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/2308.13089v1
Towards a Holistic Approach: Understanding Sociodemographic Biases in NLP Models using an Interdisciplinary Lens
Pranav Narayanan Venkit
cs.CL
The rapid growth in the usage and applications of Natural Language Processing (NLP) in various sociotechnical solutions has highlighted the need for a comprehensive understanding of bias and its impact on society. While research on bias in NLP has expanded, several challenges persist that require attention. These inclu...
http://arxiv.org/pdf/2308.13089v1
2023-08-24T21:19:48
2023-08-24T21:19:48
{ "title": [ -0.0104522705, -0.0018863678, -0.024459839, -0.027267456, -0.0236969, -0.016677856, -0.042510986, -0.031341553, 0.008552551, -0.0014591217, -0.051849365, 0.0071754456, -0.0023841858, -0.00592041, -0.017974854, -0.015792847, -0.044158936,...
http://arxiv.org/abs/2308.13081v3
Formal specification terminology for demographic agent-based models of fixed-step single-clocked simulations
Atiyah Elsheikh
cs.CL, math.DS
This document presents adequate formal terminology for the mathematical specification of a subset of Agent Based Models (ABMs) in the field of Demography. The simulation of the targeted ABMs follows a fixedstep single-clocked pattern. The proposed terminology further improves the model understanding and can act as a st...
http://arxiv.org/pdf/2308.13081v3
2023-10-20T14:05:12
2023-08-24T20:57:07
{ "title": [ -0.0016326904, -0.0016040802, -0.028121948, -0.065979004, 0.007095337, -0.00077724457, -0.03543091, -0.0063934326, -0.035980225, 0.049621582, 0.006828308, 0.018936157, 0.012077332, -0.01928711, -0.019592285, 0.02267456, -0.0022029877, ...
http://arxiv.org/abs/2308.13067v1
Causal Parrots: Large Language Models May Talk Causality But Are Not Causal
Matej Zečević, Moritz Willig, Devendra Singh Dhami, Kristian Kersting
cs.AI, cs.CL
Some argue scale is all what is needed to achieve AI, covering even causal models. We make it clear that large language models (LLMs) cannot be causal and give reason onto why sometimes we might feel otherwise. To this end, we define and exemplify a new subgroup of Structural Causal Model (SCM) that we call meta SCM wh...
http://arxiv.org/pdf/2308.13067v1
2023-08-24T20:23:13
2023-08-24T20:23:13
{ "title": [ -0.035461426, -0.01386261, -0.043792725, -0.055847168, -0.02468872, -0.01335144, -0.047454834, -0.004016876, -0.010467529, 0.009849548, -0.023101807, 0.008850098, 0.003080368, -0.02722168, -0.008644104, -0.005317688, -0.0071525574, 0...
http://arxiv.org/abs/2308.13056v2
Lexical Diversity in Kinship Across Languages and Dialects
Hadi Khalilia, Gábor Bella, Abed Alhakim Freihat, Shandy Darma, Fausto Giunchiglia
cs.CL
Languages are known to describe the world in diverse ways. Across lexicons, diversity is pervasive, appearing through phenomena such as lexical gaps and untranslatability. However, in computational resources, such as multilingual lexical databases, diversity is hardly ever represented. In this paper, we introduce a met...
http://arxiv.org/pdf/2308.13056v2
2023-10-26T12:54:30
2023-08-24T19:49:30
{ "title": [ -0.038208008, 0.013786316, -0.07354736, -0.05731201, -0.011688232, -0.020980835, -0.08148193, 0.010307312, 0.0024356842, 0.05456543, -0.02154541, 0.015625, 0.040374756, -0.029754639, -0.017730713, 0.037719727, 0.013771057, 0.05377197...
http://arxiv.org/abs/2308.13032v2
Financial News Analytics Using Fine-Tuned Llama 2 GPT Model
Bohdan M. Pavlyshenko
cs.CL, cs.AI, cs.CE, cs.IR, cs.LG
The paper considers the possibility to fine-tune Llama 2 GPT large language model (LLM) for the multitask analysis of financial news. For fine-tuning, the PEFT/LoRA based approach was used. In the study, the model was fine-tuned for the following tasks: analysing a text from financial market perspectives, highlighting ...
http://arxiv.org/pdf/2308.13032v2
2023-09-11T05:39:05
2023-08-24T18:58:10
{ "title": [ -0.023971558, -0.0018091202, -0.033447266, -0.013908386, -0.02645874, -0.004825592, -0.027511597, 0.0035114288, -0.03842163, 0.045959473, -0.020980835, 0.0095825195, 0.011947632, -0.032562256, -0.00881958, -0.027267456, 0.0063972473, ...
http://arxiv.org/abs/2308.12966v3
Qwen-VL: A Versatile Vision-Language Model for Understanding, Localization, Text Reading, and Beyond
Jinze Bai, Shuai Bai, Shusheng Yang, Shijie Wang, Sinan Tan, Peng Wang, Junyang Lin, Chang Zhou, Jingren Zhou
cs.CV, cs.CL
In this work, we introduce the Qwen-VL series, a set of large-scale vision-language models (LVLMs) designed to perceive and understand both texts and images. Starting from the Qwen-LM as a foundation, we endow it with visual capacity by the meticulously designed (i) visual receptor, (ii) input-output interface, (iii) 3...
http://arxiv.org/pdf/2308.12966v3
2023-10-13T02:41:28
2023-08-24T17:59:17
{ "title": [ -0.037475586, -0.02293396, -0.02859497, -0.028167725, -0.014289856, -0.0039978027, -0.032958984, -0.024963379, 0.015899658, 0.038024902, -0.017913818, -0.017562866, 0.027801514, -0.017852783, -0.06036377, -0.0034885406, 0.0056266785, ...
http://arxiv.org/abs/2308.12950v2
Code Llama: Open Foundation Models for Code
Baptiste Rozière, Jonas Gehring, Fabian Gloeckle, Sten Sootla, Itai Gat, Xiaoqing Ellen Tan, Yossi Adi, Jingyu Liu, Tal Remez, Jérémy Rapin, Artyom Kozhevnikov, Ivan Evtimov, Joanna Bitton, Manish Bhatt, Cristian Canton Ferrer, Aaron Grattafiori, Wenhan Xiong, Alexandre Défossez, Jade Copet, Faisal Azhar, Hugo Touvron,...
cs.CL
We release Code Llama, a family of large language models for code based on Llama 2 providing state-of-the-art performance among open models, infilling capabilities, support for large input contexts, and zero-shot instruction following ability for programming tasks. We provide multiple flavors to cover a wide range of a...
http://arxiv.org/pdf/2308.12950v2
2023-08-25T08:51:22
2023-08-24T17:39:13
{ "title": [ -0.04940796, -0.0035419464, -0.052093506, -0.048034668, -0.012901306, 0.010398865, -0.062408447, -0.004184723, -0.015083313, 0.024734497, 0.009529114, 0.014556885, 0.0076828003, -0.059692383, 0.019927979, 0.00072193146, -0.011909485, ...
http://arxiv.org/abs/2308.12898v2
Can Linguistic Knowledge Improve Multimodal Alignment in Vision-Language Pretraining?
Fei Wang, Liang Ding, Jun Rao, Ye Liu, Li Shen, Changxing Ding
cs.MM, cs.AI, cs.CL, cs.CV
The multimedia community has shown a significant interest in perceiving and representing the physical world with multimodal pretrained neural network models, and among them, the visual-language pertaining (VLP) is, currently, the most captivating topic. However, there have been few endeavors dedicated to the exploratio...
http://arxiv.org/pdf/2308.12898v2
2023-08-25T12:22:53
2023-08-24T16:17:40
{ "title": [ -0.016799927, -0.015136719, -0.04925537, -0.02998352, -0.029205322, -0.030776978, -0.03074646, 0.014793396, -0.0043678284, 0.030838013, -0.0019721985, 0.0035514832, 0.014656067, -0.034851074, -0.026565552, -0.0105896, 0.018463135, 0....
http://arxiv.org/abs/2308.12896v3
Beyond Document Page Classification: Design, Datasets, and Challenges
Jordy Van Landeghem, Sanket Biswas, Matthew B. Blaschko, Marie-Francine Moens
cs.CV, cs.CL, cs.LG
This paper highlights the need to bring document classification benchmarking closer to real-world applications, both in the nature of data tested ($X$: multi-channel, multi-paged, multi-industry; $Y$: class distributions and label set variety) and in classification tasks considered ($f$: multi-page document, page strea...
http://arxiv.org/pdf/2308.12896v3
2023-10-31T10:35:39
2023-08-24T16:16:47
{ "title": [ -0.018127441, 0.009307861, -0.049682617, -0.03933716, 0.024490356, -0.016159058, -0.025405884, -0.02407837, 0.012207031, 0.0039978027, -0.030578613, 0.018814087, -0.021972656, -0.010437012, -0.024551392, -0.009170532, -0.011474609, 0...
http://arxiv.org/abs/2308.12890v2
Large Language Models Vote: Prompting for Rare Disease Identification
David Oniani, Jordan Hilsman, Hang Dong, Fengyi Gao, Shiven Verma, Yanshan Wang
cs.CL, cs.AI
The emergence of generative Large Language Models (LLMs) emphasizes the need for accurate and efficient prompting approaches. LLMs are often applied in Few-Shot Learning (FSL) contexts, where tasks are executed with minimal training data. FSL has become popular in many Artificial Intelligence (AI) subdomains, including...
http://arxiv.org/pdf/2308.12890v2
2023-08-28T14:16:42
2023-08-24T16:09:13
{ "title": [ -0.004070282, -0.006225586, -0.017501831, -0.058380127, -0.0057029724, 0.00081539154, -0.039245605, -0.00868988, -0.0062065125, 0.007282257, -0.028015137, -0.00056123734, -0.007194519, -0.028045654, -0.0049819946, -0.013702393, -0.015747...
http://arxiv.org/abs/2308.12888v1
Inducing Causal Structure for Abstractive Text Summarization
Lu Chen, Ruqing Zhang, Wei Huang, Wei Chen, Jiafeng Guo, Xueqi Cheng
cs.CL, cs.AI
The mainstream of data-driven abstractive summarization models tends to explore the correlations rather than the causal relationships. Among such correlations, there can be spurious ones which suffer from the language prior learned from the training corpus and therefore undermine the overall effectiveness of the learne...
http://arxiv.org/pdf/2308.12888v1
2023-08-24T16:06:36
2023-08-24T16:06:36
{ "title": [ 0.00068998337, 0.004676819, -0.019317627, -0.083618164, -0.014030457, 0.022949219, -0.059539795, -0.009635925, 0.006164551, 0.027954102, -0.009239197, 0.026412964, -0.017868042, -0.010192871, -0.041137695, -0.0031356812, -0.022583008, ...
http://arxiv.org/abs/2308.12842v1
Text Similarity from Image Contents using Statistical and Semantic Analysis Techniques
Sagar Kulkarni, Sharvari Govilkar, Dhiraj Amin
cs.CL
Plagiarism detection is one of the most researched areas among the Natural Language Processing(NLP) community. A good plagiarism detection covers all the NLP methods including semantics, named entities, paraphrases etc. and produces detailed plagiarism reports. Detection of Cross Lingual Plagiarism requires deep knowle...
http://arxiv.org/pdf/2308.12842v1
2023-08-24T15:06:04
2023-08-24T15:06:04
{ "title": [ -0.029708862, 0.013511658, -0.041137695, -0.019119263, 0.02281189, -0.008514404, -0.0009775162, -0.04006958, 0.009101868, 0.045959473, -0.02658081, 0.029190063, 0.002904892, 0.0013389587, -0.05050659, 0.005744934, 0.008148193, 0.0307...
http://arxiv.org/abs/2308.12833v1
Use of LLMs for Illicit Purposes: Threats, Prevention Measures, and Vulnerabilities
Maximilian Mozes, Xuanli He, Bennett Kleinberg, Lewis D. Griffin
cs.CL, cs.CR
Spurred by the recent rapid increase in the development and distribution of large language models (LLMs) across industry and academia, much recent work has drawn attention to safety- and security-related threats and vulnerabilities of LLMs, including in the context of potentially criminal activities. Specifically, it h...
