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/2306.01016v1
PV2TEA: Patching Visual Modality to Textual-Established Information Extraction
Hejie Cui, Rongmei Lin, Nasser Zalmout, Chenwei Zhang, Jingbo Shang, Carl Yang, Xian Li
cs.CL, cs.AI, cs.CV, cs.LG, cs.MM
Information extraction, e.g., attribute value extraction, has been extensively studied and formulated based only on text. However, many attributes can benefit from image-based extraction, like color, shape, pattern, among others. The visual modality has long been underutilized, mainly due to multimodal annotation diffi...
http://arxiv.org/pdf/2306.01016v1
2023-06-01T05:39:45
2023-06-01T05:39:45
{ "title": [ -0.037017822, -0.010917664, -0.04626465, -0.03277588, -0.012878418, -0.002374649, -0.03152466, -0.026992798, 0.01864624, 0.030899048, -0.039764404, 0.018218994, 0.0070343018, 0.001367569, -0.05105591, 0.006652832, -0.010192871, 0.029...
http://arxiv.org/abs/2306.00346v1
CAISA at SemEval-2023 Task 8: Counterfactual Data Augmentation for Mitigating Class Imbalance in Causal Claim Identification
Akbar Karimi, Lucie Flek
cs.CL, cs.LG
The class imbalance problem can cause machine learning models to produce an undesirable performance on the minority class as well as the whole dataset. Using data augmentation techniques to increase the number of samples is one way to tackle this problem. We introduce a novel counterfactual data augmentation by verb re...
http://arxiv.org/pdf/2306.00346v1
2023-06-01T04:55:43
2023-06-01T04:55:43
{ "title": [ 0.042053223, 0.011161804, -0.04724121, -0.048553467, 0.030929565, -0.033233643, -0.016036987, -0.038726807, -0.02748108, 0.0023899078, -0.015899658, 0.029022217, -0.0096206665, -0.03390503, 0.013877869, -0.010894775, -0.033721924, 0....
http://arxiv.org/abs/2306.01015v2
How to Estimate Model Transferability of Pre-Trained Speech Models?
Zih-Ching Chen, Chao-Han Huck Yang, Bo Li, Yu Zhang, Nanxin Chen, Shou-Yiin Chang, Rohit Prabhavalkar, Hung-yi Lee, Tara N. Sainath
cs.CL, cs.NE, cs.SD, eess.AS
In this work, we introduce a "score-based assessment" framework for estimating the transferability of pre-trained speech models (PSMs) for fine-tuning target tasks. We leverage upon two representation theories, Bayesian likelihood estimation and optimal transport, to generate rank scores for the PSM candidates using th...
http://arxiv.org/pdf/2306.01015v2
2023-08-25T13:50:19
2023-06-01T04:52:26
{ "title": [ -0.0015087128, 0.022491455, -0.034057617, -0.017715454, -0.020721436, -0.014755249, -0.051971436, -0.027526855, -0.04067993, 0.01424408, -0.010749817, -0.014884949, 0.008590698, -0.06323242, 0.02407837, -0.025970459, 0.014625549, 0.0...
http://arxiv.org/abs/2306.00301v2
CapText: Large Language Model-based Caption Generation From Image Context and Description
Shinjini Ghosh, Sagnik Anupam
cs.LG, cs.CL
While deep-learning models have been shown to perform well on image-to-text datasets, it is difficult to use them in practice for captioning images. This is because captions traditionally tend to be context-dependent and offer complementary information about an image, while models tend to produce descriptions that desc...
http://arxiv.org/pdf/2306.00301v2
2023-06-06T03:41:05
2023-06-01T02:40:44
{ "title": [ -0.021881104, -0.0073280334, -0.033355713, -0.040527344, -0.008049011, -0.0075149536, -0.03414917, -0.01776123, 0.015930176, 0.04647827, -0.021957397, 0.028442383, 0.014907837, 0.00024747849, -0.03277588, 0.013717651, -0.011878967, 0...
http://arxiv.org/abs/2306.00288v1
Training-free Neural Architecture Search for RNNs and Transformers
Aaron Serianni, Jugal Kalita
cs.LG, cs.AI, cs.CL
Neural architecture search (NAS) has allowed for the automatic creation of new and effective neural network architectures, offering an alternative to the laborious process of manually designing complex architectures. However, traditional NAS algorithms are slow and require immense amounts of computing power. Recent res...
http://arxiv.org/pdf/2306.00288v1
2023-06-01T02:06:13
2023-06-01T02:06:13
{ "title": [ -0.015792847, -0.008232117, -0.016784668, -0.048858643, 0.0015077591, -0.0038967133, -0.01979065, -0.011161804, -0.009208679, 0.050231934, -0.011054993, -0.0023784637, 0.0026168823, -0.016998291, -0.046173096, 0.0149383545, 0.0052719116,...
http://arxiv.org/abs/2306.00253v2
AfriNames: Most ASR models "butcher" African Names
Tobi Olatunji, Tejumade Afonja, Bonaventure F. P. Dossou, Atnafu Lambebo Tonja, Chris Chinenye Emezue, Amina Mardiyyah Rufai, Sahib Singh
cs.CL, cs.CY
Useful conversational agents must accurately capture named entities to minimize error for downstream tasks, for example, asking a voice assistant to play a track from a certain artist, initiating navigation to a specific location, or documenting a laboratory result for a patient. However, where named entities such as `...
http://arxiv.org/pdf/2306.00253v2
2023-06-02T15:35:42
2023-06-01T00:17:52
{ "title": [ -0.05090332, -0.009925842, -0.010765076, 0.019119263, -0.07501221, 0.020217896, -0.07879639, -0.029632568, -0.029388428, 0.046325684, 0.010124207, -0.006362915, 0.062286377, -0.05633545, 0.045959473, 0.009986877, 0.011726379, 0.02423...
http://arxiv.org/abs/2306.00245v1
From Pixels to UI Actions: Learning to Follow Instructions via Graphical User Interfaces
Peter Shaw, Mandar Joshi, James Cohan, Jonathan Berant, Panupong Pasupat, Hexiang Hu, Urvashi Khandelwal, Kenton Lee, Kristina Toutanova
cs.LG, cs.CL, cs.CV, cs.HC
Much of the previous work towards digital agents for graphical user interfaces (GUIs) has relied on text-based representations (derived from HTML or other structured data sources), which are not always readily available. These input representations have been often coupled with custom, task-specific action spaces. This ...
http://arxiv.org/pdf/2306.00245v1
2023-05-31T23:39:18
2023-05-31T23:39:18
{ "title": [ -0.028930664, -0.0027217865, -0.028518677, -0.014389038, -0.0019798279, -0.018432617, -0.0093688965, -0.021148682, 0.01802063, 0.029647827, -0.03250122, 0.0059928894, -0.021102905, -0.0056533813, -0.06555176, -0.009857178, 0.0005135536, ...
http://arxiv.org/abs/2306.00228v1
Using Visual Cropping to Enhance Fine-Detail Question Answering of BLIP-Family Models
Jiarui Zhang, Mahyar Khayatkhoei, Prateek Chhikara, Filip Ilievski
cs.CV, cs.AI, cs.CL
Visual Question Answering is a challenging task, as it requires seamless interaction between perceptual, linguistic, and background knowledge systems. While the recent progress of visual and natural language models like BLIP has led to improved performance on this task, we lack understanding of the ability of such mode...
http://arxiv.org/pdf/2306.00228v1
2023-05-31T22:48:27
2023-05-31T22:48:27
{ "title": [ 0.01461792, -0.0076675415, -0.0071907043, -0.053131104, -0.028564453, 0.0009813309, -0.06341553, -0.034088135, 0.062286377, 0.014289856, -0.064575195, 0.0052757263, 0.0032901764, -0.0018186569, -0.06628418, -0.0038776398, 0.005783081, ...
http://arxiv.org/abs/2306.00219v2
Diffusion Brush: A Latent Diffusion Model-based Editing Tool for AI-generated Images
Peyman Gholami, Robert Xiao
cs.CV, cs.AI, cs.CL, cs.LG
Text-to-image generative models have made remarkable advancements in generating high-quality images. However, generated images often contain undesirable artifacts or other errors due to model limitations. Existing techniques to fine-tune generated images are time-consuming (manual editing), produce poorly-integrated re...
http://arxiv.org/pdf/2306.00219v2
2023-10-27T01:51:47
2023-05-31T22:27:21
{ "title": [ -0.007843018, 0.0038661957, -0.04058838, -0.022323608, 0.0035152435, -0.021118164, 0.000647068, -0.015731812, 0.016967773, 0.01838684, -0.019485474, 0.011436462, 0.010971069, -0.038513184, -0.005897522, 0.0011796951, 0.011047363, 0.0...
http://arxiv.org/abs/2306.00217v2
FEED PETs: Further Experimentation and Expansion on the Disambiguation of Potentially Euphemistic Terms
Patrick Lee, Iyanuoluwa Shode, Alain Chirino Trujillo, Yuan Zhao, Olumide Ebenezer Ojo, Diana Cuevas Plancarte, Anna Feldman, Jing Peng
cs.CL
Transformers have been shown to work well for the task of English euphemism disambiguation, in which a potentially euphemistic term (PET) is classified as euphemistic or non-euphemistic in a particular context. In this study, we expand on the task in two ways. First, we annotate PETs for vagueness, a linguistic propert...
http://arxiv.org/pdf/2306.00217v2
2023-06-06T19:17:14
2023-05-31T22:23:20
{ "title": [ 0.008308411, 0.018737793, -0.01966858, -0.0061836243, -0.00178051, -0.03225708, -0.012008667, -0.04107666, -0.01625061, 0.030929565, -0.026855469, 0.00844574, 0.04034424, -0.037506104, 0.016220093, -0.011779785, -0.017929077, 0.05273...
http://arxiv.org/abs/2306.00208v1
Strategies for improving low resource speech to text translation relying on pre-trained ASR models
Santosh Kesiraju, Marek Sarvas, Tomas Pavlicek, Cecile Macaire, Alejandro Ciuba
cs.CL, cs.SD, eess.AS
This paper presents techniques and findings for improving the performance of low-resource speech to text translation (ST). We conducted experiments on both simulated and real-low resource setups, on language pairs English - Portuguese, and Tamasheq - French respectively. Using the encoder-decoder framework for ST, our ...
http://arxiv.org/pdf/2306.00208v1
2023-05-31T21:58:07
2023-05-31T21:58:07
{ "title": [ -0.013870239, -0.011665344, -0.014274597, -0.06713867, -0.019973755, 0.0047073364, -0.048339844, 0.011131287, -0.029388428, 0.06121826, -0.0019111633, 0.0041007996, 0.027511597, -0.026748657, -0.03866577, 0.010856628, 0.017837524, 0....
