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/2304.12995v1
AudioGPT: Understanding and Generating Speech, Music, Sound, and Talking Head
Rongjie Huang, Mingze Li, Dongchao Yang, Jiatong Shi, Xuankai Chang, Zhenhui Ye, Yuning Wu, Zhiqing Hong, Jiawei Huang, Jinglin Liu, Yi Ren, Zhou Zhao, Shinji Watanabe
cs.CL, cs.AI, cs.SD, eess.AS
Large language models (LLMs) have exhibited remarkable capabilities across a variety of domains and tasks, challenging our understanding of learning and cognition. Despite the recent success, current LLMs are not capable of processing complex audio information or conducting spoken conversations (like Siri or Alexa). In...
http://arxiv.org/pdf/2304.12995v1
2023-04-25T17:05:38
2023-04-25T17:05:38
{ "title": [ -0.014450073, -0.011985779, -0.054382324, -0.04647827, 0.0021438599, -0.017089844, -0.04232788, 0.014907837, -0.029006958, 0.026779175, -0.009529114, 0.022201538, 0.031463623, -0.04360962, 0.015914917, 0.015319824, -0.014335632, 0.03...
http://arxiv.org/abs/2304.12986v2
Measuring Massive Multitask Chinese Understanding
Hui Zeng
cs.CL, cs.AI, I.2.7
The development of large-scale Chinese language models is flourishing, yet there is a lack of corresponding capability assessments. Therefore, we propose a test to measure the multitask accuracy of large Chinese language models. This test encompasses four major domains, including medicine, law, psychology, and educatio...
http://arxiv.org/pdf/2304.12986v2
2023-05-15T16:41:08
2023-04-25T16:51:53
{ "title": [ -0.034851074, 0.0030937195, -0.033050537, -0.013122559, -0.02067566, -0.012969971, -0.025634766, -0.015396118, 0.0030288696, 0.033294678, -0.018463135, -0.008453369, -0.01222229, -0.03302002, -0.043884277, -0.007255554, -0.0037155151, ...
http://arxiv.org/abs/2304.12982v1
Intent Induction from Conversations for Task-Oriented Dialogue Track at DSTC 11
James Gung, Raphael Shu, Emily Moeng, Wesley Rose, Salvatore Romeo, Yassine Benajiba, Arshit Gupta, Saab Mansour, Yi Zhang
cs.CL
With increasing demand for and adoption of virtual assistants, recent work has investigated ways to accelerate bot schema design through the automatic induction of intents or the induction of slots and dialogue states. However, a lack of dedicated benchmarks and standardized evaluation has made progress difficult to tr...
http://arxiv.org/pdf/2304.12982v1
2023-04-25T16:45:50
2023-04-25T16:45:50
{ "title": [ -0.004512787, 0.003320694, -0.06427002, -0.05206299, 0.014915466, -0.027145386, -0.01940918, 0.012283325, -0.0068588257, 0.01184082, -0.012062073, 0.0082092285, -0.030166626, -0.03753662, -0.017547607, -0.014839172, -0.02684021, 0.02...
http://arxiv.org/abs/2304.12979v1
GMNLP at SemEval-2023 Task 12: Sentiment Analysis with Phylogeny-Based Adapters
Md Mahfuz Ibn Alam, Ruoyu Xie, Fahim Faisal, Antonios Anastasopoulos
cs.CL, cs.LG
This report describes GMU's sentiment analysis system for the SemEval-2023 shared task AfriSenti-SemEval. We participated in all three sub-tasks: Monolingual, Multilingual, and Zero-Shot. Our approach uses models initialized with AfroXLMR-large, a pre-trained multilingual language model trained on African languages and...
http://arxiv.org/pdf/2304.12979v1
2023-04-25T16:39:51
2023-04-25T16:39:51
{ "title": [ -0.0037670135, -0.014038086, -0.03640747, -0.0418396, 0.010032654, -0.012702942, -0.020965576, -0.009544373, 0.025238037, 0.036621094, -0.017669678, -0.0118637085, 0.0002965927, -0.014427185, -0.057647705, 0.01159668, -0.01108551, 0....
http://arxiv.org/abs/2304.12959v1
Escaping the sentence-level paradigm in machine translation
Matt Post, Marcin Junczys-Dowmunt
cs.CL
It is well-known that document context is vital for resolving a range of translation ambiguities, and in fact the document setting is the most natural setting for nearly all translation. It is therefore unfortunate that machine translation -- both research and production -- largely remains stuck in a decades-old senten...
http://arxiv.org/pdf/2304.12959v1
2023-04-25T16:09:02
2023-04-25T16:09:02
{ "title": [ -0.020629883, -0.013435364, -0.0017108917, -0.039764404, -0.01235199, 0.0074768066, -0.014770508, -0.0070152283, 0.009513855, 0.049316406, -0.0034427643, -0.004623413, 0.0317688, -0.018707275, -0.053619385, 0.02456665, -0.011070251, ...
http://arxiv.org/abs/2304.12955v2
Nondeterministic Stacks in Neural Networks
Brian DuSell
cs.CL
Human language is full of compositional syntactic structures, and although neural networks have contributed to groundbreaking improvements in computer systems that process language, widely-used neural network architectures still exhibit limitations in their ability to process syntax. To address this issue, prior work h...
http://arxiv.org/pdf/2304.12955v2
2023-05-18T01:56:21
2023-04-25T16:00:40
{ "title": [ -0.028442383, 0.005886078, -0.023284912, -0.042999268, 0.018875122, 0.004081726, -0.014663696, -0.014945984, 0.015052795, 0.038238525, -0.0063705444, -0.015525818, -0.026901245, -0.009979248, -0.06549072, -0.008026123, 0.0024909973, ...
http://arxiv.org/abs/2304.12888v1
Out-of-distribution Evidence-aware Fake News Detection via Dual Adversarial Debiasing
Qiang Liu, Junfei Wu, Shu Wu, Liang Wang
cs.CL, cs.AI
Evidence-aware fake news detection aims to conduct reasoning between news and evidence, which is retrieved based on news content, to find uniformity or inconsistency. However, we find evidence-aware detection models suffer from biases, i.e., spurious correlations between news/evidence contents and true/fake news labels...
http://arxiv.org/pdf/2304.12888v1
2023-04-25T14:59:25
2023-04-25T14:59:25
{ "title": [ 0.017120361, 0.028762817, -0.0048599243, -0.022201538, 0.02217102, -0.0042495728, 0.006160736, -0.049835205, -0.030960083, 0.025009155, -0.026489258, 0.02420044, -0.022033691, -0.018173218, -0.020751953, -0.034179688, -0.034606934, 0...
http://arxiv.org/abs/2304.12847v1
NLP-LTU at SemEval-2023 Task 10: The Impact of Data Augmentation and Semi-Supervised Learning Techniques on Text Classification Performance on an Imbalanced Dataset
Sana Sabah Al-Azzawi, György Kovács, Filip Nilsson, Tosin Adewumi, Marcus Liwicki
cs.CL
In this paper, we propose a methodology for task 10 of SemEval23, focusing on detecting and classifying online sexism in social media posts. The task is tackling a serious issue, as detecting harmful content on social media platforms is crucial for mitigating the harm of these posts on users. Our solution for this task...
http://arxiv.org/pdf/2304.12847v1
2023-04-25T14:19:46
2023-04-25T14:19:46
{ "title": [ 0.013427734, 0.013717651, -0.05001831, -0.029769897, 0.005832672, -0.041870117, -0.0071868896, -0.00472641, 0.015960693, 0.013893127, -0.070617676, 0.043273926, -0.0057792664, 0.003742218, -0.011726379, -0.019226074, -0.036254883, 0....
http://arxiv.org/abs/2304.12836v1
Lessons Learned from a Citizen Science Project for Natural Language Processing
Jan-Christoph Klie, Ji-Ung Lee, Kevin Stowe, Gözde Gül Şahin, Nafise Sadat Moosavi, Luke Bates, Dominic Petrak, Richard Eckart de Castilho, Iryna Gurevych
cs.CL
Many Natural Language Processing (NLP) systems use annotated corpora for training and evaluation. However, labeled data is often costly to obtain and scaling annotation projects is difficult, which is why annotation tasks are often outsourced to paid crowdworkers. Citizen Science is an alternative to crowdsourcing that...
http://arxiv.org/pdf/2304.12836v1
2023-04-25T14:08:53
2023-04-25T14:08:53
{ "title": [ 0.012939453, -0.009292603, -0.02166748, -0.048187256, -0.0073547363, -0.025222778, -0.017959595, 0.010253906, -0.015655518, 0.027664185, -0.04071045, -0.0017642975, -0.014564514, -0.014183044, -0.027664185, -0.013328552, 0.0038757324, ...
http://arxiv.org/abs/2304.12833v1
A New Information Theory of Certainty for Machine Learning
Arthur Jun Zhang
cs.IT, cs.AI, cs.CL, cs.IR, cs.LG, math.IT
Claude Shannon coined entropy to quantify the uncertainty of a random distribution for communication coding theory. We observe that the uncertainty nature of entropy also limits its direct usage in mathematical modeling. Therefore we propose a new concept troenpy,as the canonical dual of entropy, to quantify the certai...
http://arxiv.org/pdf/2304.12833v1
2023-04-25T14:03:57
2023-04-25T14:03:57
{ "title": [ -0.03488159, -0.00020372868, -0.049621582, -0.05105591, 0.028366089, -0.0074882507, -0.016815186, -0.006385803, -0.0014257431, 0.011779785, -0.021591187, 0.010040283, -0.023239136, -0.022903442, -0.034973145, 0.0032157898, -0.00007230043...
http://arxiv.org/abs/2304.12814v1
A Novel Dual of Shannon Information and Weighting Scheme
Arthur Jun Zhang
cs.CL, cs.IR, cs.IT, cs.LG, math.IT
Shannon Information theory has achieved great success in not only communication technology where it was originally developed for but also many other science and engineering fields such as machine learning and artificial intelligence. Inspired by the famous weighting scheme TF-IDF, we discovered that information entropy...
http://arxiv.org/pdf/2304.12814v1
2023-04-25T13:40:50
2023-04-25T13:40:50
{ "title": [ -0.061462402, -0.0029392242, -0.060668945, -0.055999756, 0.009376526, 0.0054855347, -0.019958496, 0.008056641, 0.00844574, 0.041748047, -0.019515991, 0.009063721, -0.025772095, -0.015563965, -0.053466797, -0.013542175, -0.0029335022, ...
http://arxiv.org/abs/2304.12776v1
State Spaces Aren't Enough: Machine Translation Needs Attention
Ali Vardasbi, Telmo Pessoa Pires, Robin M. Schmidt, Stephan Peitz
cs.CL, cs.AI
Structured State Spaces for Sequences (S4) is a recently proposed sequence model with successful applications in various tasks, e.g. vision, language modeling, and audio. Thanks to its mathematical formulation, it compresses its input to a single hidden state, and is able to capture long range dependencies while avoidi...
http://arxiv.org/pdf/2304.12776v1
2023-04-25T12:54:20
2023-04-25T12:54:20
{ "title": [ -0.019470215, -0.006752014, 0.015510559, -0.017913818, -0.04171753, 0.0018377304, -0.009185791, -0.01914978, -0.013023376, 0.044281006, -0.030853271, 0.000767231, 0.021469116, -0.02519226, -0.037078857, 0.004737854, 0.012771606, 0.01...
http://arxiv.org/abs/2304.14275v1
What's in a Name? Evaluating Assembly-Part Semantic Knowledge in Language Models through User-Provided Names in CAD Files
Peter Meltzer, Joseph G. Lambourne, Daniele Grandi
cs.CL, cs.LG, cc:68T50, I.2.7; J.6
Semantic knowledge of part-part and part-whole relationships in assemblies is useful for a variety of tasks from searching design repositories to the construction of engineering knowledge bases. In this work we propose that the natural language names designers use in Computer Aided Design (CAD) software are a valuable ...
