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/2304.08821v1
TTIDA: Controllable Generative Data Augmentation via Text-to-Text and Text-to-Image Models
Yuwei Yin, Jean Kaddour, Xiang Zhang, Yixin Nie, Zhenguang Liu, Lingpeng Kong, Qi Liu
cs.CV, cs.CL, cs.LG
Data augmentation has been established as an efficacious approach to supplement useful information for low-resource datasets. Traditional augmentation techniques such as noise injection and image transformations have been widely used. In addition, generative data augmentation (GDA) has been shown to produce more divers...
http://arxiv.org/pdf/2304.08821v1
2023-04-18T08:40:30
2023-04-18T08:40:30
{ "title": [ -0.01121521, -0.0078125, -0.04711914, -0.051483154, 0.0011339188, -0.00094366074, -0.029769897, -0.012435913, 0.023010254, 0.0262146, -0.030929565, 0.019622803, 0.001411438, -0.013198853, -0.034088135, 0.0043411255, -0.006137848, 0.0...
http://arxiv.org/abs/2304.08807v2
Revisiting the Role of Similarity and Dissimilarity in Best Counter Argument Retrieval
Hongguang Shi, Shuirong Cao, Cam-Tu Nguyen
cs.CL
This paper studies the task of best counter-argument retrieval given an input argument. Following the definition that the best counter-argument addresses the same aspects as the input argument while having the opposite stance, we aim to develop an efficient and effective model for scoring counter-arguments based on sim...
http://arxiv.org/pdf/2304.08807v2
2023-04-19T03:04:28
2023-04-18T08:13:48
{ "title": [ 0.065979004, 0.0030059814, 0.008544922, -0.05984497, 0.012054443, -0.01776123, -0.0025959015, -0.048461914, -0.0018262863, 0.03213501, -0.019195557, 0.0060157776, -0.012023926, -0.00969696, -0.0574646, -0.024551392, -0.0010128021, 0....
http://arxiv.org/abs/2304.08801v2
Speaker Profiling in Multiparty Conversations
Shivani Kumar, Rishabh Gupta, Md Shad Akhtar, Tanmoy Chakraborty
cs.CL, cs.AI
In conversational settings, individuals exhibit unique behaviors, rendering a one-size-fits-all approach insufficient for generating responses by dialogue agents. Although past studies have aimed to create personalized dialogue agents using speaker persona information, they have relied on the assumption that the speake...
http://arxiv.org/pdf/2304.08801v2
2023-04-19T05:52:41
2023-04-18T08:04:46
{ "title": [ 0.00390625, 0.0362854, -0.06628418, -0.06732178, 0.023834229, -0.023986816, -0.053222656, -0.018722534, -0.017150879, 0.048614502, 0.013145447, 0.0039024353, -0.00844574, -0.0259552, -0.034973145, 0.000685215, -0.020721436, 0.0358581...
http://arxiv.org/abs/2304.08763v2
A Survey for Biomedical Text Summarization: From Pre-trained to Large Language Models
Qianqian Xie, Zheheng Luo, Benyou Wang, Sophia Ananiadou
cs.CL
The exponential growth of biomedical texts such as biomedical literature and electronic health records (EHRs), poses a significant challenge for clinicians and researchers to access clinical information efficiently. To tackle this challenge, biomedical text summarization (BTS) has been proposed as a solution to support...
http://arxiv.org/pdf/2304.08763v2
2023-07-13T04:13:17
2023-04-18T06:38:40
{ "title": [ -0.024871826, -0.0023021698, -0.017822266, -0.06878662, -0.015960693, 0.02507019, -0.07232666, -0.0068969727, 0.008049011, 0.050354004, 0.013626099, 0.008079529, 0.01133728, -0.011695862, -0.061065674, 0.025726318, -0.0012693405, 0.0...
http://arxiv.org/abs/2304.09820v1
A Two-Stage Framework with Self-Supervised Distillation For Cross-Domain Text Classification
Yunlong Feng, Bohan Li, Libo Qin, Xiao Xu, Wanxiang Che
cs.CL, cs.AI
Cross-domain text classification aims to adapt models to a target domain that lacks labeled data. It leverages or reuses rich labeled data from the different but related source domain(s) and unlabeled data from the target domain. To this end, previous work focuses on either extracting domain-invariant features or task-...
http://arxiv.org/pdf/2304.09820v1
2023-04-18T06:21:40
2023-04-18T06:21:40
{ "title": [ 0.0063056946, 0.027404785, -0.046569824, -0.06561279, 0.04031372, -0.017547607, -0.027328491, -0.012748718, -0.002822876, 0.017318726, -0.020141602, 0.020904541, -0.027023315, -0.016204834, -0.027755737, -0.010414124, -0.031402588, 0...
http://arxiv.org/abs/2304.11077v1
HeRo: RoBERTa and Longformer Hebrew Language Models
Vitaly Shalumov, Harel Haskey
cs.CL, cs.AI
In this paper, we fill in an existing gap in resources available to the Hebrew NLP community by providing it with the largest so far pre-train dataset HeDC4, a state-of-the-art pre-trained language model HeRo for standard length inputs and an efficient transformer LongHeRo for long input sequences. The HeRo model was e...
http://arxiv.org/pdf/2304.11077v1
2023-04-18T05:56:32
2023-04-18T05:56:32
{ "title": [ 0.010566711, -0.008758545, -0.04067993, -0.06530762, -0.027267456, -0.004196167, -0.06726074, 0.021865845, -0.06109619, 0.05380249, 0.027496338, -0.002002716, 0.044891357, -0.07342529, 0.035827637, 0.0015687943, -0.0112838745, 0.0175...
http://arxiv.org/abs/2304.10946v1
CancerGPT: Few-shot Drug Pair Synergy Prediction using Large Pre-trained Language Models
Tianhao Li, Sandesh Shetty, Advaith Kamath, Ajay Jaiswal, Xianqian Jiang, Ying Ding, Yejin Kim
cs.CL, cs.LG, q-bio.BM
Large pre-trained language models (LLMs) have been shown to have significant potential in few-shot learning across various fields, even with minimal training data. However, their ability to generalize to unseen tasks in more complex fields, such as biology, has yet to be fully evaluated. LLMs can offer a promising alte...
http://arxiv.org/pdf/2304.10946v1
2023-04-18T02:49:53
2023-04-18T02:49:53
{ "title": [ 0.008918762, -0.030349731, -0.029022217, -0.064208984, -0.02470398, -0.009437561, -0.05267334, 0.0035381317, -0.01689148, 0.029953003, -0.008575439, 0.012649536, -0.023040771, -0.021759033, -0.044799805, -0.008239746, -0.01914978, 0....
http://arxiv.org/abs/2306.01739v1
Comparative study on Judgment Text Classification for Transformer Based Models
Stanley Kingston, Prassanth, Shrinivas A V, Balamurugan MS, Manoj Kumar Rajagopal
cs.CL, cs.AI
This work involves the usage of various NLP models to predict the winner of a particular judgment by the means of text extraction and summarization from a judgment document. These documents are useful when it comes to legal proceedings. One such advantage is that these can be used for citations and precedence reference...
http://arxiv.org/pdf/2306.01739v1
2023-04-18T02:20:43
2023-04-18T02:20:43
{ "title": [ -0.02218628, 0.014862061, -0.043701172, -0.03466797, -0.0006322861, -0.012550354, -0.037078857, -0.011375427, 0.024047852, 0.03945923, -0.054779053, 0.0068893433, 0.0317688, -0.013061523, -0.047302246, -0.0012931824, -0.027862549, 0....
http://arxiv.org/abs/2304.08653v1
On Uncertainty Calibration and Selective Generation in Probabilistic Neural Summarization: A Benchmark Study
Polina Zablotskaia, Du Phan, Joshua Maynez, Shashi Narayan, Jie Ren, Jeremiah Liu
cs.CL, cs.LG
Modern deep models for summarization attains impressive benchmark performance, but they are prone to generating miscalibrated predictive uncertainty. This means that they assign high confidence to low-quality predictions, leading to compromised reliability and trustworthiness in real-world applications. Probabilistic d...
http://arxiv.org/pdf/2304.08653v1
2023-04-17T23:06:28
2023-04-17T23:06:28
{ "title": [ -0.023361206, 0.02128601, -0.029800415, -0.058624268, -0.013664246, 0.009399414, -0.08483887, -0.022567749, 0.004650116, 0.03463745, 0.007850647, 0.021499634, -0.022857666, -0.021591187, -0.053588867, 0.0071144104, 0.0021305084, 0.01...
http://arxiv.org/abs/2304.08649v3
Classification of US Supreme Court Cases using BERT-Based Techniques
Shubham Vatsal, Adam Meyers, John E. Ortega
cs.CL, cs.AI, cs.LG
Models based on bidirectional encoder representations from transformers (BERT) produce state of the art (SOTA) results on many natural language processing (NLP) tasks such as named entity recognition (NER), part-of-speech (POS) tagging etc. An interesting phenomenon occurs when classifying long documents such as those ...
http://arxiv.org/pdf/2304.08649v3
2023-07-24T15:33:25
2023-04-17T22:53:54
{ "title": [ 0.029266357, 0.0051612854, -0.022583008, -0.050964355, -0.0074882507, -0.0149383545, -0.021514893, -0.0063705444, -0.008735657, 0.045806885, -0.054718018, 0.017684937, 0.031707764, -0.02330017, -0.017318726, -0.0024852753, -0.018081665, ...
http://arxiv.org/abs/2304.11094v1
Effectiveness of Debiasing Techniques: An Indigenous Qualitative Analysis
Vithya Yogarajan, Gillian Dobbie, Henry Gouk
cs.CL, cs.AI
An indigenous perspective on the effectiveness of debiasing techniques for pre-trained language models (PLMs) is presented in this paper. The current techniques used to measure and debias PLMs are skewed towards the US racial biases and rely on pre-defined bias attributes (e.g. "black" vs "white"). Some require large d...
http://arxiv.org/pdf/2304.11094v1
2023-04-17T22:25:25
2023-04-17T22:25:25
{ "title": [ 0.0049743652, 0.03186035, -0.0129470825, -0.03604126, 0.013969421, 0.0010480881, -0.060333252, -0.07421875, -0.028305054, 0.043792725, -0.0143585205, -0.004837036, 0.005962372, -0.027694702, -0.047821045, -0.008743286, -0.01083374, 0...
http://arxiv.org/abs/2304.08637v1
An Evaluation on Large Language Model Outputs: Discourse and Memorization
Adrian de Wynter, Xun Wang, Alex Sokolov, Qilong Gu, Si-Qing Chen
cs.CL, cs.AI, cs.LG
We present an empirical evaluation of various outputs generated by nine of the most widely-available large language models (LLMs). Our analysis is done with off-the-shelf, readily-available tools. We find a correlation between percentage of memorized text, percentage of unique text, and overall output quality, when mea...
http://arxiv.org/pdf/2304.08637v1
2023-04-17T22:12:12
2023-04-17T22:12:12
{ "title": [ -0.026809692, -0.000787735, -0.037109375, -0.05291748, -0.015525818, -0.0069274902, -0.061401367, 0.012390137, -0.005844116, 0.03656006, -0.0030784607, -0.0072021484, 0.0020198822, -0.022247314, -0.051696777, 0.004634857, -0.019302368, ...
http://arxiv.org/abs/2304.08612v3
Bridging Discrete and Backpropagation: Straight-Through and Beyond
Liyuan Liu, Chengyu Dong, Xiaodong Liu, Bin Yu, Jianfeng Gao
cs.LG, cs.AI, cs.CL, cs.CV
Backpropagation, the cornerstone of deep learning, is limited to computing gradients for continuous variables. This limitation poses challenges for problems involving discrete latent variables. To address this issue, we propose a novel approach to approximate the gradient of parameters involved in generating discrete l...