http://arxiv.org/pdf/2308.12833v1
2023-08-24T14:45:50
2023-08-24T14:45:50
{ "title": [ -0.00687027, 0.032440186, -0.022216797, -0.018234253, -0.0044441223, -0.0074920654, -0.03744507, -0.04901123, -0.061584473, 0.0069274902, -0.0014190674, 0.014533997, -0.026062012, -0.06793213, 0.017929077, -0.017471313, -0.014427185, ...
http://arxiv.org/abs/2308.12770v1
WavMark: Watermarking for Audio Generation
Guangyu Chen, Yu Wu, Shujie Liu, Tao Liu, Xiaoyong Du, Furu Wei
cs.SD, cs.CL, eess.AS
Recent breakthroughs in zero-shot voice synthesis have enabled imitating a speaker's voice using just a few seconds of recording while maintaining a high level of realism. Alongside its potential benefits, this powerful technology introduces notable risks, including voice fraud and speaker impersonation. Unlike the con...
http://arxiv.org/pdf/2308.12770v1
2023-08-24T13:17:35
2023-08-24T13:17:35
{ "title": [ -0.011001587, -0.0030918121, -0.045166016, -0.050720215, 0.012680054, -0.0077667236, -0.017532349, -0.017044067, -0.026138306, 0.048828125, 0.01197052, 0.016433716, 0.017196655, -0.033203125, -0.03100586, 0.0077590942, -0.016174316, ...
http://arxiv.org/abs/2308.12734v1
Real-time Detection of AI-Generated Speech for DeepFake Voice Conversion
Jordan J. Bird, Ahmad Lotfi
cs.SD, cs.CL, cs.HC, cs.LG, eess.AS
There are growing implications surrounding generative AI in the speech domain that enable voice cloning and real-time voice conversion from one individual to another. This technology poses a significant ethical threat and could lead to breaches of privacy and misrepresentation, thus there is an urgent need for real-tim...
http://arxiv.org/pdf/2308.12734v1
2023-08-24T12:26:15
2023-08-24T12:26:15
{ "title": [ -0.011169434, 0.014480591, -0.046966553, -0.042053223, 0.016235352, -0.021514893, -0.00069379807, -0.011459351, -0.024108887, 0.04498291, 0.007701874, 0.0028800964, 0.025909424, -0.034851074, -0.02268982, 0.017318726, -0.013786316, 0...
http://arxiv.org/abs/2308.12711v1
Harnessing the Power of David against Goliath: Exploring Instruction Data Generation without Using Closed-Source Models
Yue Wang, Xinrui Wang, Juntao Li, Jinxiong Chang, Qishen Zhang, Zhongyi Liu, Guannan Zhang, Min Zhang
cs.CL
Instruction tuning is instrumental in enabling Large Language Models~(LLMs) to follow user instructions to complete various open-domain tasks. The success of instruction tuning depends on the availability of high-quality instruction data. Owing to the exorbitant cost and substandard quality of human annotation, recent ...
http://arxiv.org/pdf/2308.12711v1
2023-08-24T11:07:47
2023-08-24T11:07:47
{ "title": [ -0.04714966, -0.017547607, -0.048339844, -0.019500732, -0.02658081, -0.016784668, -0.03765869, -0.011520386, -0.02041626, 0.007144928, -0.039520264, 0.028640747, -0.06335449, -0.009803772, -0.033721924, -0.033172607, -0.009521484, 0....
http://arxiv.org/abs/2308.12674v1
Improving Translation Faithfulness of Large Language Models via Augmenting Instructions
Yijie Chen, Yijin Liu, Fandong Meng, Yufeng Chen, Jinan Xu, Jie Zhou
cs.CL, cs.AI
Large Language Models (LLMs) present strong general capabilities, and a current compelling challenge is stimulating their specialized capabilities, such as machine translation, through low-cost instruction tuning. The standard instruction-following data is sequentially organized as the concatenation of an instruction, ...
http://arxiv.org/pdf/2308.12674v1
2023-08-24T09:32:29
2023-08-24T09:32:29
{ "title": [ -0.03552246, -0.0126571655, -0.02192688, -0.0736084, -0.0043029785, 0.018585205, -0.078063965, 0.0072669983, 0.008544922, 0.047546387, -0.014022827, 0.013908386, 0.028579712, -0.006286621, -0.045196533, 0.033569336, 0.00006991625, 0....
http://arxiv.org/abs/2308.12648v1
From Chatter to Matter: Addressing Critical Steps of Emotion Recognition Learning in Task-oriented Dialogue
Shutong Feng, Nurul Lubis, Benjamin Ruppik, Christian Geishauser, Michael Heck, Hsien-chin Lin, Carel van Niekerk, Renato Vukovic, Milica Gašić
cs.CL
Emotion recognition in conversations (ERC) is a crucial task for building human-like conversational agents. While substantial efforts have been devoted to ERC for chit-chat dialogues, the task-oriented counterpart is largely left unattended. Directly applying chit-chat ERC models to task-oriented dialogues (ToDs) resul...
http://arxiv.org/pdf/2308.12648v1
2023-08-24T08:46:30
2023-08-24T08:46:30
{ "title": [ -0.00774765, -0.001914978, -0.042541504, -0.04562378, 0.0039901733, -0.023391724, -0.023147583, 0.019119263, 0.008239746, 0.023147583, -0.010505676, 0.00067424774, 0.0014457703, -0.03387451, -0.052520752, -0.007801056, -0.0026607513, ...
http://arxiv.org/abs/2308.12643v1
Probabilistic Method of Measuring Linguistic Productivity
Sergei Monakhov
cs.CL
In this paper I propose a new way of measuring linguistic productivity that objectively assesses the ability of an affix to be used to coin new complex words and, unlike other popular measures, is not directly dependent upon token frequency. Specifically, I suggest that linguistic productivity may be viewed as the prob...
http://arxiv.org/pdf/2308.12643v1
2023-08-24T08:36:28
2023-08-24T08:36:28
{ "title": [ -0.048675537, 0.012763977, -0.076538086, -0.010757446, -0.026031494, -0.031463623, -0.04321289, -0.0066108704, -0.029083252, 0.0020103455, 0.0021076202, -0.023422241, -0.007881165, -0.07159424, 0.008232117, -0.002439499, -0.0062065125, ...
http://arxiv.org/abs/2308.12635v1
Advancing Hungarian Text Processing with HuSpaCy: Efficient and Accurate NLP Pipelines
György Orosz, Gergő Szabó, Péter Berkecz, Zsolt Szántó, Richárd Farkas
cs.CL, cs.AI, stat.ML, 68T50, I.2.7
This paper presents a set of industrial-grade text processing models for Hungarian that achieve near state-of-the-art performance while balancing resource efficiency and accuracy. Models have been implemented in the spaCy framework, extending the HuSpaCy toolkit with several improvements to its architecture. Compared t...
http://arxiv.org/pdf/2308.12635v1
2023-08-24T08:19:51
2023-08-24T08:19:51
{ "title": [ -0.018814087, -0.0027828217, -0.040222168, -0.057495117, 0.018447876, 0.008033752, -0.046569824, -0.020431519, -0.005203247, 0.05441284, -0.02267456, 0.009475708, 0.064453125, -0.008659363, -0.0413208, 0.021209717, -0.02633667, 0.047...
http://arxiv.org/abs/2308.12604v1
PromptMRG: Diagnosis-Driven Prompts for Medical Report Generation
Haibo Jin, Haoxuan Che, Yi Lin, Hao Chen
cs.CV, cs.CL
Automatic medical report generation (MRG) is of great research value as it has the potential to relieve radiologists from the heavy burden of report writing. Despite recent advancements, accurate MRG remains challenging due to the need for precise clinical understanding and the identification of clinical findings. More...
http://arxiv.org/pdf/2308.12604v1
2023-08-24T07:10:31
2023-08-24T07:10:31
{ "title": [ -0.00730896, -0.014381409, -0.04989624, -0.086120605, 0.0033168793, 0.00995636, -0.053649902, 0.004627228, 0.0008339882, 0.007850647, -0.027648926, 0.0093688965, -0.009063721, -0.031677246, -0.020462036, -0.009407043, -0.0024471283, ...
http://arxiv.org/abs/2308.12578v1
Mind vs. Mouth: On Measuring Re-judge Inconsistency of Social Bias in Large Language Models
Yachao Zhao, Bo Wang, Dongming Zhao, Kun Huang, Yan Wang, Ruifang He, Yuexian Hou
cs.CL, cs.AI
Recent researches indicate that Pre-trained Large Language Models (LLMs) possess cognitive constructs similar to those observed in humans, prompting researchers to investigate the cognitive aspects of LLMs. This paper focuses on explicit and implicit social bias, a distinctive two-level cognitive construct in psycholog...
http://arxiv.org/pdf/2308.12578v1
2023-08-24T05:35:58
2023-08-24T05:35:58
{ "title": [ -0.028121948, 0.0012655258, -0.027938843, -0.035461426, -0.0054893494, -0.01071167, -0.03768921, -0.023620605, 0.0064430237, 0.022079468, -0.056915283, 0.010513306, 0.01121521, 0.00069475174, -0.035217285, -0.018554688, -0.030258179, ...
http://arxiv.org/abs/2308.12568v1
A Small and Fast BERT for Chinese Medical Punctuation Restoration
Tongtao Ling, Chen Liao, Zhipeng Yu, Lei Chen, Shilei Huang, Yi Liu
cs.CL
In clinical dictation, utterances after automatic speech recognition (ASR) without explicit punctuation marks may lead to the misunderstanding of dictated reports. To give a precise and understandable clinical report with ASR, automatic punctuation restoration is required. Considering a practical scenario, we propose a...
http://arxiv.org/pdf/2308.12568v1
2023-08-24T05:15:43
2023-08-24T05:15:43
{ "title": [ -0.009033203, -0.024856567, -0.008644104, -0.03793335, -0.044891357, 0.003583908, -0.031280518, 0.0037593842, -0.0440979, 0.062072754, -0.0052223206, -0.015380859, 0.08337402, -0.037261963, 0.009849548, 0.017059326, -0.00217247, 0.00...
http://arxiv.org/abs/2308.12539v1
CALM : A Multi-task Benchmark for Comprehensive Assessment of Language Model Bias
Vipul Gupta, Pranav Narayanan Venkit, Hugo Laurençon, Shomir Wilson, Rebecca J. Passonneau
cs.CL, cs.AI, cs.LG
As language models (LMs) become increasingly powerful, it is important to quantify and compare them for sociodemographic bias with potential for harm. Prior bias measurement datasets are sensitive to perturbations in their manually designed templates, therefore unreliable. To achieve reliability, we introduce the Compr...
http://arxiv.org/pdf/2308.12539v1
2023-08-24T03:53:55
2023-08-24T03:53:55
{ "title": [ -0.025482178, -0.016113281, -0.040374756, -0.052978516, -0.007858276, -0.0019483566, -0.072509766, -0.016204834, 0.021438599, 0.033843994, -0.003206253, 0.0044059753, 0.016555786, -0.021728516, -0.052520752, 0.02293396, -0.020492554, ...
http://arxiv.org/abs/2308.12531v1
CARE: Co-Attention Network for Joint Entity and Relation Extraction
Wenjun Kong, Yamei Xia
cs.CL
Joint entity and relation extraction is the fundamental task of information extraction, consisting of two subtasks: named entity recognition and relation extraction. Most existing joint extraction methods suffer from issues of feature confusion or inadequate interaction between two subtasks. In this work, we propose a ...
http://arxiv.org/pdf/2308.12531v1
2023-08-24T03:40:54
2023-08-24T03:40:54
{ "title": [ 0.00008171797, 0.0022888184, -0.049102783, -0.0871582, 0.030014038, -0.002380371, -0.03540039, 0.0072021484, 0.032104492, 0.024139404, -0.010276794, 0.0039253235, 0.018081665, -0.0074920654, -0.036102295, 0.025360107, -0.023147583, 0...