http://arxiv.org/abs/2306.00198v1
An Invariant Learning Characterization of Controlled Text Generation
Carolina Zheng, Claudia Shi, Keyon Vafa, Amir Feder, David M. Blei
cs.CL, cs.LG
Controlled generation refers to the problem of creating text that contains stylistic or semantic attributes of interest. Many approaches reduce this problem to training a predictor of the desired attribute. For example, researchers hoping to deploy a large language model to produce non-toxic content may use a toxicity ...
http://arxiv.org/pdf/2306.00198v1
2023-05-31T21:35:08
2023-05-31T21:35:08
{ "title": [ -0.03604126, 0.0072631836, -0.04220581, -0.030685425, -0.0046691895, -0.005428314, -0.036712646, -0.008918762, -0.02067566, 0.044891357, -0.020812988, 0.033599854, -0.014419556, -0.0046463013, -0.048950195, 0.002658844, -0.020126343, ...
http://arxiv.org/abs/2306.01009v1
Examining the Emergence of Deductive Reasoning in Generative Language Models
Peter Belcak, Luca A. Lanzendörfer, Roger Wattenhofer
cs.CL, cs.AI, cs.LG
We conduct a preliminary inquiry into the ability of generative transformer models to deductively reason from premises provided. We observe notable differences in the performance of models coming from different training setups and find that the deductive reasoning ability increases with scale. Further, we discover that...
http://arxiv.org/pdf/2306.01009v1
2023-05-31T21:29:49
2023-05-31T21:29:49
{ "title": [ -0.021942139, -0.015335083, -0.03717041, -0.06439209, -0.012916565, -0.002450943, -0.059051514, 0.00894928, 0.0028629303, 0.022781372, -0.014945984, -0.00076150894, 0.002418518, -0.023010254, -0.043273926, -0.0005683899, -0.024734497, ...
http://arxiv.org/abs/2306.00186v1
Factually Consistent Summarization via Reinforcement Learning with Textual Entailment Feedback
Paul Roit, Johan Ferret, Lior Shani, Roee Aharoni, Geoffrey Cideron, Robert Dadashi, Matthieu Geist, Sertan Girgin, Léonard Hussenot, Orgad Keller, Nikola Momchev, Sabela Ramos, Piotr Stanczyk, Nino Vieillard, Olivier Bachem, Gal Elidan, Avinatan Hassidim, Olivier Pietquin, Idan Szpektor
cs.CL
Despite the seeming success of contemporary grounded text generation systems, they often tend to generate factually inconsistent text with respect to their input. This phenomenon is emphasized in tasks like summarization, in which the generated summaries should be corroborated by their source article. In this work, we ...
http://arxiv.org/pdf/2306.00186v1
2023-05-31T21:04:04
2023-05-31T21:04:04
{ "title": [ 0.002292633, 0.0041389465, -0.015266418, -0.06567383, -0.013938904, 0.0076904297, -0.042816162, -0.014137268, -0.026168823, 0.039215088, -0.016342163, 0.02909851, -0.008628845, 0.004886627, -0.0368042, -0.0059165955, -0.0096588135, 0...
http://arxiv.org/abs/2306.00177v1
Contrastive Hierarchical Discourse Graph for Scientific Document Summarization
Haopeng Zhang, Xiao Liu, Jiawei Zhang
cs.CL
The extended structural context has made scientific paper summarization a challenging task. This paper proposes CHANGES, a contrastive hierarchical graph neural network for extractive scientific paper summarization. CHANGES represents a scientific paper with a hierarchical discourse graph and learns effective sentence ...
http://arxiv.org/pdf/2306.00177v1
2023-05-31T20:54:43
2023-05-31T20:54:43
{ "title": [ -0.013053894, 0.004573822, -0.028289795, -0.09423828, -0.0026626587, 0.011711121, -0.060394287, 0.005168915, 0.017562866, 0.048736572, 0.007511139, 0.022201538, 0.0023059845, -0.006706238, -0.05847168, 0.018447876, -0.030288696, 0.05...
http://arxiv.org/abs/2306.00176v1
Automated Annotation with Generative AI Requires Validation
Nicholas Pangakis, Samuel Wolken, Neil Fasching
cs.CL, cs.AI
Generative large language models (LLMs) can be a powerful tool for augmenting text annotation procedures, but their performance varies across annotation tasks due to prompt quality, text data idiosyncrasies, and conceptual difficulty. Because these challenges will persist even as LLM technology improves, we argue that ...
http://arxiv.org/pdf/2306.00176v1
2023-05-31T20:50:45
2023-05-31T20:50:45
{ "title": [ 0.022140503, -0.004547119, -0.01386261, -0.0491333, -0.055633545, -0.01158905, -0.052856445, -0.006351471, -0.022766113, 0.049224854, -0.010810852, -0.008583069, 0.011566162, -0.054779053, -0.00021111965, 0.0014314651, 0.006500244, 0...
http://arxiv.org/abs/2306.00168v2
Measuring the Robustness of Natural Language Processing Models to Domain Shifts
Nitay Calderon, Naveh Porat, Eyal Ben-David, Zorik Gekhman, Nadav Oved, Roi Reichart
cs.CL
Existing research on Domain Robustness (DR) suffers from disparate setups, lack of evaluation task variety, and reliance on challenge sets. In this paper, we pose a fundamental question: What is the state of affairs of the DR challenge in the era of Large Language Models (LLMs)? To this end, we construct a DR benchmark...
http://arxiv.org/pdf/2306.00168v2
2023-07-01T18:05:19
2023-05-31T20:25:08
{ "title": [ -0.012779236, 0.022583008, -0.041290283, -0.04385376, 0.029006958, -0.010093689, -0.04928589, -0.040374756, 0.0074272156, 0.01763916, -0.017913818, 0.002231598, 0.0021705627, -0.0075798035, -0.05142212, 0.0011816025, -0.030166626, 0....
http://arxiv.org/abs/2306.00137v1
A Sequence-to-Sequence&Set Model for Text-to-Table Generation
Tong Li, Zhihao Wang, Liangying Shao, Xuling Zheng, Xiaoli Wang, Jinsong Su
cs.CL
Recently, the text-to-table generation task has attracted increasing attention due to its wide applications. In this aspect, the dominant model formalizes this task as a sequence-to-sequence generation task and serializes each table into a token sequence during training by concatenating all rows in a top-down order. Ho...
http://arxiv.org/pdf/2306.00137v1
2023-05-31T19:28:00
2023-05-31T19:28:00
{ "title": [ -0.03152466, 0.00024747849, -0.052459717, -0.06011963, 0.00932312, 0.013824463, -0.05987549, -0.0053138733, 0.0072402954, 0.05709839, -0.01638794, 0.022109985, 0.030029297, -0.0013437271, -0.048034668, 0.0134887695, -0.004989624, 0.0...
http://arxiv.org/abs/2306.00124v1
Pre-Trained Language-Meaning Models for Multilingual Parsing and Generation
Chunliu Wang, Huiyuan Lai, Malvina Nissim, Johan Bos
cs.CL
Pre-trained language models (PLMs) have achieved great success in NLP and have recently been used for tasks in computational semantics. However, these tasks do not fully benefit from PLMs since meaning representations are not explicitly included in the pre-training stage. We introduce multilingual pre-trained language-...
http://arxiv.org/pdf/2306.00124v1
2023-05-31T19:00:33
2023-05-31T19:00:33
{ "title": [ -0.026397705, -0.011543274, -0.015357971, -0.072753906, -0.0062332153, 0.024963379, -0.08453369, 0.0045928955, -0.009597778, 0.03213501, 0.006336212, 0.007534027, -0.0017786026, -0.019729614, -0.040161133, 0.010528564, -0.017120361, ...
http://arxiv.org/abs/2306.00121v1
Multilingual Multi-Figurative Language Detection
Huiyuan Lai, Antonio Toral, Malvina Nissim
cs.CL
Figures of speech help people express abstract concepts and evoke stronger emotions than literal expressions, thereby making texts more creative and engaging. Due to its pervasive and fundamental character, figurative language understanding has been addressed in Natural Language Processing, but it's highly understudied...
http://arxiv.org/pdf/2306.00121v1
2023-05-31T18:52:41
2023-05-31T18:52:41
{ "title": [ 0.0048103333, 0.011436462, -0.02267456, -0.027191162, 0.020080566, -0.018112183, -0.018661499, -0.030563354, 0.0004556179, 0.048797607, 0.0002734661, -0.008178711, 0.038726807, -0.019454956, -0.0395813, 0.016479492, -0.022109985, 0.0...
http://arxiv.org/abs/2306.00110v1
MuseCoco: Generating Symbolic Music from Text
Peiling Lu, Xin Xu, Chenfei Kang, Botao Yu, Chengyi Xing, Xu Tan, Jiang Bian
cs.SD, cs.AI, cs.CL, cs.LG, cs.MM, eess.AS
Generating music from text descriptions is a user-friendly mode since the text is a relatively easy interface for user engagement. While some approaches utilize texts to control music audio generation, editing musical elements in generated audio is challenging for users. In contrast, symbolic music offers ease of editi...
http://arxiv.org/pdf/2306.00110v1
2023-05-31T18:34:16
2023-05-31T18:34:16
{ "title": [ -0.031585693, -0.03857422, -0.041809082, -0.0692749, -0.014205933, 0.011955261, -0.040283203, 0.017501831, -0.003282547, 0.048095703, -0.010643005, 0.003824234, 0.011161804, -0.008598328, -0.060150146, 0.00040531158, -0.013076782, 0....
http://arxiv.org/abs/2306.00107v2
MERT: Acoustic Music Understanding Model with Large-Scale Self-supervised Training
Yizhi Li, Ruibin Yuan, Ge Zhang, Yinghao Ma, Xingran Chen, Hanzhi Yin, Chenghua Lin, Anton Ragni, Emmanouil Benetos, Norbert Gyenge, Roger Dannenberg, Ruibo Liu, Wenhu Chen, Gus Xia, Yemin Shi, Wenhao Huang, Yike Guo, Jie Fu
cs.SD, cs.AI, cs.CL, cs.LG, eess.AS
Self-supervised learning (SSL) has recently emerged as a promising paradigm for training generalisable models on large-scale data in the fields of vision, text, and speech. Although SSL has been proven effective in speech and audio, its application to music audio has yet to be thoroughly explored. This is primarily due...
http://arxiv.org/pdf/2306.00107v2
2023-06-06T14:06:02
2023-05-31T18:27:43
{ "title": [ 0.0015525818, -0.008712769, -0.039978027, -0.07867432, 0.001534462, -0.012161255, -0.05117798, 0.013069153, -0.0045547485, 0.024230957, -0.020217896, -0.002943039, 0.010383606, -0.02494812, -0.034118652, -0.0060577393, -0.016647339, ...
