, cs.LG
Memory is identified as a crucial human faculty that allows for the retention of visual and linguistic information within the hippocampus and neurons in the brain, which can subsequently be retrieved to address real-world challenges that arise through a lifetime of learning. The resolution of complex AI tasks through t...
http://arxiv.org/pdf/2307.12057v2
2023-08-26T19:29:03
2023-07-05T17:05:32
{ "title": [ -0.025360107, -0.019927979, -0.026489258, -0.06970215, 0.005126953, 0.017700195, -0.06341553, -0.010498047, 0.0070991516, 0.025497437, -0.030792236, -0.007171631, -0.00002783537, -0.0068206787, -0.029876709, -0.0149383545, -0.008605957, ...
http://arxiv.org/abs/2307.02435v1
Exploring Continual Learning for Code Generation Models
Prateek Yadav, Qing Sun, Hantian Ding, Xiaopeng Li, Dejiao Zhang, Ming Tan, Xiaofei Ma, Parminder Bhatia, Ramesh Nallapati, Murali Krishna Ramanathan, Mohit Bansal, Bing Xiang
cs.LG, cs.CL, cs.SE
Large-scale code generation models such as Codex and CodeT5 have achieved impressive performance. However, libraries are upgraded or deprecated very frequently and re-training large-scale language models is computationally expensive. Therefore, Continual Learning (CL) is an important aspect that remains underexplored i...
http://arxiv.org/pdf/2307.02435v1
2023-07-05T16:58:39
2023-07-05T16:58:39
{ "title": [ -0.032196045, -0.016860962, -0.045196533, -0.057128906, -0.007598877, -0.014015198, -0.043914795, 0.02268982, 0.01701355, 0.025939941, -0.012252808, 0.003540039, -0.02885437, -0.021087646, -0.04953003, 0.0075683594, -0.0015659332, 0....
http://arxiv.org/abs/2307.03084v1
OpenDelta: A Plug-and-play Library for Parameter-efficient Adaptation of Pre-trained Models
Shengding Hu, Ning Ding, Weilin Zhao, Xingtai Lv, Zhen Zhang, Zhiyuan Liu, Maosong Sun
cs.LG, cs.AI, cs.CL
The scale of large pre-trained models (PTMs) poses significant challenges in adapting to downstream tasks due to the high optimization overhead and storage costs associated with full-parameter fine-tuning. To address this, many studies explore parameter-efficient tuning methods, also framed as "delta tuning", which upd...
http://arxiv.org/pdf/2307.03084v1
2023-07-05T16:30:14
2023-07-05T16:30:14
{ "title": [ -0.006389618, -0.0029964447, -0.039154053, -0.071899414, 0.023147583, -0.0073280334, -0.03866577, -0.032806396, 0.01121521, 0.016479492, -0.0015029907, -0.0020217896, -0.002166748, -0.033325195, -0.026901245, 0.011009216, 0.0121154785, ...
http://arxiv.org/abs/2307.02394v1
Won't Get Fooled Again: Answering Questions with False Premises
Shengding Hu, Yifan Luo, Huadong Wang, Xingyi Cheng, Zhiyuan Liu, Maosong Sun
cs.CL
Pre-trained language models (PLMs) have shown unprecedented potential in various fields, especially as the backbones for question-answering (QA) systems. However, they tend to be easily deceived by tricky questions such as "How many eyes does the sun have?". Such frailties of PLMs often allude to the lack of knowledge ...
http://arxiv.org/pdf/2307.02394v1
2023-07-05T16:09:21
2023-07-05T16:09:21
{ "title": [ 0.010635376, -0.0045204163, -0.033996582, -0.036590576, 0.023223877, -0.021392822, 0.02633667, -0.026473999, -0.033569336, 0.016799927, 0.0012693405, -0.0008993149, -0.020767212, -0.03616333, -0.022109985, -0.010810852, -0.0028934479, ...
http://arxiv.org/abs/2307.02390v1
Causal Discovery with Language Models as Imperfect Experts
Stephanie Long, Alexandre Piché, Valentina Zantedeschi, Tibor Schuster, Alexandre Drouin
cs.AI, cs.CL, cs.LG
Understanding the causal relationships that underlie a system is a fundamental prerequisite to accurate decision-making. In this work, we explore how expert knowledge can be used to improve the data-driven identification of causal graphs, beyond Markov equivalence classes. In doing so, we consider a setting where we ca...
http://arxiv.org/pdf/2307.02390v1
2023-07-05T16:01:38
2023-07-05T16:01:38
{ "title": [ -0.04840088, -0.023406982, -0.047729492, -0.0736084, -0.027648926, 0.0071792603, -0.06317139, 0.021377563, -0.00843811, 0.013298035, -0.026550293, 0.0065574646, -0.005012512, -0.01676941, -0.024093628, -0.0022640228, -0.009002686, 0....
http://arxiv.org/abs/2307.02358v1
To be or not to be: a translation reception study of a literary text translated into Dutch and Catalan using machine translation
Ana Guerberof Arenas, Antonio Toral
cs.CL, cs.HC
This article presents the results of a study involving the reception of a fictional story by Kurt Vonnegut translated from English into Catalan and Dutch in three conditions: machine-translated (MT), post-edited (PE) and translated from scratch (HT). 223 participants were recruited who rated the reading conditions usin...
http://arxiv.org/pdf/2307.02358v1
2023-07-05T15:18:52
2023-07-05T15:18:52
{ "title": [ -0.039154053, -0.021057129, 0.04296875, 0.056427002, -0.070739746, 0.012290955, -0.0057144165, -0.041137695, -0.012489319, 0.06512451, -0.0045928955, 0.03363037, 0.04699707, -0.008705139, -0.029708862, -0.0047912598, -0.026397705, 0....
http://arxiv.org/abs/2307.02340v1
MuLMS-AZ: An Argumentative Zoning Dataset for the Materials Science Domain
Timo Pierre Schrader, Teresa Bürkle, Sophie Henning, Sherry Tan, Matteo Finco, Stefan Grünewald, Maira Indrikova, Felix Hildebrand, Annemarie Friedrich
cs.CL
Scientific publications follow conventionalized rhetorical structures. Classifying the Argumentative Zone (AZ), e.g., identifying whether a sentence states a Motivation, a Result or Background information, has been proposed to improve processing of scholarly documents. In this work, we adapt and extend this idea to the...
http://arxiv.org/pdf/2307.02340v1
2023-07-05T14:55:18
2023-07-05T14:55:18
{ "title": [ 0.005897522, -0.03161621, -0.035217285, -0.06347656, -0.0317688, -0.008926392, -0.056365967, 0.0066566467, -0.015945435, 0.023483276, 0.012748718, 0.0029087067, -0.0035381317, -0.048736572, 0.003353119, 0.0009317398, -0.029541016, 0....