http://arxiv.org/abs/2308.12519v1
Large Language Model as Autonomous Decision Maker
Yining Ye, Xin Cong, Yujia Qin, Yankai Lin, Zhiyuan Liu, Maosong Sun
cs.CL
While large language models (LLMs) exhibit impressive language understanding and in-context learning abilities, their decision-making ability still heavily relies on the guidance of task-specific expert knowledge when solving real-world tasks. To unleash the potential of LLMs as autonomous decision makers, this paper p...
http://arxiv.org/pdf/2308.12519v1
2023-08-24T03:11:45
2023-08-24T03:11:45
{ "title": [ -0.017562866, -0.006126404, -0.045166016, -0.07531738, -0.0062026978, -0.0034770966, -0.05404663, 0.011787415, -0.0027503967, 0.026412964, -0.027694702, -0.015022278, -0.008728027, -0.015823364, -0.025100708, -0.012184143, -0.008415222, ...
http://arxiv.org/abs/2308.12490v1
MultiPA: a multi-task speech pronunciation assessment system for a closed and open response scenario
Yu-Wen Chen, Zhou Yu, Julia Hirschberg
cs.CL, cs.SD, eess.AS
The design of automatic speech pronunciation assessment can be categorized into closed and open response scenarios, each with strengths and limitations. A system with the ability to function in both scenarios can cater to diverse learning needs and provide a more precise and holistic assessment of pronunciation skills....
http://arxiv.org/pdf/2308.12490v1
2023-08-24T01:24:09
2023-08-24T01:24:09
{ "title": [ -0.007949829, -0.005908966, -0.07171631, -0.072753906, -0.0023498535, -0.02368164, -0.039733887, 0.038116455, -0.0044059753, 0.03869629, -0.019638062, 0.0059661865, -0.0107421875, -0.025863647, -0.05029297, -0.008583069, -0.02168274, ...
http://arxiv.org/abs/2308.12488v1
GPTEval: A Survey on Assessments of ChatGPT and GPT-4
Rui Mao, Guanyi Chen, Xulang Zhang, Frank Guerin, Erik Cambria
cs.AI, cs.CL
The emergence of ChatGPT has generated much speculation in the press about its potential to disrupt social and economic systems. Its astonishing language ability has aroused strong curiosity among scholars about its performance in different domains. There have been many studies evaluating the ability of ChatGPT and GPT...
http://arxiv.org/pdf/2308.12488v1
2023-08-24T01:17:16
2023-08-24T01:17:16
{ "title": [ -0.0090789795, 0.0015525818, -0.01058197, -0.016296387, 0.014480591, -0.0011777878, -0.009185791, -0.041778564, 0.0022659302, 0.0025959015, -0.007572174, 0.0033569336, -0.012832642, -0.04522705, -0.020904541, 0.016586304, -0.02470398, ...
http://arxiv.org/abs/2308.12477v1
American Stories: A Large-Scale Structured Text Dataset of Historical U.S. Newspapers
Melissa Dell, Jacob Carlson, Tom Bryan, Emily Silcock, Abhishek Arora, Zejiang Shen, Luca D'Amico-Wong, Quan Le, Pablo Querubin, Leander Heldring
cs.CL, cs.CV, econ.GN, q-fin.EC
Existing full text datasets of U.S. public domain newspapers do not recognize the often complex layouts of newspaper scans, and as a result the digitized content scrambles texts from articles, headlines, captions, advertisements, and other layout regions. OCR quality can also be low. This study develops a novel, deep l...
http://arxiv.org/pdf/2308.12477v1
2023-08-24T00:24:42
2023-08-24T00:24:42
{ "title": [ -0.012069702, -0.014602661, -0.030303955, -0.047576904, 0.008956909, -0.011047363, -0.011161804, -0.00077581406, 0.01878357, 0.023071289, -0.046966553, 0.039642334, 0.00091552734, 0.0030918121, -0.031219482, 0.008926392, -0.0008740425, ...
http://arxiv.org/abs/2309.12339v1
Considerations for health care institutions training large language models on electronic health records
Weipeng Zhou, Danielle Bitterman, Majid Afshar, Timothy A. Miller
cs.CY, cs.AI, cs.CL
Large language models (LLMs) like ChatGPT have excited scientists across fields; in medicine, one source of excitement is the potential applications of LLMs trained on electronic health record (EHR) data. But there are tough questions we must first answer if health care institutions are interested in having LLMs traine...
http://arxiv.org/pdf/2309.12339v1
2023-08-24T00:09:01
2023-08-24T00:09:01
{ "title": [ -0.00422287, -0.008743286, -0.04196167, -0.039367676, -0.015350342, -0.008926392, -0.028289795, 0.010566711, -0.04196167, 0.0345459, -0.0209198, 0.0006766319, 0.017852783, -0.03842163, 0.037017822, -0.016204834, 0.020141602, 0.031646...
http://arxiv.org/abs/2308.12466v1
Are ChatGPT and GPT-4 Good Poker Players? -- A Pre-Flop Analysis
Akshat Gupta
cs.CL, cs.AI, cs.GT
Since the introduction of ChatGPT and GPT-4, these models have been tested across a large number of tasks. Their adeptness across domains is evident, but their aptitude in playing games and specifically their aptitude in the realm of poker has remained unexplored. Poker is a game that requires decision making under unc...
http://arxiv.org/pdf/2308.12466v1
2023-08-23T23:16:35
2023-08-23T23:16:35
{ "title": [ 0.011764526, -0.0005841255, -0.018829346, 0.00869751, 0.010108948, -0.04220581, 0.041931152, -0.043945312, -0.031051636, 0.02999878, -0.022247314, 0.004589081, -0.00137043, -0.017120361, -0.02168274, -0.015365601, -0.016815186, 0.014...
http://arxiv.org/abs/2310.05935v1
Vulnerability Clustering and other Machine Learning Applications of Semantic Vulnerability Embeddings
Mark-Oliver Stehr, Minyoung Kim
cs.CR, cs.CL, cs.LG
Cyber-security vulnerabilities are usually published in form of short natural language descriptions (e.g., in form of MITRE's CVE list) that over time are further manually enriched with labels such as those defined by the Common Vulnerability Scoring System (CVSS). In the Vulnerability AI (Analytics and Intelligence) p...
http://arxiv.org/pdf/2310.05935v1
2023-08-23T21:39:48
2023-08-23T21:39:48
{ "title": [ 0.021209717, 0.014251709, -0.032104492, -0.033325195, -0.0056533813, -0.036956787, -0.0071754456, -0.024353027, -0.028381348, 0.047027588, -0.022338867, 0.0033302307, 0.020263672, -0.03945923, -0.0022964478, -0.013259888, -0.0019817352, ...
http://arxiv.org/abs/2308.12420v1
Evolution of ESG-focused DLT Research: An NLP Analysis of the Literature
Walter Hernandez, Kamil Tylinski, Alastair Moore, Niall Roche, Nikhil Vadgama, Horst Treiblmaier, Jiangbo Shangguan, Paolo Tasca, Jiahua Xu
cs.IR, cs.CL, cs.LG
Distributed Ledger Technologies (DLTs) have rapidly evolved, necessitating comprehensive insights into their diverse components. However, a systematic literature review that emphasizes the Environmental, Sustainability, and Governance (ESG) components of DLT remains lacking. To bridge this gap, we selected 107 seed pap...
http://arxiv.org/pdf/2308.12420v1
2023-08-23T20:42:32
2023-08-23T20:42:32
{ "title": [ -0.0049209595, 0.01235199, -0.07556152, -0.06732178, 0.019821167, -0.016204834, -0.035095215, 0.008735657, -0.03643799, 0.0069847107, -0.038482666, 0.052093506, -0.039611816, -0.0045776367, -0.010353088, -0.013397217, -0.026306152, 0...
http://arxiv.org/abs/2308.12419v1
Toward American Sign Language Processing in the Real World: Data, Tasks, and Methods
Bowen Shi
cs.CV, cs.CL
Sign language, which conveys meaning through gestures, is the chief means of communication among deaf people. Recognizing sign language in natural settings presents significant challenges due to factors such as lighting, background clutter, and variations in signer characteristics. In this thesis, I study automatic sig...
http://arxiv.org/pdf/2308.12419v1
2023-08-23T20:38:19
2023-08-23T20:38:19
{ "title": [ -0.029266357, -0.011703491, -0.025787354, -0.020599365, -0.0073623657, 0.0013551712, -0.026519775, -0.017684937, 0.004146576, 0.013130188, -0.030639648, -0.0030212402, 0.0008916855, -0.031036377, -0.031982422, -0.0051078796, 0.0011940002...
http://arxiv.org/abs/2308.12383v1
With a Little Help from your own Past: Prototypical Memory Networks for Image Captioning
Manuele Barraco, Sara Sarto, Marcella Cornia, Lorenzo Baraldi, Rita Cucchiara
cs.CV, cs.AI, cs.CL, cs.MM
Image captioning, like many tasks involving vision and language, currently relies on Transformer-based architectures for extracting the semantics in an image and translating it into linguistically coherent descriptions. Although successful, the attention operator only considers a weighted summation of projections of th...
http://arxiv.org/pdf/2308.12383v1
2023-08-23T18:53:00
2023-08-23T18:53:00
{ "title": [ 0.003320694, -0.0044670105, -0.009681702, -0.025894165, -0.0069236755, -0.017730713, 0.0071029663, -0.010009766, 0.00894928, 0.059661865, -0.013038635, 0.023010254, 0.0030517578, -0.0005893707, -0.0592041, 0.020339966, -0.0025787354, ...
http://arxiv.org/abs/2308.12372v1
Vision Transformer Adapters for Generalizable Multitask Learning
Deblina Bhattacharjee, Sabine Süsstrunk, Mathieu Salzmann
cs.CV, cs.CL
We introduce the first multitasking vision transformer adapters that learn generalizable task affinities which can be applied to novel tasks and domains. Integrated into an off-the-shelf vision transformer backbone, our adapters can simultaneously solve multiple dense vision tasks in a parameter-efficient manner, unlik...
http://arxiv.org/pdf/2308.12372v1
2023-08-23T18:40:48
2023-08-23T18:40:48
{ "title": [ -0.0048294067, 0.016464233, -0.053375244, -0.0357666, 0.002407074, -0.03665161, -0.022262573, -0.019454956, -0.01423645, 0.053955078, -0.03189087, -0.0029029846, -0.00040626526, -0.0018234253, -0.049072266, -0.009178162, 0.016433716, ...
http://arxiv.org/abs/2308.12284v1
D4: Improving LLM Pretraining via Document De-Duplication and Diversification
Kushal Tirumala, Daniel Simig, Armen Aghajanyan, Ari S. Morcos
cs.CL, cs.AI, cs.LG
Over recent years, an increasing amount of compute and data has been poured into training large language models (LLMs), usually by doing one-pass learning on as many tokens as possible randomly selected from large-scale web corpora. While training on ever-larger portions of the internet leads to consistent performance ...
http://arxiv.org/pdf/2308.12284v1
2023-08-23T17:58:14
2023-08-23T17:58:14
{ "title": [ -0.039031982, -0.0098724365, -0.0836792, -0.06311035, -0.026260376, 0.003332138, -0.083984375, 0.014427185, -0.041931152, 0.018463135, -0.0158844, 0.043884277, 0.00047802925, -0.043304443, 0.04031372, -0.015808105, -0.011741638, 0.03...
http://arxiv.org/abs/2308.12272v1
Simple is Better and Large is Not Enough: Towards Ensembling of Foundational Language Models
Nancy Tyagi, Aidin Shiri, Surjodeep Sarkar, Abhishek Kumar Umrawal, Manas Gaur
cs.CL, cs.AI
Foundational Language Models (FLMs) have advanced natural language processing (NLP) research. Current researchers are developing larger FLMs (e.g., XLNet, T5) to enable contextualized language representation, classification, and generation. While developing larger FLMs has been of significant advantage, it is also a li...
http://arxiv.org/pdf/2308.12272v1
2023-08-23T17:40:35
2023-08-23T17:40:35
{ "title": [ -0.040771484, -0.018173218, -0.04397583, -0.076416016, 0.002319336, 0.018249512, -0.08428955, 0.01486969, 0.004798889, 0.029403687, -0.01689148, 0.024475098, 0.002811432, -0.006111145, -0.0491333, 0.011199951, -0.008956909, 0.0281829...
http://arxiv.org/abs/2308.12261v1
Prompt2Model: Generating Deployable Models from Natural Language Instructions
Vijay Viswanathan, Chenyang Zhao, Amanda Bertsch, Tongshuang Wu, Graham Neubig
cs.CL
Large language models (LLMs) enable system builders today to create competent NLP systems through prompting, where they only need to describe the task in natural language and provide a few examples. However, in other ways, LLMs are a step backward from traditional special-purpose NLP models; they require extensive comp...
http://arxiv.org/pdf/2308.12261v1
2023-08-23T17:28:21
2023-08-23T17:28:21
{ "title": [ -0.046661377, -0.024246216, -0.041046143, -0.048034668, -0.016906738, 0.021743774, -0.07373047, -0.029708862, 0.00819397, 0.008911133, -0.010368347, 0.0013027191, -0.010139465, -0.020645142, -0.025878906, 0.001991272, -0.0056991577, ...
http://arxiv.org/abs/2308.12247v1
How to Protect Copyright Data in Optimization of Large Language Models?