http://arxiv.org/abs/2306.00103v1
ManagerTower: Aggregating the Insights of Uni-Modal Experts for Vision-Language Representation Learning
Xiao Xu, Bei Li, Chenfei Wu, Shao-Yen Tseng, Anahita Bhiwandiwalla, Shachar Rosenman, Vasudev Lal, Wanxiang Che, Nan Duan
cs.CV, cs.CL, cs.LG
Two-Tower Vision-Language (VL) models have shown promising improvements on various downstream VL tasks. Although the most advanced work improves performance by building bridges between encoders, it suffers from ineffective layer-by-layer utilization of uni-modal representations and cannot flexibly exploit different lev...
http://arxiv.org/pdf/2306.00103v1
2023-05-31T18:23:57
2023-05-31T18:23:57
{ "title": [ -0.004005432, -0.001039505, -0.051513672, -0.060913086, -0.006038666, -0.012931824, -0.049865723, -0.008415222, 0.0004761219, 0.046447754, -0.018630981, 0.008224487, 0.025115967, -0.006702423, -0.039215088, -0.0019025803, 0.023025513, ...
http://arxiv.org/abs/2306.00100v1
MetaXLR -- Mixed Language Meta Representation Transformation for Low-resource Cross-lingual Learning based on Multi-Armed Bandit
Liat Bezalel, Eyal Orgad
cs.CL
Transfer learning for extremely low resource languages is a challenging task as there is no large scale monolingual corpora for pre training or sufficient annotated data for fine tuning. We follow the work of MetaXL which suggests using meta learning for transfer learning from a single source language to an extremely l...
http://arxiv.org/pdf/2306.00100v1
2023-05-31T18:22:33
2023-05-31T18:22:33
{ "title": [ -0.033355713, -0.021957397, -0.038116455, -0.014122009, -0.0262146, -0.009147644, -0.028793335, -0.019058228, -0.012641907, 0.054504395, -0.019866943, 0.019927979, 0.0050086975, -0.009513855, -0.021408081, -0.0107803345, -0.013671875, ...
http://arxiv.org/abs/2305.20088v2
Improving CLIP Training with Language Rewrites
Lijie Fan, Dilip Krishnan, Phillip Isola, Dina Katabi, Yonglong Tian
cs.CV, cs.CL, cs.LG
Contrastive Language-Image Pre-training (CLIP) stands as one of the most effective and scalable methods for training transferable vision models using paired image and text data. CLIP models are trained using contrastive loss, which typically relies on data augmentations to prevent overfitting and shortcuts. However, in...
http://arxiv.org/pdf/2305.20088v2
2023-10-28T08:46:13
2023-05-31T17:59:04
{ "title": [ -0.011878967, -0.009963989, -0.0154418945, -0.08258057, -0.03756714, 0.025848389, -0.09259033, 0.008773804, -0.05444336, 0.031219482, 0.005241394, -0.019195557, 0.0016736984, -0.04336548, -0.003288269, -0.015586853, 0.016311646, -0.0...
http://arxiv.org/abs/2305.20080v1
Findings of the VarDial Evaluation Campaign 2023
Noëmi Aepli, Çağrı Çöltekin, Rob Van Der Goot, Tommi Jauhiainen, Mourhaf Kazzaz, Nikola Ljubešić, Kai North, Barbara Plank, Yves Scherrer, Marcos Zampieri
cs.CL
This report presents the results of the shared tasks organized as part of the VarDial Evaluation Campaign 2023. The campaign is part of the tenth workshop on Natural Language Processing (NLP) for Similar Languages, Varieties and Dialects (VarDial), co-located with EACL 2023. Three separate shared tasks were included th...
http://arxiv.org/pdf/2305.20080v1
2023-05-31T17:55:21
2023-05-31T17:55:21
{ "title": [ -0.0056915283, 0.021438599, -0.008514404, -0.010025024, -0.031341553, -0.0064811707, -0.055389404, -0.036071777, -0.02784729, 0.045959473, 0.029922485, -0.007865906, 0.023452759, -0.04437256, 0.01675415, 0.0025405884, 0.012031555, 0....
http://arxiv.org/abs/2306.01006v2
Scaling Evidence-based Instructional Design Expertise through Large Language Models
Gautam Yadav
cs.CL, cs.HC
This paper presents a comprehensive exploration of leveraging Large Language Models (LLMs), specifically GPT-4, in the field of instructional design. With a focus on scaling evidence-based instructional design expertise, our research aims to bridge the gap between theoretical educational studies and practical implement...
http://arxiv.org/pdf/2306.01006v2
2023-06-23T22:03:54
2023-05-31T17:54:07
{ "title": [ -0.033447266, -0.0231781, -0.03338623, -0.024124146, -0.03213501, -0.013885498, -0.03555298, -0.0061683655, -0.030395508, 0.018249512, -0.025405884, 0.015823364, -0.025314331, -0.01586914, -0.0051727295, -0.018203735, -0.009933472, 0...
http://arxiv.org/abs/2305.20076v2
Decision-Oriented Dialogue for Human-AI Collaboration
Jessy Lin, Nicholas Tomlin, Jacob Andreas, Jason Eisner
cs.CL, cs.AI
We describe a class of tasks called decision-oriented dialogues, in which AI assistants must collaborate with one or more humans via natural language to help them make complex decisions. We formalize three domains in which users face everyday decisions: (1) choosing an assignment of reviewers to conference papers, (2) ...
http://arxiv.org/pdf/2305.20076v2
2023-06-01T16:49:10
2023-05-31T17:50:02
{ "title": [ -0.0345459, -0.0069770813, -0.051116943, -0.039245605, -0.0013608932, -0.016662598, -0.010986328, 0.0024147034, -0.011566162, 0.031066895, -0.020645142, -0.00390625, -0.017059326, -0.01751709, -0.039031982, -0.017730713, -0.009460449, ...
http://arxiv.org/abs/2305.20050v1
Let's Verify Step by Step
Hunter Lightman, Vineet Kosaraju, Yura Burda, Harri Edwards, Bowen Baker, Teddy Lee, Jan Leike, John Schulman, Ilya Sutskever, Karl Cobbe
cs.LG, cs.AI, cs.CL
In recent years, large language models have greatly improved in their ability to perform complex multi-step reasoning. However, even state-of-the-art models still regularly produce logical mistakes. To train more reliable models, we can turn either to outcome supervision, which provides feedback for a final result, or ...
http://arxiv.org/pdf/2305.20050v1
2023-05-31T17:24:00
2023-05-31T17:24:00
{ "title": [ 0.03387451, -0.005958557, -0.009895325, -0.0703125, -0.016738892, 0.002254486, -0.066101074, 0.002445221, -0.04876709, 0.049926758, 0.015113831, 0.026046753, -0.0158844, -0.00843811, 0.012680054, -0.010276794, 0.03567505, 0.012863159...
http://arxiv.org/abs/2305.20046v1
Computational Language Assessment in patients with speech, language, and communication impairments
Charalambos Themistocleous
cs.CL
Speech, language, and communication symptoms enable the early detection, diagnosis, treatment planning, and monitoring of neurocognitive disease progression. Nevertheless, traditional manual neurologic assessment, the speech and language evaluation standard, is time-consuming and resource-intensive for clinicians. We a...
http://arxiv.org/pdf/2305.20046v1
2023-05-31T17:20:45
2023-05-31T17:20:45
{ "title": [ -0.03062439, 0.0063209534, -0.032806396, -0.04446411, -0.013076782, 0.00058698654, -0.041503906, 0.005004883, -0.017211914, 0.031402588, -0.017059326, 0.0014762878, -0.00598526, -0.03125, -0.034118652, -0.006801605, -0.02381897, 0.04...
http://arxiv.org/abs/2305.20045v1
ActiveAED: A Human in the Loop Improves Annotation Error Detection
Leon Weber, Barbara Plank
cs.CL, cs.LG
Manually annotated datasets are crucial for training and evaluating Natural Language Processing models. However, recent work has discovered that even widely-used benchmark datasets contain a substantial number of erroneous annotations. This problem has been addressed with Annotation Error Detection (AED) models, which ...
http://arxiv.org/pdf/2305.20045v1
2023-05-31T17:18:47
2023-05-31T17:18:47
{ "title": [ 0.025131226, 0.005130768, -0.023040771, -0.055786133, -0.0015420914, -0.034942627, -0.019424438, -0.025024414, 0.030426025, 0.042022705, -0.010055542, -0.0074424744, 0.016296387, -0.039642334, -0.045928955, 0.02067566, 0.0071792603, ...
http://arxiv.org/abs/2306.01789v2
Edit Distance based RL for RNNT decoding
Dongseong Hwang, Changwan Ryu, Khe Chai Sim
cs.SD, cs.CL, eess.AS
RNN-T is currently considered the industry standard in ASR due to its exceptional WERs in various benchmark tests and its ability to support seamless streaming and longform transcription. However, its biggest drawback lies in the significant discrepancy between its training and inference objectives. During training, RN...
http://arxiv.org/pdf/2306.01789v2
2023-07-14T16:53:12
2023-05-31T16:53:23
{ "title": [ -0.043182373, -0.00869751, -0.013656616, -0.013000488, -0.06866455, 0.015960693, -0.048614502, 0.0015821457, -0.044647217, 0.04434204, 0.0054359436, -0.057281494, 0.040252686, -0.052886963, 0.0046310425, -0.008460999, 0.009727478, -0...
http://arxiv.org/abs/2305.20019v1
Monotonic Location Attention for Length Generalization
Jishnu Ray Chowdhury, Cornelia Caragea
cs.LG, cs.AI, cs.CL
We explore different ways to utilize position-based cross-attention in seq2seq networks to enable length generalization in algorithmic tasks. We show that a simple approach of interpolating the original and reversed encoded representations combined with relative attention allows near-perfect length generalization for b...
http://arxiv.org/pdf/2305.20019v1
2023-05-31T16:48:06
2023-05-31T16:48:06
{ "title": [ 0.0017957687, 0.0032291412, -0.031021118, -0.04046631, -0.006336212, -0.020080566, -0.013061523, -0.018218994, 0.009086609, 0.04083252, -0.027893066, 0.014785767, -0.013519287, -0.00756073, -0.033966064, -0.0050735474, -0.004550934, ...
http://arxiv.org/abs/2305.20018v1
Scalable Learning of Latent Language Structure With Logical Offline Cycle Consistency
Maxwell Crouse, Ramon Astudillo, Tahira Naseem, Subhajit Chaudhury, Pavan Kapanipathi, Salim Roukos, Alexander Gray
cs.CL, cs.AI
We introduce Logical Offline Cycle Consistency Optimization (LOCCO), a scalable, semi-supervised method for training a neural semantic parser. Conceptually, LOCCO can be viewed as a form of self-learning where the semantic parser being trained is used to generate annotations for unlabeled text that are then used as new...