http://arxiv.org/abs/2307.02313v2
Utilizing ChatGPT Generated Data to Retrieve Depression Symptoms from Social Media
Ana-Maria Bucur
cs.CL
In this work, we present the contribution of the BLUE team in the eRisk Lab task on searching for symptoms of depression. The task consists of retrieving and ranking Reddit social media sentences that convey symptoms of depression from the BDI-II questionnaire. Given that synthetic data provided by LLMs have been prove...
http://arxiv.org/pdf/2307.02313v2
2023-07-06T11:08:51
2023-07-05T14:15:15
{ "title": [ 0.037628174, -0.0033550262, -0.00092697144, -0.024246216, -0.0030097961, -0.009880066, 0.008148193, -0.021560669, -0.013198853, 0.05166626, -0.025177002, 0.016738892, 0.025268555, -0.02810669, -0.023223877, -0.00047326088, -0.0017900467,...
http://arxiv.org/abs/2307.02301v1
Sumformer: Universal Approximation for Efficient Transformers
Silas Alberti, Niclas Dern, Laura Thesing, Gitta Kutyniok
cs.LG, cs.CL, stat.ML
Natural language processing (NLP) made an impressive jump with the introduction of Transformers. ChatGPT is one of the most famous examples, changing the perception of the possibilities of AI even outside the research community. However, besides the impressive performance, the quadratic time and space complexity of Tra...
http://arxiv.org/pdf/2307.02301v1
2023-07-05T13:59:35
2023-07-05T13:59:35
{ "title": [ 0.0060691833, 0.007873535, -0.029006958, -0.103271484, 0.038330078, 0.0019760132, -0.03881836, -0.019439697, -0.0058174133, 0.043640137, -0.010658264, 0.01966858, 0.030517578, -0.019729614, -0.027023315, 0.02104187, 0.0075531006, 0.0...
http://arxiv.org/abs/2307.02288v3
Performance Comparison of Large Language Models on VNHSGE English Dataset: OpenAI ChatGPT, Microsoft Bing Chat, and Google Bard
Xuan-Quy Dao
cs.CL, cs.HC
This paper presents a performance comparison of three large language models (LLMs), namely OpenAI ChatGPT, Microsoft Bing Chat (BingChat), and Google Bard, on the VNHSGE English dataset. The performance of BingChat, Bard, and ChatGPT (GPT-3.5) is 92.4\%, 86\%, and 79.2\%, respectively. The results show that BingChat is...
http://arxiv.org/pdf/2307.02288v3
2023-07-20T01:13:27
2023-07-05T13:40:57
{ "title": [ 0.0022506714, -0.0037593842, -0.026977539, 0.014305115, -0.02507019, -0.029052734, -0.0017986298, -0.004016876, -0.03378296, 0.059753418, 0.008201599, -0.021621704, 0.04675293, -0.03579712, -0.008682251, -0.013999939, -0.00048422813, ...
http://arxiv.org/abs/2307.02269v1
SpaceNLI: Evaluating the Consistency of Predicting Inferences in Space
Lasha Abzianidze, Joost Zwarts, Yoad Winter
cs.CL, 68T50, I.2.7
While many natural language inference (NLI) datasets target certain semantic phenomena, e.g., negation, tense & aspect, monotonicity, and presupposition, to the best of our knowledge, there is no NLI dataset that involves diverse types of spatial expressions and reasoning. We fill this gap by semi-automatically creatin...
http://arxiv.org/pdf/2307.02269v1
2023-07-05T13:08:18
2023-07-05T13:08:18
{ "title": [ -0.031311035, -0.019241333, -0.035186768, -0.040527344, -0.011962891, -0.0005130768, -0.047332764, -0.034057617, 0.0012807846, 0.01576233, -0.011917114, -0.0044898987, -0.016464233, -0.031707764, -0.04486084, -0.0012645721, 0.017822266, ...
http://arxiv.org/abs/2307.02185v2
Citation: A Key to Building Responsible and Accountable Large Language Models
Jie Huang, Kevin Chen-Chuan Chang
cs.CL, cs.AI, cs.CR
Large Language Models (LLMs) bring transformative benefits alongside unique challenges, including intellectual property (IP) and ethical concerns. This position paper explores a novel angle to mitigate these risks, drawing parallels between LLMs and established web systems. We identify "citation" - the acknowledgement ...
http://arxiv.org/pdf/2307.02185v2
2023-11-15T07:59:16
2023-07-05T10:25:45
{ "title": [ 0.00057792664, -0.008705139, -0.06072998, -0.09039307, -0.00003373623, -0.015899658, -0.060760498, 0.025634766, -0.028121948, 0.029968262, -0.016647339, -0.017211914, 0.037902832, -0.040039062, -0.0059013367, -0.00598526, -0.01322937, ...
http://arxiv.org/abs/2307.02179v1
Open-Source Large Language Models Outperform Crowd Workers and Approach ChatGPT in Text-Annotation Tasks
Meysam Alizadeh, Maël Kubli, Zeynab Samei, Shirin Dehghani, Juan Diego Bermeo, Maria Korobeynikova, Fabrizio Gilardi
cs.CL
This study examines the performance of open-source Large Language Models (LLMs) in text annotation tasks and compares it with proprietary models like ChatGPT and human-based services such as MTurk. While prior research demonstrated the high performance of ChatGPT across numerous NLP tasks, open-source LLMs like HugginC...
http://arxiv.org/pdf/2307.02179v1
2023-07-05T10:15:07
2023-07-05T10:15:07
{ "title": [ -0.006816864, 0.013679504, -0.01651001, -0.037841797, -0.021911621, -0.0014104843, -0.043945312, -0.009727478, -0.033599854, 0.040618896, -0.022781372, 0.015235901, 0.011779785, -0.0044441223, -0.016235352, -0.0060653687, -0.008766174, ...
http://arxiv.org/abs/2307.02157v1
Generative Job Recommendations with Large Language Model
Zhi Zheng, Zhaopeng Qiu, Xiao Hu, Likang Wu, Hengshu Zhu, Hui Xiong
cs.IR, cs.CL
The rapid development of online recruitment services has encouraged the utilization of recommender systems to streamline the job seeking process. Predominantly, current job recommendations deploy either collaborative filtering or person-job matching strategies. However, these models tend to operate as "black-box" syste...
http://arxiv.org/pdf/2307.02157v1
2023-07-05T09:58:08
2023-07-05T09:58:08
{ "title": [ -0.00007516146, -0.0126953125, -0.029037476, -0.06060791, -0.0011911392, 0.0008010864, -0.044952393, 0.0006246567, 0.00217247, 0.03692627, -0.020736694, 0.008323669, 0.005176544, -0.007270813, -0.04147339, 0.0026836395, -0.006752014, ...