http://arxiv.org/pdf/2304.08612v3
2023-10-16T16:27:02
2023-04-17T20:59:49
{ "title": [ -0.0048294067, 0.0090408325, -0.037750244, -0.06939697, 0.03945923, -0.01335144, -0.020614624, -0.021697998, 0.0071983337, 0.015106201, -0.009170532, 0.013023376, -0.027877808, -0.033416748, -0.04107666, -0.0039978027, -0.013122559, ...
http://arxiv.org/abs/2304.08582v1
Researchers eye-view of sarcasm detection in social media textual content
Swapnil Mane, Vaibhav Khatavkar
cs.CL
The enormous use of sarcastic text in all forms of communication in social media will have a physiological effect on target users. Each user has a different approach to misusing and recognising sarcasm. Sarcasm detection is difficult even for users, and this will depend on many things such as perspective, context, spec...
http://arxiv.org/pdf/2304.08582v1
2023-04-17T19:45:10
2023-04-17T19:45:10
{ "title": [ 0.009178162, 0.004558563, -0.038024902, -0.001001358, -0.013282776, -0.045898438, 0.01309967, -0.014030457, -0.0020427704, 0.028656006, -0.05029297, -0.03439331, 0.057861328, -0.029418945, -0.003753662, -0.031677246, -0.014328003, 0....
http://arxiv.org/abs/2304.08485v1
Visual Instruction Tuning
Haotian Liu, Chunyuan Li, Qingyang Wu, Yong Jae Lee
cs.CV, cs.AI, cs.CL, cs.LG
Instruction tuning large language models (LLMs) using machine-generated instruction-following data has improved zero-shot capabilities on new tasks, but the idea is less explored in the multimodal field. In this paper, we present the first attempt to use language-only GPT-4 to generate multimodal language-image instruc...
http://arxiv.org/pdf/2304.08485v1
2023-04-17T17:59:25
2023-04-17T17:59:25
{ "title": [ -0.018951416, -0.0018863678, -0.017089844, -0.015075684, -0.06274414, -0.012931824, -0.045959473, -0.020385742, 0.00041604042, 0.04486084, -0.06890869, 0.0031261444, -0.0023021698, -0.009529114, -0.0066604614, -0.012451172, 0.03640747, ...
http://arxiv.org/abs/2304.08480v1
DisCo-CLIP: A Distributed Contrastive Loss for Memory Efficient CLIP Training
Yihao Chen, Xianbiao Qi, Jianan Wang, Lei Zhang
cs.CV, cs.AI, cs.CL
We propose DisCo-CLIP, a distributed memory-efficient CLIP training approach, to reduce the memory consumption of contrastive loss when training contrastive learning models. Our approach decomposes the contrastive loss and its gradient computation into two parts, one to calculate the intra-GPU gradients and the other t...
http://arxiv.org/pdf/2304.08480v1
2023-04-17T17:58:21
2023-04-17T17:58:21
{ "title": [ 0.0041923523, 0.007194519, -0.024810791, -0.06323242, 0.014297485, 0.00086069107, -0.019866943, -0.015563965, 0.02986145, 0.03717041, -0.008041382, -0.0060195923, 0.005458832, -0.02420044, -0.054138184, 0.010383606, 0.00605011, 0.033...
http://arxiv.org/abs/2304.08467v2
Learning to Compress Prompts with Gist Tokens
Jesse Mu, Xiang Lisa Li, Noah Goodman
cs.CL
Prompting is the primary way to utilize the multitask capabilities of language models (LMs), but prompts occupy valuable space in the input context window, and repeatedly encoding the same prompt is computationally inefficient. Finetuning and distillation methods allow for specialization of LMs without prompting, but r...
http://arxiv.org/pdf/2304.08467v2
2023-07-16T00:29:16
2023-04-17T17:47:37
{ "title": [ -0.024124146, -0.015670776, -0.044799805, -0.027420044, -0.03173828, 0.0059776306, -0.03515625, 0.0054626465, 0.006454468, 0.02204895, -0.02003479, -0.02923584, 0.005634308, -0.036346436, -0.02810669, -0.021575928, 0.02180481, 0.0060...
http://arxiv.org/abs/2304.08466v1
Synthetic Data from Diffusion Models Improves ImageNet Classification
Shekoofeh Azizi, Simon Kornblith, Chitwan Saharia, Mohammad Norouzi, David J. Fleet
cs.CV, cs.AI, cs.CL, cs.LG
Deep generative models are becoming increasingly powerful, now generating diverse high fidelity photo-realistic samples given text prompts. Have they reached the point where models of natural images can be used for generative data augmentation, helping to improve challenging discriminative tasks? We show that large-sca...
http://arxiv.org/pdf/2304.08466v1
2023-04-17T17:42:29
2023-04-17T17:42:29
{ "title": [ 0.008476257, 0.01739502, -0.03677368, -0.024139404, 0.017486572, -0.025924683, 0.0006456375, -0.025222778, 0.004753113, 0.0021514893, -0.033843994, 0.0018854141, -0.040374756, -0.020339966, -0.03363037, -0.008415222, -0.0014915466, 0...
http://arxiv.org/abs/2304.08460v1
LongForm: Optimizing Instruction Tuning for Long Text Generation with Corpus Extraction
Abdullatif Köksal, Timo Schick, Anna Korhonen, Hinrich Schütze
cs.CL, cs.AI, cs.LG
Instruction tuning enables language models to generalize more effectively and better follow user intent. However, obtaining instruction data can be costly and challenging. Prior works employ methods such as expensive human annotation, crowd-sourced datasets with alignment issues, or generating noisy examples via LLMs. ...
http://arxiv.org/pdf/2304.08460v1
2023-04-17T17:36:35
2023-04-17T17:36:35
{ "title": [ -0.0541687, -0.013244629, -0.026519775, -0.046081543, -0.015151978, 0.020446777, -0.04852295, -0.009025574, -0.00573349, 0.022247314, -0.023330688, 0.015449524, -0.005947113, -0.015716553, -0.034240723, -0.0073280334, -0.023010254, 0...
http://arxiv.org/abs/2304.08453v3
Improving Autoregressive NLP Tasks via Modular Linearized Attention
Victor Agostinelli, Lizhong Chen
cs.CL, cs.AI
Various natural language processing (NLP) tasks necessitate models that are efficient and small based on their ultimate application at the edge or in other resource-constrained environments. While prior research has reduced the size of these models, increasing computational efficiency without considerable performance i...
http://arxiv.org/pdf/2304.08453v3
2023-06-24T19:23:11
2023-04-17T17:25:48
{ "title": [ 0.0038757324, 0.01737976, -0.056671143, -0.064575195, 0.013999939, -0.020355225, -0.043121338, 0.009284973, 0.007118225, 0.05090332, -0.023345947, 0.030883789, 0.010765076, 0.018676758, -0.06719971, 0.011512756, -0.007835388, 0.01358...
http://arxiv.org/abs/2304.08448v2
ImpressionGPT: An Iterative Optimizing Framework for Radiology Report Summarization with ChatGPT
Chong Ma, Zihao Wu, Jiaqi Wang, Shaochen Xu, Yaonai Wei, Zhengliang Liu, Xi Jiang, Lei Guo, Xiaoyan Cai, Shu Zhang, Tuo Zhang, Dajiang Zhu, Dinggang Shen, Tianming Liu, Xiang Li
cs.CL, cs.AI
The 'Impression' section of a radiology report is a critical basis for communication between radiologists and other physicians, and it is typically written by radiologists based on the 'Findings' section. However, writing numerous impressions can be laborious and error-prone for radiologists. Although recent studies ha...
http://arxiv.org/pdf/2304.08448v2
2023-05-03T08:09:53
2023-04-17T17:13:42
{ "title": [ -0.004096985, 0.004169464, -0.023361206, -0.075805664, -0.012062073, 0.021118164, -0.056732178, -0.0024738312, -0.0069389343, 0.038269043, -0.010803223, 0.024551392, -0.024291992, -0.01373291, -0.037963867, 0.00667572, -0.01676941, 0...
http://arxiv.org/abs/2304.08442v1
The MiniPile Challenge for Data-Efficient Language Models
Jean Kaddour
cs.CL, cs.LG
The ever-growing diversity of pre-training text corpora has equipped language models with generalization capabilities across various downstream tasks. However, such diverse datasets are often too large for academic budgets; hence, most research on Transformer architectures, training procedures, optimizers, etc. gets co...
http://arxiv.org/pdf/2304.08442v1
2023-04-17T17:03:56
2023-04-17T17:03:56
{ "title": [ -0.045318604, -0.019973755, -0.052825928, -0.09863281, 0.0093688965, 0.01776123, -0.08111572, 0.04296875, 0.013786316, 0.03414917, -0.011306763, 0.006980896, -0.001958847, -0.006378174, -0.053497314, 0.01612854, -0.012321472, 0.02340...
http://arxiv.org/abs/2304.08431v1
Prak: An automatic phonetic alignment tool for Czech
Václav Hanžl, Adléta Hanžlová
cs.CL, cs.SD, eess.AS, C.m
Labeling speech down to the identity and time boundaries of phones is a labor-intensive part of phonetic research. To simplify this work, we created a free open-source tool generating phone sequences from Czech text and time-aligning them with audio. Low architecture complexity makes the design approachable for stude...
http://arxiv.org/pdf/2304.08431v1
2023-04-17T16:51:24
2023-04-17T16:51:24
{ "title": [ -0.037384033, -0.0039634705, -0.054138184, -0.043151855, 0.010879517, 0.007587433, -0.03918457, 0.019332886, -0.006690979, 0.052734375, -0.010719299, -0.011558533, 0.070007324, -0.03982544, -0.029037476, 0.018508911, -0.017974854, 0....
http://arxiv.org/abs/2304.08369v1
New Product Development (NPD) through Social Media-based Analysis by Comparing Word2Vec and BERT Word Embeddings
Princessa Cintaqia, Matheus Inoue
cs.CL, cs.LG
This study introduces novel methods for sentiment and opinion classification of tweets to support the New Product Development (NPD) process. Two popular word embedding techniques, Word2Vec and BERT, were evaluated as inputs for classic Machine Learning and Deep Learning algorithms to identify the best-performing approa...
http://arxiv.org/pdf/2304.08369v1
2023-04-17T15:32:11
2023-04-17T15:32:11
{ "title": [ -0.023956299, -0.029403687, -0.017303467, -0.0004017353, -0.062683105, -0.016586304, -0.03074646, -0.0025253296, -0.07342529, 0.032714844, -0.041503906, 0.05807495, -0.009979248, -0.0029773712, 0.041168213, -0.042175293, -0.0048561096, ...
http://arxiv.org/abs/2304.08354v2
Tool Learning with Foundation Models
Yujia Qin, Shengding Hu, Yankai Lin, Weize Chen, Ning Ding, Ganqu Cui, Zheni Zeng, Yufei Huang, Chaojun Xiao, Chi Han, Yi Ren Fung, Yusheng Su, Huadong Wang, Cheng Qian, Runchu Tian, Kunlun Zhu, Shihao Liang, Xingyu Shen, Bokai Xu, Zhen Zhang, Yining Ye, Bowen Li, Ziwei Tang, Jing Yi, Yuzhang Zhu, Zhenning Dai, Lan Yan...
cs.CL, cs.AI, cs.LG
Humans possess an extraordinary ability to create and utilize tools, allowing them to overcome physical limitations and explore new frontiers. With the advent of foundation models, AI systems have the potential to be equally adept in tool use as humans. This paradigm, i.e., tool learning with foundation models, combine...
http://arxiv.org/pdf/2304.08354v2
2023-06-15T14:10:42
2023-04-17T15:16:10
{ "title": [ -0.06109619, -0.008415222, -0.05596924, -0.041046143, -0.02355957, 0.005130768, -0.06732178, 0.0030879974, 0.0055160522, 0.025726318, -0.01777649, -0.014305115, -0.016281128, -0.024261475, -0.036376953, -0.007007599, 0.0071372986, 0....
http://arxiv.org/abs/2304.08352v1
What Makes a Good Dataset for Symbol Description Reading?