Timothy Chu, Zhao Song, Chiwun Yang
cs.LG, cs.CL
Large language models (LLMs) and generative AI have played a transformative role in computer research and applications. Controversy has arisen as to whether these models output copyrighted data, which can occur if the data the models are trained on is copyrighted. LLMs are built on the transformer neural network archit...
http://arxiv.org/pdf/2308.12247v1
2023-08-23T16:48:04
2023-08-23T16:48:04
{ "title": [ -0.03225708, 0.0011854172, -0.033203125, -0.01939392, -0.027267456, 0.0033016205, -0.045288086, -0.008735657, -0.04324341, 0.037872314, -0.013191223, -0.012817383, 0.0085372925, -0.031341553, -0.0008997917, -0.019592285, -0.0035705566, ...
http://arxiv.org/abs/2308.12219v2
Diffusion Language Models Can Perform Many Tasks with Scaling and Instruction-Finetuning
Jiasheng Ye, Zaixiang Zheng, Yu Bao, Lihua Qian, Quanquan Gu
cs.CL, cs.AI, cs.LG
The recent surge of generative AI has been fueled by the generative power of diffusion probabilistic models and the scalable capabilities of large language models. Despite their potential, it remains elusive whether diffusion language models can solve general language tasks comparable to their autoregressive counterpar...
http://arxiv.org/pdf/2308.12219v2
2023-08-25T16:32:31
2023-08-23T16:01:12
{ "title": [ 0.014450073, -0.0062217712, -0.036376953, -0.064697266, -0.002664566, -0.012954712, -0.028305054, 0.019744873, -0.031082153, 0.040740967, -0.005695343, 0.006454468, 0.004676819, -0.033721924, 0.0121536255, -0.019348145, 0.0036144257, ...
http://arxiv.org/abs/2308.12215v1
The Challenges of Machine Learning for Trust and Safety: A Case Study on Misinformation Detection
Madelyne Xiao, Jonathan Mayer
cs.LG, cs.CL, cs.CY
We examine the disconnect between scholarship and practice in applying machine learning to trust and safety problems, using misinformation detection as a case study. We systematize literature on automated detection of misinformation across a corpus of 270 well-cited papers in the field. We then examine subsets of paper...
http://arxiv.org/pdf/2308.12215v1
2023-08-23T15:52:20
2023-08-23T15:52:20
{ "title": [ -0.012329102, 0.0076560974, -0.06738281, -0.0368042, 0.031021118, -0.039978027, 0.011161804, -0.0016460419, -0.04586792, 0.023071289, -0.018325806, 0.012077332, -0.0033111572, -0.030273438, -0.012557983, -0.016738892, -0.029006958, 0...
http://arxiv.org/abs/2308.12202v1
Curriculum Learning with Adam: The Devil Is in the Wrong Details
Lucas Weber, Jaap Jumelet, Paul Michel, Elia Bruni, Dieuwke Hupkes
cs.LG, cs.CL
Curriculum learning (CL) posits that machine learning models -- similar to humans -- may learn more efficiently from data that match their current learning progress. However, CL methods are still poorly understood and, in particular for natural language processing (NLP), have achieved only limited success. In this pape...
http://arxiv.org/pdf/2308.12202v1
2023-08-23T15:39:42
2023-08-23T15:39:42
{ "title": [ -0.022323608, -0.0076560974, -0.031677246, 0.025497437, -0.029754639, -0.029785156, 0.01763916, -0.009010315, -0.07824707, 0.015777588, 0.0004813671, 0.0015153885, 0.0010194778, -0.050598145, 0.04763794, -0.043792725, -0.0035686493, ...
http://arxiv.org/abs/2308.12157v1
Evaluation of Faithfulness Using the Longest Supported Subsequence
Anirudh Mittal, Timo Schick, Mikel Artetxe, Jane Dwivedi-Yu
cs.CL, cs.AI
As increasingly sophisticated language models emerge, their trustworthiness becomes a pivotal issue, especially in tasks such as summarization and question-answering. Ensuring their responses are contextually grounded and faithful is challenging due to the linguistic diversity and the myriad of possible answers. In thi...
http://arxiv.org/pdf/2308.12157v1
2023-08-23T14:18:44
2023-08-23T14:18:44
{ "title": [ -0.043395996, 0.014717102, -0.050720215, -0.061279297, 0.02166748, 0.0031757355, -0.042999268, -0.020019531, 0.0012741089, 0.034088135, -0.006996155, 0.000395298, 0.004169464, -0.029907227, -0.03277588, 0.005317688, 0.0028495789, 0.0...
http://arxiv.org/abs/2308.12131v1
Semantic Change Detection for the Romanian Language
Ciprian-Octavian Truică, Victor Tudose, Elena-Simona Apostol
cs.CL, cs.AI
Automatic semantic change methods try to identify the changes that appear over time in the meaning of words by analyzing their usage in diachronic corpora. In this paper, we analyze different strategies to create static and contextual word embedding models, i.e., Word2Vec and ELMo, on real-world English and Romanian da...
http://arxiv.org/pdf/2308.12131v1
2023-08-23T13:37:02
2023-08-23T13:37:02
{ "title": [ -0.030059814, -0.0043907166, -0.046783447, -0.013465881, 0.007068634, -0.012634277, -0.021896362, -0.012489319, -0.00014448166, 0.0423584, -0.03427124, -0.014907837, 0.04901123, -0.014122009, -0.03894043, 0.009231567, 0.0015945435, 0...
http://arxiv.org/abs/2308.12097v1
Instruction Position Matters in Sequence Generation with Large Language Models
Yijin Liu, Xianfeng Zeng, Fandong Meng, Jie Zhou
cs.CL
Large language models (LLMs) are capable of performing conditional sequence generation tasks, such as translation or summarization, through instruction fine-tuning. The fine-tuning data is generally sequentially concatenated from a specific task instruction, an input sentence, and the corresponding response. Considerin...
http://arxiv.org/pdf/2308.12097v1
2023-08-23T12:36:57
2023-08-23T12:36:57
{ "title": [ -0.044525146, -0.014343262, -0.043792725, -0.062042236, -0.009796143, 0.005756378, -0.074279785, 0.015792847, -0.0024471283, 0.035949707, -0.02456665, 0.004371643, 0.000243783, -0.010902405, -0.03189087, 0.008842468, -0.013946533, 0....
http://arxiv.org/abs/2308.12086v2
Out of the Cage: How Stochastic Parrots Win in Cyber Security Environments
Maria Rigaki, Ondřej Lukáš, Carlos A. Catania, Sebastian Garcia
cs.CR, cs.AI, cs.CL
Large Language Models (LLMs) have gained widespread popularity across diverse domains involving text generation, summarization, and various natural language processing tasks. Despite their inherent limitations, LLM-based designs have shown promising capabilities in planning and navigating open-world scenarios. This pap...
http://arxiv.org/pdf/2308.12086v2
2023-08-28T09:42:59
2023-08-23T12:11:27
{ "title": [ -0.0107421875, -0.0011606216, -0.03010559, -0.046203613, 0.03262329, -0.016815186, 0.0048675537, 0.0020694733, -0.055847168, 0.026550293, -0.013053894, 0.03265381, -0.02355957, -0.013671875, -0.023712158, -0.022399902, -0.009086609, ...
http://arxiv.org/abs/2308.12067v2
InstructionGPT-4: A 200-Instruction Paradigm for Fine-Tuning MiniGPT-4
Lai Wei, Zihao Jiang, Weiran Huang, Lichao Sun
cs.LG, cs.AI, cs.CL, cs.CV
Multimodal large language models are typically trained in two stages: first pre-training on image-text pairs, and then fine-tuning using supervised vision-language instruction data. Recent studies have shown that large language models can achieve satisfactory results even with a limited amount of high-quality instructi...
http://arxiv.org/pdf/2308.12067v2
2023-10-11T14:49:26
2023-08-23T11:27:30
{ "title": [ -0.026901245, -0.017044067, -0.05657959, -0.034820557, -0.0023174286, -0.031280518, -0.0046653748, 0.0063819885, -0.025131226, 0.037994385, 0.0018053055, 0.009666443, -0.018981934, -0.037963867, -0.033691406, 0.0096206665, -0.0014743805,...
http://arxiv.org/abs/2308.12060v1
FlexKBQA: A Flexible LLM-Powered Framework for Few-Shot Knowledge Base Question Answering
Zhenyu Li, Sunqi Fan, Yu Gu, Xiuxing Li, Zhichao Duan, Bowen Dong, Ning Liu, Jianyong Wang
cs.CL, cs.AI
Knowledge base question answering (KBQA) is a critical yet challenging task due to the vast number of entities within knowledge bases and the diversity of natural language questions posed by users. Unfortunately, the performance of most KBQA models tends to decline significantly in real-world scenarios where high-quali...
http://arxiv.org/pdf/2308.12060v1
2023-08-23T11:00:36
2023-08-23T11:00:36
{ "title": [ -0.047332764, -0.038330078, -0.05319214, -0.05392456, -0.042510986, 0.025787354, -0.07336426, 0.02772522, 0.02960205, 0.01626587, -0.0390625, 0.010925293, -0.022140503, -0.011276245, -0.054107666, -0.017700195, 0.017669678, -0.013732...
http://arxiv.org/abs/2308.12050v1
Aligning Language Models with Offline Reinforcement Learning from Human Feedback
Jian Hu, Li Tao, June Yang, Chandler Zhou
cs.CL, cs.AI
Learning from human preferences is crucial for language models (LMs) to effectively cater to human needs and societal values. Previous research has made notable progress by leveraging human feedback to follow instructions. However, these approaches rely primarily on online reinforcement learning (RL) techniques like Pr...
http://arxiv.org/pdf/2308.12050v1
2023-08-23T10:41:07
2023-08-23T10:41:07
{ "title": [ -0.024169922, -0.0025424957, -0.025314331, -0.030914307, -0.011352539, -0.0049705505, -0.04159546, -0.0181427, -0.023086548, 0.021118164, -0.01386261, -0.014984131, -0.013809204, -0.0143966675, -0.030197144, -0.029312134, -0.00712204, ...
http://arxiv.org/abs/2308.12045v1
CgT-GAN: CLIP-guided Text GAN for Image Captioning
Jiarui Yu, Haoran Li, Yanbin Hao, Bin Zhu, Tong Xu, Xiangnan He
cs.CV, cs.AI, cs.CL, cs.MM
The large-scale visual-language pre-trained model, Contrastive Language-Image Pre-training (CLIP), has significantly improved image captioning for scenarios without human-annotated image-caption pairs. Recent advanced CLIP-based image captioning without human annotations follows a text-only training paradigm, i.e., rec...