http://arxiv.org/abs/2307.02146v1
LOAF-M2L: Joint Learning of Wording and Formatting for Singable Melody-to-Lyric Generation
Longshen Ou, Xichu Ma, Ye Wang
cs.CL, cs.SD, eess.AS
Despite previous efforts in melody-to-lyric generation research, there is still a significant compatibility gap between generated lyrics and melodies, negatively impacting the singability of the outputs. This paper bridges the singability gap with a novel approach to generating singable lyrics by jointly Learning wOrdi...
http://arxiv.org/pdf/2307.02146v1
2023-07-05T09:42:47
2023-07-05T09:42:47
{ "title": [ -0.030441284, -0.030883789, -0.046936035, -0.06112671, -0.020111084, 0.004119873, -0.05340576, 0.03451538, -0.01449585, 0.036193848, -0.023635864, 0.0107421875, 0.0101623535, -0.015563965, -0.027954102, -0.010612488, -0.02583313, 0.0...
http://arxiv.org/abs/2307.03692v1
Becoming self-instruct: introducing early stopping criteria for minimal instruct tuning
Waseem AlShikh, Manhal Daaboul, Kirk Goddard, Brock Imel, Kiran Kamble, Parikshith Kulkarni, Melisa Russak
cs.CL, cs.AI
In this paper, we introduce the Instruction Following Score (IFS), a metric that detects language models' ability to follow instructions. The metric has a dual purpose. First, IFS can be used to distinguish between base and instruct models. We benchmark publicly available base and instruct models, and show that the rat...
http://arxiv.org/pdf/2307.03692v1
2023-07-05T09:42:25
2023-07-05T09:42:25
{ "title": [ -0.005809784, -0.024291992, -0.002374649, -0.077941895, -0.030181885, 0.0031490326, -0.049560547, 0.0085372925, -0.03604126, 0.012908936, 0.0039634705, 0.015419006, -0.033721924, -0.046966553, -0.006801605, -0.0018157959, 0.017562866, ...
http://arxiv.org/abs/2307.02127v1
Leveraging Denoised Abstract Meaning Representation for Grammatical Error Correction
Hejing Cao, Dongyan Zhao
cs.CL
Grammatical Error Correction (GEC) is the task of correcting errorful sentences into grammatically correct, semantically consistent, and coherent sentences. Popular GEC models either use large-scale synthetic corpora or use a large number of human-designed rules. The former is costly to train, while the latter requires...
http://arxiv.org/pdf/2307.02127v1
2023-07-05T09:06:56
2023-07-05T09:06:56
{ "title": [ 0.0061302185, 0.0072364807, -0.012794495, -0.040283203, -0.008422852, 0.0013332367, -0.0413208, -0.01876831, 0.041748047, 0.03652954, -0.02268982, 0.0013618469, 0.012420654, -0.0022201538, -0.07067871, 0.0063819885, -0.02671814, 0.04...
http://arxiv.org/abs/2307.02120v1
Multilingual Controllable Transformer-Based Lexical Simplification
Kim Cheng Sheang, Horacio Saggion
cs.CL
Text is by far the most ubiquitous source of knowledge and information and should be made easily accessible to as many people as possible; however, texts often contain complex words that hinder reading comprehension and accessibility. Therefore, suggesting simpler alternatives for complex words without compromising mea...
http://arxiv.org/pdf/2307.02120v1
2023-07-05T08:48:19
2023-07-05T08:48:19
{ "title": [ -0.029083252, 0.007511139, -0.032348633, -0.06695557, 0.004673004, 0.030975342, -0.08099365, -0.0011386871, -0.0030002594, 0.07080078, 0.002456665, 0.018325806, 0.031829834, -0.004371643, -0.04449463, 0.027557373, -0.00036358833, 0.0...
http://arxiv.org/abs/2307.02103v1
Do predictability factors towards signing avatars hold across cultures?
Abdelhadi Soudi, Manal El Hakkaoui, Kristof Van Laerhoven
cs.CL
Avatar technology can offer accessibility possibilities and improve the Deaf-and-Hard of Hearing sign language users access to communication, education and services, such as the healthcare system. However, sign language users acceptance of signing avatars as well as their attitudes towards them vary and depend on many ...
http://arxiv.org/pdf/2307.02103v1
2023-07-05T08:22:46
2023-07-05T08:22:46
{ "title": [ 0.0032749176, -0.0034484863, -0.0070724487, -0.011505127, -0.02468872, 0.0043678284, -0.032226562, -0.029006958, 0.013374329, 0.011726379, -0.023361206, -0.010009766, 0.009010315, -0.033294678, -0.00090932846, -0.0052986145, 0.002811432,...
http://arxiv.org/abs/2307.02087v1
Different Games in Dialogue: Combining character and conversational types in strategic choice
Alafate Abulimiti
cs.CL, 68T50, I.2.4; I.2.7
In this paper, we show that investigating the interaction of conversational type (often known as language game or speech genre) with the character types of the interlocutors is worthwhile. We present a method of calculating the decision making process for selecting dialogue moves that combines character type and conver...
http://arxiv.org/pdf/2307.02087v1
2023-07-05T07:51:47
2023-07-05T07:51:47
{ "title": [ 0.0067100525, -0.007762909, -0.013969421, -0.02859497, 0.009353638, -0.026855469, 0.020935059, -0.017929077, -0.03793335, 0.025039673, 0.011054993, 0.022781372, -0.042022705, -0.026031494, -0.016693115, -0.01675415, -0.022644043, 0.0...
http://arxiv.org/abs/2307.02083v1
Leveraging multilingual transfer for unsupervised semantic acoustic word embeddings
Christiaan Jacobs, Herman Kamper
eess.AS, cs.CL
Acoustic word embeddings (AWEs) are fixed-dimensional vector representations of speech segments that encode phonetic content so that different realisations of the same word have similar embeddings. In this paper we explore semantic AWE modelling. These AWEs should not only capture phonetics but also the meaning of a wo...
http://arxiv.org/pdf/2307.02083v1
2023-07-05T07:46:54
2023-07-05T07:46:54
{ "title": [ -0.011558533, -0.009681702, -0.03189087, -0.060913086, -0.0042304993, 0.0050811768, -0.05886841, 0.0105896, 0.004623413, 0.032714844, -0.007980347, 0.012619019, 0.016677856, -0.020233154, -0.01737976, 0.014915466, 0.0009088516, 0.045...
http://arxiv.org/abs/2307.02078v1
Graph Contrastive Topic Model
Zheheng Luo, Lei Liu, Qianqian Xie, Sophia Ananiadou
cs.CL
Existing NTMs with contrastive learning suffer from the sample bias problem owing to the word frequency-based sampling strategy, which may result in false negative samples with similar semantics to the prototypes. In this paper, we aim to explore the efficient sampling strategy and contrastive learning in NTMs to addre...