Karol Lynch, Joern Ploennigs, Bradley Eck
cs.CL, cs.AI
The usage of mathematical formulas as concise representations of a document's key ideas is common practice. Correctly interpreting these formulas, by identifying mathematical symbols and extracting their descriptions, is an important task in document understanding. This paper makes the following contributions to the ma...
http://arxiv.org/pdf/2304.08352v1
2023-04-17T15:14:27
2023-04-17T15:14:27
{ "title": [ 0.007270813, 0.006839752, -0.026046753, -0.019851685, -0.01675415, -0.01902771, -0.026489258, -0.013847351, 0.004295349, 0.023239136, -0.029937744, -0.033477783, -0.0256958, -0.019760132, -0.024383545, -0.027313232, 0.017303467, 0.01...
http://arxiv.org/abs/2304.08345v1
VALOR: Vision-Audio-Language Omni-Perception Pretraining Model and Dataset
Sihan Chen, Xingjian He, Longteng Guo, Xinxin Zhu, Weining Wang, Jinhui Tang, Jing Liu
cs.LG, cs.CL, cs.CV, cs.MM, eess.AS
In this paper, we propose a Vision-Audio-Language Omni-peRception pretraining model (VALOR) for multi-modal understanding and generation. Different from widely-studied vision-language pretraining models, VALOR jointly models relationships of vision, audio and language in an end-to-end manner. It contains three separate...
http://arxiv.org/pdf/2304.08345v1
2023-04-17T15:08:15
2023-04-17T15:08:15
{ "title": [ 0.0060806274, -0.023727417, -0.048950195, -0.04940796, -0.012298584, -0.03387451, -0.026733398, -0.0053138733, -0.015670776, 0.040618896, -0.013801575, -0.00944519, 0.03201294, -0.0390625, -0.01828003, -0.003419876, -0.006664276, 0.0...
http://arxiv.org/abs/2304.08341v1
Use of social media and Natural Language Processing (NLP) in natural hazard research
José Augusto Proença Maia Devienne
cs.CL, physics.geo-ph
Twitter is a microblogging service for sending short, public text messages (tweets) that has recently received more attention in scientific comunity. In the works of Sasaki et al. (2010) and Earle et al., (2011) the authors explored the real-time interaction on Twitter for detecting natural hazards (e.g., earthquakes, ...
http://arxiv.org/pdf/2304.08341v1
2023-04-17T15:03:05
2023-04-17T15:03:05
{ "title": [ 0.004333496, 0.004878998, -0.05206299, -0.032928467, 0.0034160614, -0.03845215, 0.0040512085, 0.002418518, -0.03805542, 0.034362793, -0.050109863, 0.028518677, 0.007888794, -0.015289307, 0.005290985, -0.014450073, -0.00894165, 0.0275...
http://arxiv.org/abs/2304.08327v1
LED: A Dataset for Life Event Extraction from Dialogs
Yi-Pei Chen, An-Zi Yen, Hen-Hsen Huang, Hideki Nakayama, Hsin-Hsi Chen
cs.CL
Lifelogging has gained more attention due to its wide applications, such as personalized recommendations or memory assistance. The issues of collecting and extracting personal life events have emerged. People often share their life experiences with others through conversations. However, extracting life events from conv...
http://arxiv.org/pdf/2304.08327v1
2023-04-17T14:46:59
2023-04-17T14:46:59
{ "title": [ 0.0067825317, -0.020462036, -0.050354004, -0.09869385, 0.0063171387, -0.010803223, -0.054016113, 0.017913818, 0.02861023, 0.036621094, 0.0031814575, 0.025314331, -0.0013027191, -0.017990112, -0.054870605, 0.030715942, -0.012786865, 0...
http://arxiv.org/abs/2304.08315v3
Thorny Roses: Investigating the Dual Use Dilemma in Natural Language Processing
Lucie-Aimée Kaffee, Arnav Arora, Zeerak Talat, Isabelle Augenstein
cs.CL, cs.AI
Dual use, the intentional, harmful reuse of technology and scientific artefacts, is a problem yet to be well-defined within the context of Natural Language Processing (NLP). However, as NLP technologies continue to advance and become increasingly widespread in society, their inner workings have become increasingly opaq...
http://arxiv.org/pdf/2304.08315v3
2023-10-30T11:14:50
2023-04-17T14:37:43
{ "title": [ -0.013832092, -0.00541687, -0.04095459, -0.043029785, 0.011802673, -0.016235352, -0.0055503845, -0.00598526, -0.02659607, 0.031051636, -0.040618896, 0.022766113, -0.01776123, 0.0017900467, -0.03475952, -0.012710571, -0.0115356445, 0....
http://arxiv.org/abs/2304.08295v1
Interactive and Explainable Region-guided Radiology Report Generation
Tim Tanida, Philip Müller, Georgios Kaissis, Daniel Rueckert
cs.CV, cs.CL, cs.LG
The automatic generation of radiology reports has the potential to assist radiologists in the time-consuming task of report writing. Existing methods generate the full report from image-level features, failing to explicitly focus on anatomical regions in the image. We propose a simple yet effective region-guided report...
http://arxiv.org/pdf/2304.08295v1
2023-04-17T14:12:09
2023-04-17T14:12:09
{ "title": [ -0.017105103, -0.00166893, -0.034332275, -0.06994629, -0.0057640076, 0.006801605, -0.041625977, 0.008407593, -0.0178833, 0.012039185, -0.017868042, 0.0056877136, -0.048187256, -0.026397705, -0.03479004, -0.012451172, 0.0036010742, -0...
http://arxiv.org/abs/2304.08216v2
Context-Dependent Embedding Utterance Representations for Emotion Recognition in Conversations
Patrícia Pereira, Helena Moniz, Isabel Dias, Joao Paulo Carvalho
cs.CL, cs.AI
Emotion Recognition in Conversations (ERC) has been gaining increasing importance as conversational agents become more and more common. Recognizing emotions is key for effective communication, being a crucial component in the development of effective and empathetic conversational agents. Knowledge and understanding of ...
http://arxiv.org/pdf/2304.08216v2
2023-06-03T07:43:12
2023-04-17T12:37:57
{ "title": [ 0.019638062, 0.009239197, -0.03378296, -0.06854248, 0.015098572, -0.015342712, -0.0345459, 0.0063171387, 0.03567505, 0.03704834, -0.03112793, 0.00019657612, 0.028442383, -0.012878418, -0.047302246, 0.012420654, -0.011878967, 0.031051...
http://arxiv.org/abs/2304.08205v1
VECO 2.0: Cross-lingual Language Model Pre-training with Multi-granularity Contrastive Learning
Zhen-Ru Zhang, Chuanqi Tan, Songfang Huang, Fei Huang
cs.CL
Recent studies have demonstrated the potential of cross-lingual transferability by training a unified Transformer encoder for multiple languages. In addition to involving the masked language model objective, existing cross-lingual pre-training works leverage sentence-level contrastive learning or plugs in extra cross-a...
http://arxiv.org/pdf/2304.08205v1
2023-04-17T12:23:41
2023-04-17T12:23:41
{ "title": [ -0.007095337, -0.013015747, -0.024307251, -0.0758667, -0.0012130737, 0.008735657, -0.07128906, -0.0087890625, 0.019348145, 0.02583313, -0.005908966, 0.0073280334, 0.015296936, -0.032806396, -0.03353882, 0.014823914, -0.018844604, 0.0...
http://arxiv.org/abs/2304.08177v2
Efficient and Effective Text Encoding for Chinese LLaMA and Alpaca
Yiming Cui, Ziqing Yang, Xin Yao
cs.CL, cs.HC, cs.LG
Large Language Models (LLMs), such as ChatGPT and GPT-4, have dramatically transformed natural language processing research and shown promising strides towards Artificial General Intelligence (AGI). Nonetheless, the high costs associated with training and deploying LLMs present substantial obstacles to transparent, acc...
http://arxiv.org/pdf/2304.08177v2
2023-06-15T06:23:19
2023-04-17T11:39:53
{ "title": [ -0.053863525, -0.003168106, -0.021774292, 0.019592285, -0.03894043, -0.0019054413, -0.026016235, -0.021072388, -0.021408081, 0.05569458, 0.0009908676, -0.0064697266, 0.04623413, -0.045166016, 0.0071487427, 0.006214142, 0.00018203259, ...
http://arxiv.org/abs/2304.08137v1
Political corpus creation through automatic speech recognition on EU debates
Hugo de Vos, Suzan Verberne
cs.CL
In this paper, we present a transcribed corpus of the LIBE committee of the EU parliament, totalling 3.6 Million running words. The meetings of parliamentary committees of the EU are a potentially valuable source of information for political scientists but the data is not readily available because only disclosed as spe...
http://arxiv.org/pdf/2304.08137v1
2023-04-17T10:41:59
2023-04-17T10:41:59
{ "title": [ 0.019302368, 0.008712769, -0.03668213, -0.036254883, -0.00605011, -0.01789856, -0.027374268, -0.013557434, -0.020996094, 0.05230713, -0.019866943, -0.001156807, 0.014213562, -0.011924744, -0.05380249, -0.017105103, -0.0017614365, 0.0...
http://arxiv.org/abs/2304.08115v1
An Empirical Study of Multitask Learning to Improve Open Domain Dialogue Systems
Mehrdad Farahani, Richard Johansson
cs.CL
Autoregressive models used to generate responses in open-domain dialogue systems often struggle to take long-term context into account and to maintain consistency over a dialogue. Previous research in open-domain dialogue generation has shown that the use of \emph{auxiliary tasks} can introduce inductive biases that en...
http://arxiv.org/pdf/2304.08115v1
2023-04-17T09:44:56
2023-04-17T09:44:56
{ "title": [ 0.011703491, 0.009300232, -0.050048828, -0.059326172, 0.01713562, -0.028533936, -0.041503906, 0.0037345886, -0.023651123, 0.033325195, -0.0017147064, 0.028900146, -0.036224365, -0.02293396, -0.033569336, -0.024841309, -0.016693115, 0...
http://arxiv.org/abs/2304.08109v2
A Comparative Study between Full-Parameter and LoRA-based Fine-Tuning on Chinese Instruction Data for Instruction Following Large Language Model
Xianghui Sun, Yunjie Ji, Baochang Ma, Xiangang Li
cs.CL
Recently, the instruction-tuning of large language models is a crucial area of research in the field of natural language processing. Due to resource and cost limitations, several researchers have employed parameter-efficient tuning techniques, such as LoRA, for instruction tuning, and have obtained encouraging results ...
http://arxiv.org/pdf/2304.08109v2
2023-04-18T03:08:18
2023-04-17T09:36:36
{ "title": [ -0.044067383, -0.035339355, -0.012413025, -0.032043457, -0.054779053, 0.008682251, -0.040283203, 0.014640808, -0.02331543, 0.03656006, -0.00049591064, 0.01159668, -0.015991211, -0.028778076, -0.016326904, -0.019851685, -0.012878418, ...
http://arxiv.org/abs/2304.08103v2
Low-code LLM: Visual Programming over LLMs
Yuzhe Cai, Shaoguang Mao, Wenshan Wu, Zehua Wang, Yaobo Liang, Tao Ge, Chenfei Wu, Wang You, Ting Song, Yan Xia, Jonathan Tien, Nan Duan
cs.CL, cs.HC
Effectively utilizing LLMs for complex tasks is challenging, often involving a time-consuming and uncontrollable prompt engineering process. This paper introduces a novel human-LLM interaction framework, Low-code LLM. It incorporates six types of simple low-code visual programming interactions, all supported by clickin...
http://arxiv.org/pdf/2304.08103v2
2023-04-20T06:42:24
2023-04-17T09:27:40
{ "title": [ -0.054107666, -0.033599854, -0.048583984, -0.06008911, -0.037200928, 0.012969971, -0.066833496, 0.013664246, 0.01739502, 0.018081665, -0.049438477, 0.0037441254, -0.011459351, -0.027389526, -0.022415161, -0.0066375732, 0.0037212372, ...