http://arxiv.org/pdf/2304.14275v1
2023-04-25T12:30:01
2023-04-25T12:30:01
{ "title": [ -0.08319092, -0.012748718, -0.07611084, -0.031829834, -0.026916504, -0.013366699, -0.08508301, 0.017807007, 0.0035514832, 0.029388428, -0.018173218, 0.022460938, 0.014373779, -0.019515991, -0.01826477, 0.009292603, 0.0079574585, 0.01...
http://arxiv.org/abs/2304.12764v1
Test-Time Adaptation with Perturbation Consistency Learning
Yi Su, Yixin Ji, Juntao Li, Hai Ye, Min Zhang
cs.CL
Currently, pre-trained language models (PLMs) do not cope well with the distribution shift problem, resulting in models trained on the training set failing in real test scenarios. To address this problem, the test-time adaptation (TTA) shows great potential, which updates model parameters to suit the test data at the t...
http://arxiv.org/pdf/2304.12764v1
2023-04-25T12:29:22
2023-04-25T12:29:22
{ "title": [ -0.03591919, -0.0052719116, -0.045776367, -0.04208374, 0.013694763, -0.006385803, -0.013038635, 0.009567261, -0.005695343, 0.032714844, -0.035858154, 0.008590698, -0.030914307, -0.013175964, -0.044830322, -0.024414062, 0.007896423, 0...
http://arxiv.org/abs/2304.12730v2
CitePrompt: Using Prompts to Identify Citation Intent in Scientific Papers
Avishek Lahiri, Debarshi Kumar Sanyal, Imon Mukherjee
cs.CL
Citations in scientific papers not only help us trace the intellectual lineage but also are a useful indicator of the scientific significance of the work. Citation intents prove beneficial as they specify the role of the citation in a given context. In this paper, we present CitePrompt, a framework which uses the hithe...
http://arxiv.org/pdf/2304.12730v2
2023-05-03T11:47:55
2023-04-25T11:19:52
{ "title": [ -0.00053691864, 0.0061683655, -0.051116943, -0.022872925, 0.012107849, -0.03201294, -0.005393982, -0.03353882, -0.007965088, 0.013511658, -0.009147644, 0.0071792603, 0.017700195, -0.032318115, -0.024414062, -0.00077342987, -0.0124053955,...
http://arxiv.org/abs/2304.12706v1
What does BERT learn about prosody?
Sofoklis Kakouros, Johannah O'Mahony
cs.CL
Language models have become nearly ubiquitous in natural language processing applications achieving state-of-the-art results in many tasks including prosody. As the model design does not define predetermined linguistic targets during training but rather aims at learning generalized representations of the language, anal...
http://arxiv.org/pdf/2304.12706v1
2023-04-25T10:34:56
2023-04-25T10:34:56
{ "title": [ -0.009162903, -0.024475098, -0.014961243, -0.05227661, -0.044830322, 0.017562866, -0.046844482, 0.04650879, -0.0020923615, 0.038970947, 0.006641388, -0.019744873, 0.027053833, -0.030685425, 0.0007739067, 0.010597229, 0.0044136047, -0...
http://arxiv.org/abs/2304.12674v1
Compressing Sentence Representation with maximum Coding Rate Reduction
Domagoj Ševerdija, Tomislav Prusina, Antonio Jovanović, Luka Borozan, Jurica Maltar, Domagoj Matijević
cs.CL, cs.LG, 68T50, 62M45, I.2.7; I.2.0; I.5.3
In most natural language inference problems, sentence representation is needed for semantic retrieval tasks. In recent years, pre-trained large language models have been quite effective for computing such representations. These models produce high-dimensional sentence embeddings. An evident performance gap between larg...
http://arxiv.org/pdf/2304.12674v1
2023-04-25T09:23:43
2023-04-25T09:23:43
{ "title": [ -0.036010742, -0.018569946, -0.029876709, -0.04385376, -0.0018014908, 0.014442444, -0.02645874, 0.019073486, 0.017318726, 0.05404663, 0.008613586, -0.021240234, 0.015174866, -0.014457703, -0.070495605, 0.020721436, 0.0069770813, 0.02...
http://arxiv.org/abs/2304.12633v2
PUNR: Pre-training with User Behavior Modeling for News Recommendation
Guangyuan Ma, Hongtao Liu, Xing Wu, Wanhui Qian, Zhepeng Lv, Qing Yang, Songlin Hu
cs.IR, cs.CL
News recommendation aims to predict click behaviors based on user behaviors. How to effectively model the user representations is the key to recommending preferred news. Existing works are mostly focused on improvements in the supervised fine-tuning stage. However, there is still a lack of PLM-based unsupervised pre-tr...
http://arxiv.org/pdf/2304.12633v2
2023-10-30T08:13:42
2023-04-25T08:03:52
{ "title": [ 0.008628845, 0.006450653, -0.024047852, -0.038269043, 0.008399963, -0.0154953, -0.001083374, -0.027053833, -0.019897461, 0.027359009, -0.014457703, 0.038848877, -0.022766113, -0.013366699, -0.02130127, 0.002527237, -0.020736694, 0.02...
http://arxiv.org/abs/2304.12631v1
Explain like I am BM25: Interpreting a Dense Model's Ranked-List with a Sparse Approximation
Michael Llordes, Debasis Ganguly, Sumit Bhatia, Chirag Agarwal
cs.IR, cs.CL
Neural retrieval models (NRMs) have been shown to outperform their statistical counterparts owing to their ability to capture semantic meaning via dense document representations. These models, however, suffer from poor interpretability as they do not rely on explicit term matching. As a form of local per-query explanat...
http://arxiv.org/pdf/2304.12631v1
2023-04-25T07:58:38
2023-04-25T07:58:38
{ "title": [ -0.011550903, -0.01461792, 0.005142212, -0.011756897, -0.037475586, -0.005508423, -0.02619934, -0.032348633, -0.0069770813, 0.004425049, -0.038604736, 0.006980896, -0.06112671, -0.018157959, -0.016479492, -0.05493164, 0.0029850006, -...
http://arxiv.org/abs/2304.12569v1
KINLP at SemEval-2023 Task 12: Kinyarwanda Tweet Sentiment Analysis
Antoine Nzeyimana
cs.CL, I.2.7
This paper describes the system entered by the author to the SemEval-2023 Task 12: Sentiment analysis for African languages. The system focuses on the Kinyarwanda language and uses a language-specific model. Kinyarwanda morphology is modeled in a two tier transformer architecture and the transformer model is pre-traine...
http://arxiv.org/pdf/2304.12569v1
2023-04-25T04:30:03
2023-04-25T04:30:03
{ "title": [ -0.018966675, -0.005214691, -0.02293396, -0.007686615, -0.036499023, 0.005493164, -0.031829834, -0.018951416, -0.033813477, 0.08654785, -0.01889038, 0.021957397, 0.01965332, -0.014961243, -0.028686523, -0.02684021, 0.003396988, 0.034...
http://arxiv.org/abs/2304.12519v2
GlyphDiffusion: Text Generation as Image Generation
Junyi Li, Wayne Xin Zhao, Jian-Yun Nie, Ji-Rong Wen
cs.CL, cs.CV, cs.LG
Diffusion models have become a new generative paradigm for text generation. Considering the discrete categorical nature of text, in this paper, we propose GlyphDiffusion, a novel diffusion approach for text generation via text-guided image generation. Our key idea is to render the target text as a glyph image containin...
http://arxiv.org/pdf/2304.12519v2
2023-05-08T07:44:48
2023-04-25T02:14:44
{ "title": [ -0.009254456, -0.009597778, -0.016067505, -0.043426514, -0.003932953, 0.008956909, -0.03237915, -0.03387451, 0.027664185, 0.044067383, -0.015853882, -0.0019130707, 0.014190674, -0.013153076, -0.0592041, 0.01612854, 0.0000141859055, 0...
http://arxiv.org/abs/2304.12443v1
Understanding and Predicting Human Label Variation in Natural Language Inference through Explanation
Nan-Jiang Jiang, Chenhao Tan, Marie-Catherine de Marneffe
cs.CL
Human label variation (Plank 2022), or annotation disagreement, exists in many natural language processing (NLP) tasks. To be robust and trusted, NLP models need to identify such variation and be able to explain it. To this end, we created the first ecologically valid explanation dataset with diverse reasoning, LiveNLI...
http://arxiv.org/pdf/2304.12443v1
2023-04-24T20:45:09
2023-04-24T20:45:09
{ "title": [ -0.030899048, -0.0072517395, -0.03643799, -0.069885254, -0.007396698, 0.0011844635, -0.049835205, 0.00894928, -0.022659302, -0.00011098385, -0.013572693, 0.00011855364, -0.013381958, -0.025283813, -0.014923096, -0.011779785, -0.014076233...
http://arxiv.org/abs/2304.12425v1
TIGTEC : Token Importance Guided TExt Counterfactuals
Milan Bhan, Jean-Noel Vittaut, Nicolas Chesneau, Marie-Jeanne Lesot
cs.LG, cs.CL, stat.ME
Counterfactual examples explain a prediction by highlighting changes of instance that flip the outcome of a classifier. This paper proposes TIGTEC, an efficient and modular method for generating sparse, plausible and diverse counterfactual explanations for textual data. TIGTEC is a text editing heuristic that targets a...
http://arxiv.org/pdf/2304.12425v1
2023-04-24T20:11:58
2023-04-24T20:11:58
{ "title": [ 0.016799927, -0.011627197, -0.037384033, -0.021408081, -0.03338623, -0.024230957, -0.030426025, -0.02885437, -0.003112793, 0.043884277, -0.032043457, 0.010253906, 0.016021729, -0.013595581, -0.027328491, -0.012016296, -0.016677856, 0...
http://arxiv.org/abs/2304.12410v2
PEFT-Ref: A Modular Reference Architecture and Typology for Parameter-Efficient Finetuning Techniques
Mohammed Sabry, Anya Belz
cs.CL, cs.AI, cs.LG
Recent parameter-efficient finetuning (PEFT) techniques aim to improve over the considerable cost of fully finetuning large pretrained language models (PLM). As different PEFT techniques proliferate, it is becoming difficult to compare them, in particular in terms of (i) the structure and functionality they add to the ...
http://arxiv.org/pdf/2304.12410v2
2023-10-19T15:08:05
2023-04-24T19:43:11
{ "title": [ -0.009635925, -0.017745972, -0.034820557, -0.11352539, 0.008338928, 0.019546509, -0.066589355, 0.008956909, -0.028884888, 0.05505371, 0.008872986, 0.025558472, 0.056030273, -0.032165527, -0.015823364, 0.037719727, -0.016220093, 0.031...
http://arxiv.org/abs/2304.12404v1
Semantic Tokenizer for Enhanced Natural Language Processing
Sandeep Mehta, Darpan Shah, Ravindra Kulkarni, Cornelia Caragea
cs.CL
Traditionally, NLP performance improvement has been focused on improving models and increasing the number of model parameters. NLP vocabulary construction has remained focused on maximizing the number of words represented through subword regularization. We present a novel tokenizer that uses semantics to drive vocabula...
http://arxiv.org/pdf/2304.12404v1
2023-04-24T19:33:41
2023-04-24T19:33:41
{ "title": [ -0.026107788, -0.010658264, -0.059448242, -0.0748291, 0.029388428, 0.0045928955, -0.040863037, 0.00010287762, 0.0007491112, 0.03656006, -0.024276733, 0.010688782, -0.00056123734, -0.00699234, -0.04675293, 0.007419586, -0.010353088, 0...