http://arxiv.org/pdf/2307.02078v1
2023-07-05T07:39:47
2023-07-05T07:39:47
{ "title": [ 0.016326904, 0.023971558, -0.025634766, -0.012481689, -0.044036865, -0.032073975, -0.02571106, -0.039642334, 0.02041626, 0.040740967, -0.02861023, 0.014709473, 0.01701355, -0.044830322, 0.0141067505, 0.0036716461, 0.009094238, 0.0241...
http://arxiv.org/abs/2307.02075v1
Combating Confirmation Bias: A Unified Pseudo-Labeling Framework for Entity Alignment
Qijie Ding, Jie Yin, Daokun Zhang, Junbin Gao
cs.AI, cs.CL, cs.LG
Entity alignment (EA) aims at identifying equivalent entity pairs across different knowledge graphs (KGs) that refer to the same real-world identity. To systematically combat confirmation bias for pseudo-labeling-based entity alignment, we propose a Unified Pseudo-Labeling framework for Entity Alignment (UPL-EA) that e...
http://arxiv.org/pdf/2307.02075v1
2023-07-05T07:32:34
2023-07-05T07:32:34
{ "title": [ -0.0135650635, 0.0022392273, -0.05722046, -0.078430176, 0.032226562, -0.011444092, -0.033569336, -0.0054740906, 0.013961792, 0.029067993, 0.018814087, -0.0009236336, 0.027740479, -0.027893066, -0.03074646, 0.016281128, -0.012130737, ...
http://arxiv.org/abs/2307.02054v3
Emoji Prediction in Tweets using BERT
Muhammad Osama Nusrat, Zeeshan Habib, Mehreen Alam, Saad Ahmed Jamal
cs.CL, cs.AI
In recent years, the use of emojis in social media has increased dramatically, making them an important element in understanding online communication. However, predicting the meaning of emojis in a given text is a challenging task due to their ambiguous nature. In this study, we propose a transformer-based approach for...
http://arxiv.org/pdf/2307.02054v3
2023-08-26T11:02:16
2023-07-05T06:38:52
{ "title": [ 0.011985779, -0.012557983, -0.0067329407, -0.048187256, -0.019897461, 0.0061836243, -0.03744507, 0.004623413, -0.0048332214, 0.07489014, -0.022766113, 0.004207611, 0.05859375, -0.004043579, -0.050323486, 0.012504578, -0.014862061, 0....
http://arxiv.org/abs/2307.02053v1
Flacuna: Unleashing the Problem Solving Power of Vicuna using FLAN Fine-Tuning
Deepanway Ghosal, Yew Ken Chia, Navonil Majumder, Soujanya Poria
cs.CL
Recently, the release of INSTRUCTEVAL has provided valuable insights into the performance of large language models (LLMs) that utilize encoder-decoder or decoder-only architecture. Interestingly, despite being introduced four years ago, T5-based LLMs, such as FLAN-T5, continue to outperform the latest decoder-based LLM...
http://arxiv.org/pdf/2307.02053v1
2023-07-05T06:36:54
2023-07-05T06:36:54
{ "title": [ -0.016113281, -0.0044174194, -0.062805176, -0.024368286, -0.0073242188, -0.012916565, -0.01914978, 0.003156662, 0.0040740967, 0.02218628, -0.0028858185, -0.04864502, 0.041534424, -0.04940796, -0.011871338, -0.022583008, -0.0024986267, ...
http://arxiv.org/abs/2307.02051v1
Flowchase: a Mobile Application for Pronunciation Training
Noé Tits, Zoé Broisson
eess.AS, cs.AI, cs.CL, cs.HC, cs.SD
In this paper, we present a solution for providing personalized and instant feedback to English learners through a mobile application, called Flowchase, that is connected to a speech technology able to segment and analyze speech segmental and supra-segmental features. The speech processing pipeline receives linguistic ...
http://arxiv.org/pdf/2307.02051v1
2023-07-05T06:32:42
2023-07-05T06:32:42
{ "title": [ -0.013931274, -0.024627686, -0.047607422, -0.0541687, -0.03137207, -0.009170532, -0.041534424, 0.039367676, -0.038116455, 0.05834961, -0.007888794, 0.016738892, 0.008285522, -0.0309906, -0.031188965, -0.01386261, -0.022338867, 0.0628...
http://arxiv.org/abs/2307.02047v2
CAME: Confidence-guided Adaptive Memory Efficient Optimization
Yang Luo, Xiaozhe Ren, Zangwei Zheng, Zhuo Jiang, Xin Jiang, Yang You
cs.CL
Adaptive gradient methods, such as Adam and LAMB, have demonstrated excellent performance in the training of large language models. Nevertheless, the need for adaptivity requires maintaining second-moment estimates of the per-parameter gradients, which entails a high cost of extra memory overheads. To solve this proble...
http://arxiv.org/pdf/2307.02047v2
2023-08-07T06:21:31
2023-07-05T06:05:36
{ "title": [ -0.055603027, -0.005115509, -0.04864502, -0.05026245, 0.0055236816, 0.0016088486, -0.032226562, 0.0053596497, 0.0007162094, 0.03878784, 0.010696411, 0.018371582, -0.012954712, -0.026824951, -0.026245117, 0.02973938, 0.0008263588, 0.0...
http://arxiv.org/abs/2307.02046v2
Recommender Systems in the Era of Large Language Models (LLMs)
Wenqi Fan, Zihuai Zhao, Jiatong Li, Yunqing Liu, Xiaowei Mei, Yiqi Wang, Zhen Wen, Fei Wang, Xiangyu Zhao, Jiliang Tang, Qing Li
cs.IR, cs.AI, cs.CL
With the prosperity of e-commerce and web applications, Recommender Systems (RecSys) have become an important component of our daily life, providing personalized suggestions that cater to user preferences. While Deep Neural Networks (DNNs) have made significant advancements in enhancing recommender systems by modeling ...
http://arxiv.org/pdf/2307.02046v2
2023-08-05T05:49:13
2023-07-05T06:03:40
{ "title": [ -0.021453857, -0.015426636, -0.045684814, -0.047821045, -0.0019607544, -0.0128479, -0.024032593, 0.009437561, -0.019470215, 0.026977539, -0.028549194, 0.016448975, -0.0015668869, -0.020645142, -0.0116119385, 0.0005288124, -0.0054244995, ...