http://arxiv.org/abs/2304.08085v1
InstructUIE: Multi-task Instruction Tuning for Unified Information Extraction
Xiao Wang, Weikang Zhou, Can Zu, Han Xia, Tianze Chen, Yuansen Zhang, Rui Zheng, Junjie Ye, Qi Zhang, Tao Gui, Jihua Kang, Jingsheng Yang, Siyuan Li, Chunsai Du
cs.CL, cs.AI
Large language models have unlocked strong multi-task capabilities from reading instructive prompts. However, recent studies have shown that existing large models still have difficulty with information extraction tasks. For example, gpt-3.5-turbo achieved an F1 score of 18.22 on the Ontonotes dataset, which is signific...
http://arxiv.org/pdf/2304.08085v1
2023-04-17T09:00:50
2023-04-17T09:00:50
{ "title": [ -0.05593872, -0.019241333, -0.04284668, -0.06567383, -0.019210815, -0.0020370483, -0.057891846, 0.0008339882, -0.0057525635, 0.03781128, -0.02961731, 0.025177002, -0.012466431, 0.0027999878, -0.034332275, -0.0023994446, -0.023361206, ...
http://arxiv.org/abs/2304.11060v2
SkillGPT: a RESTful API service for skill extraction and standardization using a Large Language Model
Nan Li, Bo Kang, Tijl De Bie
cs.CL, cs.AI
We present SkillGPT, a tool for skill extraction and standardization (SES) from free-style job descriptions and user profiles with an open-source Large Language Model (LLM) as backbone. Most previous methods for similar tasks either need supervision or rely on heavy data-preprocessing and feature engineering. Directly ...
http://arxiv.org/pdf/2304.11060v2
2023-10-18T08:38:41
2023-04-17T08:43:20
{ "title": [ -0.013008118, -0.0044288635, -0.06439209, -0.048919678, 0.011459351, -0.024093628, -0.024230957, -0.003873825, 0.0010976791, 0.03643799, 0.0067710876, 0.0026988983, 0.016036987, -0.019943237, -0.035583496, 0.0057868958, -0.019851685, ...
http://arxiv.org/abs/2304.08047v1
ERTIM@MC2: Diversified Argumentative Tweets Retrieval
Kévin Deturck, Parantapa Goswami, Damien Nouvel, Frédérique Segond
cs.CL
In this paper, we present our participation to CLEF MC2 2018 edition for the task 2 Mining opinion argumentation. It consists in detecting the most argumentative and diverse Tweets about some festivals in English and French from a massive multilingual collection. We measure argumentativity of a Tweet computing the amou...
http://arxiv.org/pdf/2304.08047v1
2023-04-17T08:06:17
2023-04-17T08:06:17
{ "title": [ 0.024871826, -0.005306244, -0.013343811, -0.06262207, 0.004333496, -0.0024738312, -0.020309448, -0.005821228, -0.008705139, 0.039764404, -0.031173706, 0.019729614, 0.0039596558, 0.003036499, -0.054626465, -0.0137786865, -0.018371582, ...
http://arxiv.org/abs/2304.07995v1
From Zero to Hero: Examining the Power of Symbolic Tasks in Instruction Tuning
Qian Liu, Fan Zhou, Zhengbao Jiang, Longxu Dou, Min Lin
cs.CL, cs.AI
Fine-tuning language models on tasks with instructions has demonstrated potential in facilitating zero-shot generalization to unseen tasks. In this paper, we introduce a straightforward yet effective method for enhancing instruction tuning by employing symbolic tasks. Compared to crowdsourced human tasks or model-gener...
http://arxiv.org/pdf/2304.07995v1
2023-04-17T05:29:42
2023-04-17T05:29:42
{ "title": [ -0.029449463, -0.011474609, -0.020950317, -0.014038086, -0.016220093, -0.028076172, -0.014160156, -0.0115737915, -0.026245117, 0.04031372, -0.0011825562, 0.013839722, -0.020690918, -0.011123657, -0.04522705, -0.008148193, -0.009712219, ...
http://arxiv.org/abs/2304.10548v1
Supporting Qualitative Analysis with Large Language Models: Combining Codebook with GPT-3 for Deductive Coding
Ziang Xiao, Xingdi Yuan, Q. Vera Liao, Rania Abdelghani, Pierre-Yves Oudeyer
cs.CL, cs.AI, cs.HC
Qualitative analysis of textual contents unpacks rich and valuable information by assigning labels to the data. However, this process is often labor-intensive, particularly when working with large datasets. While recent AI-based tools demonstrate utility, researchers may not have readily available AI resources and expe...
http://arxiv.org/pdf/2304.10548v1
2023-04-17T04:52:43
2023-04-17T04:52:43
{ "title": [ -0.010665894, 0.010383606, -0.04425049, -0.03427124, -0.0048294067, -0.010978699, -0.055419922, -0.03753662, 0.02508545, 0.021743774, -0.036468506, -0.011520386, 0.02154541, -0.01928711, -0.030319214, -0.005317688, -0.026000977, 0.03...
http://arxiv.org/abs/2304.07987v4
Chinese Open Instruction Generalist: A Preliminary Release
Ge Zhang, Yemin Shi, Ruibo Liu, Ruibin Yuan, Yizhi Li, Siwei Dong, Yu Shu, Zhaoqun Li, Zekun Wang, Chenghua Lin, Wenhao Huang, Jie Fu
cs.CL, cs.AI
Instruction tuning is widely recognized as a key technique for building generalist language models, which has attracted the attention of researchers and the public with the release of InstructGPT~\citep{ouyang2022training} and ChatGPT\footnote{\url{https://chat.openai.com/}}. Despite impressive progress in English-orie...
http://arxiv.org/pdf/2304.07987v4
2023-04-25T01:50:19
2023-04-17T04:45:06
{ "title": [ -0.018707275, -0.0012750626, -0.011268616, -0.023773193, -0.016326904, -0.0043792725, -0.024047852, -0.023361206, -0.036224365, 0.03616333, 0.00028848648, 0.016815186, 0.022781372, -0.05996704, 0.031402588, 0.0051231384, 0.0066604614, ...
http://arxiv.org/abs/2304.07957v1
A Question-Answering Approach to Key Value Pair Extraction from Form-like Document Images
Kai Hu, Zhuoyuan Wu, Zhuoyao Zhong, Weihong Lin, Lei Sun, Qiang Huo
cs.CL
In this paper, we present a new question-answering (QA) based key-value pair extraction approach, called KVPFormer, to robustly extracting key-value relationships between entities from form-like document images. Specifically, KVPFormer first identifies key entities from all entities in an image with a Transformer encod...
http://arxiv.org/pdf/2304.07957v1
2023-04-17T02:55:31
2023-04-17T02:55:31
{ "title": [ -0.022323608, 0.0034656525, -0.0032196045, -0.014717102, -0.010635376, -0.0020389557, -0.032562256, -0.029083252, 0.059173584, 0.0077781677, -0.016082764, -0.032318115, -0.017974854, -0.021697998, -0.06890869, -0.016464233, 0.003540039, ...
http://arxiv.org/abs/2304.11085v1
Testing the Reliability of ChatGPT for Text Annotation and Classification: A Cautionary Remark
Michael V. Reiss
cs.CL
Recent studies have demonstrated promising potential of ChatGPT for various text annotation and classification tasks. However, ChatGPT is non-deterministic which means that, as with human coders, identical input can lead to different outputs. Given this, it seems appropriate to test the reliability of ChatGPT. Therefor...
http://arxiv.org/pdf/2304.11085v1
2023-04-17T00:41:19
2023-04-17T00:41:19
{ "title": [ -0.010803223, -0.007888794, -0.005973816, -0.016159058, -0.0051498413, 0.0032482147, -0.017349243, -0.03543091, -0.01889038, 0.028274536, -0.010757446, 0.043670654, -0.036376953, -0.0074501038, -0.05014038, -0.007507324, -0.045928955, ...
http://arxiv.org/abs/2304.07880v4
Sabiá: Portuguese Large Language Models
Ramon Pires, Hugo Abonizio, Thales Sales Almeida, Rodrigo Nogueira
cs.CL, cs.AI
As the capabilities of language models continue to advance, it is conceivable that "one-size-fits-all" model will remain as the main paradigm. For instance, given the vast number of languages worldwide, many of which are low-resource, the prevalent practice is to pretrain a single model on multiple languages. In this p...
http://arxiv.org/pdf/2304.07880v4
2023-11-09T10:36:32
2023-04-16T20:11:19
{ "title": [ -0.055267334, -0.020523071, -0.041992188, -0.06124878, -0.009254456, 0.027572632, -0.07910156, 0.011978149, 0.002773285, 0.038604736, -0.0105896, -0.0054359436, 0.0284729, -0.01953125, -0.033325195, 0.023925781, -0.010955811, 0.04882...
http://arxiv.org/abs/2304.07869v2
Neural Machine Translation For Low Resource Languages
Vakul Goyle, Parvathy Krishnaswamy, Kannan Girija Ravikumar, Utsa Chattopadhyay, Kartikay Goyle
cs.CL, cs.AI
Neural Machine translation is a challenging task due to the inherent complex nature and the fluidity that natural languages bring. Nonetheless, in recent years, it has achieved state-of-the-art performance in several language pairs. Although, a lot of traction can be seen in the areas of multilingual neural machine tra...
http://arxiv.org/pdf/2304.07869v2
2023-04-18T01:36:29
2023-04-16T19:27:48
{ "title": [ -0.032409668, -0.010383606, -0.005794525, -0.05215454, -0.006916046, 0.02079773, -0.047088623, 0.0006246567, -0.009414673, 0.053131104, 0.00032663345, -0.013534546, 0.016082764, -0.020736694, -0.05456543, 0.014671326, -0.0036182404, ...
http://arxiv.org/abs/2304.07854v1
Towards Better Instruction Following Language Models for Chinese: Investigating the Impact of Training Data and Evaluation
Yunjie Ji, Yan Gong, Yong Deng, Yiping Peng, Qiang Niu, Baochang Ma, Xiangang Li
cs.CL
Recently, significant public efforts have been directed towards developing low-cost models with capabilities akin to ChatGPT, thereby fostering the growth of open-source conversational models. However, there remains a scarcity of comprehensive and in-depth evaluations of these models' performance. In this study, we exa...
http://arxiv.org/pdf/2304.07854v1
2023-04-16T18:37:39
2023-04-16T18:37:39
{ "title": [ -0.025512695, 0.00447464, -0.027038574, -0.020889282, -0.005470276, -0.012626648, -0.024719238, -0.018997192, -0.013313293, 0.010658264, -0.010864258, -0.0013618469, -0.0045051575, -0.039520264, -0.022720337, -0.009811401, -0.032562256, ...
http://arxiv.org/abs/2304.07849v3
ChatPLUG: Open-Domain Generative Dialogue System with Internet-Augmented Instruction Tuning for Digital Human
Junfeng Tian, Hehong Chen, Guohai Xu, Ming Yan, Xing Gao, Jianhai Zhang, Chenliang Li, Jiayi Liu, Wenshen Xu, Haiyang Xu, Qi Qian, Wei Wang, Qinghao Ye, Jiejing Zhang, Ji Zhang, Fei Huang, Jingren Zhou
cs.CL
In this paper, we present ChatPLUG, a Chinese open-domain dialogue system for digital human applications that instruction finetunes on a wide range of dialogue tasks in a unified internet-augmented format. Different from other open-domain dialogue models that focus on large-scale pre-training and scaling up model size ...
http://arxiv.org/pdf/2304.07849v3
2023-05-15T16:17:15
2023-04-16T18:16:35
{ "title": [ -0.016159058, -0.01838684, -0.034973145, -0.050567627, -0.018936157, -0.012298584, -0.039916992, -0.006427765, 0.0031719208, 0.024414062, -0.010864258, -0.00018560886, 0.0061569214, -0.025482178, -0.03945923, -0.010231018, -0.02305603, ...