http://arxiv.org/pdf/2305.20018v1
2023-05-31T16:47:20
2023-05-31T16:47:20
{ "title": [ -0.02583313, -0.0158844, -0.059295654, -0.07727051, -0.006629944, 0.001953125, -0.057128906, 0.046447754, 0.0010299683, 0.023620605, -0.031311035, 0.012176514, -0.017318726, -0.014701843, -0.028030396, -0.016677856, -0.00096702576, 0...
http://arxiv.org/abs/2305.20010v1
Human or Not? A Gamified Approach to the Turing Test
Daniel Jannai, Amos Meron, Barak Lenz, Yoav Levine, Yoav Shoham
cs.AI, cs.CL, cs.CY, cs.HC, 68T50, I.2.7
We present "Human or Not?", an online game inspired by the Turing test, that measures the capability of AI chatbots to mimic humans in dialog, and of humans to tell bots from other humans. Over the course of a month, the game was played by over 1.5 million users who engaged in anonymous two-minute chat sessions with ei...
http://arxiv.org/pdf/2305.20010v1
2023-05-31T16:32:22
2023-05-31T16:32:22
{ "title": [ 0.0058898926, -0.0040779114, -0.01084137, -0.022247314, 0.005783081, -0.019989014, 0.013938904, -0.035614014, -0.03527832, 0.025924683, -0.014350891, 0.013877869, -0.033569336, -0.017288208, -0.0395813, -0.01966858, -0.01828003, 0.01...
http://arxiv.org/abs/2305.19999v3
Beam Tree Recursive Cells
Jishnu Ray Chowdhury, Cornelia Caragea
cs.LG, cs.AI, cs.CL
We propose Beam Tree Recursive Cell (BT-Cell) - a backpropagation-friendly framework to extend Recursive Neural Networks (RvNNs) with beam search for latent structure induction. We further extend this framework by proposing a relaxation of the hard top-k operators in beam search for better propagation of gradient signa...
http://arxiv.org/pdf/2305.19999v3
2023-06-20T11:50:33
2023-05-31T16:20:04
{ "title": [ -0.008010864, -0.018661499, -0.034423828, -0.020996094, -0.06530762, -0.0061187744, -0.07861328, -0.0049552917, 0.0009813309, 0.05529785, -0.030776978, -0.040740967, 0.038726807, -0.015342712, -0.02319336, -0.009895325, 0.02406311, 0...
http://arxiv.org/abs/2305.19998v1
Efficient Shapley Values Estimation by Amortization for Text Classification
Chenghao Yang, Fan Yin, He He, Kai-Wei Chang, Xiaofei Ma, Bing Xiang
cs.CL, cs.LG
Despite the popularity of Shapley Values in explaining neural text classification models, computing them is prohibitive for large pretrained models due to a large number of model evaluations. In practice, Shapley Values are often estimated with a small number of stochastic model evaluations. However, we show that the e...
http://arxiv.org/pdf/2305.19998v1
2023-05-31T16:19:13
2023-05-31T16:19:13
{ "title": [ -0.02619934, 0.011795044, -0.039916992, -0.041412354, 0.011428833, 0.004852295, -0.020446777, -0.012229919, 0.024429321, 0.018463135, -0.0602417, 0.000518322, -0.029815674, 0.004928589, -0.06536865, -0.034606934, -0.027023315, 0.0454...
http://arxiv.org/abs/2305.19981v2
MedNgage: A Dataset for Understanding Engagement in Patient-Nurse Conversations
Yan Wang, Heidi Ann Scharf Donovan, Sabit Hassan, Mailhe Alikhani
cs.CL
Patients who effectively manage their symptoms often demonstrate higher levels of engagement in conversations and interventions with healthcare practitioners. This engagement is multifaceted, encompassing cognitive and socio-affective dimensions. Consequently, it is crucial for AI systems to understand the engagement i...
http://arxiv.org/pdf/2305.19981v2
2023-06-20T16:52:56
2023-05-31T16:06:07
{ "title": [ 0.010131836, 0.007247925, -0.03817749, -0.06384277, 0.001660347, -0.017654419, -0.038604736, -0.009338379, -0.034240723, 0.03466797, -0.02659607, 0.024963379, -0.020248413, -0.023864746, -0.027404785, -0.021347046, -0.0072746277, 0.0...
http://arxiv.org/abs/2305.19974v1
Correcting Semantic Parses with Natural Language through Dynamic Schema Encoding
Parker Glenn, Parag Pravin Dakle, Preethi Raghavan
cs.CL
In addressing the task of converting natural language to SQL queries, there are several semantic and syntactic challenges. It becomes increasingly important to understand and remedy the points of failure as the performance of semantic parsing systems improve. We explore semantic parse correction with natural language f...
http://arxiv.org/pdf/2305.19974v1
2023-05-31T16:01:57
2023-05-31T16:01:57
{ "title": [ -0.017959595, -0.010154724, -0.03778076, -0.07623291, 0.020965576, 0.006603241, -0.046051025, -0.0005373955, -0.0011606216, 0.015472412, -0.014022827, -0.009750366, -0.008384705, -0.01876831, -0.037506104, -0.016677856, -0.021530151, ...
http://arxiv.org/abs/2305.19972v1
ViLaS: Integrating Vision and Language into Automatic Speech Recognition
Minglun Han, Feilong Chen, Ziyi Ni, Linghui Meng, Jing Shi, Shuang Xu, Bo Xu
eess.AS, cs.AI, cs.CL
Employing additional multimodal information to improve automatic speech recognition (ASR) performance has been proven effective in previous works. However, many of these works focus only on the utilization of visual cues from human lip motion. In fact, context-dependent visual and linguistic cues can also be used to im...
http://arxiv.org/pdf/2305.19972v1
2023-05-31T16:01:20
2023-05-31T16:01:20
{ "title": [ -0.025604248, -0.018859863, -0.038024902, -0.055877686, -0.000037014484, -0.005092621, -0.027175903, 0.00945282, 0.013114929, 0.030395508, -0.05050659, -0.007347107, 0.0048179626, -0.0040740967, -0.064331055, -0.018127441, 0.0064353943, ...
http://arxiv.org/abs/2305.19936v1
Metropolis-Hastings algorithm in joint-attention naming game: Experimental semiotics study
Ryota Okumura, Tadahiro Taniguchi, Yosinobu Hagiwara, Akira Taniguchi
cs.CL, cs.HC
In this study, we explore the emergence of symbols during interactions between individuals through an experimental semiotic study. Previous studies investigate how humans organize symbol systems through communication using artificially designed subjective experiments. In this study, we have focused on a joint attention...
http://arxiv.org/pdf/2305.19936v1
2023-05-31T15:20:54
2023-05-31T15:20:54
{ "title": [ -0.0154418945, 0.0071258545, -0.013679504, 0.033996582, -0.026184082, -0.02935791, 0.023712158, -0.031066895, -0.010856628, 0.04916382, 0.013343811, 0.002538681, 0.051452637, -0.022659302, -0.018157959, -0.005554199, -0.020324707, 0....
http://arxiv.org/abs/2305.19933v1
Speaking the Language of Your Listener: Audience-Aware Adaptation via Plug-and-Play Theory of Mind
Ece Takmaz, Nicolo' Brandizzi, Mario Giulianelli, Sandro Pezzelle, Raquel Fernández
cs.CL, cs.AI, cs.CV
Dialogue participants may have varying levels of knowledge about the topic under discussion. In such cases, it is essential for speakers to adapt their utterances by taking their audience into account. Yet, it is an open question how such adaptation can be modelled in computational agents. In this paper, we model a vis...
http://arxiv.org/pdf/2305.19933v1
2023-05-31T15:17:28
2023-05-31T15:17:28
{ "title": [ 0.008888245, -0.021377563, -0.007347107, -0.034851074, -0.002161026, -0.0077590942, -0.009231567, -0.023605347, -0.024047852, 0.018035889, 0.0124053955, 0.0069847107, -0.016159058, -0.03326416, -0.03894043, 0.002368927, 0.012329102, ...
http://arxiv.org/abs/2305.19928v4
Supplementary Features of BiLSTM for Enhanced Sequence Labeling
Conglei Xu, Kun Shen, Hongguang Sun
cs.CL
Sequence labeling tasks require the computation of sentence representations for each word within a given sentence. A prevalent method incorporates a Bi-directional Long Short-Term Memory (BiLSTM) layer to enhance the sequence structure information. However, empirical evidence Li (2020) suggests that the capacity of BiL...
http://arxiv.org/pdf/2305.19928v4
2023-06-23T13:47:17
2023-05-31T15:05:25
{ "title": [ 0.00015938282, -0.0027389526, -0.02659607, -0.11810303, 0.010604858, 0.015640259, -0.07873535, 0.004940033, -0.0021038055, 0.024261475, -0.0013227463, -0.005382538, -0.009429932, -0.041259766, -0.01586914, 0.006263733, -0.0067634583, ...
http://arxiv.org/abs/2305.19926v2
ChatGPT an ENFJ, Bard an ISTJ: Empirical Study on Personalities of Large Language Models
Jen-tse Huang, Wenxuan Wang, Man Ho Lam, Eric John Li, Wenxiang Jiao, Michael R. Lyu
cs.CL
Large Language Models (LLMs) have made remarkable advancements in the field of artificial intelligence, significantly reshaping the human-computer interaction. We not only focus on the performance of LLMs, but also explore their features from a psychological perspective, acknowledging the importance of understanding th...
http://arxiv.org/pdf/2305.19926v2
2023-06-07T13:45:30
2023-05-31T15:03:28
{ "title": [ -0.03933716, -0.02217102, -0.05340576, 0.0034046173, -0.03125, -0.023727417, -0.025238037, 0.01676941, -0.014816284, 0.049194336, 0.0044021606, 0.020065308, 0.05429077, -0.045043945, 0.0024547577, 0.025985718, -0.035369873, 0.0701293...
http://arxiv.org/abs/2305.19915v4
Source Code Data Augmentation for Deep Learning: A Survey
Terry Yue Zhuo, Zhou Yang, Zhensu Sun, Yufei Wang, Li Li, Xiaoning Du, Zhenchang Xing, David Lo
cs.CL, cs.AI, cs.SE
The increasingly popular adoption of deep learning models in many critical source code tasks motivates the development of data augmentation (DA) techniques to enhance training data and improve various capabilities (e.g., robustness and generalizability) of these models. Although a series of DA methods have been propose...
http://arxiv.org/pdf/2305.19915v4
2023-11-13T17:34:53
2023-05-31T14:47:44
{ "title": [ -0.014892578, -0.013160706, -0.054626465, -0.07757568, 0.017944336, -0.0105896, -0.025909424, 0.028198242, 0.013961792, 0.01436615, -0.038879395, 0.013572693, -0.019104004, -0.014198303, -0.027694702, 0.0052986145, 0.002275467, 0.012...