http://arxiv.org/abs/2307.02028v2
EHRSHOT: An EHR Benchmark for Few-Shot Evaluation of Foundation Models
Michael Wornow, Rahul Thapa, Ethan Steinberg, Jason A. Fries, Nigam H. Shah
cs.LG, cs.AI, cs.CL
While the general machine learning (ML) community has benefited from public datasets, tasks, and models, the progress of ML in healthcare has been hampered by a lack of such shared assets. The success of foundation models creates new challenges for healthcare ML by requiring access to shared pretrained models to valida...
http://arxiv.org/pdf/2307.02028v2
2023-11-02T09:33:35
2023-07-05T05:24:59
{ "title": [ -0.022918701, -0.00819397, -0.048614502, -0.067993164, -0.0004529953, 0.010299683, -0.06149292, -0.010353088, 0.0032215118, 0.02357483, -0.0028743744, 0.03137207, -0.0058631897, -0.038970947, -0.025131226, 0.008529663, 0.0045051575, ...
http://arxiv.org/abs/2307.02018v1
Comparative Analysis of GPT-4 and Human Graders in Evaluating Praise Given to Students in Synthetic Dialogues
Dollaya Hirunyasiri, Danielle R. Thomas, Jionghao Lin, Kenneth R. Koedinger, Vincent Aleven
cs.CL, cs.AI, cs.HC
Research suggests that providing specific and timely feedback to human tutors enhances their performance. However, it presents challenges due to the time-consuming nature of assessing tutor performance by human evaluators. Large language models, such as the AI-chatbot ChatGPT, hold potential for offering constructive f...
http://arxiv.org/pdf/2307.02018v1
2023-07-05T04:14:01
2023-07-05T04:14:01
{ "title": [ -0.008781433, -0.0014028549, -0.030639648, -0.0033340454, -0.021942139, -0.027175903, 0.02078247, 0.0067253113, -0.018585205, 0.023391724, -0.019638062, -0.009399414, -0.020385742, -0.017791748, -0.051574707, -0.044433594, -0.064331055, ...
http://arxiv.org/abs/2307.03678v1
Evaluating the Effectiveness of Large Language Models in Representing Textual Descriptions of Geometry and Spatial Relations
Yuhan Ji, Song Gao
cs.CL, cs.AI, cs.LG, I.2
This research focuses on assessing the ability of large language models (LLMs) in representing geometries and their spatial relations. We utilize LLMs including GPT-2 and BERT to encode the well-known text (WKT) format of geometries and then feed their embeddings into classifiers and regressors to evaluate the effectiv...
http://arxiv.org/pdf/2307.03678v1
2023-07-05T03:50:08
2023-07-05T03:50:08
{ "title": [ -0.031921387, -0.026443481, -0.021728516, -0.030212402, -0.025375366, -0.006790161, -0.043273926, -0.019699097, 0.026748657, 0.038635254, -0.016189575, -0.020080566, 0.031051636, -0.013160706, -0.056488037, 0.012802124, 0.00013768673, ...
http://arxiv.org/abs/2307.02009v1
Using Data Augmentations and VTLN to Reduce Bias in Dutch End-to-End Speech Recognition Systems
Tanvina Patel, Odette Scharenborg
cs.CL
Speech technology has improved greatly for norm speakers, i.e., adult native speakers of a language without speech impediments or strong accents. However, non-norm or diverse speaker groups show a distinct performance gap with norm speakers, which we refer to as bias. In this work, we aim to reduce bias against differe...
http://arxiv.org/pdf/2307.02009v1
2023-07-05T03:39:40
2023-07-05T03:39:40
{ "title": [ 0.025634766, -0.0061798096, -0.04486084, -0.07397461, -0.0042419434, -0.01020813, -0.027618408, 0.027328491, -0.0234375, 0.035217285, -0.035827637, 0.0070343018, 0.0023460388, -0.019592285, -0.034942627, -0.014816284, -0.018615723, 0...
http://arxiv.org/abs/2307.02006v1
PULSAR at MEDIQA-Sum 2023: Large Language Models Augmented by Synthetic Dialogue Convert Patient Dialogues to Medical Records
Viktor Schlegel, Hao Li, Yuping Wu, Anand Subramanian, Thanh-Tung Nguyen, Abhinav Ramesh Kashyap, Daniel Beck, Xiaojun Zeng, Riza Theresa Batista-Navarro, Stefan Winkler, Goran Nenadic
cs.CL
This paper describes PULSAR, our system submission at the ImageClef 2023 MediQA-Sum task on summarising patient-doctor dialogues into clinical records. The proposed framework relies on domain-specific pre-training, to produce a specialised language model which is trained on task-specific natural data augmented by synth...
http://arxiv.org/pdf/2307.02006v1
2023-07-05T03:31:12
2023-07-05T03:31:12
{ "title": [ 0.011619568, -0.024673462, -0.03060913, -0.03668213, -0.027786255, -0.014389038, -0.045898438, -0.00034952164, -0.015357971, 0.015335083, -0.0025959015, -0.025772095, 0.010864258, -0.042144775, 0.0022830963, -0.03982544, 0.023513794, ...
http://arxiv.org/abs/2307.01972v1
Open-Domain Hierarchical Event Schema Induction by Incremental Prompting and Verification
Sha Li, Ruining Zhao, Manling Li, Heng Ji, Chris Callison-Burch, Jiawei Han
cs.CL
Event schemas are a form of world knowledge about the typical progression of events. Recent methods for event schema induction use information extraction systems to construct a large number of event graph instances from documents, and then learn to generalize the schema from such instances. In contrast, we propose to t...
http://arxiv.org/pdf/2307.01972v1
2023-07-05T01:00:44
2023-07-05T01:00:44
{ "title": [ 0.01083374, 0.0096206665, -0.05203247, -0.1159668, 0.044830322, 0.00762558, -0.06903076, -0.010063171, 0.015007019, 0.0011796951, -0.010887146, 0.0067710876, 0.002527237, -0.044128418, 0.0013561249, 0.009979248, -0.0015468597, -0.001...
http://arxiv.org/abs/2307.01933v1
Concept2Box: Joint Geometric Embeddings for Learning Two-View Knowledge Graphs
Zijie Huang, Daheng Wang, Binxuan Huang, Chenwei Zhang, Jingbo Shang, Yan Liang, Zhengyang Wang, Xian Li, Christos Faloutsos, Yizhou Sun, Wei Wang
cs.AI, cs.CG, cs.CL, cs.SC
Knowledge graph embeddings (KGE) have been extensively studied to embed large-scale relational data for many real-world applications. Existing methods have long ignored the fact many KGs contain two fundamentally different views: high-level ontology-view concepts and fine-grained instance-view entities. They usually em...
http://arxiv.org/pdf/2307.01933v1
2023-07-04T21:37:39
2023-07-04T21:37:39
{ "title": [ -0.029953003, -0.019515991, -0.05255127, -0.07080078, -0.00982666, -0.0007343292, -0.054229736, 0.011558533, 0.013793945, 0.018615723, -0.01789856, 0.014511108, -0.016830444, -0.0077323914, -0.033813477, 0.0028839111, 0.0029582977, 0...