http://arxiv.org/abs/2304.07830v3
The language of sounds unheard: Exploring musical timbre semantics of large language models
Kai Siedenburg, Charalampos Saitis
cs.CL, cs.SD, eess.AS
Semantic dimensions of sound have been playing a central role in understanding the nature of auditory sensory experience as well as the broader relation between perception, language, and meaning. Accordingly, and given the recent proliferation of large language models (LLMs), here we asked whether such models exhibit a...
http://arxiv.org/pdf/2304.07830v3
2023-05-04T10:05:26
2023-04-16T16:50:25
{ "title": [ -0.01838684, -0.018310547, -0.03503418, -0.072631836, -0.015945435, -0.00091838837, -0.07098389, 0.018829346, 0.010925293, 0.032043457, -0.009529114, -0.01121521, 0.026931763, -0.028121948, -0.037963867, 0.014030457, -0.009300232, 0....
http://arxiv.org/abs/2304.07810v2
VISAR: A Human-AI Argumentative Writing Assistant with Visual Programming and Rapid Draft Prototyping
Zheng Zhang, Jie Gao, Ranjodh Singh Dhaliwal, Toby Jia-Jun Li
cs.HC, cs.AI, cs.CL, cs.LG
In argumentative writing, writers must brainstorm hierarchical writing goals, ensure the persuasiveness of their arguments, and revise and organize their plans through drafting. Recent advances in large language models (LLMs) have made interactive text generation through a chat interface (e.g., ChatGPT) possible. Howev...
http://arxiv.org/pdf/2304.07810v2
2023-07-27T20:24:42
2023-04-16T15:29:03
{ "title": [ -0.031021118, -0.030334473, -0.030914307, -0.033172607, -0.046966553, 0.0066375732, -0.03326416, -0.008171082, -0.0016956329, 0.017852783, -0.030685425, 0.0178833, 0.01146698, -0.020706177, -0.025604248, -0.024261475, -0.014808655, 0...
http://arxiv.org/abs/2304.07805v1
EasyNER: A Customizable Easy-to-Use Pipeline for Deep Learning- and Dictionary-based Named Entity Recognition from Medical Text
Rafsan Ahmed, Petter Berntsson, Alexander Skafte, Salma Kazemi Rashed, Marcus Klang, Adam Barvesten, Ola Olde, William Lindholm, Antton Lamarca Arrizabalaga, Pierre Nugues, Sonja Aits
q-bio.QM, cs.CL, 92-04, 92-08, 68T50, J.3; I.2.7; H.3.3
Medical research generates a large number of publications with the PubMed database already containing >35 million research articles. Integration of the knowledge scattered across this large body of literature could provide key insights into physiological mechanisms and disease processes leading to novel medical interve...
http://arxiv.org/pdf/2304.07805v1
2023-04-16T15:17:56
2023-04-16T15:17:56
{ "title": [ -0.021591187, 0.021453857, -0.043029785, -0.06384277, 0.02053833, 0.0071411133, -0.04815674, -0.016464233, -0.004558563, 0.024673462, -0.01878357, 0.034210205, 0.010284424, -0.01411438, -0.006614685, 0.008934021, 0.008262634, 0.04324...
http://arxiv.org/abs/2304.07781v1
It's All in the Embedding! Fake News Detection Using Document Embeddings
Ciprian-Octavian Truică, Elena-Simona Apostol
cs.CL, cs.AI
With the current shift in the mass media landscape from journalistic rigor to social media, personalized social media is becoming the new norm. Although the digitalization progress of the media brings many advantages, it also increases the risk of spreading disinformation, misinformation, and malformation through the u...
http://arxiv.org/pdf/2304.07781v1
2023-04-16T13:30:06
2023-04-16T13:30:06
{ "title": [ 0.0022659302, 0.0135269165, -0.0309906, -0.049468994, 0.026947021, -0.022750854, -0.008270264, -0.029830933, -0.0390625, 0.011329651, -0.028381348, 0.018615723, -0.01902771, -0.020980835, -0.0076828003, -0.025039673, -0.01725769, 0.0...
http://arxiv.org/abs/2304.07778v1
SikuGPT: A Generative Pre-trained Model for Intelligent Information Processing of Ancient Texts from the Perspective of Digital Humanities
Liu Chang, Wang Dongbo, Zhao Zhixiao, Hu Die, Wu Mengcheng, Lin Litao, Shen Si, Li Bin, Liu Jiangfeng, Zhang Hai, Zhao Lianzheng
cs.CL
The rapid advance in artificial intelligence technology has facilitated the prosperity of digital humanities research. Against such backdrop, research methods need to be transformed in the intelligent processing of ancient texts, which is a crucial component of digital humanities research, so as to adapt to new develop...
http://arxiv.org/pdf/2304.07778v1
2023-04-16T13:25:24
2023-04-16T13:25:24
{ "title": [ 0.003862381, 0.0075531006, -0.027572632, -0.019561768, 0.0071907043, -0.020309448, 0.011756897, -0.02218628, 0.047332764, 0.039398193, -0.018539429, 0.011177063, 0.02355957, -0.021362305, -0.059143066, 0.025390625, -0.0074806213, 0.0...
http://arxiv.org/abs/2304.07774v1
Syntactic Complexity Identification, Measurement, and Reduction Through Controlled Syntactic Simplification
Muhammad Salman, Armin Haller, Sergio J. Rodríguez Méndez
cs.CL, cs.IR
Text simplification is one of the domains in Natural Language Processing (NLP) that offers an opportunity to understand the text in a simplified manner for exploration. However, it is always hard to understand and retrieve knowledge from unstructured text, which is usually in the form of compound and complex sentences....
http://arxiv.org/pdf/2304.07774v1
2023-04-16T13:13:58
2023-04-16T13:13:58
{ "title": [ -0.02381897, 0.011062622, -0.04776001, -0.06970215, 0.020507812, -0.0059280396, -0.059143066, -0.0059013367, -0.021759033, 0.040374756, 0.0041542053, 0.021835327, -0.0006580353, -0.03289795, -0.027999878, 0.0071868896, -0.00029468536, ...
http://arxiv.org/abs/2304.07772v2
A Comprehensive Evaluation of the Copy Mechanism for Natural Language to SPARQL Query Generation
Samuel Reyd, Amal Zouaq, Papa Abdou Karim Karou Diallo
cs.CL, cs.LG
In recent years, the field of neural machine translation (NMT) for SPARQL query generation has witnessed a significant growth. Recently, the incorporation of the copy mechanism with traditional encoder-decoder architectures and the use of pre-trained encoder-decoders have set new performance benchmarks. This paper pres...
http://arxiv.org/pdf/2304.07772v2
2023-05-01T13:37:56
2023-04-16T13:12:26
{ "title": [ -0.009017944, -0.015670776, -0.043548584, -0.068237305, 0.009933472, 0.012435913, -0.03604126, 0.013092041, -0.0000056028366, 0.027862549, -0.025009155, 0.009307861, 0.0051727295, -0.017196655, -0.030654907, -0.0028305054, -0.014854431, ...
http://arxiv.org/abs/2304.07759v1
MisRoBÆRTa: Transformers versus Misinformation
Ciprian-Octavian Truică, Elena-Simona Apostol
cs.CL, cs.AI
Misinformation is considered a threat to our democratic values and principles. The spread of such content on social media polarizes society and undermines public discourse by distorting public perceptions and generating social unrest while lacking the rigor of traditional journalism. Transformers and transfer learning ...
http://arxiv.org/pdf/2304.07759v1
2023-04-16T12:14:38
2023-04-16T12:14:38
{ "title": [ 0.05618286, -0.011657715, 0.012107849, -0.02973938, -0.028289795, -0.012840271, 0.015457153, -0.033843994, -0.06286621, 0.035491943, -0.02319336, 0.016021729, -0.009437561, -0.047943115, 0.013412476, -0.035339355, 0.001912117, 0.0726...
http://arxiv.org/abs/2304.09826v2
Fairness in AI and Its Long-Term Implications on Society
Ondrej Bohdal, Timothy Hospedales, Philip H. S. Torr, Fazl Barez
cs.CY, cs.AI, cs.CL, cs.CV, cs.LG
Successful deployment of artificial intelligence (AI) in various settings has led to numerous positive outcomes for individuals and society. However, AI systems have also been shown to harm parts of the population due to biased predictions. AI fairness focuses on mitigating such biases to ensure AI decision making is n...
http://arxiv.org/pdf/2304.09826v2
2023-07-19T19:30:52
2023-04-16T11:22:59
{ "title": [ 0.016708374, 0.01927185, -0.036254883, -0.03540039, 0.008331299, -0.040008545, 0.02357483, -0.0011882782, -0.010299683, 0.029830933, -0.066711426, -0.00033426285, 0.017654419, -0.013435364, -0.008163452, -0.008644104, -0.0006322861, ...
http://arxiv.org/abs/2304.07741v2
Canvas: End-to-End Kernel Architecture Search in Neural Networks
Chenggang Zhao, Genghan Zhang, Mingyu Gao
cs.LG, cs.AI, cs.CL, cs.DC, cs.PF
The demands for higher performance and accuracy in neural networks (NNs) never end. Existing tensor compilation and Neural Architecture Search (NAS) techniques orthogonally optimize the two goals but actually share many similarities in their concrete strategies. We exploit such opportunities by combining the two into o...
http://arxiv.org/pdf/2304.07741v2
2023-04-18T06:02:01
2023-04-16T10:05:42
{ "title": [ -0.033233643, -0.010368347, -0.021850586, -0.023620605, 0.009498596, -0.0026474, -0.0126571655, -0.022857666, 0.011352539, 0.037017822, -0.01109314, -0.00039196014, -0.0035305023, -0.013412476, -0.055358887, 0.011886597, 0.000023841858, ...
http://arxiv.org/abs/2304.07699v1
USNID: A Framework for Unsupervised and Semi-supervised New Intent Discovery
Hanlei Zhang, Hua Xu, Xin Wang, Fei Long, Kai Gao
cs.CL
New intent discovery is of great value to natural language processing, allowing for a better understanding of user needs and providing friendly services. However, most existing methods struggle to capture the complicated semantics of discrete text representations when limited or no prior knowledge of labeled data is av...
http://arxiv.org/pdf/2304.07699v1
2023-04-16T05:30:42
2023-04-16T05:30:42
{ "title": [ -0.011528015, -0.00009083748, -0.05130005, -0.052337646, 0.02229309, -0.016235352, -0.033996582, -0.017166138, -0.019134521, 0.013923645, -0.020751953, 0.03149414, -0.012321472, -0.022079468, -0.0069618225, -0.001203537, -0.0206604, ...
http://arxiv.org/abs/2304.09102v1
Solving Math Word Problems by Combining Language Models With Symbolic Solvers
Joy He-Yueya, Gabriel Poesia, Rose E. Wang, Noah D. Goodman
cs.CL, cs.AI
Automatically generating high-quality step-by-step solutions to math word problems has many applications in education. Recently, combining large language models (LLMs) with external tools to perform complex reasoning and calculation has emerged as a promising direction for solving math word problems, but prior approach...
http://arxiv.org/pdf/2304.09102v1
2023-04-16T04:16:06
2023-04-16T04:16:06
{ "title": [ -0.021591187, -0.00415802, -0.019882202, -0.030639648, -0.013633728, -0.00472641, -0.030593872, -0.021911621, -0.002500534, 0.036132812, -0.017868042, 0.0038223267, 0.0024204254, -0.012145996, -0.025985718, 0.0093307495, 0.002603531, ...
http://arxiv.org/abs/2304.07687v2
MLRegTest: A Benchmark for the Machine Learning of Regular Languages
Sam van der Poel, Dakotah Lambert, Kalina Kostyszyn, Tiantian Gao, Rahul Verma, Derek Andersen, Joanne Chau, Emily Peterson, Cody St. Clair, Paul Fodor, Chihiro Shibata, Jeffrey Heinz
cs.LG, cs.CL, cs.FL
Evaluating machine learning (ML) systems on their ability to learn known classifiers allows fine-grained examination of the patterns they can learn, which builds confidence when they are applied to the learning of unknown classifiers. This article presents a new benchmark for ML systems on sequence classification calle...