http://arxiv.org/abs/2305.19911v1
Neuron to Graph: Interpreting Language Model Neurons at Scale
Alex Foote, Neel Nanda, Esben Kran, Ioannis Konstas, Shay Cohen, Fazl Barez
cs.LG, cs.CL
Advances in Large Language Models (LLMs) have led to remarkable capabilities, yet their inner mechanisms remain largely unknown. To understand these models, we need to unravel the functions of individual neurons and their contribution to the network. This paper introduces a novel automated approach designed to scale in...
http://arxiv.org/pdf/2305.19911v1
2023-05-31T14:44:33
2023-05-31T14:44:33
{ "title": [ -0.034454346, -0.01134491, -0.030075073, -0.022018433, -0.017120361, -0.0066947937, -0.02671814, -0.012954712, -0.0146865845, 0.013328552, -0.012626648, -0.007873535, -0.0066566467, -0.03793335, -0.03112793, 0.004863739, -0.0069618225, ...
http://arxiv.org/abs/2305.19905v1
How to Plant Trees in Language Models: Data and Architectural Effects on the Emergence of Syntactic Inductive Biases
Aaron Mueller, Tal Linzen
cs.CL
Accurate syntactic representations are essential for robust generalization in natural language. Recent work has found that pre-training can teach language models to rely on hierarchical syntactic features - as opposed to incorrect linear features - when performing tasks after fine-tuning. We test what aspects of pre-tr...
http://arxiv.org/pdf/2305.19905v1
2023-05-31T14:38:14
2023-05-31T14:38:14
{ "title": [ -0.014984131, -0.0071868896, -0.023361206, -0.030212402, -0.015266418, -0.016494751, -0.01638794, -0.0064888, -0.02029419, 0.016052246, -0.056396484, 0.021987915, -0.013198853, -0.00289917, -0.031829834, -0.030059814, -0.024932861, 0...
http://arxiv.org/abs/2305.19902v1
AQE: Argument Quadruplet Extraction via a Quad-Tagging Augmented Generative Approach
Jia Guo, Liying Cheng, Wenxuan Zhang, Stanley Kok, Xin Li, Lidong Bing
cs.CL
Argument mining involves multiple sub-tasks that automatically identify argumentative elements, such as claim detection, evidence extraction, stance classification, etc. However, each subtask alone is insufficient for a thorough understanding of the argumentative structure and reasoning process. To learn a complete vie...
http://arxiv.org/pdf/2305.19902v1
2023-05-31T14:35:53
2023-05-31T14:35:53
{ "title": [ 0.02508545, -0.014152527, -0.026809692, -0.119384766, 0.001247406, 0.009666443, -0.054992676, 0.016479492, -0.010047913, 0.030456543, 0.03302002, 0.005683899, -0.012756348, -0.050567627, -0.035705566, 0.011695862, -0.020812988, 0.050...
http://arxiv.org/abs/2305.19894v2
Med-UniC: Unifying Cross-Lingual Medical Vision-Language Pre-Training by Diminishing Bias
Zhongwei Wan, Che Liu, Mi Zhang, Jie Fu, Benyou Wang, Sibo Cheng, Lei Ma, César Quilodrán-Casas, Rossella Arcucci
cs.CL, cs.AI, cs.CV
The scarcity of data presents a critical obstacle to the efficacy of medical visionlanguage pre-training (VLP). A potential solution lies in the combination of datasets from various language communities. Nevertheless, the main challenge stems from the complexity of integrating diverse syntax and semantics, language-spe...
http://arxiv.org/pdf/2305.19894v2
2023-09-25T18:58:36
2023-05-31T14:28:19
{ "title": [ -0.00065755844, -0.013153076, -0.030349731, -0.0670166, -0.00092697144, -0.007827759, -0.045928955, -0.01902771, -0.009094238, 0.051635742, 0.0057258606, 0.0065307617, 0.04135132, -0.0368042, -0.017166138, 0.0109939575, 0.015182495, ...
http://arxiv.org/abs/2305.19857v1
TPDM: Selectively Removing Positional Information for Zero-shot Translation via Token-Level Position Disentangle Module
Xingran Chen, Ge Zhang, Jie Fu
cs.CL
Due to Multilingual Neural Machine Translation's (MNMT) capability of zero-shot translation, many works have been carried out to fully exploit the potential of MNMT in zero-shot translation. It is often hypothesized that positional information may hinder the MNMT from outputting a robust encoded representation for deco...
http://arxiv.org/pdf/2305.19857v1
2023-05-31T13:48:45
2023-05-31T13:48:45
{ "title": [ 0.0017061234, -0.016296387, -0.034423828, -0.019195557, -0.044555664, -0.024902344, -0.053009033, -0.021270752, 0.028121948, 0.061523438, 0.0038223267, -0.005130768, 0.050689697, -0.04989624, -0.009811401, 0.026565552, 0.00017726421, ...
http://arxiv.org/abs/2305.19847v1
How Does Pretraining Improve Discourse-Aware Translation?
Zhihong Huang, Longyue Wang, Siyou Liu, Derek F. Wong
cs.CL, cs.AI
Pretrained language models (PLMs) have produced substantial improvements in discourse-aware neural machine translation (NMT), for example, improved coherence in spoken language translation. However, the underlying reasons for their strong performance have not been well explained. To bridge this gap, we introduce a prob...
http://arxiv.org/pdf/2305.19847v1
2023-05-31T13:36:51
2023-05-31T13:36:51
{ "title": [ 0.02947998, 0.012779236, -0.004776001, -0.05026245, -0.018829346, 0.0026168823, -0.05178833, -0.0057373047, -0.032196045, 0.034362793, 0.01838684, 0.010093689, 0.007369995, -0.040008545, 0.013160706, 0.0035362244, 0.01008606, 0.01824...
http://arxiv.org/abs/2305.19845v1
Guiding Computational Stance Detection with Expanded Stance Triangle Framework
Zhengyuan Liu, Yong Keong Yap, Hai Leong Chieu, Nancy F. Chen
cs.CL
Stance detection determines whether the author of a piece of text is in favor of, against, or neutral towards a specified target, and can be used to gain valuable insights into social media. The ubiquitous indirect referral of targets makes this task challenging, as it requires computational solutions to model semantic...
http://arxiv.org/pdf/2305.19845v1
2023-05-31T13:33:29
2023-05-31T13:33:29
{ "title": [ -0.028778076, 0.0026683807, -0.03152466, -0.047180176, -0.013137817, -0.009933472, -0.041900635, -0.019378662, 0.036987305, 0.041534424, -0.051483154, -0.016952515, -0.006504059, -0.0040130615, -0.06750488, -0.011276245, 0.0053520203, ...
http://arxiv.org/abs/2305.19840v1
BEIR-PL: Zero Shot Information Retrieval Benchmark for the Polish Language
Konrad Wojtasik, Vadim Shishkin, Kacper Wołowiec, Arkadiusz Janz, Maciej Piasecki
cs.IR, cs.AI, cs.CL
The BEIR dataset is a large, heterogeneous benchmark for Information Retrieval (IR) in zero-shot settings, garnering considerable attention within the research community. However, BEIR and analogous datasets are predominantly restricted to the English language. Our objective is to establish extensive large-scale resour...
http://arxiv.org/pdf/2305.19840v1
2023-05-31T13:29:07
2023-05-31T13:29:07
{ "title": [ -0.011917114, -0.02418518, -0.022750854, -0.040496826, -0.013290405, -0.007751465, -0.007091522, -0.0004811287, 0.033935547, 0.057373047, -0.0149002075, 0.0012187958, 0.024780273, -0.010551453, -0.06689453, 0.027526855, -0.010749817, ...
http://arxiv.org/abs/2305.19835v1
Deliberate then Generate: Enhanced Prompting Framework for Text Generation
Bei Li, Rui Wang, Junliang Guo, Kaitao Song, Xu Tan, Hany Hassan, Arul Menezes, Tong Xiao, Jiang Bian, JingBo Zhu
cs.CL, cs.AI
Large language models (LLMs) have shown remarkable success across a wide range of natural language generation tasks, where proper prompt designs make great impacts. While existing prompting methods are normally restricted to providing correct information, in this paper, we encourage the model to deliberate by proposing...
http://arxiv.org/pdf/2305.19835v1
2023-05-31T13:23:04
2023-05-31T13:23:04
{ "title": [ -0.026748657, -0.008811951, -0.046813965, -0.0635376, 0.0018148422, 0.0037517548, -0.056549072, -0.01020813, 0.00472641, 0.021209717, -0.018829346, 0.018112183, 0.0036640167, -0.019104004, -0.036712646, 0.008544922, -0.008956909, 0.0...
http://arxiv.org/abs/2305.19821v1
LMCap: Few-shot Multilingual Image Captioning by Retrieval Augmented Language Model Prompting
Rita Ramos, Bruno Martins, Desmond Elliott
cs.CL, cs.CV
Multilingual image captioning has recently been tackled by training with large-scale machine translated data, which is an expensive, noisy, and time-consuming process. Without requiring any multilingual caption data, we propose LMCap, an image-blind few-shot multilingual captioning model that works by prompting a langu...
http://arxiv.org/pdf/2305.19821v1
2023-05-31T13:03:17
2023-05-31T13:03:17
{ "title": [ -0.0013828278, -0.012557983, -0.00920105, -0.018722534, -0.021484375, -0.0028076172, -0.027191162, -0.022705078, -0.001162529, 0.04458618, -0.019714355, 0.025512695, 0.016571045, -0.01399231, -0.022460938, 0.0077934265, 0.0038528442, ...
http://arxiv.org/abs/2305.19783v1
IDAS: Intent Discovery with Abstractive Summarization
Maarten De Raedt, Fréderic Godin, Thomas Demeester, Chris Develder
cs.CL
Intent discovery is the task of inferring latent intents from a set of unlabeled utterances, and is a useful step towards the efficient creation of new conversational agents. We show that recent competitive methods in intent discovery can be outperformed by clustering utterances based on abstractive summaries, i.e., "l...
http://arxiv.org/pdf/2305.19783v1
2023-05-31T12:19:40
2023-05-31T12:19:40
{ "title": [ -0.016586304, 0.0031833649, -0.058410645, -0.07446289, 0.015472412, 0.0017232895, -0.046447754, -0.006919861, -0.0031929016, 0.027740479, -0.017242432, 0.032196045, -0.02381897, -0.010818481, -0.035369873, 0.0057258606, -0.017990112, ...
http://arxiv.org/abs/2305.19769v1
Attention-Based Methods For Audio Question Answering
Parthasaarathy Sudarsanam, Tuomas Virtanen
cs.CL, cs.LG, cs.SD, eess.AS
Audio question answering (AQA) is the task of producing natural language answers when a system is provided with audio and natural language questions. In this paper, we propose neural network architectures based on self-attention and cross-attention for the AQA task. The self-attention layers extract powerful audio and ...