http://arxiv.org/pdf/2308.12045v1
2023-08-23T10:25:37
2023-08-23T10:25:37
{ "title": [ -0.0096588135, 0.005092621, -0.026138306, -0.038482666, 0.004863739, -0.009925842, -0.02154541, -0.011360168, 0.0059928894, 0.039489746, -0.022613525, 0.024887085, 0.0050621033, -0.0054740906, -0.042755127, 0.0049209595, -0.014175415, ...
http://arxiv.org/abs/2308.12043v1
IncreLoRA: Incremental Parameter Allocation Method for Parameter-Efficient Fine-tuning
Feiyu Zhang, Liangzhi Li, Junhao Chen, Zhouqiang Jiang, Bowen Wang, Yiming Qian
cs.CL, cs.AI, cs.LG
With the increasing size of pre-trained language models (PLMs), fine-tuning all the parameters in the model is not efficient, especially when there are a large number of downstream tasks, which incur significant training and storage costs. Many parameter-efficient fine-tuning (PEFT) approaches have been proposed, among...
http://arxiv.org/pdf/2308.12043v1
2023-08-23T10:08:10
2023-08-23T10:08:10
{ "title": [ -0.029006958, -0.017913818, -0.05126953, -0.09136963, 0.011909485, 0.005252838, -0.04916382, 0.014152527, -0.035308838, 0.025039673, -0.0017271042, 0.0209198, 0.0019874573, -0.025299072, -0.009597778, 0.007080078, -0.010871887, 0.028...
http://arxiv.org/abs/2308.12039v1
Hybrid Retrieval and Multi-stage Text Ranking Solution at TREC 2022 Deep Learning Track
Guangwei Xu, Yangzhao Zhang, Longhui Zhang, Dingkun Long, Pengjun Xie, Ruijie Guo
cs.IR, cs.CL
Large-scale text retrieval technology has been widely used in various practical business scenarios. This paper presents our systems for the TREC 2022 Deep Learning Track. We explain the hybrid text retrieval and multi-stage text ranking method adopted in our solution. The retrieval stage combined the two structures of ...
http://arxiv.org/pdf/2308.12039v1
2023-08-23T09:56:59
2023-08-23T09:56:59
{ "title": [ -0.002412796, 0.0019626617, -0.024871826, -0.034851074, 0.0034236908, -0.025344849, 0.008300781, 0.0030288696, 0.018615723, 0.030441284, -0.037597656, -0.0018749237, 0.025024414, -0.012420654, -0.01309967, -0.008781433, -0.009468079, ...
http://arxiv.org/abs/2308.12038v1
Large Multilingual Models Pivot Zero-Shot Multimodal Learning across Languages
Jinyi Hu, Yuan Yao, Chongyi Wang, Shan Wang, Yinxu Pan, Qianyu Chen, Tianyu Yu, Hanghao Wu, Yue Zhao, Haoye Zhang, Xu Han, Yankai Lin, Jiao Xue, Dahai Li, Zhiyuan Liu, Maosong Sun
cs.CL, cs.CV
Recently there has been a significant surge in multimodal learning in terms of both image-to-text and text-to-image generation. However, the success is typically limited to English, leaving other languages largely behind. Building a competitive counterpart in other languages is highly challenging due to the low-resourc...
http://arxiv.org/pdf/2308.12038v1
2023-08-23T09:55:41
2023-08-23T09:55:41
{ "title": [ -0.013191223, -0.01828003, -0.026733398, -0.018692017, -0.04296875, -0.020690918, -0.05316162, 0.0052375793, 0.0043029785, 0.045837402, -0.009300232, -0.0074920654, 0.030899048, -0.046142578, 0.004432678, 0.009864807, 0.0024929047, 0...
http://arxiv.org/abs/2308.12033v1
PREFER: Prompt Ensemble Learning via Feedback-Reflect-Refine
Chenrui Zhang, Lin Liu, Jinpeng Wang, Chuyuan Wang, Xiao Sun, Hongyu Wang, Mingchen Cai
cs.CL, cs.AI
As an effective tool for eliciting the power of Large Language Models (LLMs), prompting has recently demonstrated unprecedented abilities across a variety of complex tasks. To further improve the performance, prompt ensemble has attracted substantial interest for tackling the hallucination and instability of LLMs. Howe...
http://arxiv.org/pdf/2308.12033v1
2023-08-23T09:46:37
2023-08-23T09:46:37
{ "title": [ 0.0012693405, -0.016738892, -0.04547119, -0.03894043, -0.03375244, -0.00548172, -0.039978027, -0.0033073425, 0.00036358833, 0.006843567, -0.058532715, -0.0018320084, -0.023895264, -0.017440796, -0.015525818, -0.04135132, 0.005924225, ...
http://arxiv.org/abs/2308.12032v3
From Quantity to Quality: Boosting LLM Performance with Self-Guided Data Selection for Instruction Tuning
Ming Li, Yong Zhang, Zhitao Li, Jiuhai Chen, Lichang Chen, Ning Cheng, Jianzong Wang, Tianyi Zhou, Jing Xiao
cs.CL
In the realm of Large Language Models, the balance between instruction data quality and quantity has become a focal point. Recognizing this, we introduce a self-guided methodology for LLMs to autonomously discern and select cherry samples from vast open-source datasets, effectively minimizing manual curation and potent...
http://arxiv.org/pdf/2308.12032v3
2023-09-15T20:33:44
2023-08-23T09:45:29
{ "title": [ -0.08557129, -0.012496948, -0.0546875, -0.03982544, -0.036834717, 0.0029792786, -0.06829834, 0.027938843, -0.021774292, 0.014762878, -0.027923584, 0.041992188, -0.039611816, -0.02041626, -0.016845703, -0.009498596, -0.0013990402, 0.0...
http://arxiv.org/abs/2308.12030v2
Prompt-Based Length Controlled Generation with Reinforcement Learning
Renlong Jie, Xiaojun Meng, Lifeng Shang, Xin Jiang, Qun Liu
cs.CL, cs.AI, cs.LG
Large language models (LLMs) like ChatGPT and GPT-4 have attracted great attention given their surprising performance on a wide range of NLP tasks. Length controlled generation of LLMs emerges as an important topic, which enables users to fully leverage the capability of LLMs in more real-world scenarios like generatin...
http://arxiv.org/pdf/2308.12030v2
2023-09-30T07:54:22
2023-08-23T09:43:10
{ "title": [ -0.033843994, -0.017471313, -0.0362854, -0.019943237, -0.026824951, -0.009963989, -0.019729614, -0.004573822, -0.04220581, 0.019104004, -0.0022735596, 0.026153564, -0.040649414, -0.033477783, 0.0050811768, -0.015151978, -0.003929138, ...
http://arxiv.org/abs/2308.12025v1
Knowledge-injected Prompt Learning for Chinese Biomedical Entity Normalization
Songhua Yang, Chenghao Zhang, Hongfei Xu, Yuxiang Jia
cs.CL
The Biomedical Entity Normalization (BEN) task aims to align raw, unstructured medical entities to standard entities, thus promoting data coherence and facilitating better downstream medical applications. Recently, prompt learning methods have shown promising results in this task. However, existing research falls short...
http://arxiv.org/pdf/2308.12025v1
2023-08-23T09:32:40
2023-08-23T09:32:40
{ "title": [ -0.018661499, 0.0003335476, -0.04473877, -0.06304932, -0.0011262894, -0.001821518, -0.04385376, 0.005706787, 0.004196167, 0.008399963, -0.024047852, 0.0025749207, -0.0014400482, -0.03930664, -0.011100769, -0.0068473816, 0.0011224747, ...
http://arxiv.org/abs/2308.12022v1
Reranking Passages with Coarse-to-Fine Neural Retriever using List-Context Information
Hongyin Zhu
cs.CL
Passage reranking is a crucial task in many applications, particularly when dealing with large-scale documents. Traditional neural architectures are limited in retrieving the best passage for a question because they usually match the question to each passage separately, seldom considering contextual information in othe...
http://arxiv.org/pdf/2308.12022v1
2023-08-23T09:29:29
2023-08-23T09:29:29
{ "title": [ 0.011154175, 0.019973755, -0.0076026917, -0.032836914, 0.010902405, 0.0037326813, -0.006259918, -0.038024902, 0.022125244, 0.04776001, -0.024963379, 0.013114929, 0.021362305, 0.0013532639, -0.042999268, 0.011405945, 0.0045394897, -0....
http://arxiv.org/abs/2308.12014v2
From Instructions to Intrinsic Human Values -- A Survey of Alignment Goals for Big Models
Jing Yao, Xiaoyuan Yi, Xiting Wang, Jindong Wang, Xing Xie
cs.AI, cs.CL, cs.CY
Big models, exemplified by Large Language Models (LLMs), are models typically pre-trained on massive data and comprised of enormous parameters, which not only obtain significantly improved performance across diverse tasks but also present emergent capabilities absent in smaller models. However, the growing intertwining...
http://arxiv.org/pdf/2308.12014v2
2023-09-04T03:32:05
2023-08-23T09:11:13
{ "title": [ -0.032562256, 0.0041770935, -0.0569458, -0.05267334, 0.0038719177, -0.008392334, -0.04559326, -0.014122009, 0.012519836, 0.038024902, 0.00055742264, -0.003967285, -0.013404846, -0.028076172, -0.033416748, 0.0068969727, 0.0076522827, ...
http://arxiv.org/abs/2308.12008v1
Graecia capta ferum victorem cepit. Detecting Latin Allusions to Ancient Greek Literature
Frederick Riemenschneider, Anette Frank
cs.CL, I.2.7
Intertextual allusions hold a pivotal role in Classical Philology, with Latin authors frequently referencing Ancient Greek texts. Until now, the automatic identification of these intertextual references has been constrained to monolingual approaches, seeking parallels solely within Latin or Greek texts. In this study, ...
http://arxiv.org/pdf/2308.12008v1
2023-08-23T08:54:05
2023-08-23T08:54:05
{ "title": [ -0.0003452301, 0.007534027, -0.022521973, 0.0725708, -0.021102905, -0.04586792, 0.032928467, -0.062408447, 0.019943237, 0.0112838745, -0.022476196, 0.004814148, 0.038238525, -0.042999268, 0.034240723, -0.0095825195, 0.0046920776, 0.0...
http://arxiv.org/abs/2308.11995v1
Topical-Chat: Towards Knowledge-Grounded Open-Domain Conversations
Karthik Gopalakrishnan, Behnam Hedayatnia, Qinlang Chen, Anna Gottardi, Sanjeev Kwatra, Anu Venkatesh, Raefer Gabriel, Dilek Hakkani-Tur
cs.CL, cs.AI
Building socialbots that can have deep, engaging open-domain conversations with humans is one of the grand challenges of artificial intelligence (AI). To this end, bots need to be able to leverage world knowledge spanning several domains effectively when conversing with humans who have their own world knowledge. Existi...
http://arxiv.org/pdf/2308.11995v1
2023-08-23T08:33:14
2023-08-23T08:33:14
{ "title": [ -0.0053215027, -0.0012178421, -0.037384033, -0.08795166, 0.013114929, 0.009239197, -0.058135986, 0.0060806274, 0.02722168, 0.0413208, -0.0009584427, 0.0030212402, -0.006549835, -0.018585205, -0.060943604, 0.001832962, -0.00806427, 0....
http://arxiv.org/abs/2308.11971v1
EVE: Efficient Vision-Language Pre-training with Masked Prediction and Modality-Aware MoE
Junyi Chen, Longteng Guo, Jia Sun, Shuai Shao, Zehuan Yuan, Liang Lin, Dongyu Zhang
cs.CV, cs.CL, cs.LG, cs.MM
Building scalable vision-language models to learn from diverse, multimodal data remains an open challenge. In this paper, we introduce an Efficient Vision-languagE foundation model, namely EVE, which is one unified multimodal Transformer pre-trained solely by one unified pre-training task. Specifically, EVE encodes bot...