http://arxiv.org/pdf/2304.07687v2
2023-11-09T01:29:24
2023-04-16T03:49:50
{ "title": [ -0.03765869, -0.025375366, -0.055541992, -0.09088135, 0.010787964, 0.008392334, -0.05770874, 0.033172607, -0.011260986, 0.016601562, -0.024215698, 0.037841797, -0.029251099, -0.0048713684, -0.025161743, -0.007980347, -0.0026435852, 0...
http://arxiv.org/abs/2304.07666v2
ArguGPT: evaluating, understanding and identifying argumentative essays generated by GPT models
Yikang Liu, Ziyin Zhang, Wanyang Zhang, Shisen Yue, Xiaojing Zhao, Xinyuan Cheng, Yiwen Zhang, Hai Hu
cs.CL
AI generated content (AIGC) presents considerable challenge to educators around the world. Instructors need to be able to detect such text generated by large language models, either with the naked eye or with the help of some tools. There is also growing need to understand the lexical, syntactic and stylistic features ...
http://arxiv.org/pdf/2304.07666v2
2023-09-23T14:05:58
2023-04-16T01:50:26
{ "title": [ 0.019897461, 0.000082314014, -0.009864807, -0.02684021, 0.00091028214, -0.0043907166, -0.0066184998, -0.054870605, -0.029129028, -0.012832642, -0.0012044907, 0.004611969, -0.033599854, -0.03338623, -0.016845703, -0.030029297, -0.02966308...
http://arxiv.org/abs/2304.07637v1
TransDocs: Optical Character Recognition with word to word translation
Abhishek Bamotra, Phani Krishna Uppala
cs.CV, cs.CL, cs.LG
While OCR has been used in various applications, its output is not always accurate, leading to misfit words. This research work focuses on improving the optical character recognition (OCR) with ML techniques with integration of OCR with long short-term memory (LSTM) based sequence to sequence deep learning models to pe...
http://arxiv.org/pdf/2304.07637v1
2023-04-15T21:40:14
2023-04-15T21:40:14
{ "title": [ -0.060333252, -0.020767212, -0.0076560974, -0.024612427, -0.025680542, 0.014335632, -0.03149414, -0.018737793, 0.0061569214, 0.04196167, -0.040618896, 0.014389038, 0.020706177, -0.0072021484, -0.04171753, 0.0009236336, 0.00907135, 0....
http://arxiv.org/abs/2304.07633v1
Detecting Out-of-Context Multimodal Misinformation with interpretable neural-symbolic model
Yizhou Zhang, Loc Trinh, Defu Cao, Zijun Cui, Yan Liu
cs.CL, cs.LG
Recent years have witnessed the sustained evolution of misinformation that aims at manipulating public opinions. Unlike traditional rumors or fake news editors who mainly rely on generated and/or counterfeited images, text and videos, current misinformation creators now more tend to use out-of-context multimedia conten...
http://arxiv.org/pdf/2304.07633v1
2023-04-15T21:11:55
2023-04-15T21:11:55
{ "title": [ 0.028503418, 0.032958984, -0.019180298, -0.031707764, 0.014526367, -0.016021729, -0.012275696, -0.036193848, -0.03918457, 0.03845215, 0.004673004, 0.0027656555, 0.0027103424, -0.029891968, -0.022354126, -0.028076172, -0.031082153, 0....
http://arxiv.org/abs/2305.01759v1
Evaluation of Speaker Anonymization on Emotional Speech
Hubert Nourtel, Pierre Champion, Denis Jouvet, Anthony Larcher, Marie Tahon
eess.AS, cs.AI, cs.CL
Speech data carries a range of personal information, such as the speaker's identity and emotional state. These attributes can be used for malicious purposes. With the development of virtual assistants, a new generation of privacy threats has emerged. Current studies have addressed the topic of preserving speech privacy...
http://arxiv.org/pdf/2305.01759v1
2023-04-15T20:50:29
2023-04-15T20:50:29
{ "title": [ 0.01864624, 0.018249512, -0.03829956, -0.0574646, 0.010116577, -0.023483276, -0.0234375, -0.007118225, -0.02708435, 0.040374756, -0.0013418198, 0.007129669, 0.014770508, -0.035125732, -0.014930725, 0.004573822, -0.019058228, 0.032501...
http://arxiv.org/abs/2307.02518v1
Analyzing the Performance of ChatGPT in Cardiology and Vascular Pathologies
Walid Hariri
cs.CL, cs.CY
The article aims to analyze the performance of ChatGPT, a large language model developed by OpenAI, in the context of cardiology and vascular pathologies. The study evaluated the accuracy of ChatGPT in answering challenging multiple-choice questions (QCM) using a dataset of 190 questions from the Siamois-QCM platform. ...
http://arxiv.org/pdf/2307.02518v1
2023-04-15T20:08:48
2023-04-15T20:08:48
{ "title": [ 0.008590698, -0.005874634, -0.0013198853, -0.012870789, -0.0025119781, -0.009735107, -0.008575439, -0.029937744, -0.024551392, 0.032348633, -0.00819397, 0.030853271, 0.007972717, -0.02748108, -0.015037537, 0.01361084, -0.0062828064, ...
http://arxiv.org/abs/2304.07625v1
Neural Approaches to Entity-Centric Information Extraction
Klim Zaporojets
cs.CL
Artificial Intelligence (AI) has huge impact on our daily lives with applications such as voice assistants, facial recognition, chatbots, autonomously driving cars, etc. Natural Language Processing (NLP) is a cross-discipline of AI and Linguistics, dedicated to study the understanding of the text. This is a very challe...
http://arxiv.org/pdf/2304.07625v1
2023-04-15T20:07:37
2023-04-15T20:07:37
{ "title": [ -0.024536133, 0.010566711, -0.036315918, -0.050720215, 0.018356323, -0.0045700073, -0.009559631, 0.007080078, 0.0017337799, 0.04660034, -0.013748169, 0.007221222, 0.007965088, -0.0027618408, -0.047058105, 0.010978699, -0.01537323, 0....
http://arxiv.org/abs/2304.07619v4
Can ChatGPT Forecast Stock Price Movements? Return Predictability and Large Language Models
Alejandro Lopez-Lira, Yuehua Tang
q-fin.ST, cs.CL
We examine the potential of ChatGPT and other large language models in predicting stock market returns using news headlines. We use ChatGPT to assess whether each headline is good, bad, or neutral for firms' stock prices. We document a significantly positive correlation between ChatGPT scores and subsequent daily stock...
http://arxiv.org/pdf/2304.07619v4
2023-09-09T15:20:54
2023-04-15T19:22:37
{ "title": [ -0.029922485, -0.005596161, -0.02848816, 0.0012712479, -0.021072388, -0.0024166107, -0.027618408, -0.021713257, -0.020446777, 0.044281006, -0.0024089813, -0.011749268, 0.007499695, -0.014625549, -0.039733887, -0.012367249, -0.0067710876,...
http://arxiv.org/abs/2304.07611v1
A CTC Alignment-based Non-autoregressive Transformer for End-to-end Automatic Speech Recognition
Ruchao Fan, Wei Chu, Peng Chang, Abeer Alwan
cs.CL, eess.AS
Recently, end-to-end models have been widely used in automatic speech recognition (ASR) systems. Two of the most representative approaches are connectionist temporal classification (CTC) and attention-based encoder-decoder (AED) models. Autoregressive transformers, variants of AED, adopt an autoregressive mechanism for...
http://arxiv.org/pdf/2304.07611v1
2023-04-15T18:34:29
2023-04-15T18:34:29
{ "title": [ -0.021514893, 0.0103302, -0.06951904, -0.067993164, 0.0003130436, -0.004196167, -0.06921387, 0.016998291, 0.0034732819, 0.060638428, -0.04031372, 0.010391235, 0.029937744, 0.0007829666, -0.06488037, 0.003894806, -0.0070228577, 0.0228...
http://arxiv.org/abs/2304.07499v1
Robust Educational Dialogue Act Classifiers with Low-Resource and Imbalanced Datasets
Jionghao Lin, Wei Tan, Ngoc Dang Nguyen, David Lang, Lan Du, Wray Buntine, Richard Beare, Guanliang Chen, Dragan Gasevic
cs.CL, cs.AI, cs.HC, cs.LG
Dialogue acts (DAs) can represent conversational actions of tutors or students that take place during tutoring dialogues. Automating the identification of DAs in tutoring dialogues is significant to the design of dialogue-based intelligent tutoring systems. Many prior studies employ machine learning models to classify ...
http://arxiv.org/pdf/2304.07499v1
2023-04-15T08:01:59
2023-04-15T08:01:59
{ "title": [ 0.0059051514, 0.017959595, -0.043670654, -0.012519836, 0.022277832, -0.03375244, -0.00016236305, -0.03756714, 0.017654419, 0.009559631, -0.022338867, 0.0077781677, -0.039855957, -0.020553589, -0.05142212, -0.033233643, -0.030670166, ...
http://arxiv.org/abs/2304.07438v3
Tractable Control for Autoregressive Language Generation
Honghua Zhang, Meihua Dang, Nanyun Peng, Guy Van den Broeck
cs.CL, cs.AI
Despite the success of autoregressive large language models in text generation, it remains a major challenge to generate text that satisfies complex constraints: sampling from the conditional distribution ${\Pr}(\text{text} | \alpha)$ is intractable for even the simplest lexical constraints $\alpha$. To overcome this c...
http://arxiv.org/pdf/2304.07438v3
2023-06-07T08:23:55
2023-04-15T00:19:44
{ "title": [ -0.03274536, -0.010879517, -0.050628662, -0.06890869, 0.0021305084, 0.011650085, -0.074645996, 0.01576233, -0.0184021, 0.026870728, -0.0138549805, 0.022888184, -0.019683838, -0.009170532, -0.046722412, -0.0040245056, 0.0010499954, 0....
http://arxiv.org/abs/2304.07437v1
Medical Question Summarization with Entity-driven Contrastive Learning
Sibo Wei, Wenpeng Lu, Xueping Peng, Shoujin Wang, Yi-Fei Wang, Weiyu Zhang
cs.CL, cs.AI
By summarizing longer consumer health questions into shorter and essential ones, medical question answering (MQA) systems can more accurately understand consumer intentions and retrieve suitable answers. However, medical question summarization is very challenging due to obvious distinctions in health trouble descriptio...
http://arxiv.org/pdf/2304.07437v1
2023-04-15T00:19:03
2023-04-15T00:19:03
{ "title": [ 0.015205383, 0.01965332, -0.014846802, -0.06719971, 0.013572693, 0.0016508102, -0.020141602, -0.022766113, 0.012832642, 0.039978027, -0.00053453445, 0.02482605, -0.010925293, -0.012649536, -0.05645752, 0.01411438, 0.0024299622, 0.017...
http://arxiv.org/abs/2304.07396v2
Improving Patient Pre-screening for Clinical Trials: Assisting Physicians with Large Language Models
Danny M. den Hamer, Perry Schoor, Tobias B. Polak, Daniel Kapitan
cs.LG, cs.AI, cs.CL, I.2.1; I.2.7
Physicians considering clinical trials for their patients are met with the laborious process of checking many text based eligibility criteria. Large Language Models (LLMs) have shown to perform well for clinical information extraction and clinical reasoning, including medical tests, but not yet in real-world scenarios....
http://arxiv.org/pdf/2304.07396v2
2023-06-29T12:59:16
2023-04-14T21:19:46
{ "title": [ -0.014328003, 0.0057411194, -0.019638062, -0.050750732, -0.0068893433, 0.0010795593, -0.053863525, -0.033813477, -0.020187378, 0.025314331, -0.021865845, 0.011795044, -0.018997192, -0.022888184, -0.01914978, -0.021850586, 0.0066070557, ...