http://arxiv.org/abs/2304.12397v1
On the Challenges of Using Black-Box APIs for Toxicity Evaluation in Research
Luiza Pozzobon, Beyza Ermis, Patrick Lewis, Sara Hooker
cs.CL, cs.AI
Perception of toxicity evolves over time and often differs between geographies and cultural backgrounds. Similarly, black-box commercially available APIs for detecting toxicity, such as the Perspective API, are not static, but frequently retrained to address any unattended weaknesses and biases. We evaluate the implica...
http://arxiv.org/pdf/2304.12397v1
2023-04-24T19:11:51
2023-04-24T19:11:51
{ "title": [ 0.0018892288, -0.0028495789, -0.025115967, -0.08325195, 0.020751953, -0.010650635, -0.04034424, -0.00062179565, -0.023345947, 0.025543213, -0.03277588, 0.022842407, -0.037078857, 0.004535675, -0.049865723, -0.014915466, -0.016677856, ...
http://arxiv.org/abs/2304.12395v2
Extreme Classification for Answer Type Prediction in Question Answering
Vinay Setty
cs.CL, cs.AI, cs.IR
Semantic answer type prediction (SMART) is known to be a useful step towards effective question answering (QA) systems. The SMART task involves predicting the top-$k$ knowledge graph (KG) types for a given natural language question. This is challenging due to the large number of types in KGs. In this paper, we propose ...
http://arxiv.org/pdf/2304.12395v2
2023-04-26T19:52:57
2023-04-24T19:08:51
{ "title": [ 0.012626648, 0.00843811, -0.03152466, -0.069885254, 0.03616333, -0.014312744, 0.003019333, 0.007980347, 0.019821167, 0.018417358, -0.009819031, 0.010429382, -0.014198303, -0.018615723, -0.045898438, 0.009918213, -0.0014543533, 0.0498...
http://arxiv.org/abs/2304.12370v1
Better Question-Answering Models on a Budget
Yudhanjaya Wijeratne, Ishan Marikar
cs.CL
Low-rank adaptation (LoRA) and question-answer datasets from large language models have made it much easier for much smaller models to be finetuned to the point where they display sophisticated conversational abilities. In this paper, we present Eluwa, a family of LoRA models that use the Stanford Alpaca dataset and ma...
http://arxiv.org/pdf/2304.12370v1
2023-04-24T18:06:27
2023-04-24T18:06:27
{ "title": [ 0.027160645, -0.03668213, -0.010604858, -0.0748291, -0.059020996, -0.0019159317, -0.04727173, 0.026367188, -0.02168274, 0.021270752, -0.05618286, -0.014450073, 0.005153656, -0.020477295, 0.012039185, -0.023010254, 0.026062012, 0.0032...
http://arxiv.org/abs/2304.12272v1
AMR Parsing with Instruction Fine-tuned Pre-trained Language Models
Young-Suk Lee, Ramón Fernandez Astudillo, Radu Florian, Tahira Naseem, Salim Roukos
cs.CL, cs.AI
Instruction fine-tuned language models on a collection of instruction annotated datasets (FLAN) have shown highly effective to improve model performance and generalization to unseen tasks. However, a majority of standard parsing tasks including abstract meaning representation (AMR), universal dependency (UD), semantic ...
http://arxiv.org/pdf/2304.12272v1
2023-04-24T17:12:17
2023-04-24T17:12:17
{ "title": [ -0.024597168, -0.011238098, -0.029632568, -0.06707764, -0.040771484, -0.0051879883, -0.09436035, 0.016830444, -0.05279541, 0.04373169, -0.014587402, 0.0051116943, 0.015655518, -0.032592773, -0.0018405914, 0.0019721985, 0.0012397766, ...
http://arxiv.org/abs/2304.12269v1
Enriching Source Code with Contextual Data for Code Completion Models: An Empirical Study
Tim van Dam, Maliheh Izadi, Arie van Deursen
cs.CL
Transformer-based pre-trained models have recently achieved great results in solving many software engineering tasks including automatic code completion which is a staple in a developer's toolkit. While many have striven to improve the code-understanding abilities of such models, the opposite -- making the code easier ...
http://arxiv.org/pdf/2304.12269v1
2023-04-24T17:09:14
2023-04-24T17:09:14
{ "title": [ -0.010864258, -0.015716553, -0.042999268, -0.09259033, 0.021026611, 0.009300232, -0.05783081, 0.015640259, 0.011199951, 0.03829956, -0.059051514, 0.029174805, -0.000028133392, 0.0073547363, -0.05078125, -0.014854431, -0.026748657, 0....
http://arxiv.org/abs/2304.12244v2
WizardLM: Empowering Large Language Models to Follow Complex Instructions
Can Xu, Qingfeng Sun, Kai Zheng, Xiubo Geng, Pu Zhao, Jiazhan Feng, Chongyang Tao, Daxin Jiang
cs.CL, cs.AI
Training large language models (LLMs) with open-domain instruction following data brings colossal success. However, manually creating such instruction data is very time-consuming and labor-intensive. Moreover, humans may struggle to produce high-complexity instructions. In this paper, we show an avenue for creating lar...
http://arxiv.org/pdf/2304.12244v2
2023-06-10T13:18:25
2023-04-24T16:31:06
{ "title": [ -0.029693604, -0.023376465, -0.051330566, -0.08355713, -0.0009455681, 0.0060272217, -0.072387695, 0.0072784424, 0.0037574768, 0.03366089, 0.0007505417, 0.0016269684, 0.020889282, -0.023849487, -0.033111572, 0.0256958, -0.008125305, 0...
http://arxiv.org/abs/2304.12206v2
PAXQA: Generating Cross-lingual Question Answering Examples at Training Scale
Bryan Li, Chris Callison-Burch
cs.CL
Existing question answering (QA) systems owe much of their success to large, high-quality training data. Such annotation efforts are costly, and the difficulty compounds in the cross-lingual setting. Therefore, prior cross-lingual QA work has focused on releasing evaluation datasets, and then applying zero-shot methods...
http://arxiv.org/pdf/2304.12206v2
2023-10-17T15:46:54
2023-04-24T15:46:26
{ "title": [ -0.018249512, -0.018371582, -0.025772095, -0.047790527, -0.0009841919, 0.010894775, -0.03857422, -0.012008667, 0.023147583, 0.018157959, -0.012908936, 0.009063721, -0.0054512024, -0.0126571655, -0.0463562, 0.005493164, -0.0131073, 0....
http://arxiv.org/abs/2304.12102v1
Unlocking Context Constraints of LLMs: Enhancing Context Efficiency of LLMs with Self-Information-Based Content Filtering
Yucheng Li
cs.CL
Large language models (LLMs) have received significant attention by achieving remarkable performance across various tasks. However, their fixed context length poses challenges when processing long documents or maintaining extended conversations. This paper proposes a method called \textit{Selective Context} that employ...
http://arxiv.org/pdf/2304.12102v1
2023-04-24T13:55:47
2023-04-24T13:55:47
{ "title": [ -0.033081055, 0.008514404, -0.054840088, -0.035461426, -0.006587982, -0.015792847, -0.01689148, -0.00289917, 0.0037384033, 0.036346436, -0.0385437, 0.030853271, 0.015319824, -0.03488159, -0.0013790131, 0.0034370422, -0.012367249, 0.0...
http://arxiv.org/abs/2304.14276v1
AI, write an essay for me: A large-scale comparison of human-written versus ChatGPT-generated essays
Steffen Herbold, Annette Hautli-Janisz, Ute Heuer, Zlata Kikteva, Alexander Trautsch
cs.CL, cs.LG
Background: Recently, ChatGPT and similar generative AI models have attracted hundreds of millions of users and become part of the public discourse. Many believe that such models will disrupt society and will result in a significant change in the education system and information generation in the future. So far, this b...
http://arxiv.org/pdf/2304.14276v1
2023-04-24T12:58:28
2023-04-24T12:58:28
{ "title": [ -0.004600525, -0.0031986237, -0.0064926147, 0.016204834, -0.012413025, -0.017944336, 0.022399902, -0.044525146, -0.01473999, 0.026672363, -0.0262146, 0.030471802, -0.0074920654, 0.0018024445, -0.020446777, -0.013290405, -0.03213501, ...
http://arxiv.org/abs/2304.12036v3
Generating Post-hoc Explanations for Skip-gram-based Node Embeddings by Identifying Important Nodes with Bridgeness
Hogun Park, Jennifer Neville
cs.LG, cs.CL, cs.SI
Node representation learning in a network is an important machine learning technique for encoding relational information in a continuous vector space while preserving the inherent properties and structures of the network. Recently, unsupervised node embedding methods such as DeepWalk, LINE, struc2vec, PTE, UserItem2vec...
http://arxiv.org/pdf/2304.12036v3
2023-05-15T09:31:29
2023-04-24T12:25:35
{ "title": [ -0.010177612, -0.012481689, -0.019180298, -0.075683594, -0.010231018, 0.018737793, -0.05001831, 0.013702393, -0.0037193298, -0.02015686, -0.023162842, -0.009750366, -0.020980835, -0.030181885, 0.0037651062, -0.023620605, -0.028381348, ...
http://arxiv.org/abs/2304.12026v1
SocialDial: A Benchmark for Socially-Aware Dialogue Systems
Haolan Zhan, Zhuang Li, Yufei Wang, Linhao Luo, Tao Feng, Xiaoxi Kang, Yuncheng Hua, Lizhen Qu, Lay-Ki Soon, Suraj Sharma, Ingrid Zukerman, Zhaleh Semnani-Azad, Gholamreza Haffari
cs.CL
Dialogue systems have been widely applied in many scenarios and are now more powerful and ubiquitous than ever before. With large neural models and massive available data, current dialogue systems have access to more knowledge than any people in their life. However, current dialogue systems still do not perform at a hu...
http://arxiv.org/pdf/2304.12026v1
2023-04-24T11:55:22
2023-04-24T11:55:22
{ "title": [ -0.016830444, -0.00027894974, -0.030548096, -0.067871094, 0.018325806, 0.0021247864, -0.04650879, -0.0021266937, 0.02029419, 0.017074585, -0.0181427, 0.010215759, -0.02281189, -0.010803223, -0.047790527, 0.0003387928, -0.015617371, 0...
http://arxiv.org/abs/2304.12008v1
CHEAT: A Large-scale Dataset for Detecting ChatGPT-writtEn AbsTracts
Peipeng Yu, Jiahan Chen, Xuan Feng, Zhihua Xia
cs.CL
The powerful ability of ChatGPT has caused widespread concern in the academic community. Malicious users could synthesize dummy academic content through ChatGPT, which is extremely harmful to academic rigor and originality. The need to develop ChatGPT-written content detection algorithms call for large-scale datasets. ...
http://arxiv.org/pdf/2304.12008v1
2023-04-24T11:19:33
2023-04-24T11:19:33
{ "title": [ 0.011146545, 0.00970459, -0.009002686, -0.01058197, -0.00068569183, -0.009887695, -0.0062828064, -0.044677734, -0.020233154, 0.025115967, -0.020446777, 0.04095459, -0.009513855, -0.00071430206, -0.029693604, 0.0022850037, -0.0309906, ...
http://arxiv.org/abs/2304.12940v2
Topological properties and organizing principles of semantic networks
Gabriel Budel, Ying Jin, Piet Van Mieghem, Maksim Kitsak
cs.CL, physics.soc-ph
Interpreting natural language is an increasingly important task in computer algorithms due to the growing availability of unstructured textual data. Natural Language Processing (NLP) applications rely on semantic networks for structured knowledge representation. The fundamental properties of semantic networks must be t...
http://arxiv.org/pdf/2304.12940v2
2023-08-17T14:51:25
2023-04-24T11:12:21
{ "title": [ -0.028045654, -0.007881165, -0.056243896, -0.06173706, 0.012664795, -0.0044555664, -0.026229858, 0.0076828003, -0.009918213, 0.033325195, -0.00699234, 0.005382538, -0.018585205, -0.016784668, -0.041046143, 0.007797241, 0.0043907166, ...