http://arxiv.org/pdf/2308.11971v1
2023-08-23T07:36:30
2023-08-23T07:36:30
{ "title": [ -0.022033691, -0.012191772, -0.06008911, -0.07733154, -0.0030231476, -0.01007843, -0.051574707, 0.01776123, 0.00021350384, 0.020355225, -0.0121154785, 0.011543274, 0.037750244, -0.036468506, -0.0017232895, 0.0037117004, 0.0068969727, ...
http://arxiv.org/abs/2308.11940v2
Audio Generation with Multiple Conditional Diffusion Model
Zhifang Guo, Jianguo Mao, Rui Tao, Long Yan, Kazushige Ouchi, Hong Liu, Xiangdong Wang
cs.SD, cs.CL, cs.LG, eess.AS
Text-based audio generation models have limitations as they cannot encompass all the information in audio, leading to restricted controllability when relying solely on text. To address this issue, we propose a novel model that enhances the controllability of existing pre-trained text-to-audio models by incorporating ad...
http://arxiv.org/pdf/2308.11940v2
2023-10-10T03:35:35
2023-08-23T06:21:46
{ "title": [ -0.014854431, 0.0005455017, -0.07006836, -0.07171631, 0.015106201, -0.00004863739, -0.05203247, 0.016571045, 0.0052948, 0.04550171, -0.014251709, -0.0008187294, 0.01701355, -0.026153564, -0.03555298, 0.0075912476, -0.020339966, 0.020...
http://arxiv.org/abs/2308.11923v1
Audio Difference Captioning Utilizing Similarity-Discrepancy Disentanglement
Daiki Takeuchi, Yasunori Ohishi, Daisuke Niizumi, Noboru Harada, Kunio Kashino
eess.AS, cs.CL, cs.LG, cs.SD
We proposed Audio Difference Captioning (ADC) as a new extension task of audio captioning for describing the semantic differences between input pairs of similar but slightly different audio clips. The ADC solves the problem that conventional audio captioning sometimes generates similar captions for similar audio clips,...
http://arxiv.org/pdf/2308.11923v1
2023-08-23T05:13:25
2023-08-23T05:13:25
{ "title": [ -0.0033359528, 0.008300781, -0.052642822, -0.041870117, 0.008125305, -0.03253174, 0.000436306, -0.0104599, -0.007003784, 0.05722046, -0.023223877, 0.02041626, 0.02998352, -0.0043945312, -0.021881104, 0.0135269165, -0.019821167, 0.044...
http://arxiv.org/abs/2308.12923v1
Diagnosing Infeasible Optimization Problems Using Large Language Models
Hao Chen, Gonzalo E. Constante-Flores, Can Li
cs.HC, cs.CL, cs.LG, math.OC
Decision-making problems can be represented as mathematical optimization models, finding wide applications in fields such as economics, engineering and manufacturing, transportation, and health care. Optimization models are mathematical abstractions of the problem of making the best decision while satisfying a set of r...
http://arxiv.org/pdf/2308.12923v1
2023-08-23T04:34:05
2023-08-23T04:34:05
{ "title": [ -0.023788452, -0.01424408, -0.0064582825, -0.024963379, -0.020828247, 0.009849548, -0.046966553, -0.029052734, 0.023117065, 0.019302368, -0.0096206665, -0.0072250366, -0.022415161, -0.019638062, -0.030578613, 0.002210617, -0.0063591003, ...
http://arxiv.org/abs/2308.11891v2
Bridging the Gap: Deciphering Tabular Data Using Large Language Model
Hengyuan Zhang, Peng Chang, Zongcheng Ji
cs.CL, cs.AI
In the realm of natural language processing, the understanding of tabular data has perpetually stood as a focal point of scholarly inquiry. The emergence of expansive language models, exemplified by the likes of ChatGPT, has ushered in a wave of endeavors wherein researchers aim to harness these models for tasks relate...
http://arxiv.org/pdf/2308.11891v2
2023-08-28T14:07:12
2023-08-23T03:38:21
{ "title": [ -0.052734375, -0.009666443, -0.050750732, -0.068237305, -0.007217407, 0.010978699, -0.09411621, -0.003358841, 0.008094788, 0.037384033, -0.010566711, -0.026504517, 0.016571045, -0.012107849, -0.06213379, -0.006538391, -0.0052604675, ...
http://arxiv.org/abs/2308.11878v1
Cabrita: closing the gap for foreign languages
Celio Larcher, Marcos Piau, Paulo Finardi, Pedro Gengo, Piero Esposito, Vinicius Caridá
cs.CL, cs.AI, cs.LG
The strategy of training the model from scratch in a specific language or domain serves two essential purposes: i) enhancing performance in the particular linguistic or domain context, and ii) ensuring effective tokenization. The main limitation inherent to this approach lies in the associated cost, which can reach six...
http://arxiv.org/pdf/2308.11878v1
2023-08-23T02:49:35
2023-08-23T02:49:35
{ "title": [ -0.0066108704, -0.00447464, -0.024368286, -0.011413574, -0.055480957, -0.0026073456, -0.06036377, -0.042419434, -0.03955078, 0.045074463, -0.0020751953, -0.017715454, 0.018310547, -0.041503906, -0.012962341, -0.024856567, -0.010681152, ...
http://arxiv.org/abs/2308.11827v1
Exploring the Effectiveness of GPT Models in Test-Taking: A Case Study of the Driver's License Knowledge Test
Saba Rahimi, Tucker Balch, Manuela Veloso
cs.CL, cs.AI, cs.LG
Large language models such as Open AI's Generative Pre-trained Transformer (GPT) models are proficient at answering questions, but their knowledge is confined to the information present in their training data. This limitation renders them ineffective when confronted with questions about recent developments or non-publi...
http://arxiv.org/pdf/2308.11827v1
2023-08-22T23:18:53
2023-08-22T23:18:53
{ "title": [ 0.00844574, 0.0071868896, -0.02255249, -0.0016708374, -0.0024929047, -0.039367676, 0.010475159, -0.043395996, -0.013496399, 0.019760132, -0.035461426, 0.0076026917, -0.015312195, -0.01348114, -0.05014038, -0.025177002, -0.020751953, ...
http://arxiv.org/abs/2308.11807v1
Towards an On-device Agent for Text Rewriting
Yun Zhu, Yinxiao Liu, Felix Stahlberg, Shankar Kumar, Yu-hui Chen, Liangchen Luo, Lei Shu, Renjie Liu, Jindong Chen, Lei Meng
cs.CL
Large Language Models (LLMs) have demonstrated impressive capabilities for text rewriting. Nonetheless, the large sizes of these models make them impractical for on-device inference, which would otherwise allow for enhanced privacy and economical inference. Creating a smaller yet potent language model for text rewritin...
http://arxiv.org/pdf/2308.11807v1
2023-08-22T22:18:38
2023-08-22T22:18:38
{ "title": [ -0.039276123, -0.03387451, -0.01675415, -0.059051514, -0.015571594, 0.025421143, -0.06549072, 0.018157959, -0.024551392, 0.04647827, -0.02607727, 0.022354126, 0.019363403, 0.0005350113, -0.0357666, 0.0024738312, -0.02758789, 0.036712...
http://arxiv.org/abs/2308.11780v1
Few-shot Anomaly Detection in Text with Deviation Learning
Anindya Sundar Das, Aravind Ajay, Sriparna Saha, Monowar Bhuyan
cs.LG, cs.CL
Most current methods for detecting anomalies in text concentrate on constructing models solely relying on unlabeled data. These models operate on the presumption that no labeled anomalous examples are available, which prevents them from utilizing prior knowledge of anomalies that are typically present in small numbers ...
http://arxiv.org/pdf/2308.11780v1
2023-08-22T20:40:21
2023-08-22T20:40:21
{ "title": [ -0.004245758, -0.0029239655, -0.026885986, -0.02305603, -0.0055999756, -0.012580872, -0.010032654, -0.0018787384, -0.013008118, 0.016845703, -0.040924072, 0.034942627, -0.0008544922, -0.00044608116, -0.024429321, -0.024017334, -0.0141983...
http://arxiv.org/abs/2308.11773v2
Identifying depression-related topics in smartphone-collected free-response speech recordings using an automatic speech recognition system and a deep learning topic model
Yuezhou Zhang, Amos A Folarin, Judith Dineley, Pauline Conde, Valeria de Angel, Shaoxiong Sun, Yatharth Ranjan, Zulqarnain Rashid, Callum Stewart, Petroula Laiou, Heet Sankesara, Linglong Qian, Faith Matcham, Katie M White, Carolin Oetzmann, Femke Lamers, Sara Siddi, Sara Simblett, Björn W. Schuller, Srinivasan Vairava...
cs.CL, cs.CY, cs.SD, eess.AS, q-bio.QM
Language use has been shown to correlate with depression, but large-scale validation is needed. Traditional methods like clinic studies are expensive. So, natural language processing has been employed on social media to predict depression, but limitations remain-lack of validated labels, biased user samples, and no con...
http://arxiv.org/pdf/2308.11773v2
2023-09-05T11:59:25
2023-08-22T20:30:59
{ "title": [ 0.014877319, -0.001461029, -0.005207062, -0.07006836, -0.0116119385, 0.0020046234, -0.06036377, -0.000518322, 0.00491333, 0.034301758, -0.007904053, -0.0019702911, 0.0033664703, -0.038269043, -0.028762817, -0.009567261, -0.013374329, ...
http://arxiv.org/abs/2308.11764v4
Halo: Estimation and Reduction of Hallucinations in Open-Source Weak Large Language Models
Mohamed Elaraby, Mengyin Lu, Jacob Dunn, Xueying Zhang, Yu Wang, Shizhu Liu, Pingchuan Tian, Yuping Wang, Yuxuan Wang
cs.CL, cs.AI
Large Language Models (LLMs) have revolutionized Natural Language Processing (NLP). Although convenient for research and practical applications, open-source LLMs with fewer parameters often suffer from severe hallucinations compared to their larger counterparts. This paper focuses on measuring and reducing hallucinatio...
http://arxiv.org/pdf/2308.11764v4
2023-09-13T18:01:36
2023-08-22T20:12:49
{ "title": [ -0.014480591, 0.0013904572, -0.031280518, -0.052093506, 0.000385046, -0.005832672, -0.049102783, -0.02027893, 0.0072364807, 0.035369873, 0.0024299622, -0.004711151, 0.036132812, -0.036834717, -0.05291748, 0.029312134, -0.006252289, 0...
http://arxiv.org/abs/2308.11730v1
Knowledge Graph Prompting for Multi-Document Question Answering
Yu Wang, Nedim Lipka, Ryan A. Rossi, Alexa Siu, Ruiyi Zhang, Tyler Derr
cs.CL, cs.AI, cs.IR, cs.LG
The 'pre-train, prompt, predict' paradigm of large language models (LLMs) has achieved remarkable success in open-domain question answering (OD-QA). However, few works explore this paradigm in the scenario of multi-document question answering (MD-QA), a task demanding a thorough understanding of the logical association...
http://arxiv.org/pdf/2308.11730v1
2023-08-22T18:41:31
2023-08-22T18:41:31
{ "title": [ -0.025253296, -0.014198303, -0.03729248, -0.07269287, -0.02243042, 0.0067367554, -0.05947876, 0.018798828, 0.005508423, 0.025909424, -0.035339355, 0.0113220215, 0.00010108948, -0.0015506744, -0.043884277, -0.009712219, 0.0055770874, ...
http://arxiv.org/abs/2308.11696v3
Efficient Benchmarking (of Language Models)
Yotam Perlitz, Elron Bandel, Ariel Gera, Ofir Arviv, Liat Ein-Dor, Eyal Shnarch, Noam Slonim, Michal Shmueli-Scheuer, Leshem Choshen
cs.CL, cs.AI, cs.CV, cs.LG
The increasing versatility of language models LMs has given rise to a new class of benchmarks that comprehensively assess a broad range of capabilities. Such benchmarks are associated with massive computational costs reaching thousands of GPU hours per model. However the efficiency aspect of these evaluation efforts ha...