http://arxiv.org/abs/2304.07382v1
Zero-Shot Multi-Label Topic Inference with Sentence Encoders
Souvika Sarkar, Dongji Feng, Shubhra Kanti Karmaker Santu
cs.CL, cs.IR
Sentence encoders have indeed been shown to achieve superior performances for many downstream text-mining tasks and, thus, claimed to be fairly general. Inspired by this, we performed a detailed study on how to leverage these sentence encoders for the "zero-shot topic inference" task, where the topics are defined/provi...
http://arxiv.org/pdf/2304.07382v1
2023-04-14T20:27:09
2023-04-14T20:27:09
{ "title": [ -0.009254456, -0.004558563, -0.03555298, -0.060943604, -0.006816864, -0.004459381, -0.03967285, 0.002500534, 0.024978638, 0.025283813, -0.013717651, 0.0385437, -0.013954163, -0.004501343, -0.01914978, 0.028289795, -0.009559631, 0.039...
http://arxiv.org/abs/2304.07333v1
The Self-Perception and Political Biases of ChatGPT
Jérôme Rutinowski, Sven Franke, Jan Endendyk, Ina Dormuth, Markus Pauly
cs.CY, cs.AI, cs.CL, cs.HC
This contribution analyzes the self-perception and political biases of OpenAI's Large Language Model ChatGPT. Taking into account the first small-scale reports and studies that have emerged, claiming that ChatGPT is politically biased towards progressive and libertarian points of view, this contribution aims to provide...
http://arxiv.org/pdf/2304.07333v1
2023-04-14T18:06:13
2023-04-14T18:06:13
{ "title": [ 0.015159607, 0.0056762695, -0.023483276, -0.006515503, -0.022415161, -0.036346436, 0.016021729, -0.018005371, -0.03363037, 0.039093018, -0.030014038, 0.019256592, -0.027496338, -0.023605347, -0.00063848495, -0.0037155151, -0.01739502, ...
http://arxiv.org/abs/2304.07327v2
OpenAssistant Conversations -- Democratizing Large Language Model Alignment
Andreas Köpf, Yannic Kilcher, Dimitri von Rütte, Sotiris Anagnostidis, Zhi-Rui Tam, Keith Stevens, Abdullah Barhoum, Nguyen Minh Duc, Oliver Stanley, Richárd Nagyfi, Shahul ES, Sameer Suri, David Glushkov, Arnav Dantuluri, Andrew Maguire, Christoph Schuhmann, Huu Nguyen, Alexander Mattick
cs.CL, cs.AI, I.2
Aligning large language models (LLMs) with human preferences has proven to drastically improve usability and has driven rapid adoption as demonstrated by ChatGPT. Alignment techniques such as supervised fine-tuning (SFT) and reinforcement learning from human feedback (RLHF) greatly reduce the required skill and domain ...
http://arxiv.org/pdf/2304.07327v2
2023-10-31T11:38:07
2023-04-14T18:01:29
{ "title": [ -0.034118652, -0.00065135956, -0.051849365, -0.0680542, 0.005821228, 0.00919342, -0.07574463, -0.01259613, -0.008590698, 0.041015625, 0.013145447, 0.0049095154, 0.0055999756, -0.027633667, -0.0335083, 0.010513306, -0.0023517609, 0.00...
http://arxiv.org/abs/2304.07258v2
Learn What Is Possible, Then Choose What Is Best: Disentangling One-To-Many Relations in Language Through Text-based Games
Benjamin Towle, Ke Zhou
cs.CL, cs.AI, cs.LG
Language models pre-trained on large self-supervised corpora, followed by task-specific fine-tuning has become the dominant paradigm in NLP. These pre-training datasets often have a one-to-many structure--e.g. in dialogue there are many valid responses for a given context. However, only some of these responses will be ...
http://arxiv.org/pdf/2304.07258v2
2023-04-26T10:35:13
2023-04-14T17:11:26
{ "title": [ 0.020721436, -0.017959595, -0.017623901, -0.04864502, -0.032073975, -0.014099121, -0.023773193, 0.0039482117, 0.019744873, -0.008575439, -0.035980225, 0.012207031, -0.016586304, -0.033203125, 0.0047569275, -0.054748535, -0.007549286, ...
http://arxiv.org/abs/2304.07242v1
Covidia: COVID-19 Interdisciplinary Academic Knowledge Graph
Cheng Deng, Jiaxin Ding, Luoyi Fu, Weinan Zhang, Xinbing Wang, Chenghu Zhou
cs.IR, cs.CL, H.3.3; I.2.7
The pandemic of COVID-19 has inspired extensive works across different research fields. Existing literature and knowledge platforms on COVID-19 only focus on collecting papers on biology and medicine, neglecting the interdisciplinary efforts, which hurdles knowledge sharing and research collaborations between fields to...
http://arxiv.org/pdf/2304.07242v1
2023-04-14T16:45:38
2023-04-14T16:45:38
{ "title": [ -0.008331299, -0.0005235672, -0.06951904, -0.052001953, -0.012145996, -0.026168823, -0.041168213, 0.01852417, -0.06286621, 0.038360596, -0.02444458, 0.011459351, 0.039215088, -0.03741455, 0.01537323, -0.027297974, 0.022781372, 0.0329...
http://arxiv.org/abs/2304.07235v1
Optimal inference of a generalised Potts model by single-layer transformers with factored attention
Riccardo Rende, Federica Gerace, Alessandro Laio, Sebastian Goldt
cond-mat.dis-nn, cond-mat.stat-mech, cs.CL, stat.ML
Transformers are the type of neural networks that has revolutionised natural language processing and protein science. Their key building block is a mechanism called self-attention which is trained to predict missing words in sentences. Despite the practical success of transformers in applications it remains unclear wha...
http://arxiv.org/pdf/2304.07235v1
2023-04-14T16:32:56
2023-04-14T16:32:56
{ "title": [ 0.010818481, 0.0043907166, -0.013755798, -0.057769775, -0.010772705, 0.011695862, -0.04196167, -0.017486572, 0.00095796585, 0.020050049, 0.0012178421, 0.01687622, -0.01927185, -0.0256958, -0.032165527, 0.008880615, -0.00762558, 0.049...
http://arxiv.org/abs/2304.09103v1
ChatGPT: Applications, Opportunities, and Threats
Aram Bahrini, Mohammadsadra Khamoshifar, Hossein Abbasimehr, Robert J. Riggs, Maryam Esmaeili, Rastin Mastali Majdabadkohne, Morteza Pasehvar
cs.CY, cs.CL
Developed by OpenAI, ChatGPT (Conditional Generative Pre-trained Transformer) is an artificial intelligence technology that is fine-tuned using supervised machine learning and reinforcement learning techniques, allowing a computer to generate natural language conversation fully autonomously. ChatGPT is built on the tra...
http://arxiv.org/pdf/2304.09103v1
2023-04-14T16:25:03
2023-04-14T16:25:03
{ "title": [ -0.0021915436, -0.01512146, -0.012886047, -0.027511597, 0.007331848, 0.0065956116, -0.0029621124, -0.023391724, -0.023757935, 0.018951416, 0.0018320084, 0.02267456, -0.026504517, -0.03201294, -0.02154541, 0.014724731, -0.021560669, 0...
http://arxiv.org/abs/2304.07183v1
Just Tell Me: Prompt Engineering in Business Process Management
Kiran Busch, Alexander Rochlitzer, Diana Sola, Henrik Leopold
cs.AI, cs.CL, cs.LG
GPT-3 and several other language models (LMs) can effectively address various natural language processing (NLP) tasks, including machine translation and text summarization. Recently, they have also been successfully employed in the business process management (BPM) domain, e.g., for predictive process monitoring and pr...
http://arxiv.org/pdf/2304.07183v1
2023-04-14T14:55:19
2023-04-14T14:55:19
{ "title": [ -0.03466797, -0.0076942444, -0.06011963, -0.03387451, 0.0060653687, -0.016082764, -0.039154053, 0.005619049, -0.024261475, 0.011932373, -0.016174316, -0.010055542, -0.014060974, -0.045776367, -0.01927185, -0.009490967, -0.0004196167, ...
http://arxiv.org/abs/2304.08244v2
API-Bank: A Comprehensive Benchmark for Tool-Augmented LLMs
Minghao Li, Yingxiu Zhao, Bowen Yu, Feifan Song, Hangyu Li, Haiyang Yu, Zhoujun Li, Fei Huang, Yongbin Li
cs.CL, cs.AI
Recent research has demonstrated that Large Language Models (LLMs) can enhance their capabilities by utilizing external tools. However, three pivotal questions remain unanswered: (1) How effective are current LLMs in utilizing tools? (2) How can we enhance LLMs' ability to utilize tools? (3) What obstacles need to be o...
http://arxiv.org/pdf/2304.08244v2
2023-10-25T06:54:12
2023-04-14T14:05:32
{ "title": [ -0.08062744, -0.025909424, -0.06274414, -0.0635376, -0.034362793, 0.02243042, -0.09655762, 0.031280518, 0.0015878677, 0.038238525, -0.036712646, 0.026763916, 0.0048179626, -0.010131836, -0.028121948, 0.004306793, -0.01574707, 0.00590...
http://arxiv.org/abs/2304.07130v1
OPI at SemEval 2023 Task 9: A Simple But Effective Approach to Multilingual Tweet Intimacy Analysis
Sławomir Dadas
cs.CL
This paper describes our submission to the SemEval 2023 multilingual tweet intimacy analysis shared task. The goal of the task was to assess the level of intimacy of Twitter posts in ten languages. The proposed approach consists of several steps. First, we perform in-domain pre-training to create a language model adapt...
http://arxiv.org/pdf/2304.07130v1
2023-04-14T13:49:28
2023-04-14T13:49:28
{ "title": [ 0.006187439, -0.027526855, 0.008361816, -0.032440186, -0.022445679, 0.013328552, -0.021347046, -0.010749817, 0.010231018, 0.039611816, -0.016983032, 0.0357666, 0.023910522, -0.02128601, 0.0029029846, 0.022338867, 0.0009069443, 0.0162...
http://arxiv.org/abs/2304.07127v1
OPI at SemEval 2023 Task 1: Image-Text Embeddings and Multimodal Information Retrieval for Visual Word Sense Disambiguation
Sławomir Dadas
cs.CL
The goal of visual word sense disambiguation is to find the image that best matches the provided description of the word's meaning. It is a challenging problem, requiring approaches that combine language and image understanding. In this paper, we present our submission to SemEval 2023 visual word sense disambiguation s...
http://arxiv.org/pdf/2304.07127v1
2023-04-14T13:45:59
2023-04-14T13:45:59
{ "title": [ 0.010032654, 0.0038337708, -0.023239136, -0.018569946, -0.015075684, -0.011878967, -0.012321472, -0.02130127, 0.042541504, 0.036254883, -0.062164307, 0.028442383, 0.011245728, 0.010688782, -0.04550171, 0.00009840727, -0.0035171509, 0...
http://arxiv.org/abs/2304.07125v1
Keeping the Questions Conversational: Using Structured Representations to Resolve Dependency in Conversational Question Answering
Munazza Zaib, Quan Z. Sheng, Wei Emma Zhang, Adnan Mahmood
cs.CL, cs.IR
Having an intelligent dialogue agent that can engage in conversational question answering (ConvQA) is now no longer limited to Sci-Fi movies only and has, in fact, turned into a reality. These intelligent agents are required to understand and correctly interpret the sequential turns provided as the context of the given...
http://arxiv.org/pdf/2304.07125v1
2023-04-14T13:42:32
2023-04-14T13:42:32
{ "title": [ 0.0041275024, -0.0019035339, -0.033355713, -0.08532715, -0.004245758, 0.0041923523, -0.062469482, 0.009750366, 0.016296387, 0.018997192, -0.021972656, 0.005870819, -0.015281677, -0.010299683, -0.054351807, -0.0060157776, 0.0008211136, ...