http://arxiv.org/abs/2304.11960v2
ThreatCrawl: A BERT-based Focused Crawler for the Cybersecurity Domain
Philipp Kuehn, Mike Schmidt, Markus Bayer, Christian Reuter
cs.CR, cs.CL, cs.LG
Publicly available information contains valuable information for Cyber Threat Intelligence (CTI). This can be used to prevent attacks that have already taken place on other systems. Ideally, only the initial attack succeeds and all subsequent ones are detected and stopped. But while there are different standards to exc...
http://arxiv.org/pdf/2304.11960v2
2023-04-26T13:25:45
2023-04-24T09:53:33
{ "title": [ 0.008201599, -0.019134521, -0.037322998, -0.05810547, 0.0054969788, -0.021850586, -0.0013341904, 0.024658203, -0.036315918, 0.020370483, -0.013458252, 0.019073486, -0.007083893, -0.021347046, -0.015060425, -0.020111084, -0.020996094, ...
http://arxiv.org/abs/2304.11957v4
Benchmarking ChatGPT-4 on ACR Radiation Oncology In-Training (TXIT) Exam and Red Journal Gray Zone Cases: Potentials and Challenges for AI-Assisted Medical Education and Decision Making in Radiation Oncology
Yixing Huang, Ahmed Gomaa, Sabine Semrau, Marlen Haderlein, Sebastian Lettmaier, Thomas Weissmann, Johanna Grigo, Hassen Ben Tkhayat, Benjamin Frey, Udo S. Gaipl, Luitpold V. Distel, Andreas Maier, Rainer Fietkau, Christoph Bert, Florian Putz
physics.med-ph, cs.CL
The potential of large language models in medicine for education and decision making purposes has been demonstrated as they achieve decent scores on medical exams such as the United States Medical Licensing Exam (USMLE) and the MedQA exam. In this work, we evaluate the performance of ChatGPT-4 in the specialized field ...
http://arxiv.org/pdf/2304.11957v4
2023-08-21T09:20:48
2023-04-24T09:50:39
{ "title": [ 0.00043296814, -0.00049209595, 0.03945923, 0.027999878, -0.01965332, 0.00970459, 0.031463623, -0.04156494, -0.040649414, 0.035247803, 0.015159607, 0.042785645, -0.040374756, -0.02330017, 0.00027823448, -0.007888794, -0.010620117, 0.0...
http://arxiv.org/abs/2304.11943v1
Constructing Tree-based Index for Efficient and Effective Dense Retrieval
Haitao Li, Qingyao Ai, Jingtao Zhan, Jiaxin Mao, Yiqun Liu, Zheng Liu, Zhao Cao
cs.IR, cs.CL
Recent studies have shown that Dense Retrieval (DR) techniques can significantly improve the performance of first-stage retrieval in IR systems. Despite its empirical effectiveness, the application of DR is still limited. In contrast to statistic retrieval models that rely on highly efficient inverted index solutions, ...
http://arxiv.org/pdf/2304.11943v1
2023-04-24T09:25:39
2023-04-24T09:25:39
{ "title": [ -0.011924744, -0.008598328, -0.025314331, -0.05923462, 0.012718201, 0.010749817, -0.0065994263, -0.004611969, 0.0368042, 0.041229248, -0.04196167, -0.0025234222, 0.004180908, -0.0037021637, -0.074523926, 0.00035119057, 0.0053215027, ...
http://arxiv.org/abs/2304.12331v1
USTEP: Structuration des logs en flux gr{â}ce {à} un arbre de recherche {é}volutif
Arthur Vervaet, Raja Chiky, Mar Callau-Zori
cs.CL, cs.IR
Logs record valuable system information at runtime. They are widely used by data-driven approaches for development and monitoring purposes. Parsing log messages to structure their format is a classic preliminary step for log-mining tasks. As they appear upstream, parsing operations can become a processing time bottlene...
http://arxiv.org/pdf/2304.12331v1
2023-04-24T09:12:00
2023-04-24T09:12:00
{ "title": [ 0.0055007935, 0.007293701, -0.046081543, -0.03945923, -0.012817383, -0.003643036, -0.06530762, -0.020004272, -0.025939941, 0.0047416687, 0.0362854, -0.021530151, 0.012321472, -0.08355713, 0.022064209, -0.0039482117, 0.017318726, -0.0...
http://arxiv.org/abs/2304.11924v1
KInITVeraAI at SemEval-2023 Task 3: Simple yet Powerful Multilingual Fine-Tuning for Persuasion Techniques Detection
Timo Hromadka, Timotej Smolen, Tomas Remis, Branislav Pecher, Ivan Srba
cs.CL, cs.LG
This paper presents the best-performing solution to the SemEval 2023 Task 3 on the subtask 3 dedicated to persuasion techniques detection. Due to a high multilingual character of the input data and a large number of 23 predicted labels (causing a lack of labelled data for some language-label combinations), we opted for...
http://arxiv.org/pdf/2304.11924v1
2023-04-24T09:06:43
2023-04-24T09:06:43
{ "title": [ 0.022277832, 0.0019741058, -0.0075912476, -0.01928711, 0.0030097961, -0.017242432, -0.023712158, -0.047729492, -0.0395813, 0.03375244, -0.00844574, 0.033721924, -0.01977539, -0.014022827, -0.0058288574, 0.000620842, -0.024551392, 0.0...
http://arxiv.org/abs/2304.11922v1
Generating Topic Pages for Scientific Concepts Using Scientific Publications
Hosein Azarbonyad, Zubair Afzal, George Tsatsaronis
cs.CL, cs.DL, cs.IR
In this paper, we describe Topic Pages, an inventory of scientific concepts and information around them extracted from a large collection of scientific books and journals. The main aim of Topic Pages is to provide all the necessary information to the readers to understand scientific concepts they come across while read...
http://arxiv.org/pdf/2304.11922v1
2023-04-24T09:03:14
2023-04-24T09:03:14
{ "title": [ -0.017425537, 0.0015907288, -0.038208008, -0.026428223, -0.011154175, -0.018508911, -0.02267456, -0.03152466, -0.0022907257, 0.030426025, 0.01802063, 0.030517578, -0.014350891, -0.021331787, -0.02319336, 0.022155762, -0.0058250427, 0...
http://arxiv.org/abs/2304.12329v1
Pre-trained Embeddings for Entity Resolution: An Experimental Analysis [Experiment, Analysis & Benchmark]
Alexandros Zeakis, George Papadakis, Dimitrios Skoutas, Manolis Koubarakis
cs.DB, cs.CL
Many recent works on Entity Resolution (ER) leverage Deep Learning techniques involving language models to improve effectiveness. This is applied to both main steps of ER, i.e., blocking and matching. Several pre-trained embeddings have been tested, with the most popular ones being fastText and variants of the BERT mod...
http://arxiv.org/pdf/2304.12329v1
2023-04-24T08:53:54
2023-04-24T08:53:54
{ "title": [ 0.018188477, 0.0065612793, -0.047027588, -0.101135254, 0.046203613, -0.015090942, -0.036376953, 0.0047035217, -0.0073509216, 0.03640747, 0.001241684, 0.014045715, 0.032592773, -0.01663208, -0.017745972, 0.017959595, -0.006641388, 0.0...
http://arxiv.org/abs/2304.11913v2
Development of a Trust-Aware User Simulator for Statistical Proactive Dialog Modeling in Human-AI Teams
Matthias Kraus, Ron Riekenbrauck, Wolfgang Minker
cs.AI, cs.CL
The concept of a Human-AI team has gained increasing attention in recent years. For effective collaboration between humans and AI teammates, proactivity is crucial for close coordination and effective communication. However, the design of adequate proactivity for AI-based systems to support humans is still an open ques...
http://arxiv.org/pdf/2304.11913v2
2023-06-18T18:56:01
2023-04-24T08:42:51
{ "title": [ -0.019683838, 0.012298584, -0.076538086, -0.029922485, 0.0009074211, -0.0357666, -0.020568848, -0.0001809597, -0.0031814575, 0.042022705, 0.009239197, 0.00806427, 0.031951904, -0.045043945, -0.028305054, 0.010910034, -0.014503479, -0...
http://arxiv.org/abs/2304.12998v1
ChatLLM Network: More brains, More intelligence
Rui Hao, Linmei Hu, Weijian Qi, Qingliu Wu, Yirui Zhang, Liqiang Nie
cs.AI, cs.CL
Dialogue-based language models mark a huge milestone in the field of artificial intelligence, by their impressive ability to interact with users, as well as a series of challenging tasks prompted by customized instructions. However, the prevalent large-scale dialogue-based language models like ChatGPT still have room f...
http://arxiv.org/pdf/2304.12998v1
2023-04-24T08:29:14
2023-04-24T08:29:14
{ "title": [ 0.0013046265, -0.014892578, -0.033935547, -0.025680542, -0.03753662, -0.028869629, -0.048706055, -0.028030396, -0.026367188, 0.036987305, 0.019561768, -0.00566864, -0.0181427, -0.04425049, -0.032287598, -0.0071640015, -0.0025634766, ...
http://arxiv.org/abs/2304.13731v2
Text-to-Audio Generation using Instruction-Tuned LLM and Latent Diffusion Model
Deepanway Ghosal, Navonil Majumder, Ambuj Mehrish, Soujanya Poria
eess.AS, cs.AI, cs.CL, cs.SD
The immense scale of the recent large language models (LLM) allows many interesting properties, such as, instruction- and chain-of-thought-based fine-tuning, that has significantly improved zero- and few-shot performance in many natural language processing (NLP) tasks. Inspired by such successes, we adopt such an instr...
http://arxiv.org/pdf/2304.13731v2
2023-05-29T12:09:08
2023-04-24T07:45:28
{ "title": [ -0.054534912, -0.015914917, -0.04663086, -0.02255249, -0.032714844, -0.0046157837, -0.04425049, 0.01473999, -0.022537231, 0.05340576, -0.019744873, 0.031463623, 0.009498596, -0.01725769, -0.022964478, -0.0022468567, -0.027908325, 0.0...
http://arxiv.org/abs/2304.11872v1
Generation-driven Contrastive Self-training for Zero-shot Text Classification with Instruction-tuned GPT
Ruohong Zhang, Yau-Shian Wang, Yiming Yang
cs.CL, cs.AI
Moreover, GPT-based zero-shot classification models tend to make independent predictions over test instances, which can be sub-optimal as the instance correlations and the decision boundaries in the target space are ignored. To address these difficulties and limitations, we propose a new approach to zero-shot text clas...
http://arxiv.org/pdf/2304.11872v1
2023-04-24T07:35:38
2023-04-24T07:35:38
{ "title": [ 0.0060424805, -0.000022172928, -0.006877899, -0.035308838, -0.022262573, -0.015029907, -0.0056762695, -0.008140564, 0.0121536255, 0.039093018, -0.016296387, 0.024337769, -0.02848816, -0.012123108, -0.05279541, -0.0020122528, -0.009628296...
http://arxiv.org/abs/2304.11791v1
Directed Acyclic Transformer Pre-training for High-quality Non-autoregressive Text Generation
Fei Huang, Pei Ke, Minlie Huang
cs.CL
Non-AutoRegressive (NAR) text generation models have drawn much attention because of their significantly faster decoding speed and good generation quality in machine translation. However, in a wider range of text generation tasks, existing NAR models lack proper pre-training, making them still far behind the pre-traine...
http://arxiv.org/pdf/2304.11791v1
2023-04-24T02:30:33
2023-04-24T02:30:33
{ "title": [ -0.0067443848, 0.0046844482, -0.06323242, -0.06915283, 0.0004518032, -0.014053345, -0.06793213, 0.013076782, -0.012702942, 0.046936035, -0.014152527, 0.022018433, 0.021728516, -0.01890564, -0.031402588, -0.00091457367, -0.00018393993, ...