http://arxiv.org/pdf/2305.19769v1
2023-05-31T12:00:51
2023-05-31T12:00:51
{ "title": [ 0.008323669, 0.0070648193, -0.02027893, -0.06774902, 0.011428833, -0.0042915344, -0.031829834, 0.012565613, 0.020523071, 0.04888916, -0.026992798, -0.02029419, 0.021392822, -0.00042033195, -0.0602417, 0.009788513, -0.0051727295, 0.01...
http://arxiv.org/abs/2306.01004v1
AoM: Detecting Aspect-oriented Information for Multimodal Aspect-Based Sentiment Analysis
Ru Zhou, Wenya Guo, Xumeng Liu, Shenglong Yu, Ying Zhang, Xiaojie Yuan
cs.CL, cs.AI
Multimodal aspect-based sentiment analysis (MABSA) aims to extract aspects from text-image pairs and recognize their sentiments. Existing methods make great efforts to align the whole image to corresponding aspects. However, different regions of the image may relate to different aspects in the same sentence, and coarse...
http://arxiv.org/pdf/2306.01004v1
2023-05-31T11:50:43
2023-05-31T11:50:43
{ "title": [ -0.006324768, 0.00016772747, -0.034576416, -0.06414795, 0.008056641, 0.0024108887, -0.03189087, -0.0020923615, 0.020767212, 0.052825928, 0.0019893646, 0.003709793, 0.010894775, -0.01133728, -0.076049805, 0.019210815, -0.019180298, 0....
http://arxiv.org/abs/2305.19761v1
Recursive Metropolis-Hastings Naming Game: Symbol Emergence in a Multi-agent System based on Probabilistic Generative Models
Jun Inukai, Tadahiro Taniguchi, Akira Taniguchi, Yoshinobu Hagiwara
cs.CL, cs.LG, cs.MA
In the studies on symbol emergence and emergent communication in a population of agents, a computational model was employed in which agents participate in various language games. Among these, the Metropolis-Hastings naming game (MHNG) possesses a notable mathematical property: symbol emergence through MHNG is proven to...
http://arxiv.org/pdf/2305.19761v1
2023-05-31T11:46:13
2023-05-31T11:46:13
{ "title": [ -0.021087646, 0.023483276, -0.019195557, 0.00579834, -0.0071983337, -0.005050659, -0.008293152, -0.0284729, -0.003074646, 0.029541016, 0.0012969971, 0.02281189, 0.000009357929, -0.024795532, -0.020706177, -0.006641388, -0.012573242, ...
http://arxiv.org/abs/2305.19759v1
Simple yet Effective Code-Switching Language Identification with Multitask Pre-Training and Transfer Learning
Shuyue Stella Li, Cihan Xiao, Tianjian Li, Bismarck Odoom
cs.CL, eess.AS
Code-switching, also called code-mixing, is the linguistics phenomenon where in casual settings, multilingual speakers mix words from different languages in one utterance. Due to its spontaneous nature, code-switching is extremely low-resource, which makes it a challenging problem for language and speech processing tas...
http://arxiv.org/pdf/2305.19759v1
2023-05-31T11:43:16
2023-05-31T11:43:16
{ "title": [ -0.002937317, 0.018554688, -0.045013428, -0.017547607, 0.013946533, -0.024032593, -0.011405945, -0.031036377, -0.04083252, 0.037628174, 0.005619049, 0.013336182, -0.00084877014, -0.04208374, -0.0075950623, -0.010940552, -0.005558014, ...
http://arxiv.org/abs/2305.19757v1
Automatic Discrimination of Human and Neural Machine Translation in Multilingual Scenarios
Malina Chichirau, Rik van Noord, Antonio Toral
cs.CL
We tackle the task of automatically discriminating between human and machine translations. As opposed to most previous work, we perform experiments in a multilingual setting, considering multiple languages and multilingual pretrained language models. We show that a classifier trained on parallel data with a single sour...
http://arxiv.org/pdf/2305.19757v1
2023-05-31T11:41:24
2023-05-31T11:41:24
{ "title": [ -0.018371582, 0.009544373, -0.007701874, -0.026626587, -0.0016212463, 0.0049057007, -0.023727417, -0.018127441, -0.028442383, 0.06286621, 0.0019302368, 0.00015234947, 0.022872925, -0.016677856, -0.047180176, 0.014709473, -0.009185791, ...
http://arxiv.org/abs/2305.19754v1
Sentence Simplification Using Paraphrase Corpus for Initialization
Kang Liu, Jipeng Qiang
cs.CL
Neural sentence simplification method based on sequence-to-sequence framework has become the mainstream method for sentence simplification (SS) task. Unfortunately, these methods are currently limited by the scarcity of parallel SS corpus. In this paper, we focus on how to reduce the dependence on parallel corpus by le...
http://arxiv.org/pdf/2305.19754v1
2023-05-31T11:39:10
2023-05-31T11:39:10
{ "title": [ -0.029800415, -0.0077819824, -0.019927979, -0.04574585, -0.036743164, 0.01637268, -0.06286621, -0.013648987, -0.014877319, 0.06951904, -0.009269714, 0.0021038055, 0.028198242, -0.018722534, -0.040039062, 0.007850647, 0.0011377335, 0....
http://arxiv.org/abs/2305.19750v1
Text-to-Speech Pipeline for Swiss German -- A comparison
Tobias Bollinger, Jan Deriu, Manfred Vogel
cs.CL, cs.SD, eess.AS
In this work, we studied the synthesis of Swiss German speech using different Text-to-Speech (TTS) models. We evaluated the TTS models on three corpora, and we found, that VITS models performed best, hence, using them for further testing. We also introduce a new method to evaluate TTS models by letting the discriminato...
http://arxiv.org/pdf/2305.19750v1
2023-05-31T11:33:18
2023-05-31T11:33:18
{ "title": [ -0.0154953, 0.007080078, -0.046813965, -0.051086426, -0.0075645447, -0.005176544, -0.04812622, 0.018661499, -0.058807373, 0.06530762, -0.0137786865, 0.0067367554, 0.056671143, -0.038879395, -0.0065345764, 0.00082588196, -0.0031490326, ...
http://arxiv.org/abs/2305.19748v2
UKP-SQuARE: An Interactive Tool for Teaching Question Answering
Haishuo Fang, Haritz Puerto, Iryna Gurevych
cs.CL
The exponential growth of question answering (QA) has made it an indispensable topic in any Natural Language Processing (NLP) course. Additionally, the breadth of QA derived from this exponential growth makes it an ideal scenario for teaching related NLP topics such as information retrieval, explainability, and adversa...
http://arxiv.org/pdf/2305.19748v2
2023-06-02T14:18:44
2023-05-31T11:29:04
{ "title": [ -0.013183594, -0.017974854, -0.021133423, -0.036254883, -0.00422287, 0.0071754456, 0.005054474, -0.018615723, 0.03012085, 0.0211792, -0.025924683, -0.0064086914, 0.006504059, -0.010169983, -0.04812622, -0.001420021, -0.016693115, 0.0...
http://arxiv.org/abs/2305.19747v1
Analyzing Text Representations by Measuring Task Alignment
Cesar Gonzalez-Gutierrez, Audi Primadhanty, Francesco Cazzaro, Ariadna Quattoni
cs.CL
Textual representations based on pre-trained language models are key, especially in few-shot learning scenarios. What makes a representation good for text classification? Is it due to the geometric properties of the space or because it is well aligned with the task? We hypothesize the second claim. To test it, we devel...
http://arxiv.org/pdf/2305.19747v1
2023-05-31T11:20:48
2023-05-31T11:20:48
{ "title": [ -0.03164673, 0.0044021606, -0.048309326, -0.019104004, -0.0030136108, -0.013183594, -0.035583496, -0.018875122, 0.0046539307, 0.04849243, 0.004863739, 0.0073280334, -0.0026721954, -0.008232117, -0.05731201, 0.011268616, 0.0025482178, ...
http://arxiv.org/abs/2305.19734v1
Knowledge Base Question Answering for Space Debris Queries
Paul Darm, Antonio Valerio Miceli-Barone, Shay B. Cohen, Annalisa Riccardi
cs.AI, cs.CL, cs.DB, I.2.7
Space agencies execute complex satellite operations that need to be supported by the technical knowledge contained in their extensive information systems. Knowledge bases (KB) are an effective way of storing and accessing such information at scale. In this work we present a system, developed for the European Space Agen...
http://arxiv.org/pdf/2305.19734v1
2023-05-31T10:55:41
2023-05-31T10:55:41
{ "title": [ 0.0002951622, 0.0021133423, -0.030029297, -0.093811035, 0.007205963, 0.0051612854, -0.051208496, 0.0057296753, 0.0127334595, 0.033355713, -0.04006958, -0.0070381165, -0.002275467, -0.011253357, -0.044036865, -0.012825012, 0.020507812, ...
http://arxiv.org/abs/2305.19713v2
Red Teaming Language Model Detectors with Language Models
Zhouxing Shi, Yihan Wang, Fan Yin, Xiangning Chen, Kai-Wei Chang, Cho-Jui Hsieh
cs.CL, cs.LG
The prevalence and strong capability of large language models (LLMs) present significant safety and ethical risks if exploited by malicious users. To prevent the potentially deceptive usage of LLMs, recent works have proposed algorithms to detect LLM-generated text and protect LLMs. In this paper, we investigate the ro...
http://arxiv.org/pdf/2305.19713v2
2023-10-19T05:56:52
2023-05-31T10:08:37
{ "title": [ -0.036376953, 0.011886597, -0.03010559, -0.02772522, -0.020309448, -0.006439209, -0.05657959, -0.03173828, -0.0112838745, 0.031204224, -0.022705078, -0.005973816, 0.0026493073, -0.027755737, -0.023391724, 0.0053482056, -0.0135650635, ...
http://arxiv.org/abs/2305.19709v1
XPhoneBERT: A Pre-trained Multilingual Model for Phoneme Representations for Text-to-Speech
Linh The Nguyen, Thinh Pham, Dat Quoc Nguyen
cs.CL, cs.SD, eess.AS
We present XPhoneBERT, the first multilingual model pre-trained to learn phoneme representations for the downstream text-to-speech (TTS) task. Our XPhoneBERT has the same model architecture as BERT-base, trained using the RoBERTa pre-training approach on 330M phoneme-level sentences from nearly 100 languages and locale...
http://arxiv.org/pdf/2305.19709v1
2023-05-31T10:05:33
2023-05-31T10:05:33
{ "title": [ -0.006538391, -0.007980347, -0.015731812, -0.07342529, -0.0016593933, 0.018203735, -0.06518555, 0.013206482, -0.0046806335, 0.053741455, -0.00022125244, 0.010665894, 0.02168274, -0.019561768, -0.04107666, 0.014053345, -0.017547607, 0...