http://arxiv.org/abs/2304.07101v1
Task-oriented Document-Grounded Dialog Systems by HLTPR@RWTH for DSTC9 and DSTC10
David Thulke, Nico Daheim, Christian Dugast, Hermann Ney
cs.CL, cs.AI, cs.LG
This paper summarizes our contributions to the document-grounded dialog tasks at the 9th and 10th Dialog System Technology Challenges (DSTC9 and DSTC10). In both iterations the task consists of three subtasks: first detect whether the current turn is knowledge seeking, second select a relevant knowledge document, and t...
http://arxiv.org/pdf/2304.07101v1
2023-04-14T12:46:29
2023-04-14T12:46:29
{ "title": [ 0.008255005, -0.000012099743, -0.05810547, -0.0574646, -0.010253906, -0.030273438, -0.04360962, 0.007156372, 0.00642395, 0.049072266, -0.03488159, -0.030792236, 0.051849365, -0.030319214, -0.018920898, -0.021255493, -0.00064229965, -...
http://arxiv.org/abs/2304.07065v1
SEA: A Scalable Entity Alignment System
Junyang Wu, Tianyi Li, Lu Chen, Yunjun Gao, Ziheng Wei
cs.CL, cs.AI, cs.IR
Entity alignment (EA) aims to find equivalent entities in different knowledge graphs (KGs). State-of-the-art EA approaches generally use Graph Neural Networks (GNNs) to encode entities. However, most of them train the models and evaluate the results in a fullbatch fashion, which prohibits EA from being scalable on larg...
http://arxiv.org/pdf/2304.07065v1
2023-04-14T11:39:07
2023-04-14T11:39:07
{ "title": [ -0.023834229, -0.012924194, -0.068603516, -0.09484863, 0.040283203, 0.005012512, -0.041992188, 0.0093307495, 0.023727417, 0.034820557, -0.0040130615, -0.007545471, 0.037017822, -0.020004272, -0.043823242, 0.018798828, -0.020614624, 0...
http://arxiv.org/abs/2304.08247v2
MedAlpaca -- An Open-Source Collection of Medical Conversational AI Models and Training Data
Tianyu Han, Lisa C. Adams, Jens-Michalis Papaioannou, Paul Grundmann, Tom Oberhauser, Alexander Löser, Daniel Truhn, Keno K. Bressem
cs.CL, cs.AI
As large language models (LLMs) like OpenAI's GPT series continue to make strides, we witness the emergence of artificial intelligence applications in an ever-expanding range of fields. In medicine, these LLMs hold considerable promise for improving medical workflows, diagnostics, patient care, and education. Yet, ther...
http://arxiv.org/pdf/2304.08247v2
2023-10-04T23:28:00
2023-04-14T11:28:08
{ "title": [ -0.008003235, -0.009628296, -0.05105591, -0.0076675415, -0.007713318, -0.028366089, -0.0053596497, -0.0028572083, -0.001958847, 0.05633545, 0.00092697144, -0.006198883, 0.067993164, -0.04232788, 0.013069153, 0.007843018, 0.01361084, ...
http://arxiv.org/abs/2304.07022v1
Label Dependencies-aware Set Prediction Networks for Multi-label Text Classification
Han Quanjie, Du Xinkai, Sun Yalin, Lv Chao
cs.CL
Multi-label text classification aims to extract all the related labels from a sentence, which can be viewed as a sequence generation problem. However, the labels in training dataset are unordered. We propose to treat it as a direct set prediction problem and don't need to consider the order of labels. Besides, in order...
http://arxiv.org/pdf/2304.07022v1
2023-04-14T09:31:17
2023-04-14T09:31:17
{ "title": [ -0.022415161, 0.00012373924, -0.051116943, -0.06591797, 0.005809784, -0.0008735657, -0.042785645, 0.013023376, -0.0020446777, 0.029281616, -0.010025024, 0.023498535, -0.021148682, -0.0022506714, -0.049865723, -0.0053901672, -0.02003479, ...
http://arxiv.org/abs/2304.07007v1
Dialogue Games for Benchmarking Language Understanding: Motivation, Taxonomy, Strategy
David Schlangen
cs.CL
How does one measure "ability to understand language"? If it is a person's ability that is being measured, this is a question that almost never poses itself in an unqualified manner: Whatever formal test is applied, it takes place on the background of the person's language use in daily social practice, and what is meas...
http://arxiv.org/pdf/2304.07007v1
2023-04-14T09:11:36
2023-04-14T09:11:36
{ "title": [ -0.035858154, -0.012329102, -0.02494812, -0.042510986, -0.017700195, -0.0039405823, -0.037078857, 0.0008330345, -0.022735596, 0.0041618347, 0.010513306, 0.013969421, -0.05606079, -0.034301758, -0.037719727, -0.017944336, -0.014808655, ...
http://arxiv.org/abs/2304.07002v1
SimpLex: a lexical text simplification architecture
Ciprian-Octavian Truică, Andrei-Ionut Stan, Elena-Simona Apostol
cs.CL, cs.AI
Text simplification (TS) is the process of generating easy-to-understand sentences from a given sentence or piece of text. The aim of TS is to reduce both the lexical (which refers to vocabulary complexity and meaning) and syntactic (which refers to the sentence structure) complexity of a given text or sentence without...
http://arxiv.org/pdf/2304.07002v1
2023-04-14T08:52:31
2023-04-14T08:52:31
{ "title": [ -0.04144287, -0.014541626, -0.040252686, -0.07299805, 0.012565613, 0.023391724, -0.062194824, -0.006549835, 0.0154800415, 0.061187744, -0.008781433, 0.027160645, 0.011474609, 0.0046920776, -0.06341553, 0.020477295, -0.00058078766, 0....
http://arxiv.org/abs/2304.06975v1
HuaTuo: Tuning LLaMA Model with Chinese Medical Knowledge
Haochun Wang, Chi Liu, Nuwa Xi, Zewen Qiang, Sendong Zhao, Bing Qin, Ting Liu
cs.CL
Large Language Models (LLMs), such as the LLaMA model, have demonstrated their effectiveness in various general-domain natural language processing (NLP) tasks. Nevertheless, LLMs have not yet performed optimally in biomedical domain tasks due to the need for medical expertise in the responses. In response to this chall...
http://arxiv.org/pdf/2304.06975v1
2023-04-14T07:54:17
2023-04-14T07:54:17
{ "title": [ -0.038360596, -0.015342712, -0.041625977, -0.0041160583, -0.025115967, -0.010734558, -0.047424316, -0.044158936, -0.02368164, 0.035980225, 0.03062439, 0.0024280548, 0.015548706, -0.05911255, -0.004470825, -0.00013625622, 0.0046577454, ...
http://arxiv.org/abs/2304.06962v1
Prompt Engineering and Calibration for Zero-Shot Commonsense Reasoning
Chenkai Ma
cs.CL, cs.AI
Prompt engineering and calibration make large language models excel at reasoning tasks, including multiple choice commonsense reasoning. From a practical perspective, we investigate and evaluate these strategies on smaller language models. Through experiments on five commonsense reasoning benchmarks, we find that each ...
http://arxiv.org/pdf/2304.06962v1
2023-04-14T07:07:42
2023-04-14T07:07:42
{ "title": [ 0.0059776306, -0.037475586, -0.026657104, -0.038208008, -0.021865845, -0.018859863, -0.007522583, 0.00096702576, 0.03677368, 0.00459671, -0.015464783, -0.012580872, -0.03463745, -0.041503906, -0.032104492, 0.004459381, -0.003396988, ...
http://arxiv.org/abs/2304.08242v1
The Deep Latent Position Topic Model for Clustering and Representation of Networks with Textual Edges
Rémi Boutin, Pierre Latouche, Charles Bouveyron
cs.LG, cs.CL, cs.SI, stat.ME
Numerical interactions leading to users sharing textual content published by others are naturally represented by a network where the individuals are associated with the nodes and the exchanged texts with the edges. To understand those heterogeneous and complex data structures, clustering nodes into homogeneous groups a...
http://arxiv.org/pdf/2304.08242v1
2023-04-14T07:01:57
2023-04-14T07:01:57
{ "title": [ -0.025100708, 0.0036945343, -0.04272461, -0.033691406, 0.012237549, 0.0022182465, -0.026306152, -0.01878357, 0.006549835, 0.025619507, -0.0023670197, 0.039001465, -0.017227173, -0.020599365, -0.011146545, 0.018630981, -0.009895325, 0...
http://arxiv.org/abs/2304.06939v3
Multimodal C4: An Open, Billion-scale Corpus of Images Interleaved with Text
Wanrong Zhu, Jack Hessel, Anas Awadalla, Samir Yitzhak Gadre, Jesse Dodge, Alex Fang, Youngjae Yu, Ludwig Schmidt, William Yang Wang, Yejin Choi
cs.CV, cs.CL
In-context vision and language models like Flamingo support arbitrarily interleaved sequences of images and text as input. This format not only enables few-shot learning via interleaving independent supervised (image, text) examples, but also, more complex prompts involving interaction between images, e.g., "What do im...
http://arxiv.org/pdf/2304.06939v3
2023-10-28T04:19:41
2023-04-14T06:17:46
{ "title": [ -0.021057129, 0.009277344, -0.066711426, -0.047088623, 0.010856628, -0.023986816, -0.03186035, -0.024398804, 0.0028572083, 0.041168213, -0.042541504, 0.02973938, 0.000106692314, -0.008728027, -0.013908386, -0.002521515, 0.002664566, ...
http://arxiv.org/abs/2304.06910v1
HCAM -- Hierarchical Cross Attention Model for Multi-modal Emotion Recognition
Soumya Dutta, Sriram Ganapathy
eess.AS, cs.CL, cs.SD
Emotion recognition in conversations is challenging due to the multi-modal nature of the emotion expression. We propose a hierarchical cross-attention model (HCAM) approach to multi-modal emotion recognition using a combination of recurrent and co-attention neural network models. The input to the model consists of two ...
http://arxiv.org/pdf/2304.06910v1
2023-04-14T03:25:00
2023-04-14T03:25:00
{ "title": [ -0.008369446, 0.010131836, -0.045135498, -0.04940796, 0.009544373, -0.015052795, -0.035827637, -0.01902771, 0.029418945, 0.048217773, -0.020111084, -0.013267517, 0.046051025, -0.020492554, -0.053253174, 0.017181396, -0.013244629, 0.0...
http://arxiv.org/abs/2304.06875v3
Research without Re-search: Maximal Update Parametrization Yields Accurate Loss Prediction across Scales
Yiqun Yao, Yequan Wang
cs.CL, cs.LG
As language models scale up, it becomes increasingly expensive to verify research ideas because conclusions on small models do not trivially transfer to large ones. A possible solution is to establish a generic system that directly predicts some metrics for large models solely based on the results and hyperparameters f...
http://arxiv.org/pdf/2304.06875v3
2023-09-03T06:55:28
2023-04-14T00:45:01
{ "title": [ -0.0181427, -0.009124756, 0.003080368, -0.04748535, -0.04159546, 0.01979065, -0.08325195, -0.032348633, -0.026473999, 0.039886475, 0.0234375, 0.007762909, -0.0090408325, -0.034423828, -0.034210205, 0.0025806427, 0.035369873, 0.008987...
http://arxiv.org/abs/2304.08243v1
Stochastic Code Generation
Swapnil Sharma, Nikita Anand, Kranthi Kiran G. V
cs.CL, cs.LG
Large language models pre-trained for code generation can generate high-quality short code but often struggle with generating coherent long code and understanding higher-level or system-level specifications. This issue is also observed in language modeling for long text generation, and one proposed solution is the use ...
http://arxiv.org/pdf/2304.08243v1
2023-04-14T00:01:05
2023-04-14T00:01:05
{ "title": [ -0.055114746, -0.012519836, -0.056396484, -0.04562378, -0.029693604, -0.0059432983, -0.07104492, 0.022384644, -0.019088745, 0.026977539, -0.015777588, 0.018051147, -0.028182983, -0.042266846, 0.025146484, 0.015014648, 0.01737976, 0.0...