http://arxiv.org/abs/2304.11766v3
NAIST-SIC-Aligned: Automatically-Aligned English-Japanese Simultaneous Interpretation Corpus
Jinming Zhao, Yuka Ko, Kosuke Doi, Ryo Fukuda, Katsuhito Sudoh, Satoshi Nakamura
cs.CL
It remains a question that how simultaneous interpretation (SI) data affects simultaneous machine translation (SiMT). Research has been limited due to the lack of a large-scale training corpus. In this work, we aim to fill in the gap by introducing NAIST-SIC-Aligned, which is an automatically-aligned parallel English-J...
http://arxiv.org/pdf/2304.11766v3
2023-06-06T06:02:42
2023-04-23T23:03:58
{ "title": [ -0.0395813, -0.010635376, -0.040130615, -0.052124023, -0.021621704, 0.018981934, -0.074279785, -0.006942749, 0.0056266785, 0.04510498, -0.04083252, 0.001912117, 0.04272461, -0.017791748, -0.052368164, 0.009269714, -0.008911133, 0.026...
http://arxiv.org/abs/2304.11721v1
A Lightweight Constrained Generation Alternative for Query-focused Summarization
Zhichao Xu, Daniel Cohen
cs.IR, cs.CL
Query-focused summarization (QFS) aims to provide a summary of a document that satisfies information need of a given query and is useful in various IR applications, such as abstractive snippet generation. Current QFS approaches typically involve injecting additional information, e.g. query-answer relevance or fine-grai...
http://arxiv.org/pdf/2304.11721v1
2023-04-23T18:43:48
2023-04-23T18:43:48
{ "title": [ -0.027832031, -0.012527466, -0.052368164, -0.103149414, -0.004173279, 0.016036987, -0.075927734, 0.031585693, 0.009231567, 0.052703857, 0.0016021729, 0.008125305, -0.0017023087, -0.009979248, -0.057678223, 0.0055007935, -0.0016460419, ...
http://arxiv.org/abs/2304.11693v1
Studying the Impact of Semi-Cooperative Drivers on Overall Highway Flow
Noam Buckman, Sertac Karaman, Daniela Rus
cs.CL, cs.AI, cs.MA
Semi-cooperative behaviors are intrinsic properties of human drivers and should be considered for autonomous driving. In addition, new autonomous planners can consider the social value orientation (SVO) of human drivers to generate socially-compliant trajectories. Yet the overall impact on traffic flow for this new cla...
http://arxiv.org/pdf/2304.11693v1
2023-04-23T16:01:36
2023-04-23T16:01:36
{ "title": [ 0.025497437, 0.026428223, -0.04812622, -0.022155762, -0.005130768, -0.06665039, 0.03338623, -0.0013790131, -0.044311523, 0.060455322, -0.02418518, 0.015686035, -0.012756348, -0.012954712, -0.02104187, -0.025741577, 0.010894775, -0.01...
http://arxiv.org/abs/2304.11679v2
Domain Mastery Benchmark: An Ever-Updating Benchmark for Evaluating Holistic Domain Knowledge of Large Language Model--A Preliminary Release
Zhouhong Gu, Xiaoxuan Zhu, Haoning Ye, Lin Zhang, Zhuozhi Xiong, Zihan Li, Qianyu He, Sihang Jiang, Hongwei Feng, Yanghua Xiao
cs.CL
Domain knowledge refers to the in-depth understanding, expertise, and familiarity with a specific subject, industry, field, or area of special interest. The existing benchmarks are all lack of an overall design for domain knowledge evaluation. Holding the belief that the real ability of domain language understanding ca...
http://arxiv.org/pdf/2304.11679v2
2023-08-10T05:27:58
2023-04-23T15:11:49
{ "title": [ -0.019744873, -0.005252838, -0.052093506, -0.06933594, -0.0127334595, -0.021484375, -0.061828613, 0.0059928894, 0.024230957, 0.027572632, 0.015655518, 0.0154418945, 0.009460449, -0.0390625, -0.028442383, 0.016586304, -0.018600464, 0....
http://arxiv.org/abs/2304.11666v1
Hold the Suspect! : An Analysis on Media Framing of Itaewon Halloween Crowd Crush
TaeYoung Kang
cs.CL, cs.CY, stat.AP
Based on the 10.9K articles from top 40 news providers of South Korea, this paper analyzed the media framing of Itaewon Halloween Crowd Crush during the first 72 hours after the incident. By adopting word-vector embedding and clustering, we figured out that conservative media focused on political parties' responses and...
http://arxiv.org/pdf/2304.11666v1
2023-04-23T14:22:25
2023-04-23T14:22:25
{ "title": [ 0.038024902, 0.01184845, -0.010475159, -0.008087158, 0.00055360794, -0.029556274, 0.016204834, -0.03036499, -0.055236816, 0.009056091, -0.025177002, 0.05480957, -0.02130127, -0.037750244, 0.0178833, -0.030181885, -0.0016374588, -0.02...
http://arxiv.org/abs/2304.11664v1
IslamicPCQA: A Dataset for Persian Multi-hop Complex Question Answering in Islamic Text Resources
Arash Ghafouri, Hasan Naderi, Mohammad Aghajani asl, Mahdi Firouzmandi
cs.CL, cs.AI
Nowadays, one of the main challenges for Question Answering Systems is to answer complex questions using various sources of information. Multi-hop questions are a type of complex questions that require multi-step reasoning to answer. In this article, the IslamicPCQA dataset is introduced. This is the first Persian data...
http://arxiv.org/pdf/2304.11664v1
2023-04-23T14:20:58
2023-04-23T14:20:58
{ "title": [ -0.002292633, -0.006088257, -0.025848389, -0.055267334, 0.0109939575, -0.010055542, -0.005882263, 0.019866943, 0.024932861, 0.025726318, -0.031555176, 0.010063171, -0.006629944, -0.020996094, -0.06304932, -0.0006489754, -0.0005264282, ...
http://arxiv.org/abs/2304.11657v2
Enhancing Chain-of-Thoughts Prompting with Iterative Bootstrapping in Large Language Models
Jiashuo Sun, Yi Luo, Yeyun Gong, Chen Lin, Yelong Shen, Jian Guo, Nan Duan
cs.CL, cs.AI
Large language models (LLMs) can achieve highly effective performance on various reasoning tasks by incorporating step-by-step chain-of-thought (CoT) prompting as demonstrations. However, the reasoning chains of demonstrations generated by LLMs are prone to errors, which can subsequently lead to incorrect reasoning dur...
http://arxiv.org/pdf/2304.11657v2
2023-10-12T13:57:58
2023-04-23T13:54:39
{ "title": [ -0.012107849, -0.004096985, -0.028442383, -0.04940796, -0.024963379, -0.005039215, -0.068481445, -0.006378174, 0.02003479, 0.0017232895, -0.037597656, 0.0032157898, 0.022994995, -0.02949524, -0.02355957, 0.01058197, -0.0045280457, 0....
http://arxiv.org/abs/2304.11633v1
Evaluating ChatGPT's Information Extraction Capabilities: An Assessment of Performance, Explainability, Calibration, and Faithfulness
Bo Li, Gexiang Fang, Yang Yang, Quansen Wang, Wei Ye, Wen Zhao, Shikun Zhang
cs.CL
The capability of Large Language Models (LLMs) like ChatGPT to comprehend user intent and provide reasonable responses has made them extremely popular lately. In this paper, we focus on assessing the overall ability of ChatGPT using 7 fine-grained information extraction (IE) tasks. Specially, we present the systematica...
http://arxiv.org/pdf/2304.11633v1
2023-04-23T12:33:18
2023-04-23T12:33:18
{ "title": [ -0.021240234, -0.0058670044, -0.015052795, -0.02508545, -0.0053520203, 0.0181427, -0.021255493, -0.030166626, -0.0158844, 0.016052246, -0.022399902, 0.041870117, -0.032073975, 0.0020198822, -0.02458191, 0.006855011, -0.029190063, 0.0...
http://arxiv.org/abs/2304.11618v1
Modality-Aware Negative Sampling for Multi-modal Knowledge Graph Embedding
Yichi Zhang, Mingyang Chen, Wen Zhang
cs.CL, cs.AI
Negative sampling (NS) is widely used in knowledge graph embedding (KGE), which aims to generate negative triples to make a positive-negative contrast during training. However, existing NS methods are unsuitable when multi-modal information is considered in KGE models. They are also inefficient due to their complex des...
http://arxiv.org/pdf/2304.11618v1
2023-04-23T11:22:17
2023-04-23T11:22:17
{ "title": [ -0.0069847107, -0.0028476715, -0.05633545, -0.05307007, -0.010032654, -0.022583008, -0.053985596, -0.004508972, 0.03189087, 0.028701782, -0.02671814, 0.040252686, 0.028656006, -0.019302368, -0.022628784, 0.017669678, -0.00022053719, ...
http://arxiv.org/abs/2304.11567v1
Differentiate ChatGPT-generated and Human-written Medical Texts
Wenxiong Liao, Zhengliang Liu, Haixing Dai, Shaochen Xu, Zihao Wu, Yiyang Zhang, Xiaoke Huang, Dajiang Zhu, Hongmin Cai, Tianming Liu, Xiang Li
cs.CL, cs.AI
Background: Large language models such as ChatGPT are capable of generating grammatically perfect and human-like text content, and a large number of ChatGPT-generated texts have appeared on the Internet. However, medical texts such as clinical notes and diagnoses require rigorous validation, and erroneous medical conte...
http://arxiv.org/pdf/2304.11567v1
2023-04-23T07:38:07
2023-04-23T07:38:07
{ "title": [ 0.0046081543, 0.004096985, -0.00699234, -0.0035152435, -0.06311035, 0.003314972, -0.01612854, -0.034606934, -0.042236328, 0.04309082, -0.0006942749, 0.029525757, 0.00067186356, -0.039276123, 0.01889038, 0.003686905, -0.0029144287, 0....
http://arxiv.org/abs/2304.11556v1
Divide and Prompt: Chain of Thought Prompting for Text-to-SQL
Xiping Liu, Zhao Tan
cs.CL, cs.AI
Chain-of-thought (CoT) prompting combined with large language models (LLMs) have achieved encouraging results on complex reasoning tasks. Text-to-SQL is a critical semantic parsing task that converts natural language questions into SQL statements, involving a complex reasoning process. However, there is little work abo...
http://arxiv.org/pdf/2304.11556v1
2023-04-23T06:52:35
2023-04-23T06:52:35
{ "title": [ 0.0082092285, -0.020568848, -0.0390625, -0.06304932, -0.0018863678, -0.011619568, -0.01625061, 0.013832092, 0.031585693, 0.029815674, -0.07354736, 0.01109314, 0.02281189, -0.008842468, -0.04864502, -0.020309448, 0.0059814453, -0.0008...
http://arxiv.org/abs/2304.11534v2
Graph Neural Networks for Text Classification: A Survey
Kunze Wang, Yihao Ding, Soyeon Caren Han
cs.CL
Text Classification is the most essential and fundamental problem in Natural Language Processing. While numerous recent text classification models applied the sequential deep learning technique, graph neural network-based models can directly deal with complex structured text data and exploit global information. Many re...
http://arxiv.org/pdf/2304.11534v2
2023-04-27T13:42:25
2023-04-23T04:21:50
{ "title": [ -0.037902832, 0.0019102097, -0.029220581, -0.023422241, -0.013504028, 0.0051231384, -0.0385437, -0.0071105957, -0.0058174133, 0.028945923, 0.000197649, 0.0149383545, -0.021240234, -0.015777588, -0.043273926, -0.00076150894, -0.020645142,...
http://arxiv.org/abs/2304.11520v3
Processing Natural Language on Embedded Devices: How Well Do Modern Models Perform?