http://arxiv.org/abs/2305.19707v1
Building Extractive Question Answering System to Support Human-AI Health Coaching Model for Sleep Domain
Iva Bojic, Qi Chwen Ong, Shafiq Joty, Josip Car
cs.CL
Non-communicable diseases (NCDs) are a leading cause of global deaths, necessitating a focus on primary prevention and lifestyle behavior change. Health coaching, coupled with Question Answering (QA) systems, has the potential to transform preventive healthcare. This paper presents a human-Artificial Intelligence (AI) ...
http://arxiv.org/pdf/2305.19707v1
2023-05-31T10:03:18
2023-05-31T10:03:18
{ "title": [ -0.025787354, -0.013381958, -0.027816772, -0.008476257, -0.0116119385, -0.0048217773, 0.003850937, -0.0075645447, -0.015396118, 0.014953613, -0.021118164, -0.005722046, -0.030929565, -0.018615723, -0.04385376, -0.036346436, 0.011512756, ...
http://arxiv.org/abs/2305.19689v1
Assessing Word Importance Using Models Trained for Semantic Tasks
Dávid Javorský, Ondřej Bojar, François Yvon
cs.CL
Many NLP tasks require to automatically identify the most significant words in a text. In this work, we derive word significance from models trained to solve semantic task: Natural Language Inference and Paraphrase Identification. Using an attribution method aimed to explain the predictions of these models, we derive i...
http://arxiv.org/pdf/2305.19689v1
2023-05-31T09:34:26
2023-05-31T09:34:26
{ "title": [ -0.032928467, 0.0023040771, -0.064208984, -0.038269043, -0.0048446655, -0.013618469, -0.0440979, -0.0051460266, 0.005332947, 0.029220581, -0.020523071, 0.0012350082, -0.0030078888, -0.0059432983, -0.049102783, 0.00052785873, -0.018844604...
http://arxiv.org/abs/2305.19664v1
Unveiling Cross Modality Bias in Visual Question Answering: A Causal View with Possible Worlds VQA
Ali Vosoughi, Shijian Deng, Songyang Zhang, Yapeng Tian, Chenliang Xu, Jiebo Luo
cs.CV, cs.CL, cs.MM
To increase the generalization capability of VQA systems, many recent studies have tried to de-bias spurious language or vision associations that shortcut the question or image to the answer. Despite these efforts, the literature fails to address the confounding effect of vision and language simultaneously. As a result...
http://arxiv.org/pdf/2305.19664v1
2023-05-31T09:02:58
2023-05-31T09:02:58
{ "title": [ -0.013755798, -0.009101868, -0.03704834, -0.01109314, -0.027389526, -0.027038574, 0.0018482208, -0.01953125, 0.013153076, 0.010063171, -0.031433105, -0.01927185, 0.010894775, -0.031066895, -0.039886475, -0.03086853, 0.005317688, 0.04...
http://arxiv.org/abs/2305.19650v1
Adverbs, Surprisingly
Dmitry Nikolaev, Collin F. Baker, Miriam R. L. Petruck, Sebastian Padó
cs.CL
This paper begins with the premise that adverbs are neglected in computational linguistics. This view derives from two analyses: a literature review and a novel adverb dataset to probe a state-of-the-art language model, thereby uncovering systematic gaps in accounts for adverb meaning. We suggest that using Frame Seman...
http://arxiv.org/pdf/2305.19650v1
2023-05-31T08:30:08
2023-05-31T08:30:08
{ "title": [ 0.017974854, -0.013519287, -0.005722046, -0.023834229, -0.049346924, -0.012619019, -0.008674622, -0.00680542, 0.010025024, 0.017807007, -0.034332275, 0.014228821, -0.006919861, -0.03768921, 0.04827881, -0.0074272156, 0.0072517395, 0....
http://arxiv.org/abs/2305.19607v1
Adversarial Clean Label Backdoor Attacks and Defenses on Text Classification Systems
Ashim Gupta, Amrith Krishna
cs.CL, cs.CR
Clean-label (CL) attack is a form of data poisoning attack where an adversary modifies only the textual input of the training data, without requiring access to the labeling function. CL attacks are relatively unexplored in NLP, as compared to label flipping (LF) attacks, where the latter additionally requires access to...
http://arxiv.org/pdf/2305.19607v1
2023-05-31T07:23:46
2023-05-31T07:23:46
{ "title": [ -0.02017212, 0.020339966, -0.030807495, -0.033599854, 0.02810669, -0.0085372925, -0.0055274963, -0.02255249, -0.05630493, 0.03591919, 0.016860962, 0.031280518, -0.043701172, -0.02557373, -0.04171753, -0.030685425, -0.016189575, 0.050...
http://arxiv.org/abs/2305.19597v1
What does the Failure to Reason with "Respectively" in Zero/Few-Shot Settings Tell Us about Language Models?
Ruixiang Cui, Seolhwa Lee, Daniel Hershcovich, Anders Søgaard
cs.CL, cs.AI
Humans can effortlessly understand the coordinate structure of sentences such as "Niels Bohr and Kurt Cobain were born in Copenhagen and Seattle, respectively". In the context of natural language inference (NLI), we examine how language models (LMs) reason with respective readings (Gawron and Kehler, 2004) from two per...
http://arxiv.org/pdf/2305.19597v1
2023-05-31T06:45:09
2023-05-31T06:45:09
{ "title": [ -0.013282776, -0.04208374, 0.012306213, 0.010131836, -0.08306885, 0.0043296814, -0.0262146, -0.003408432, 0.013122559, 0.025314331, -0.027572632, -0.042022705, 0.01713562, -0.027191162, -0.031066895, -0.029464722, -0.010269165, 0.067...
http://arxiv.org/abs/2305.19589v1
SLABERT Talk Pretty One Day: Modeling Second Language Acquisition with BERT
Aditya Yadavalli, Alekhya Yadavalli, Vera Tobin
cs.CL
Second language acquisition (SLA) research has extensively studied cross-linguistic transfer, the influence of linguistic structure of a speaker's native language [L1] on the successful acquisition of a foreign language [L2]. Effects of such transfer can be positive (facilitating acquisition) or negative (impeding acqu...
http://arxiv.org/pdf/2305.19589v1
2023-05-31T06:22:07
2023-05-31T06:22:07
{ "title": [ -0.019866943, -0.025268555, -0.035186768, -0.008300781, -0.043060303, -0.01776123, -0.030853271, -0.00038146973, -0.039794922, 0.02609253, -0.008323669, 0.035949707, 0.01914978, -0.044433594, 0.03387451, -0.0013074875, -0.037109375, ...
http://arxiv.org/abs/2305.19585v1
LAIT: Efficient Multi-Segment Encoding in Transformers with Layer-Adjustable Interaction
Jeremiah Milbauer, Annie Louis, Mohammad Javad Hosseini, Alex Fabrikant, Donald Metzler, Tal Schuster
cs.CL, cs.LG
Transformer encoders contextualize token representations by attending to all other tokens at each layer, leading to quadratic increase in compute effort with the input length. In practice, however, the input text of many NLP tasks can be seen as a sequence of related segments (e.g., the sequence of sentences within a p...
http://arxiv.org/pdf/2305.19585v1
2023-05-31T06:09:59
2023-05-31T06:09:59
{ "title": [ -0.013664246, 0.00592041, -0.021102905, -0.06335449, 0.020950317, 0.011177063, -0.013801575, -0.008049011, -0.008964539, 0.04663086, -0.0035915375, 0.012908936, 0.018112183, -0.019622803, -0.021987915, 0.019104004, -0.001991272, 0.03...
http://arxiv.org/abs/2305.19584v1
The Tag-Team Approach: Leveraging CLS and Language Tagging for Enhancing Multilingual ASR
Kaousheik Jayakumar, Vrunda N. Sukhadia, A Arunkumar, S. Umesh
cs.CL, eess.AS
Building a multilingual Automated Speech Recognition (ASR) system in a linguistically diverse country like India can be a challenging task due to the differences in scripts and the limited availability of speech data. This problem can be solved by exploiting the fact that many of these languages are phonetically simila...
http://arxiv.org/pdf/2305.19584v1
2023-05-31T06:09:11
2023-05-31T06:09:11
{ "title": [ -0.014373779, -0.0070991516, -0.05456543, -0.0869751, -0.0115737915, 0.0012521744, -0.07287598, 0.024368286, -0.04824829, 0.046661377, -0.006767273, 0.017349243, 0.0033092499, -0.02456665, -0.022033691, -0.008728027, -0.0018529892, 0...
http://arxiv.org/abs/2305.19567v4
DC CoMix TTS: An End-to-End Expressive TTS with Discrete Code Collaborated with Mixer
Yerin Choi, Myoung-Wan Koo
cs.SD, cs.CL, eess.AS
Despite the huge successes made in neutral TTS, content-leakage remains a challenge. In this paper, we propose a new input representation and simple architecture to achieve improved prosody modeling. Inspired by the recent success in the use of discrete code in TTS, we introduce discrete code to the input of the refere...
http://arxiv.org/pdf/2305.19567v4
2023-06-28T11:42:15
2023-05-31T05:36:22
{ "title": [ 0.008262634, -0.0020160675, -0.071777344, -0.040893555, -0.011985779, -0.03744507, -0.018447876, 0.03152466, 0.0023498535, 0.03265381, -0.009162903, -0.030578613, 0.07537842, -0.0569458, 0.0017948151, -0.0051078796, -0.015617371, 0.0...
http://arxiv.org/abs/2305.19563v1
Zero-Shot Automatic Pronunciation Assessment
Hongfu Liu, Mingqian Shi, Ye Wang
cs.SD, cs.CL, cs.LG, eess.AS
Automatic Pronunciation Assessment (APA) is vital for computer-assisted language learning. Prior methods rely on annotated speech-text data to train Automatic Speech Recognition (ASR) models or speech-score data to train regression models. In this work, we propose a novel zero-shot APA method based on the pre-trained a...
http://arxiv.org/pdf/2305.19563v1
2023-05-31T05:17:17
2023-05-31T05:17:17
{ "title": [ 0.0042037964, -0.0044555664, -0.03894043, -0.051605225, -0.031158447, -0.019897461, -0.014625549, 0.038848877, 0.022506714, 0.05529785, -0.024993896, -0.0068740845, 0.02684021, -0.037506104, -0.036193848, 0.028427124, 0.0010309219, 0...
http://arxiv.org/abs/2305.19555v2
Large Language Models Are Not Strong Abstract Reasoners
Gaël Gendron, Qiming Bao, Michael Witbrock, Gillian Dobbie
cs.CL, cs.LG, I.2.2; I.2.3; I.2.7; I.5.1
Large Language Models have shown tremendous performance on a large variety of natural language processing tasks, ranging from text comprehension to common sense reasoning. However, the mechanisms responsible for this success remain opaque, and it is unclear whether LLMs can achieve human-like cognitive capabilities or ...