http://arxiv.org/abs/2307.02503v1
Natural Language Generation and Understanding of Big Code for AI-Assisted Programming: A Review
Man Fai Wong, Shangxin Guo, Ching Nam Hang, Siu Wai Ho, Chee Wei Tan
cs.SE, cs.AI, cs.CL
This paper provides a comprehensive review of the literature concerning the utilization of Natural Language Processing (NLP) techniques, with a particular focus on transformer-based large language models (LLMs) trained using Big Code, within the domain of AI-assisted programming tasks. LLMs, augmented with software nat...
http://arxiv.org/pdf/2307.02503v1
2023-07-04T21:26:51
2023-07-04T21:26:51
{ "title": [ -0.05480957, -0.04083252, -0.030334473, -0.05935669, -0.025238037, 0.016098022, -0.044799805, 0.02949524, 0.0028705597, 0.0146484375, -0.020385742, 0.00095415115, -0.009750366, -0.029800415, -0.025497437, -0.007091522, -0.0124435425, ...
http://arxiv.org/abs/2307.02502v1
Math Agents: Computational Infrastructure, Mathematical Embedding, and Genomics
Melanie Swan, Takashi Kido, Eric Roland, Renato P. dos Santos
q-bio.OT, cs.AI, cs.CL, 68R12, I.2; J.3
The advancement in generative AI could be boosted with more accessible mathematics. Beyond human-AI chat, large language models (LLMs) are emerging in programming, algorithm discovery, and theorem proving, yet their genomics application is limited. This project introduces Math Agents and mathematical embedding as fresh...
http://arxiv.org/pdf/2307.02502v1
2023-07-04T20:16:32
2023-07-04T20:16:32
{ "title": [ -0.026184082, -0.0033740997, -0.027114868, -0.03302002, 0.0041007996, -0.016296387, 0.004459381, -0.008285522, 0.0038814545, 0.04019165, -0.019073486, 0.015151978, 0.007797241, -0.0096206665, -0.033843994, 0.008483887, 0.005870819, 0...
http://arxiv.org/abs/2307.01900v1
Concept-Based Explanations to Test for False Causal Relationships Learned by Abusive Language Classifiers
Isar Nejadgholi, Svetlana Kiritchenko, Kathleen C. Fraser, Esma Balkır
cs.CL, cs.AI
Classifiers tend to learn a false causal relationship between an over-represented concept and a label, which can result in over-reliance on the concept and compromised classification accuracy. It is imperative to have methods in place that can compare different models and identify over-reliances on specific concepts. W...
http://arxiv.org/pdf/2307.01900v1
2023-07-04T19:57:54
2023-07-04T19:57:54
{ "title": [ 0.0069389343, -0.027236938, -0.0043144226, -0.055480957, -0.023284912, -0.010032654, -0.0038166046, -0.000058233738, -0.028961182, -0.014968872, 0.0066070557, 0.01083374, -0.0335083, -0.04559326, 0.007701874, -0.035247803, -0.02281189, ...
http://arxiv.org/abs/2307.01896v2
Transformed Protoform Reconstruction
Young Min Kim, Kalvin Chang, Chenxuan Cui, David Mortensen
cs.CL
Protoform reconstruction is the task of inferring what morphemes or words appeared like in the ancestral languages of a set of daughter languages. Meloni et al. (2021) achieved the state-of-the-art on Latin protoform reconstruction with an RNN-based encoder-decoder with attention model. We update their model with the s...
http://arxiv.org/pdf/2307.01896v2
2023-07-06T03:51:57
2023-07-04T19:40:20
{ "title": [ 0.0025291443, 0.004196167, -0.03564453, -0.046569824, -0.05508423, -0.0061416626, -0.07385254, -0.009239197, 0.017089844, 0.047851562, -0.012191772, -0.028152466, 0.06781006, -0.06210327, 0.019302368, 0.01889038, 0.041290283, -0.0184...
http://arxiv.org/abs/2307.01881v1
ProPILE: Probing Privacy Leakage in Large Language Models
Siwon Kim, Sangdoo Yun, Hwaran Lee, Martin Gubri, Sungroh Yoon, Seong Joon Oh
cs.CR, cs.CL
The rapid advancement and widespread use of large language models (LLMs) have raised significant concerns regarding the potential leakage of personally identifiable information (PII). These models are often trained on vast quantities of web-collected data, which may inadvertently include sensitive personal data. This p...
http://arxiv.org/pdf/2307.01881v1
2023-07-04T18:53:47
2023-07-04T18:53:47
{ "title": [ -0.03475952, -0.017196655, -0.047546387, -0.06439209, 0.0073776245, -0.00046277046, -0.03161621, 0.022781372, -0.016235352, 0.020126343, -0.0151901245, 0.022644043, 0.004142761, -0.013931274, -0.024917603, 0.004108429, -0.02482605, 0...
http://arxiv.org/abs/2307.01878v1
KDSTM: Neural Semi-supervised Topic Modeling with Knowledge Distillation
Weijie Xu, Xiaoyu Jiang, Jay Desai, Bin Han, Fuqin Yan, Francis Iannacci
cs.CL, cs.AI
In text classification tasks, fine tuning pretrained language models like BERT and GPT-3 yields competitive accuracy; however, both methods require pretraining on large text datasets. In contrast, general topic modeling methods possess the advantage of analyzing documents to extract meaningful patterns of words without...
http://arxiv.org/pdf/2307.01878v1
2023-07-04T18:49:19
2023-07-04T18:49:19
{ "title": [ -0.02394104, 0.01159668, -0.044647217, -0.05206299, 0.0033626556, 0.003118515, -0.065979004, -0.024871826, 0.040496826, 0.017196655, -0.010948181, 0.011306763, -0.0039787292, -0.020568848, -0.031463623, 0.008926392, 0.0000346303, -0....
http://arxiv.org/abs/2307.01870v1
Exploring Non-Verbal Predicates in Semantic Role Labeling: Challenges and Opportunities
Riccardo Orlando, Simone Conia, Roberto Navigli
cs.CL, cs.AI
Although we have witnessed impressive progress in Semantic Role Labeling (SRL), most of the research in the area is carried out assuming that the majority of predicates are verbs. Conversely, predicates can also be expressed using other parts of speech, e.g., nouns and adjectives. However, non-verbal predicates appear ...
http://arxiv.org/pdf/2307.01870v1
2023-07-04T18:28:59
2023-07-04T18:28:59
{ "title": [ -0.0022525787, -0.021057129, -0.020477295, -0.06970215, -0.011924744, -0.0008254051, -0.058410645, -0.003944397, -0.019439697, 0.012893677, 0.0036525726, 0.021118164, -0.014472961, -0.020385742, -0.016067505, -0.0058517456, -0.031829834,...