Souvika Sarkar, Mohammad Fakhruddin Babar, Md Mahadi Hassan, Monowar Hasan, Shubhra Kanti Karmaker Santu
cs.CL, cs.SY, eess.SY
Voice-controlled systems are becoming ubiquitous in many IoT-specific applications such as home/industrial automation, automotive infotainment, and healthcare. While cloud-based voice services (\eg Alexa, Siri) can leverage high-performance computing servers, some use cases (\eg robotics, automotive infotainment) may r...
http://arxiv.org/pdf/2304.11520v3
2023-09-12T21:15:12
2023-04-23T03:01:39
{ "title": [ -0.04949951, -0.014389038, -0.013618469, -0.019470215, -0.011405945, 0.0004067421, -0.03387451, -0.0062294006, -0.008323669, 0.026687622, -0.03665161, 0.000055015087, 0.0031642914, -0.006046295, -0.047698975, -0.002117157, -0.013046265, ...
http://arxiv.org/abs/2304.11501v1
Translationese Reduction using Abstract Meaning Representation
Shira Wein, Nathan Schneider
cs.CL
Translated texts or utterances bear several hallmarks distinct from texts originating in the language. This phenomenon, known as translationese, is well-documented, and when found in training or test sets can affect model performance. Still, work to mitigate the effect of translationese in human translated text is unde...
http://arxiv.org/pdf/2304.11501v1
2023-04-23T00:04:14
2023-04-23T00:04:14
{ "title": [ -0.024276733, -0.0051498413, -0.016357422, -0.044677734, -0.008094788, 0.006828308, -0.052703857, -0.008255005, 0.019012451, 0.06561279, 0.0141067505, -0.0027599335, 0.025741577, -0.02519226, -0.05911255, 0.03314209, 0.0025520325, 0....
http://arxiv.org/abs/2304.11490v3
Boosting Theory-of-Mind Performance in Large Language Models via Prompting
Shima Rahimi Moghaddam, Christopher J. Honey
cs.AI, cs.CL
Large language models (LLMs) excel in many tasks in 2023, but they still face challenges in complex reasoning. Theory-of-mind (ToM) tasks, which require understanding agents' beliefs, goals, and mental states, are essential for common-sense reasoning involving humans, making it crucial to enhance LLM performance in thi...
http://arxiv.org/pdf/2304.11490v3
2023-04-26T04:02:04
2023-04-22T22:50:50
{ "title": [ -0.010574341, -0.026535034, 0.0012254715, -0.048675537, -0.024887085, 0.004676819, -0.044708252, -0.0061149597, -0.0023651123, 0.010231018, -0.026565552, 0.0076026917, 0.02116394, -0.028427124, -0.035949707, -0.00068187714, -0.010139465,...
http://arxiv.org/abs/2304.11485v1
Understanding Lexical Biases when Identifying Gang-related Social Media Communications
Dhiraj Murthy, Constantine Caramanis, Koustav Rudra
cs.CL, cs.AI, cs.LG, cs.SI, J.4; K.4.2; I.2.7
Individuals involved in gang-related activity use mainstream social media including Facebook and Twitter to express taunts and threats as well as grief and memorializing. However, identifying the impact of gang-related activity in order to serve community member needs through social media sources has a unique set of ch...
http://arxiv.org/pdf/2304.11485v1
2023-04-22T21:51:49
2023-04-22T21:51:49
{ "title": [ -0.024597168, -0.0024700165, -0.035491943, -0.00705719, -0.0035057068, -0.014595032, -0.0023822784, -0.009819031, -0.03704834, 0.03201294, -0.039489746, 0.034820557, -0.0047073364, 0.0011940002, -0.013694763, -0.02017212, -0.024749756, ...
http://arxiv.org/abs/2304.11473v2
(Vector) Space is Not the Final Frontier: Product Search as Program Synthesis
Jacopo Tagliabue, Ciro Greco
cs.IR, cs.CL, cs.LG
As ecommerce continues growing, huge investments in ML and NLP for Information Retrieval are following. While the vector space model dominated retrieval modelling in product search - even as vectorization itself greatly changed with the advent of deep learning -, our position paper argues in a contrarian fashion that p...
http://arxiv.org/pdf/2304.11473v2
2023-06-11T17:35:02
2023-04-22T20:00:06
{ "title": [ -0.020751953, -0.014015198, -0.0006723404, -0.025894165, -0.0069618225, -0.014457703, -0.006187439, -0.034301758, -0.012680054, 0.012779236, -0.03793335, 0.004585266, -0.05697632, -0.0060424805, -0.056610107, -0.036712646, 0.0015382767, ...
http://arxiv.org/abs/2304.11461v1
Recurrent Neural Networks and Long Short-Term Memory Networks: Tutorial and Survey
Benyamin Ghojogh, Ali Ghodsi
cs.LG, cs.CL, cs.NE, cs.SD, eess.AS
This is a tutorial paper on Recurrent Neural Network (RNN), Long Short-Term Memory Network (LSTM), and their variants. We start with a dynamical system and backpropagation through time for RNN. Then, we discuss the problems of gradient vanishing and explosion in long-term dependencies. We explain close-to-identity weig...
http://arxiv.org/pdf/2304.11461v1
2023-04-22T18:22:10
2023-04-22T18:22:10
{ "title": [ -0.015716553, -0.011489868, -0.036193848, -0.05557251, 0.03503418, 0.008262634, -0.011108398, -0.011558533, 0.0022068024, 0.03616333, 0.002067566, 0.0012149811, -0.005016327, -0.025283813, -0.04562378, 0.015991211, -0.006023407, 0.02...
http://arxiv.org/abs/2304.11434v1
L3Cube-IndicSBERT: A simple approach for learning cross-lingual sentence representations using multilingual BERT
Samruddhi Deode, Janhavi Gadre, Aditi Kajale, Ananya Joshi, Raviraj Joshi
cs.CL, cs.LG
The multilingual Sentence-BERT (SBERT) models map different languages to common representation space and are useful for cross-language similarity and mining tasks. We propose a simple yet effective approach to convert vanilla multilingual BERT models into multilingual sentence BERT models using synthetic corpus. We sim...
http://arxiv.org/pdf/2304.11434v1
2023-04-22T15:45:40
2023-04-22T15:45:40
{ "title": [ -0.010559082, -0.0158844, -0.016998291, -0.047332764, -0.01423645, 0.013679504, -0.076171875, -0.012763977, -0.011489868, 0.05847168, -0.023895264, 0.022964478, 0.034851074, 0.0017614365, -0.031829834, 0.016311646, -0.023757935, 0.06...
http://arxiv.org/abs/2304.11410v1
A bounded rationality account of dependency length minimization in Hindi
Sidharth Ranjan, Titus von der Malsburg
cs.CL, econ.TH
The principle of DEPENDENCY LENGTH MINIMIZATION, which seeks to keep syntactically related words close in a sentence, is thought to universally shape the structure of human languages for effective communication. However, the extent to which dependency length minimization is applied in human language systems is not yet ...
http://arxiv.org/pdf/2304.11410v1
2023-04-22T13:53:50
2023-04-22T13:53:50
{ "title": [ -0.047180176, -0.013008118, -0.058044434, -0.077819824, -0.029006958, 0.011367798, -0.08496094, 0.042114258, -0.050567627, 0.008758545, 0.007259369, 0.017196655, -0.047332764, -0.029678345, -0.020904541, -0.019119263, -0.017440796, 0...
http://arxiv.org/abs/2304.11406v2
LaMP: When Large Language Models Meet Personalization
Alireza Salemi, Sheshera Mysore, Michael Bendersky, Hamed Zamani
cs.CL
This paper highlights the importance of personalization in the current state of natural language understanding and generation and introduces the LaMP benchmark -- a novel benchmark for training and evaluating language models for producing personalized outputs. LaMP offers a comprehensive evaluation framework with diver...
http://arxiv.org/pdf/2304.11406v2
2023-05-19T20:34:59
2023-04-22T13:42:04
{ "title": [ -0.019012451, -0.010154724, -0.036102295, -0.070495605, 0.0012998581, 0.009819031, -0.040863037, 0.0046424866, -0.016479492, 0.03451538, -0.001531601, 0.016189575, 0.008735657, -0.016479492, -0.022323608, 0.018310547, -0.016601562, 0...
http://arxiv.org/abs/2304.11389v2
Transformer-Based Language Model Surprisal Predicts Human Reading Times Best with About Two Billion Training Tokens
Byung-Doh Oh, William Schuler
cs.CL
Recent psycholinguistic studies have drawn conflicting conclusions about the relationship between the quality of a language model and the ability of its surprisal estimates to predict human reading times, which has been speculated to be due to the large gap in both the amount of training data and model capacity across ...
http://arxiv.org/pdf/2304.11389v2
2023-10-22T20:03:54
2023-04-22T12:50:49
{ "title": [ -0.012191772, -0.002439499, -0.032989502, -0.0037441254, -0.04663086, -0.02999878, -0.053466797, -0.022964478, -0.020614624, 0.0149383545, -0.048431396, 0.00037050247, 0.0021247864, -0.03781128, 0.031143188, -0.03878784, 0.00033330917, ...
http://arxiv.org/abs/2304.11370v1
SAILER: Structure-aware Pre-trained Language Model for Legal Case Retrieval
Haitao Li, Qingyao Ai, Jia Chen, Qian Dong, Yueyue Wu, Yiqun Liu, Chong Chen, Qi Tian
cs.IR, cs.CL
Legal case retrieval, which aims to find relevant cases for a query case, plays a core role in the intelligent legal system. Despite the success that pre-training has achieved in ad-hoc retrieval tasks, effective pre-training strategies for legal case retrieval remain to be explored. Compared with general documents, le...
http://arxiv.org/pdf/2304.11370v1
2023-04-22T10:47:01
2023-04-22T10:47:01
{ "title": [ -0.006549835, -0.020355225, -0.034118652, -0.06341553, -0.0099105835, 0.01058197, -0.048553467, 0.0102005005, 0.016586304, 0.036193848, -0.04043579, -0.013771057, 0.046691895, 0.0020427704, -0.054840088, 0.0024642944, -0.016082764, 0...
http://arxiv.org/abs/2304.13557v1
"I'm" Lost in Translation: Pronoun Missteps in Crowdsourced Data Sets
Katie Seaborn, Yeongdae Kim
cs.CL, cs.CY, cs.HC, cs.LG
As virtual assistants continue to be taken up globally, there is an ever-greater need for these speech-based systems to communicate naturally in a variety of languages. Crowdsourcing initiatives have focused on multilingual translation of big, open data sets for use in natural language processing (NLP). Yet, language t...
http://arxiv.org/pdf/2304.13557v1
2023-04-22T09:27:32
2023-04-22T09:27:32
{ "title": [ -0.02722168, 0.0021705627, -0.031097412, -0.018051147, -0.0065612793, -0.01058197, -0.02017212, -0.029144287, -0.0025577545, 0.029281616, -0.028945923, 0.01826477, 0.0017881393, 0.0025539398, -0.055847168, -0.0067977905, -0.01386261, ...
http://arxiv.org/abs/2304.11350v2
Romanian Multiword Expression Detection Using Multilingual Adversarial Training and Lateral Inhibition
Andrei-Marius Avram, Verginica Barbu Mititelu, Dumitru-Clementin Cercel
cs.CL
Multiword expressions are a key ingredient for developing large-scale and linguistically sound natural language processing technology. This paper describes our improvements in automatically identifying Romanian multiword expressions on the corpus released for the PARSEME v1.2 shared task. Our approach assumes a multili...