http://arxiv.org/pdf/2308.11696v3
2023-09-18T16:25:23
2023-08-22T17:59:30
{ "title": [ -0.010620117, -0.0061836243, -0.025344849, -0.051574707, -0.043670654, -0.013969421, -0.05935669, 0.0033016205, -0.017150879, 0.0524292, -0.008460999, 0.0128479, 0.014152527, -0.04449463, -0.007888794, 0.008964539, 0.010284424, 0.028...
http://arxiv.org/abs/2308.11606v2
StoryBench: A Multifaceted Benchmark for Continuous Story Visualization
Emanuele Bugliarello, Hernan Moraldo, Ruben Villegas, Mohammad Babaeizadeh, Mohammad Taghi Saffar, Han Zhang, Dumitru Erhan, Vittorio Ferrari, Pieter-Jan Kindermans, Paul Voigtlaender
cs.CV, cs.CL
Generating video stories from text prompts is a complex task. In addition to having high visual quality, videos need to realistically adhere to a sequence of text prompts whilst being consistent throughout the frames. Creating a benchmark for video generation requires data annotated over time, which contrasts with the ...
http://arxiv.org/pdf/2308.11606v2
2023-10-12T17:50:38
2023-08-22T17:53:55
{ "title": [ -0.028381348, -0.009384155, -0.049438477, -0.056671143, -0.0030231476, -0.0075263977, -0.030151367, 0.0025844574, 0.007270813, 0.034057617, -0.017486572, 0.027648926, -0.022613525, -0.015991211, -0.047058105, 0.0013341904, -0.0073280334,...
http://arxiv.org/abs/2308.11601v2
Tryage: Real-time, intelligent Routing of User Prompts to Large Language Models
Surya Narayanan Hari, Matt Thomson
cs.LG, cs.AI, cs.CL, cs.MA
The introduction of the transformer architecture and the self-attention mechanism has led to an explosive production of language models trained on specific downstream tasks and data domains. With over 200, 000 models in the Hugging Face ecosystem, users grapple with selecting and optimizing models to suit multifaceted ...
http://arxiv.org/pdf/2308.11601v2
2023-08-23T17:34:17
2023-08-22T17:48:24
{ "title": [ -0.015945435, -0.029174805, -0.026748657, -0.058929443, -0.00409317, -0.0067825317, -0.040496826, -0.003955841, -0.010520935, 0.022735596, -0.01272583, 0.010566711, -0.0064201355, -0.024230957, -0.03414917, -0.002368927, -0.00274086, ...
http://arxiv.org/abs/2308.11596v3
SeamlessM4T: Massively Multilingual & Multimodal Machine Translation
Seamless Communication, Loïc Barrault, Yu-An Chung, Mariano Cora Meglioli, David Dale, Ning Dong, Paul-Ambroise Duquenne, Hady Elsahar, Hongyu Gong, Kevin Heffernan, John Hoffman, Christopher Klaiber, Pengwei Li, Daniel Licht, Jean Maillard, Alice Rakotoarison, Kaushik Ram Sadagopan, Guillaume Wenzek, Ethan Ye, Bapi Ak...
cs.CL, I.2.7
What does it take to create the Babel Fish, a tool that can help individuals translate speech between any two languages? While recent breakthroughs in text-based models have pushed machine translation coverage beyond 200 languages, unified speech-to-speech translation models have yet to achieve similar strides. More sp...
http://arxiv.org/pdf/2308.11596v3
2023-10-25T03:52:07
2023-08-22T17:44:18
{ "title": [ -0.04446411, -0.020050049, -0.036132812, -0.055267334, 0.0028762817, 0.016082764, -0.036132812, 0.0069999695, -0.0074386597, 0.0602417, 0.00894165, 0.008010864, 0.03189087, -0.021850586, -0.037994385, 0.022415161, -0.0044403076, 0.03...
http://arxiv.org/abs/2308.11563v1
Using ChatGPT as a CAT tool in Easy Language translation
Silvana Deilen, Sergio Hernández Garrido, Ekaterina Lapshinova-Koltunski, Christiane Maaß
cs.CL
This study sets out to investigate the feasibility of using ChatGPT to translate citizen-oriented administrative texts into German Easy Language, a simplified, controlled language variety that is adapted to the needs of people with reading impairments. We use ChatGPT to translate selected texts from websites of German ...
http://arxiv.org/pdf/2308.11563v1
2023-08-22T16:59:31
2023-08-22T16:59:31
{ "title": [ -0.015655518, -0.022094727, -0.005027771, -0.0056610107, -0.049682617, 0.008918762, -0.018341064, -0.015319824, -0.008834839, 0.062683105, -0.0037155151, -0.0031585693, 0.038513184, -0.04095459, -0.015838623, 0.022415161, -0.00031638145,...
http://arxiv.org/abs/2308.11537v1
BELB: a Biomedical Entity Linking Benchmark
Samuele Garda, Leon Weber-Genzel, Robert Martin, Ulf Leser
cs.CL
Biomedical entity linking (BEL) is the task of grounding entity mentions to a knowledge base. It plays a vital role in information extraction pipelines for the life sciences literature. We review recent work in the field and find that, as the task is absent from existing benchmarks for biomedical text mining, different...
http://arxiv.org/pdf/2308.11537v1
2023-08-22T16:05:18
2023-08-22T16:05:18
{ "title": [ -0.014038086, -0.009056091, -0.05331421, -0.10144043, 0.017578125, 0.0020923615, -0.060333252, 0.006187439, 0.009223938, 0.026016235, 0.020126343, -0.000032782555, 0.020095825, -0.046661377, -0.011184692, 0.02986145, -0.002243042, 0....
http://arxiv.org/abs/2308.11531v2
Empowering Refugee Claimants and their Lawyers: Using Machine Learning to Examine Decision-Making in Refugee Law
Claire Barale
cs.CL
Our project aims at helping and supporting stakeholders in refugee status adjudications, such as lawyers, judges, governing bodies, and claimants, in order to make better decisions through data-driven intelligence and increase the understanding and transparency of the refugee application process for all involved partie...
http://arxiv.org/pdf/2308.11531v2
2023-09-21T14:19:37
2023-08-22T15:59:21
{ "title": [ -0.01309967, -0.017822266, -0.03668213, 0.00207901, -0.028930664, -0.030410767, 0.004196167, -0.01121521, -0.005844116, 0.010437012, -0.04647827, -0.006134033, 0.0033550262, -0.026885986, -0.021102905, -0.03164673, -0.0001296997, 0.0...
http://arxiv.org/abs/2308.11507v2
Unsupervised Prototype Adapter for Vision-Language Models
Yi Zhang, Ce Zhang, Xueting Hu, Zhihai He
cs.CV, cs.CL
Recently, large-scale pre-trained vision-language models (e.g. CLIP and ALIGN) have demonstrated remarkable effectiveness in acquiring transferable visual representations. To leverage the valuable knowledge encoded within these models for downstream tasks, several fine-tuning approaches, including prompt tuning methods...
http://arxiv.org/pdf/2308.11507v2
2023-08-25T00:07:50
2023-08-22T15:28:49
{ "title": [ -0.019180298, -0.005558014, -0.041412354, -0.049804688, -0.004295349, -0.0060310364, -0.048614502, -0.0076026917, 0.01687622, 0.01713562, -0.028457642, -0.020812988, 0.009475708, -0.020812988, -0.03012085, -0.0053863525, 0.0115356445, ...
http://arxiv.org/abs/2308.11490v2
Can Authorship Representation Learning Capture Stylistic Features?
Andrew Wang, Cristina Aggazzotti, Rebecca Kotula, Rafael Rivera Soto, Marcus Bishop, Nicholas Andrews
cs.CL, cs.LG
Automatically disentangling an author's style from the content of their writing is a longstanding and possibly insurmountable problem in computational linguistics. At the same time, the availability of large text corpora furnished with author labels has recently enabled learning authorship representations in a purely d...
http://arxiv.org/pdf/2308.11490v2
2023-08-24T20:52:01
2023-08-22T15:10:45
{ "title": [ -0.043121338, 0.012588501, -0.045318604, -0.008644104, -0.026519775, -0.01159668, -0.045654297, -0.01878357, -0.005142212, 0.039215088, 0.0068092346, 0.00039434433, 0.040802002, -0.025421143, -0.019744873, 0.019454956, -0.006340027, ...
http://arxiv.org/abs/2308.11683v1
Learning to generate and corr- uh I mean repair language in real-time
Arash Eshghi, Arash Ashrafzadeh
cs.CL
In conversation, speakers produce language incrementally, word by word, while continuously monitoring the appropriateness of their own contribution in the dynamically unfolding context of the conversation; and this often leads them to repair their own utterance on the fly. This real-time language processing capacity is...
http://arxiv.org/pdf/2308.11683v1
2023-08-22T15:09:55
2023-08-22T15:09:55
{ "title": [ 0.023925781, -0.03756714, -0.022979736, -0.081726074, -0.03173828, -0.0066719055, -0.041992188, 0.032348633, -0.02835083, 0.019073486, -0.018615723, 0.01739502, -0.023605347, -0.019836426, 0.0020999908, -0.034820557, 0.0036964417, 0....
http://arxiv.org/abs/2308.11483v1
Large Language Models Sensitivity to The Order of Options in Multiple-Choice Questions
Pouya Pezeshkpour, Estevam Hruschka
cs.CL, cs.AI, cs.LG
Large Language Models (LLMs) have demonstrated remarkable capabilities in various NLP tasks. However, previous works have shown these models are sensitive towards prompt wording, and few-shot demonstrations and their order, posing challenges to fair assessment of these models. As these models become more powerful, it b...
http://arxiv.org/pdf/2308.11483v1
2023-08-22T14:54:59
2023-08-22T14:54:59
{ "title": [ -0.022964478, 0.0047454834, -0.028533936, -0.0024299622, -0.020187378, -0.030380249, -0.0069122314, -0.028549194, 0.0041160583, 0.02671814, -0.030822754, -0.008277893, -0.009788513, -0.013183594, -0.048553467, -0.010902405, -0.013298035,...
http://arxiv.org/abs/2308.11466v2
SONAR: Sentence-Level Multimodal and Language-Agnostic Representations
Paul-Ambroise Duquenne, Holger Schwenk, Benoît Sagot
cs.CL
We introduce SONAR, a new multilingual and multimodal fixed-size sentence embedding space. Our single text encoder, covering 200 languages, substantially outperforms existing sentence embeddings such as LASER3 and LabSE on the xsim and xsim++ multilingual similarity search tasks. Speech segments can be embedded in the ...
http://arxiv.org/pdf/2308.11466v2
2023-08-23T10:46:16
2023-08-22T14:25:15
{ "title": [ -0.0070533752, -0.0066452026, -0.020507812, -0.05441284, 0.012184143, -0.0018968582, -0.031707764, -0.017791748, 0.015640259, 0.036193848, 0.0000590086, -0.012184143, -0.00022041798, -0.021072388, -0.052368164, 0.014778137, -0.0149383545...
http://arxiv.org/abs/2308.11447v1
Aspect-oriented Opinion Alignment Network for Aspect-Based Sentiment Classification
Xueyi Liu, Rui Hou, Yanglei Gan, Da Luo, Changlin Li, Xiaojun Shi, Qiao Liu
cs.CL, cs.AI, I.2.7
Aspect-based sentiment classification is a crucial problem in fine-grained sentiment analysis, which aims to predict the sentiment polarity of the given aspect according to its context. Previous works have made remarkable progress in leveraging attention mechanism to extract opinion words for different aspects. However...
http://arxiv.org/pdf/2308.11447v1
2023-08-22T13:55:36
2023-08-22T13:55:36
{ "title": [ -0.0032539368, -0.00014901161, -0.044769287, -0.05480957, 0.0060806274, -0.0029563904, -0.023727417, 0.009056091, 0.010398865, 0.055999756, -0.01373291, -0.018249512, -0.0014543533, -0.00440979, -0.0793457, -0.0024032593, -0.015388489, ...
http://arxiv.org/abs/2308.12381v1
Inferring gender from name: a large scale performance evaluation study
Kriste Krstovski, Yao Lu, Ye Xu
cs.CL, cs.AI, cs.LG
A person's gender is a crucial piece of information when performing research across a wide range of scientific disciplines, such as medicine, sociology, political science, and economics, to name a few. However, in increasing instances, especially given the proliferation of big data, gender information is not readily av...