http://arxiv.org/abs/2307.05329v1
Decoding the Popularity of TV Series: A Network Analysis Perspective
Melody Yu
cs.SI, cs.CL, cs.NI
In this paper, we analyze the character networks extracted from three popular television series and explore the relationship between a TV show episode's character network metrics and its review from IMDB. Character networks are graphs created from the plot of a TV show that represents the interactions of characters in ...
http://arxiv.org/pdf/2307.05329v1
2023-07-04T18:16:58
2023-07-04T18:16:58
{ "title": [ -0.015548706, 0.007896423, -0.05645752, -0.021484375, 0.019073486, -0.031555176, 0.005634308, -0.0015935898, -0.034362793, 0.033447266, -0.009925842, 0.021774292, -0.023742676, -0.018341064, -0.036315918, -0.0029621124, -0.0017566681, ...
http://arxiv.org/abs/2307.01784v1
The Inner Sentiments of a Thought
Chris Gagne, Peter Dayan
cs.CL, cs.AI
Transformer-based large-scale language models (LLMs) are able to generate highly realistic text. They are duly able to express, and at least implicitly represent, a wide range of sentiments and color, from the obvious, such as valence and arousal to the subtle, such as determination and admiration. We provide a first e...
http://arxiv.org/pdf/2307.01784v1
2023-07-04T15:44:37
2023-07-04T15:44:37
{ "title": [ 0.02394104, -0.041381836, 0.037506104, 0.011886597, -0.07128906, -0.0049438477, 0.01586914, -0.027664185, 0.04284668, 0.029312134, -0.08300781, -0.025161743, 0.016693115, -0.00680542, -0.022460938, -0.024353027, 0.043914795, -0.00459...
http://arxiv.org/abs/2307.01764v1
Knowledge-Aware Audio-Grounded Generative Slot Filling for Limited Annotated Data
Guangzhi Sun, Chao Zhang, Ivan Vulić, Paweł Budzianowski, Philip C. Woodland
cs.CL
Manually annotating fine-grained slot-value labels for task-oriented dialogue (ToD) systems is an expensive and time-consuming endeavour. This motivates research into slot-filling methods that operate with limited amounts of labelled data. Moreover, the majority of current work on ToD is based solely on text as the inp...
http://arxiv.org/pdf/2307.01764v1
2023-07-04T15:05:42
2023-07-04T15:05:42
{ "title": [ 0.0040397644, -0.00217247, -0.0413208, -0.07330322, -0.01007843, -0.011505127, -0.052703857, 0.014640808, 0.0146484375, 0.03744507, -0.0552063, 0.0026664734, -0.003068924, -0.012207031, -0.06097412, -0.025009155, -0.006958008, 0.0153...
http://arxiv.org/abs/2307.01715v2
Align With Purpose: Optimize Desired Properties in CTC Models with a General Plug-and-Play Framework
Eliya Segev, Maya Alroy, Ronen Katsir, Noam Wies, Ayana Shenhav, Yael Ben-Oren, David Zar, Oren Tadmor, Jacob Bitterman, Amnon Shashua, Tal Rosenwein
cs.CL, cs.LG, cs.SD, eess.AS
Connectionist Temporal Classification (CTC) is a widely used criterion for training supervised sequence-to-sequence (seq2seq) models. It enables learning the relations between input and output sequences, termed alignments, by marginalizing over perfect alignments (that yield the ground truth), at the expense of imperfe...
http://arxiv.org/pdf/2307.01715v2
2023-07-06T07:02:45
2023-07-04T13:34:47
{ "title": [ -0.006801605, -0.007965088, -0.06951904, -0.043304443, -0.039764404, -0.017684937, -0.07446289, -0.0041046143, -0.05319214, 0.025939941, -0.01763916, 0.023651123, 0.011245728, -0.04232788, 0.04058838, -0.018844604, 0.0007452965, -0.0...
http://arxiv.org/abs/2307.01709v1
Dipping PLMs Sauce: Bridging Structure and Text for Effective Knowledge Graph Completion via Conditional Soft Prompting
Chen Chen, Yufei Wang, Aixin Sun, Bing Li, Kwok-Yan Lam
cs.CL
Knowledge Graph Completion (KGC) often requires both KG structural and textual information to be effective. Pre-trained Language Models (PLMs) have been used to learn the textual information, usually under the fine-tune paradigm for the KGC task. However, the fine-tuned PLMs often overwhelmingly focus on the textual in...
http://arxiv.org/pdf/2307.01709v1
2023-07-04T13:24:04
2023-07-04T13:24:04
{ "title": [ -0.037078857, -0.024505615, -0.053497314, -0.087402344, -0.020584106, 0.02394104, -0.0791626, 0.02053833, -0.0025596619, 0.015594482, -0.04458618, 0.035736084, -0.002735138, 0.0010490417, -0.018737793, -0.016799927, 0.0104904175, 0.0...
http://arxiv.org/abs/2307.01693v1
Racial Bias Trends in the Text of US Legal Opinions
Rohan Jinturkar
cs.CL
Although there is widespread recognition of racial bias in US law, it is unclear how such bias appears in the language of law, namely judicial opinions, and whether it varies across time period or region. Building upon approaches for measuring implicit racial bias in large-scale corpora, we approximate GloVe word embed...
http://arxiv.org/pdf/2307.01693v1
2023-07-04T12:56:23
2023-07-04T12:56:23
{ "title": [ 0.022705078, -0.006629944, -0.02960205, -0.009208679, -0.02885437, -0.036132812, 0.007534027, 0.0014705658, -0.0033855438, 0.040039062, -0.059051514, 0.009483337, 0.033691406, -0.012275696, -0.0054473877, -0.01876831, -0.0059432983, ...
http://arxiv.org/abs/2307.01680v1
Robust Hate Speech Detection in Social Media: A Cross-Dataset Empirical Evaluation
Dimosthenis Antypas, Jose Camacho-Collados
cs.CL, I.2.7
The automatic detection of hate speech online is an active research area in NLP. Most of the studies to date are based on social media datasets that contribute to the creation of hate speech detection models trained on them. However, data creation processes contain their own biases, and models inherently learn from the...
http://arxiv.org/pdf/2307.01680v1
2023-07-04T12:22:40
2023-07-04T12:22:40
{ "title": [ 0.027313232, 0.02230835, -0.045562744, -0.03656006, 0.04425049, -0.03390503, 0.00724411, -0.018295288, -0.009880066, 0.029769897, -0.0340271, 0.028015137, 0.0058555603, -0.0007352829, -0.043060303, -0.0039520264, -0.018417358, 0.0697...