http://arxiv.org/pdf/2304.11350v2
2023-05-08T16:54:03
2023-04-22T09:10:49
{ "title": [ -0.0018663406, 0.005329132, -0.016983032, 0.0041122437, -0.006214142, -0.003856659, 0.0027542114, -0.02571106, -0.03866577, 0.045562744, 0.02130127, -0.0009622574, 0.020904541, -0.036895752, -0.029724121, -0.009986877, -0.013595581, ...
http://arxiv.org/abs/2304.11340v1
Semantic Specialization for Knowledge-based Word Sense Disambiguation
Sakae Mizuki, Naoaki Okazaki
cs.CL
A promising approach for knowledge-based Word Sense Disambiguation (WSD) is to select the sense whose contextualized embeddings computed for its definition sentence are closest to those computed for a target word in a given sentence. This approach relies on the similarity of the \textit{sense} and \textit{context} embe...
http://arxiv.org/pdf/2304.11340v1
2023-04-22T07:40:23
2023-04-22T07:40:23
{ "title": [ -0.03390503, -0.016296387, -0.031280518, -0.061950684, -0.007762909, 0.0040359497, -0.040649414, 0.01361084, 0.0073394775, 0.057128906, -0.029815674, 0.025756836, 0.012252808, 0.0048179626, -0.054626465, 0.0070152283, -0.00957489, 0....
http://arxiv.org/abs/2304.11164v1
Dialectical language model evaluation: An initial appraisal of the commonsense spatial reasoning abilities of LLMs
Anthony G Cohn, Jose Hernandez-Orallo
cs.CL, cs.AI
Language models have become very popular recently and many claims have been made about their abilities, including for commonsense reasoning. Given the increasingly better results of current language models on previous static benchmarks for commonsense reasoning, we explore an alternative dialectical evaluation. The goa...
http://arxiv.org/pdf/2304.11164v1
2023-04-22T06:28:46
2023-04-22T06:28:46
{ "title": [ -0.009010315, -0.026107788, -0.006122589, 0.018569946, -0.05407715, -0.0077056885, -0.010154724, -0.038970947, 0.02432251, 0.0368042, -0.018218994, -0.02268982, 0.037628174, -0.034423828, -0.037353516, -0.0076293945, -0.009971619, 0....
http://arxiv.org/abs/2304.12810v1
Transcending the "Male Code": Implicit Masculine Biases in NLP Contexts
Katie Seaborn, Shruti Chandra, Thibault Fabre
cs.CL, cs.CY, cs.HC, cs.LG
Critical scholarship has elevated the problem of gender bias in data sets used to train virtual assistants (VAs). Most work has focused on explicit biases in language, especially against women, girls, femme-identifying people, and genderqueer folk; implicit associations through word embeddings; and limited models of ge...
http://arxiv.org/pdf/2304.12810v1
2023-04-22T03:53:24
2023-04-22T03:53:24
{ "title": [ -0.011016846, 0.003791809, -0.03277588, -0.031799316, -0.0015506744, -0.023849487, -0.013702393, -0.010131836, 0.0029144287, 0.041107178, -0.03765869, -0.0026626587, 0.03062439, -0.00006484985, -0.042022705, -0.004447937, -0.04574585, ...
http://arxiv.org/abs/2304.11292v1
On the Identification of the Energy related Issues from the App Reviews
Noshin Nawal
cs.AI, cs.CL, cs.LG
The energy inefficiency of the apps can be a major issue for the app users which is discussed on App Stores extensively. Previous research has shown the importance of investigating the energy related app reviews to identify the major causes or categories of energy related user feedback. However, there is no study that ...
http://arxiv.org/pdf/2304.11292v1
2023-04-22T01:54:30
2023-04-22T01:54:30
{ "title": [ 0.049194336, 0.0070266724, 0.03086853, 0.03164673, -0.010955811, -0.013450623, 0.05545044, -0.04196167, -0.045806885, 0.045074463, 0.025131226, -0.016983032, 0.01927185, -0.076538086, 0.0127334595, -0.009819031, -0.011360168, 0.04367...
http://arxiv.org/abs/2304.12809v1
Can Voice Assistants Sound Cute? Towards a Model of Kawaii Vocalics
Katie Seaborn, Somang Nam, Julia Keckeis, Tatsuya Itagaki
cs.HC, cs.AI, cs.CL, cs.CY, cs.SD
The Japanese notion of "kawaii" or expressions of cuteness, vulnerability, and/or charm is a global cultural export. Work has explored kawaii-ness as a design feature and factor of user experience in the visual appearance, nonverbal behaviour, and sound of robots and virtual characters. In this initial work, we conside...
http://arxiv.org/pdf/2304.12809v1
2023-04-22T01:47:19
2023-04-22T01:47:19
{ "title": [ -0.03262329, -0.015159607, -0.05206299, -0.026473999, -0.019073486, -0.011924744, -0.017105103, 0.014183044, -0.031799316, 0.07287598, -0.0055351257, 0.004272461, 0.032592773, -0.019073486, -0.04736328, -0.0037784576, -0.0015792847, ...
http://arxiv.org/abs/2304.11276v1
The Role of AI in Human-AI Creative Writing for Hong Kong Secondary Students
Hengky Susanto, David James Woo, Kai Guo
cs.CL
The recent advancement in Natural Language Processing (NLP) capability has led to the development of language models (e.g., ChatGPT) that is capable of generating human-like language. In this study, we explore how language models can be utilized to help the ideation aspect of creative writing. Our empirical findings sh...
http://arxiv.org/pdf/2304.11276v1
2023-04-21T23:50:09
2023-04-21T23:50:09
{ "title": [ -0.022354126, -0.009346008, -0.015296936, 0.036132812, -0.037322998, -0.013931274, 0.03390503, -0.033569336, -0.03793335, 0.051696777, -0.04336548, -0.003932953, 0.023483276, -0.0032253265, -0.011764526, -0.047058105, -0.01386261, 0....
http://arxiv.org/abs/2304.14415v1
Generative AI Perceptions: A Survey to Measure the Perceptions of Faculty, Staff, and Students on Generative AI Tools in Academia
Sara Amani, Lance White, Trini Balart, Laksha Arora, Dr. Kristi J. Shryock, Dr. Kelly Brumbelow, Dr. Karan L. Watson
cs.HC, cs.AI, cs.CL, cs.CY
ChatGPT is a natural language processing tool that can engage in human-like conversations and generate coherent and contextually relevant responses to various prompts. ChatGPT is capable of understanding natural text that is input by a user and generating appropriate responses in various forms. This tool represents a m...
http://arxiv.org/pdf/2304.14415v1
2023-04-21T23:08:39
2023-04-21T23:08:39
{ "title": [ -0.046905518, 0.017547607, -0.034362793, 0.0579834, -0.00008094311, -0.0231781, 0.068603516, -0.036743164, 0.029418945, 0.017608643, -0.07733154, -0.013191223, 0.00034165382, -0.006706238, -0.03945923, -0.03265381, -0.0018024445, -0....
http://arxiv.org/abs/2304.11163v1
ChatGPT, Large Language Technologies, and the Bumpy Road of Benefiting Humanity
Atoosa Kasirzadeh
cs.CY, cs.CL
The allure of emerging AI technologies is undoubtedly thrilling. However, the promise that AI technologies will benefit all of humanity is empty so long as we lack a nuanced understanding of what humanity is supposed to be in the face of widening global inequality and pressing existential threats. Going forward, it is ...
http://arxiv.org/pdf/2304.11163v1
2023-04-21T22:53:45
2023-04-21T22:53:45
{ "title": [ -0.02633667, -0.01600647, -0.046173096, -0.032684326, -0.003932953, -0.014175415, -0.007709503, 0.007118225, -0.039764404, 0.028198242, -0.022994995, 0.025787354, -0.012863159, -0.008773804, -0.0146484375, -0.0016918182, -0.008155823, ...
http://arxiv.org/abs/2304.11257v1
Who's the Best Detective? LLMs vs. MLs in Detecting Incoherent Fourth Grade Math Answers
Felipe Urrutia, Roberto Araya
cs.CL, cs.AI, cs.LG
Written answers to open-ended questions can have a higher long-term effect on learning than multiple-choice questions. However, it is critical that teachers immediately review the answers, and ask to redo those that are incoherent. This can be a difficult task and can be time-consuming for teachers. A possible solution...
http://arxiv.org/pdf/2304.11257v1
2023-04-21T21:25:30
2023-04-21T21:25:30
{ "title": [ -0.027130127, -0.02684021, -0.033233643, 0.010681152, -0.051818848, -0.016357422, 0.0064735413, 0.020324707, -0.012542725, 0.017196655, -0.0017700195, 0.0027103424, -0.006088257, -0.03955078, 0.008605957, -0.03869629, -0.0035037994, ...
http://arxiv.org/abs/2304.11256v2
UBC-DLNLP at SemEval-2023 Task 12: Impact of Transfer Learning on African Sentiment Analysis
Gagan Bhatia, Ife Adebara, AbdelRahim Elmadany, Muhammad Abdul-Mageed
cs.CL
We describe our contribution to the SemEVAl 2023 AfriSenti-SemEval shared task, where we tackle the task of sentiment analysis in 14 different African languages. We develop both monolingual and multilingual models under a full supervised setting (subtasks A and B). We also develop models for the zero-shot setting (subt...
http://arxiv.org/pdf/2304.11256v2
2023-04-25T07:44:37
2023-04-21T21:25:14
{ "title": [ 0.00040721893, -0.012588501, -0.011878967, -0.01222229, -0.018005371, -0.008575439, -0.025405884, -0.037384033, -0.0056610107, 0.035827637, -0.03186035, 0.025344849, 0.014717102, -0.012046814, -0.020004272, 0.0076141357, -0.0063323975, ...
http://arxiv.org/abs/2304.11223v1
A Group-Specific Approach to NLP for Hate Speech Detection
Karina Halevy
cs.CL
Automatic hate speech detection is an important yet complex task, requiring knowledge of common sense, stereotypes of protected groups, and histories of discrimination, each of which may constantly evolve. In this paper, we propose a group-specific approach to NLP for online hate speech detection. The approach consists...
http://arxiv.org/pdf/2304.11223v1
2023-04-21T19:08:49
2023-04-21T19:08:49
{ "title": [ 0.0073623657, 0.0034008026, -0.03878784, -0.03729248, 0.026504517, -0.027877808, -0.0005187988, -0.0040512085, -0.0059394836, 0.030441284, -0.01826477, 0.008346558, 0.0027008057, -0.006198883, -0.055908203, -0.0030517578, -0.02696228, ...
http://arxiv.org/abs/2304.11220v2
Learn What NOT to Learn: Towards Generative Safety in Chatbots
Leila Khalatbari, Yejin Bang, Dan Su, Willy Chung, Saeed Ghadimi, Hossein Sameti, Pascale Fung
cs.CL
Conversational models that are generative and open-domain are particularly susceptible to generating unsafe content since they are trained on web-based social data. Prior approaches to mitigating this issue have drawbacks, such as disrupting the flow of conversation, limited generalization to unseen toxic input context...
http://arxiv.org/pdf/2304.11220v2
2023-04-25T08:16:47
2023-04-21T18:59:06
{ "title": [ 0.0062942505, -0.014427185, -0.02229309, -0.06744385, 0.005256653, 0.00073719025, -0.020141602, 0.0033550262, -0.012794495, 0.040985107, -0.03640747, 0.041900635, -0.003490448, -0.0042152405, -0.03189087, -0.0057525635, -0.008155823, ...
http://arxiv.org/abs/2304.11158v2
Emergent and Predictable Memorization in Large Language Models
Stella Biderman, USVSN Sai Prashanth, Lintang Sutawika, Hailey Schoelkopf, Quentin Anthony, Shivanshu Purohit, Edward Raff
cs.CL
Memorization, or the tendency of large language models (LLMs) to output entire sequences from their training data verbatim, is a key concern for safely deploying language models. In particular, it is vital to minimize a model's memorization of sensitive datapoints such as those containing personal identifiable informat...