http://arxiv.org/pdf/2308.12381v1
2023-08-22T13:38:45
2023-08-22T13:38:45
{ "title": [ -0.044647217, -0.0032634735, -0.07098389, -0.06866455, 0.022521973, -0.011047363, -0.0362854, 0.0060920715, 0.02168274, 0.024398804, -0.017166138, -0.0021514893, 0.025238037, -0.0072517395, -0.04498291, 0.006210327, -0.021026611, 0.0...
http://arxiv.org/abs/2308.11432v2
A Survey on Large Language Model based Autonomous Agents
Lei Wang, Chen Ma, Xueyang Feng, Zeyu Zhang, Hao Yang, Jingsen Zhang, Zhiyuan Chen, Jiakai Tang, Xu Chen, Yankai Lin, Wayne Xin Zhao, Zhewei Wei, Ji-Rong Wen
cs.AI, cs.CL
Autonomous agents have long been a prominent research focus in both academic and industry communities. Previous research in this field often focuses on training agents with limited knowledge within isolated environments, which diverges significantly from human learning processes, and thus makes the agents hard to achie...
http://arxiv.org/pdf/2308.11432v2
2023-09-07T04:42:48
2023-08-22T13:30:37
{ "title": [ -0.039245605, -0.00598526, -0.028076172, -0.056427002, 0.010055542, 0.007698059, -0.040527344, 0.00819397, -0.0013170242, 0.03439331, -0.019317627, -0.0027313232, 0.0059890747, -0.009010315, -0.03515625, 0.00041627884, -0.020812988, ...
http://arxiv.org/abs/2308.11411v1
Extracting Relational Triples Based on Graph Recursive Neural Network via Dynamic Feedback Forest Algorithm
Hongyin Zhu
cs.CL
Extracting relational triples (subject, predicate, object) from text enables the transformation of unstructured text data into structured knowledge. The named entity recognition (NER) and the relation extraction (RE) are two foundational subtasks in this knowledge generation pipeline. The integration of subtasks poses ...
http://arxiv.org/pdf/2308.11411v1
2023-08-22T13:00:13
2023-08-22T13:00:13
{ "title": [ -0.03805542, -0.0006313324, -0.053375244, -0.06262207, -0.024230957, 0.004432678, -0.07531738, 0.0090408325, 0.019989014, 0.052642822, -0.0124053955, 0.018341064, -0.021850586, -0.001039505, -0.07525635, 0.0020694733, 0.0035076141, 0...
http://arxiv.org/abs/2308.11380v1
Convoifilter: A case study of doing cocktail party speech recognition
Thai-Binh Nguyen, Alexander Waibel
cs.SD, cs.CL, eess.AS
This paper presents an end-to-end model designed to improve automatic speech recognition (ASR) for a particular speaker in a crowded, noisy environment. The model utilizes a single-channel speech enhancement module that isolates the speaker's voice from background noise, along with an ASR module. Through this approach,...
http://arxiv.org/pdf/2308.11380v1
2023-08-22T12:09:30
2023-08-22T12:09:30
{ "title": [ 0.037078857, -0.0053901672, -0.06951904, -0.08520508, 0.0054244995, -0.033447266, -0.06555176, 0.02229309, -0.0025501251, 0.047302246, -0.03643799, -0.0011196136, 0.03817749, -0.024230957, -0.01663208, -0.012008667, -0.014160156, 0.0...
http://arxiv.org/abs/2308.11351v1
M3PS: End-to-End Multi-Grained Multi-Modal Attribute-Aware Product Summarization in E-commerce
Tao Chen, Ze Lin, Hui Li, Jiayi Ji, Yiyi Zhou, Guanbin Li, Rongrong Ji
cs.MM, cs.CL
Given the long textual product information and the product image, Multi-Modal Product Summarization (MMPS) aims to attract customers' interest and increase their desire to purchase by highlighting product characteristics with a short textual summary. Existing MMPS methods have achieved promising performance. Neverthele...
http://arxiv.org/pdf/2308.11351v1
2023-08-22T11:00:09
2023-08-22T11:00:09
{ "title": [ -0.010047913, -0.0052108765, -0.046875, -0.082336426, 0.014450073, -0.00040769577, -0.056518555, 0.0039405823, 0.00680542, 0.054016113, 0.0014715195, 0.0079956055, 0.048583984, -0.01448822, -0.05065918, 0.013557434, -0.0011949539, 0....
http://arxiv.org/abs/2308.11284v1
LEAP: Efficient and Automated Test Method for NLP Software
Mingxuan Xiao, Yan Xiao, Hai Dong, Shunhui Ji, Pengcheng Zhang
cs.SE, cs.CL
The widespread adoption of DNNs in NLP software has highlighted the need for robustness. Researchers proposed various automatic testing techniques for adversarial test cases. However, existing methods suffer from two limitations: weak error-discovering capabilities, with success rates ranging from 0% to 24.6% for BERT-...
http://arxiv.org/pdf/2308.11284v1
2023-08-22T08:51:10
2023-08-22T08:51:10
{ "title": [ -0.03274536, -0.01939392, -0.055480957, -0.06347656, 0.0037384033, -0.003168106, -0.038391113, 0.0143585205, -0.004940033, 0.01828003, -0.038208008, 0.013893127, 0.018630981, -0.011123657, -0.032836914, -0.0029697418, -0.01828003, 0....
http://arxiv.org/abs/2308.11276v1
Music Understanding LLaMA: Advancing Text-to-Music Generation with Question Answering and Captioning
Shansong Liu, Atin Sakkeer Hussain, Chenshuo Sun, Ying Shan
cs.SD, cs.AI, cs.CL, cs.MM, eess.AS
Text-to-music generation (T2M-Gen) faces a major obstacle due to the scarcity of large-scale publicly available music datasets with natural language captions. To address this, we propose the Music Understanding LLaMA (MU-LLaMA), capable of answering music-related questions and generating captions for music files. Our m...
http://arxiv.org/pdf/2308.11276v1
2023-08-22T08:43:33
2023-08-22T08:43:33
{ "title": [ -0.020355225, -0.013381958, -0.048553467, -0.06286621, -0.0046463013, -0.0015687943, -0.035614014, 0.025970459, -0.0034637451, 0.03744507, -0.042663574, 0.02645874, 0.026443481, -0.01927185, -0.0234375, -0.0007157326, -0.016113281, 0...
http://arxiv.org/abs/2308.11257v2
HopPG: Self-Iterative Program Generation for Multi-Hop Question Answering over Heterogeneous Knowledge
Yingyao Wang, Yongwei Zhou, Chaoqun Duan, Junwei Bao, Tiejun Zhao
cs.CL
The semantic parsing-based method is an important research branch for knowledge-based question answering. It usually generates executable programs lean upon the question and then conduct them to reason answers over a knowledge base. Benefit from this inherent mechanism, it has advantages in the performance and the inte...
http://arxiv.org/pdf/2308.11257v2
2023-09-10T17:05:36
2023-08-22T08:00:50
{ "title": [ -0.034240723, -0.014732361, -0.04397583, -0.046051025, -0.011741638, 0.0064315796, -0.050323486, 0.025405884, 0.016662598, 0.034088135, -0.030975342, 0.006580353, -0.0020427704, -0.005344391, -0.051971436, -0.0079956055, 0.004676819, ...
http://arxiv.org/abs/2308.11224v2
Evaluating Large Language Models on Graphs: Performance Insights and Comparative Analysis
Chang Liu, Bo Wu
cs.AI, cs.CL
Large Language Models (LLMs) have garnered considerable interest within both academic and industrial. Yet, the application of LLMs to graph data remains under-explored. In this study, we evaluate the capabilities of four LLMs in addressing several analytical problems with graph data. We employ four distinct evaluation ...
http://arxiv.org/pdf/2308.11224v2
2023-09-09T03:14:10
2023-08-22T06:32:07
{ "title": [ -0.0418396, -0.025466919, -0.03894043, -0.070739746, -0.012512207, 0.016235352, -0.07910156, 0.019058228, -0.00019490719, 0.026367188, 0.0016231537, -0.0021800995, -0.0018482208, -0.025360107, -0.035369873, 0.014167786, -0.012306213, ...
http://arxiv.org/abs/2308.11189v1
Diversity Measures: Domain-Independent Proxies for Failure in Language Model Queries
Noel Ngu, Nathaniel Lee, Paulo Shakarian
cs.CL, cs.AI, cs.LG
Error prediction in large language models often relies on domain-specific information. In this paper, we present measures for quantification of error in the response of a large language model based on the diversity of responses to a given prompt - hence independent of the underlying application. We describe how three s...
http://arxiv.org/pdf/2308.11189v1
2023-08-22T04:49:23
2023-08-22T04:49:23
{ "title": [ -0.010696411, 0.0026073456, -0.04257202, -0.058135986, 0.020553589, -0.022506714, -0.03363037, -0.005306244, 0.020462036, 0.041992188, -0.018798828, -0.004753113, 0.0287323, -0.017547607, -0.06225586, 0.011413574, -0.01789856, 0.0269...
http://arxiv.org/abs/2308.11171v1
ViCo: Engaging Video Comment Generation with Human Preference Rewards
Yuchong Sun, Bei Liu, Xu Chen, Ruihua Song, Jianlong Fu
cs.CV, cs.CL
Engaging video comments play an important role in video social media, as they are the carrier of feelings, thoughts, or humor of the audience. Preliminary works have made initial exploration for video comment generation by adopting caption-style encoder-decoder models. However, comment generation presents some unique c...
http://arxiv.org/pdf/2308.11171v1
2023-08-22T04:01:01
2023-08-22T04:01:01
{ "title": [ 0.00089263916, 0.0022773743, -0.031433105, -0.04348755, 0.008598328, -0.015960693, -0.012382507, -0.011726379, -0.001499176, 0.039489746, -0.02279663, 0.0046691895, -0.01600647, -0.0014381409, -0.06213379, -0.012367249, -0.024353027, ...
http://arxiv.org/abs/2308.11148v2
LLaMA-Reviewer: Advancing Code Review Automation with Large Language Models through Parameter-Efficient Fine-Tuning
Junyi Lu, Lei Yu, Xiaojia Li, Li Yang, Chun Zuo
cs.SE, cs.CL, cs.LG
The automation of code review activities, a long-standing pursuit in software engineering, has been primarily addressed by numerous domain-specific pre-trained models. Despite their success, these models frequently demand extensive resources for pre-training from scratch. In contrast, Large Language Models (LLMs) provi...
http://arxiv.org/pdf/2308.11148v2
2023-09-05T02:28:49
2023-08-22T03:10:40
{ "title": [ -0.052368164, -0.033172607, -0.029754639, -0.047088623, -0.033599854, 0.006454468, -0.04916382, 0.014785767, -0.007675171, 0.032073975, -0.03930664, 0.02482605, 0.015167236, -0.0112838745, -0.012008667, -0.002664566, -0.008522034, 0....
http://arxiv.org/abs/2308.11138v1
NLP-based detection of systematic anomalies among the narratives of consumer complaints
Peiheng Gao, Ning Sun, Xuefeng Wang, Chen Yang, Ričardas Zitikis
stat.ME, cs.CL, q-fin.RM, stat.ML
We develop an NLP-based procedure for detecting systematic nonmeritorious consumer complaints, simply called systematic anomalies, among complaint narratives. While classification algorithms are used to detect pronounced anomalies, in the case of smaller and frequent systematic anomalies, the algorithms may falter due ...
http://arxiv.org/pdf/2308.11138v1
2023-08-22T02:39:42
2023-08-22T02:39:42
{ "title": [ -0.004886627, 0.0013084412, -0.031204224, -0.033996582, -0.0013170242, -0.014259338, -0.018356323, -0.0107421875, -0.000954628, 0.0128479, -0.033569336, 0.017440796, -0.00027418137, -0.0067977905, -0.023483276, -0.012702942, -0.036956787...