http://arxiv.org/abs/2307.01673v1
Disentanglement in a GAN for Unconditional Speech Synthesis
Matthew Baas, Herman Kamper
eess.AS, cs.CL, cs.SD
Can we develop a model that can synthesize realistic speech directly from a latent space, without explicit conditioning? Despite several efforts over the last decade, previous adversarial and diffusion-based approaches still struggle to achieve this, even on small-vocabulary datasets. To address this, we propose AudioS...
http://arxiv.org/pdf/2307.01673v1
2023-07-04T12:06:07
2023-07-04T12:06:07
{ "title": [ -0.017059326, 0.014968872, -0.06500244, -0.05441284, 0.0049552917, -0.0118255615, -0.04095459, 0.016281128, -0.014022827, 0.04498291, -0.0154418945, 0.002193451, 0.044128418, -0.03616333, -0.0072631836, 0.008659363, -0.027023315, 0.0...
http://arxiv.org/abs/2307.01672v1
Boosting Norwegian Automatic Speech Recognition
Javier de la Rosa, Rolv-Arild Braaten, Per Egil Kummervold, Freddy Wetjen, Svein Arne Brygfjeld
cs.CL
In this paper, we present several baselines for automatic speech recognition (ASR) models for the two official written languages in Norway: Bokm{\aa}l and Nynorsk. We compare the performance of models of varying sizes and pre-training approaches on multiple Norwegian speech datasets. Additionally, we measure the perfor...
http://arxiv.org/pdf/2307.01672v1
2023-07-04T12:05:15
2023-07-04T12:05:15
{ "title": [ 0.025222778, -0.00029826164, -0.034423828, -0.08215332, -0.004310608, -0.006134033, -0.050048828, 0.0067634583, -0.03668213, 0.05505371, -0.045684814, 0.015792847, 0.014877319, -0.0077400208, -0.007865906, 0.004814148, 0.012519836, 0...
http://arxiv.org/abs/2307.01664v1
Unified Conversational Models with System-Initiated Transitions between Chit-Chat and Task-Oriented Dialogues
Ye Liu, Stefan Ultes, Wolfgang Minker, Wolfgang Maier
cs.CL
Spoken dialogue systems (SDSs) have been separately developed under two different categories, task-oriented and chit-chat. The former focuses on achieving functional goals and the latter aims at creating engaging social conversations without special goals. Creating a unified conversational model that can engage in both...
http://arxiv.org/pdf/2307.01664v1
2023-07-04T11:53:23
2023-07-04T11:53:23
{ "title": [ -0.0009393692, 0.005256653, -0.035247803, -0.08129883, 0.0070648193, -0.003370285, -0.07128906, 0.00011777878, 0.0059776306, 0.05596924, -0.006778717, -0.0071640015, 0.016677856, -0.018081665, -0.057250977, 0.0032482147, -0.013595581, ...
http://arxiv.org/abs/2307.02499v1
mPLUG-DocOwl: Modularized Multimodal Large Language Model for Document Understanding
Jiabo Ye, Anwen Hu, Haiyang Xu, Qinghao Ye, Ming Yan, Yuhao Dan, Chenlin Zhao, Guohai Xu, Chenliang Li, Junfeng Tian, Qian Qi, Ji Zhang, Fei Huang
cs.CL, cs.AI
Document understanding refers to automatically extract, analyze and comprehend information from various types of digital documents, such as a web page. Existing Multi-model Large Language Models (MLLMs), including mPLUG-Owl, have demonstrated promising zero-shot capabilities in shallow OCR-free text recognition, indica...
http://arxiv.org/pdf/2307.02499v1
2023-07-04T11:28:07
2023-07-04T11:28:07
{ "title": [ -0.035949707, -0.014633179, -0.042816162, -0.07904053, -0.008468628, 0.013175964, -0.06896973, 0.01600647, 0.008460999, 0.036315918, -0.014778137, -0.003932953, 0.015563965, -0.007007599, -0.047210693, -0.0019950867, -0.017501831, 0....
http://arxiv.org/abs/2307.01644v1
Insert-expansions for Tool-enabled Conversational Agents
Andreas Göldi, Roman Rietsche
cs.HC, cs.AI, cs.CL, H.5
This paper delves into an advanced implementation of Chain-of-Thought-Prompting in Large Language Models, focusing on the use of tools (or "plug-ins") within the explicit reasoning paths generated by this prompting method. We find that tool-enabled conversational agents often become sidetracked, as additional context f...
http://arxiv.org/pdf/2307.01644v1
2023-07-04T10:57:31
2023-07-04T10:57:31
{ "title": [ -0.014831543, -0.01499176, -0.031143188, -0.048095703, -0.004901886, 0.0020389557, -0.029556274, 0.00573349, 0.014373779, 0.046325684, -0.0385437, 0.005378723, 0.01979065, -0.0032787323, -0.057922363, 0.0049324036, -0.015701294, 0.01...
http://arxiv.org/abs/2307.01640v1
Chain of Thought Prompting Elicits Knowledge Augmentation
Dingjun Wu, Jing Zhang, Xinmei Huang
cs.CL
The knowledge-augmented deep learning paradigm refers to a paradigm in which domain knowledge is identified and integrated into deep models. Conventional methods typically employ task-specific approaches to gather external knowledge from various sources. In contrast, large language models are extensively pre-trained an...
http://arxiv.org/pdf/2307.01640v1
2023-07-04T10:51:16
2023-07-04T10:51:16
{ "title": [ 0.018478394, -0.0031738281, -0.05130005, -0.03994751, -0.0053138733, -0.03652954, -0.00894928, -0.0013017654, -0.014633179, 0.0032958984, -0.045410156, 0.0022525787, -0.018661499, -0.028167725, -0.018920898, -0.014511108, 0.032043457, ...
http://arxiv.org/abs/2307.01609v1
A Language Model for Grammatical Error Correction in L2 Russian
Nikita Remnev, Sergei Obiedkov, Ekaterina Rakhilina, Ivan Smirnov, Anastasia Vyrenkova
cs.CL
Grammatical error correction is one of the fundamental tasks in Natural Language Processing. For the Russian language, most of the spellcheckers available correct typos and other simple errors with high accuracy, but often fail when faced with non-native (L2) writing, since the latter contains errors that are not typic...
http://arxiv.org/pdf/2307.01609v1
2023-07-04T09:50:13
2023-07-04T09:50:13
{ "title": [ -0.00023889542, 0.0036354065, -0.0068511963, -0.033477783, -0.00491333, 0.0045280457, -0.024337769, -0.013031006, -0.028701782, 0.036468506, -0.015838623, 0.0032367706, 0.015426636, -0.028457642, -0.032165527, -0.0068244934, -0.04498291,...