http://arxiv.org/pdf/2304.11158v2
2023-05-31T19:09:45
2023-04-21T17:58:31
{ "title": [ -0.03314209, -0.0067863464, -0.034423828, -0.05770874, -0.012229919, -0.006149292, -0.04006958, 0.0141067505, -0.0035953522, 0.050354004, -0.014984131, 0.011856079, 0.006755829, -0.0046920776, -0.046691895, 0.0077667236, -0.007167816, ...
http://arxiv.org/abs/2305.03512v1
Building Multimodal AI Chatbots
Min Young Lee
cs.CL
This work aims to create a multimodal AI system that chats with humans and shares relevant photos. While earlier works were limited to dialogues about specific objects or scenes within images, recent works have incorporated images into open-domain dialogues. However, their response generators are unimodal, accepting te...
http://arxiv.org/pdf/2305.03512v1
2023-04-21T16:43:54
2023-04-21T16:43:54
{ "title": [ -0.0154953, -0.0127334595, -0.026260376, -0.03237915, -0.03933716, -0.002632141, -0.02798462, 0.009101868, -0.018798828, 0.053497314, -0.029510498, 0.0008378029, 0.014289856, -0.020446777, -0.012702942, -0.012535095, -0.00004428625, ...
http://arxiv.org/abs/2304.14347v1
The Dark Side of ChatGPT: Legal and Ethical Challenges from Stochastic Parrots and Hallucination
Zihao Li
cs.CY, cs.CL, cs.LG
With the launch of ChatGPT, Large Language Models (LLMs) are shaking up our whole society, rapidly altering the way we think, create and live. For instance, the GPT integration in Bing has altered our approach to online searching. While nascent LLMs have many advantages, new legal and ethical risks are also emerging, s...
http://arxiv.org/pdf/2304.14347v1
2023-04-21T16:40:54
2023-04-21T16:40:54
{ "title": [ 0.0043907166, -0.010009766, -0.02444458, -0.009208679, 0.001408577, -0.024291992, 0.0010375977, -0.017990112, -0.016281128, 0.03262329, -0.040740967, 0.035583496, -0.0033569336, -0.006702423, -0.022613525, -0.00774765, -0.016784668, ...
http://arxiv.org/abs/2304.11111v1
Inducing anxiety in large language models increases exploration and bias
Julian Coda-Forno, Kristin Witte, Akshay K. Jagadish, Marcel Binz, Zeynep Akata, Eric Schulz
cs.CL, cs.AI, cs.LG
Large language models are transforming research on machine learning while galvanizing public debates. Understanding not only when these models work well and succeed but also why they fail and misbehave is of great societal relevance. We propose to turn the lens of computational psychiatry, a framework used to computati...
http://arxiv.org/pdf/2304.11111v1
2023-04-21T16:29:43
2023-04-21T16:29:43
{ "title": [ -0.010307312, -0.0076293945, -0.016220093, -0.020965576, -0.01737976, -0.018692017, -0.016708374, -0.023666382, 0.0069351196, 0.013702393, -0.04385376, 0.0020198822, 0.00166893, -0.023635864, -0.037506104, -0.008110046, -0.015930176, ...
http://arxiv.org/abs/2304.11107v1
ChatABL: Abductive Learning via Natural Language Interaction with ChatGPT
Tianyang Zhong, Yaonai Wei, Li Yang, Zihao Wu, Zhengliang Liu, Xiaozheng Wei, Wenjun Li, Junjie Yao, Chong Ma, Xiang Li, Dajiang Zhu, Xi Jiang, Junwei Han, Dinggang Shen, Tianming Liu, Tuo Zhang
cs.CL, cs.AI
Large language models (LLMs) such as ChatGPT have recently demonstrated significant potential in mathematical abilities, providing valuable reasoning paradigm consistent with human natural language. However, LLMs currently have difficulty in bridging perception, language understanding and reasoning capabilities due to ...
http://arxiv.org/pdf/2304.11107v1
2023-04-21T16:23:47
2023-04-21T16:23:47
{ "title": [ -0.011833191, -0.021636963, -0.0034179688, -0.006706238, -0.013427734, 0.00011700392, 0.015670776, -0.033813477, -0.037597656, -0.0042381287, -0.021408081, 0.028549194, -0.022460938, -0.027236938, 0.023468018, -0.012664795, -0.03112793, ...
http://arxiv.org/abs/2304.11015v3
DIN-SQL: Decomposed In-Context Learning of Text-to-SQL with Self-Correction
Mohammadreza Pourreza, Davood Rafiei
cs.CL, cs.AI, cs.DB, cs.HC
There is currently a significant gap between the performance of fine-tuned models and prompting approaches using Large Language Models (LLMs) on the challenging task of text-to-SQL, as evaluated on datasets such as Spider. To improve the performance of LLMs in the reasoning process, we study how decomposing the task in...
http://arxiv.org/pdf/2304.11015v3
2023-11-02T20:30:12
2023-04-21T15:02:18
{ "title": [ -0.01637268, -0.015037537, -0.040649414, -0.061462402, 0.014480591, -0.0010709763, -0.012496948, 0.015838623, 0.032043457, 0.03768921, -0.05441284, 0.009727478, 0.0065689087, 0.0035686493, -0.05947876, -0.016784668, -0.010108948, 0.0...
http://arxiv.org/abs/2304.10996v1
BERT Based Clinical Knowledge Extraction for Biomedical Knowledge Graph Construction and Analysis
Ayoub Harnoune, Maryem Rhanoui, Mounia Mikram, Siham Yousfi, Zineb Elkaimbillah, Bouchra El Asri
cs.CL, cs.AI
Background : Knowledge is evolving over time, often as a result of new discoveries or changes in the adopted methods of reasoning. Also, new facts or evidence may become available, leading to new understandings of complex phenomena. This is particularly true in the biomedical field, where scientists and physicians are ...
http://arxiv.org/pdf/2304.10996v1
2023-04-21T14:45:33
2023-04-21T14:45:33
{ "title": [ -0.040924072, -0.021148682, -0.06585693, -0.091796875, -0.0008983612, 0.019500732, -0.06652832, 0.028427124, -0.011856079, 0.031097412, 0.0010719299, 0.0010051727, -0.0002541542, -0.025848389, -0.040008545, 0.0010442734, -0.013641357, ...
http://arxiv.org/abs/2304.10994v1
Information Extraction from Documents: Question Answering vs Token Classification in real-world setups
Laurent Lam, Pirashanth Ratnamogan, Joël Tang, William Vanhuffel, Fabien Caspani
cs.CL
Research in Document Intelligence and especially in Document Key Information Extraction (DocKIE) has been mainly solved as Token Classification problem. Recent breakthroughs in both natural language processing (NLP) and computer vision helped building document-focused pre-training methods, leveraging a multimodal under...
http://arxiv.org/pdf/2304.10994v1
2023-04-21T14:43:42
2023-04-21T14:43:42
{ "title": [ -0.028778076, -0.014266968, -0.039642334, -0.03994751, -0.006374359, -0.011619568, -0.019470215, -0.009010315, 0.0033874512, 0.005329132, -0.04360962, -0.00548172, -0.017425537, -0.0035133362, -0.042816162, -0.026443481, -0.01637268, ...
http://arxiv.org/abs/2304.10977v1
Evaluating Transformer Language Models on Arithmetic Operations Using Number Decomposition
Matteo Muffo, Aldo Cocco, Enrico Bertino
cs.CL, stat.ML
In recent years, Large Language Models such as GPT-3 showed remarkable capabilities in performing NLP tasks in the zero and few shot settings. On the other hand, the experiments highlighted the difficulty of GPT-3 in carrying out tasks that require a certain degree of reasoning, such as arithmetic operations. In this p...
http://arxiv.org/pdf/2304.10977v1
2023-04-21T14:21:52
2023-04-21T14:21:52
{ "title": [ -0.026367188, 0.0037517548, -0.0390625, -0.05722046, 0.013893127, -0.006439209, -0.043640137, 0.012504578, -0.00010794401, 0.045806885, -0.0005273819, -0.00995636, 0.033050537, -0.021209717, -0.036895752, 0.01499939, -0.013137817, 0....
http://arxiv.org/abs/2304.10973v1
LEIA: Linguistic Embeddings for the Identification of Affect
Segun Taofeek Aroyehun, Lukas Malik, Hannah Metzler, Nikolas Haimerl, Anna Di Natale, David Garcia
cs.CL, cs.AI, cs.CY, cs.LG
The wealth of text data generated by social media has enabled new kinds of analysis of emotions with language models. These models are often trained on small and costly datasets of text annotations produced by readers who guess the emotions expressed by others in social media posts. This affects the quality of emotion ...
http://arxiv.org/pdf/2304.10973v1
2023-04-21T14:17:10
2023-04-21T14:17:10
{ "title": [ -0.008987427, -0.014953613, -0.022766113, -0.05697632, -0.0047798157, -0.0008831024, -0.03878784, -0.011383057, 0.03479004, 0.015731812, -0.019104004, -0.0154953, 0.020324707, -0.026565552, -0.045898438, 0.010719299, -0.011184692, 0....
http://arxiv.org/abs/2304.10859v2
Text2Time: Transformer-based Article Time Period Prediction
Karthick Prasad Gunasekaran, B Chase Babrich, Saurabh Shirodkar, Hee Hwang
cs.CL
The task of predicting the publication period of text documents, such as news articles, is an important but less studied problem in the field of natural language processing. Predicting the year of a news article can be useful in various contexts, such as historical research, sentiment analysis, and media monitoring. In...
http://arxiv.org/pdf/2304.10859v2
2023-04-24T03:56:03
2023-04-21T10:05:03
{ "title": [ -0.027297974, -0.0064582825, -0.055236816, -0.052734375, 0.007255554, -0.009338379, -0.050048828, -0.030700684, 0.007926941, 0.043304443, -0.015533447, 0.01309967, 0.016784668, -0.010002136, -0.04727173, 0.016021729, -0.00016391277, ...
http://arxiv.org/abs/2304.10844v1
Better Sign Language Translation with Monolingual Data
Ru Peng, Yawen Zeng, Junbo Zhao
cs.CL
Sign language translation (SLT) systems, which are often decomposed into video-to-gloss (V2G) recognition and gloss-to-text (G2T) translation through the pivot gloss, heavily relies on the availability of large-scale parallel G2T pairs. However, the manual annotation of pivot gloss, which is a sequence of transcribed w...
http://arxiv.org/pdf/2304.10844v1
2023-04-21T09:39:54
2023-04-21T09:39:54
{ "title": [ -0.031311035, 0.005382538, -0.05581665, -0.035614014, -0.033172607, -0.015914917, -0.05834961, -0.0028514862, 0.002784729, 0.0395813, -0.048919678, 0.001991272, -0.000623703, -0.02331543, -0.013183594, -0.016357422, 0.011390686, 0.00...
http://arxiv.org/abs/2304.10813v1
Tokenization Tractability for Human and Machine Learning Model: An Annotation Study
Tatsuya Hiraoka, Tomoya Iwakura
cs.CL
Is tractable tokenization for humans also tractable for machine learning models? This study investigates relations between tractable tokenization for humans (e.g., appropriateness and readability) and one for models of machine learning (e.g., performance on an NLP task). We compared six tokenization methods on the Japa...
http://arxiv.org/pdf/2304.10813v1
2023-04-21T08:49:06
2023-04-21T08:49:06
{ "title": [ -0.015426636, 0.007270813, -0.02835083, -0.04397583, 0.0016803741, -0.0030841827, -0.039489746, -0.01651001, -0.02507019, 0.02545166, -0.024169922, 0.0040359497, -0.011650085, -0.0096206665, -0.04321289, -0.021774292, -0.018371582, 0...