http://arxiv.org/pdf/2305.19555v2
2023-10-16T23:25:59
2023-05-31T04:50:29
{ "title": [ -0.015823364, -0.013031006, -0.02355957, -0.060180664, -0.032806396, 0.002811432, -0.03955078, 0.0057754517, -0.007621765, 0.032196045, -0.03237915, -0.043426514, -0.009925842, -0.0345459, -0.028045654, -0.01890564, 0.0036392212, 0.0...
http://arxiv.org/abs/2305.19549v1
Accurate and Structured Pruning for Efficient Automatic Speech Recognition
Huiqiang Jiang, Li Lyna Zhang, Yuang Li, Yu Wu, Shijie Cao, Ting Cao, Yuqing Yang, Jinyu Li, Mao Yang, Lili Qiu
cs.CL
Automatic Speech Recognition (ASR) has seen remarkable advancements with deep neural networks, such as Transformer and Conformer. However, these models typically have large model sizes and high inference costs, posing a challenge to deploy on resource-limited devices. In this paper, we propose a novel compression strat...
http://arxiv.org/pdf/2305.19549v1
2023-05-31T04:31:16
2023-05-31T04:31:16
{ "title": [ -0.010269165, -0.005756378, -0.03479004, -0.09295654, 0.013168335, 0.02078247, -0.056488037, 0.030410767, 0.0020503998, 0.028717041, -0.031341553, 0.013885498, 0.0012521744, -0.010238647, -0.04623413, -0.008224487, 0.0017194748, 0.03...
http://arxiv.org/abs/2305.19512v2
Fine-grained Text Style Transfer with Diffusion-Based Language Models
Yiwei Lyu, Tiange Luo, Jiacheng Shi, Todd C. Hollon, Honglak Lee
cs.CL, cs.AI, cs.LG
Diffusion probabilistic models have shown great success in generating high-quality images controllably, and researchers have tried to utilize this controllability into text generation domain. Previous works on diffusion-based language models have shown that they can be trained without external knowledge (such as pre-tr...
http://arxiv.org/pdf/2305.19512v2
2023-06-12T02:13:16
2023-05-31T02:51:26
{ "title": [ -0.028579712, -0.011360168, -0.043518066, -0.034362793, 0.011108398, 0.004501343, -0.017456055, 0.0058631897, 0.013534546, 0.044891357, -0.0034713745, 0.02003479, 0.023910522, -0.014099121, -0.03970337, 0.02494812, -0.004890442, 0.02...
http://arxiv.org/abs/2305.19500v1
Exploring Lottery Prompts for Pre-trained Language Models
Yulin Chen, Ning Ding, Xiaobin Wang, Shengding Hu, Hai-Tao Zheng, Zhiyuan Liu, Pengjun Xie
cs.CL
Consistently scaling pre-trained language models (PLMs) imposes substantial burdens on model adaptation, necessitating more efficient alternatives to conventional fine-tuning. Given the advantage of prompting in the zero-shot setting and the observed performance fluctuation among different prompts, we explore the insta...
http://arxiv.org/pdf/2305.19500v1
2023-05-31T02:17:04
2023-05-31T02:17:04
{ "title": [ 0.01928711, -0.010055542, 0.0002515316, -0.044158936, -0.0059661865, -0.014007568, -0.011192322, -0.0071983337, 0.004245758, 0.029190063, -0.028381348, 0.0016937256, 0.022659302, -0.019943237, -0.019989014, 0.015899658, -0.004436493, ...
http://arxiv.org/abs/2305.19497v1
Towards Flow Graph Prediction of Open-Domain Procedural Texts
Keisuke Shirai, Hirotaka Kameko, Shinsuke Mori
cs.CL
Machine comprehension of procedural texts is essential for reasoning about the steps and automating the procedures. However, this requires identifying entities within a text and resolving the relationships between the entities. Previous work focused on the cooking domain and proposed a framework to convert a recipe tex...
http://arxiv.org/pdf/2305.19497v1
2023-05-31T02:15:15
2023-05-31T02:15:15
{ "title": [ -0.023391724, -0.010643005, -0.025268555, -0.097229004, -0.026885986, 0.018600464, -0.109191895, 0.004535675, -0.015777588, 0.053741455, 0.007030487, 0.029174805, 0.0050086975, -0.005847931, -0.048980713, 0.0025291443, -0.0118255615, ...
http://arxiv.org/abs/2305.19474v1
Ethical Considerations for Machine Translation of Indigenous Languages: Giving a Voice to the Speakers
Manuel Mager, Elisabeth Mager, Katharina Kann, Ngoc Thang Vu
cs.CL
In recent years machine translation has become very successful for high-resource language pairs. This has also sparked new interest in research on the automatic translation of low-resource languages, including Indigenous languages. However, the latter are deeply related to the ethnic and cultural groups that speak (or ...
http://arxiv.org/pdf/2305.19474v1
2023-05-31T01:04:20
2023-05-31T01:04:20
{ "title": [ -0.048431396, -0.007659912, -0.010696411, 0.004573822, -0.04437256, 0.0013523102, -0.017333984, -0.01209259, -0.022201538, 0.054748535, -0.017333984, 0.0028705597, 0.042022705, -0.0256958, -0.04272461, 0.009292603, -0.012138367, 0.04...
http://arxiv.org/abs/2305.19472v2
PlaSma: Making Small Language Models Better Procedural Knowledge Models for (Counterfactual) Planning
Faeze Brahman, Chandra Bhagavatula, Valentina Pyatkin, Jena D. Hwang, Xiang Lorraine Li, Hirona J. Arai, Soumya Sanyal, Keisuke Sakaguchi, Xiang Ren, Yejin Choi
cs.CL, cs.AI, cs.LG
Procedural planning, which entails decomposing a high-level goal into a sequence of temporally ordered steps, is an important yet intricate task for machines. It involves integrating common-sense knowledge to reason about complex contextualized situations that are often counterfactual, e.g. "scheduling a doctor's appoi...
http://arxiv.org/pdf/2305.19472v2
2023-07-26T19:45:19
2023-05-31T00:55:40
{ "title": [ -0.039398193, -0.021438599, -0.04699707, -0.045135498, -0.024261475, 0.007904053, -0.07006836, -0.00868988, -0.0023002625, 0.01638794, -0.03945923, 0.007835388, -0.0109939575, -0.014198303, -0.019332886, -0.015655518, 0.011230469, 0....
http://arxiv.org/abs/2305.19466v2
The Impact of Positional Encoding on Length Generalization in Transformers
Amirhossein Kazemnejad, Inkit Padhi, Karthikeyan Natesan Ramamurthy, Payel Das, Siva Reddy
cs.CL, cs.AI, cs.LG
Length generalization, the ability to generalize from small training context sizes to larger ones, is a critical challenge in the development of Transformer-based language models. Positional encoding (PE) has been identified as a major factor influencing length generalization, but the exact impact of different PE schem...
http://arxiv.org/pdf/2305.19466v2
2023-11-06T19:48:10
2023-05-31T00:29:55
{ "title": [ -0.026016235, -0.0020961761, -0.017807007, -0.023956299, -0.017364502, -0.0024795532, -0.027008057, -0.01966858, -0.0061912537, 0.060577393, -0.019104004, 0.016647339, -0.002122879, -0.008590698, -0.04486084, -0.0023517609, 0.0009274483,...
http://arxiv.org/abs/2306.00024v1
Self-Verification Improves Few-Shot Clinical Information Extraction
Zelalem Gero, Chandan Singh, Hao Cheng, Tristan Naumann, Michel Galley, Jianfeng Gao, Hoifung Poon
cs.CL, cs.LG
Extracting patient information from unstructured text is a critical task in health decision-support and clinical research. Large language models (LLMs) have shown the potential to accelerate clinical curation via few-shot in-context learning, in contrast to supervised learning which requires much more costly human anno...
http://arxiv.org/pdf/2306.00024v1
2023-05-30T22:05:11
2023-05-30T22:05:11
{ "title": [ 0.024459839, 0.00007838011, -0.042144775, -0.074157715, -0.0025691986, -0.014793396, -0.04055786, 0.0046920776, -0.05331421, 0.027648926, -0.007209778, 0.014770508, -0.017044067, -0.040374756, -0.012161255, -0.014335632, 0.0152282715, ...
http://arxiv.org/abs/2305.19426v1
ScoNe: Benchmarking Negation Reasoning in Language Models With Fine-Tuning and In-Context Learning
Jingyuan Selena She, Christopher Potts, Samuel R. Bowman, Atticus Geiger
cs.CL, cs.LG
A number of recent benchmarks seek to assess how well models handle natural language negation. However, these benchmarks lack the controlled example paradigms that would allow us to infer whether a model had learned how negation morphemes semantically scope. To fill these analytical gaps, we present the Scoped Negation...
http://arxiv.org/pdf/2305.19426v1
2023-05-30T21:43:11
2023-05-30T21:43:11
{ "title": [ -0.009986877, -0.0256958, -0.0011301041, -0.08166504, -0.016220093, 0.02458191, -0.05718994, 0.0029067993, 0.007297516, 0.024658203, -0.048736572, 0.011161804, 0.016525269, -0.012016296, -0.02809143, -0.0029525757, -0.02255249, 0.054...
http://arxiv.org/abs/2305.19419v1
Hierarchical Multi-Instance Multi-Label Learning for Detecting Propaganda Techniques
Anni Chen, Bhuwan Dhingra
cs.CL
Since the introduction of the SemEval 2020 Task 11 (Martino et al., 2020a), several approaches have been proposed in the literature for classifying propaganda based on the rhetorical techniques used to influence readers. These methods, however, classify one span at a time, ignoring dependencies from the labels of other...
http://arxiv.org/pdf/2305.19419v1
2023-05-30T21:23:19
2023-05-30T21:23:19
{ "title": [ 0.04373169, 0.027389526, -0.020812988, -0.047088623, 0.032714844, -0.03741455, -0.0006661415, -0.049438477, -0.055633545, 0.014816284, -0.0029678345, 0.0418396, -0.006668091, -0.032165527, 0.024368286, -0.012893677, -0.03540039, 0.05...
http://arxiv.org/abs/2305.19409v1
Examining risks of racial biases in NLP tools for child protective services
Anjalie Field, Amanda Coston, Nupoor Gandhi, Alexandra Chouldechova, Emily Putnam-Hornstein, David Steier, Yulia Tsvetkov
cs.CL
Although much literature has established the presence of demographic bias in natural language processing (NLP) models, most work relies on curated bias metrics that may not be reflective of real-world applications. At the same time, practitioners are increasingly using algorithmic tools in high-stakes settings, with pa...
http://arxiv.org/pdf/2305.19409v1
2023-05-30T21:00:47
2023-05-30T21:00:47
{ "title": [ 0.018585205, -0.005970001, -0.01713562, -0.035980225, -0.016647339, -0.004600525, -0.0025959015, -0.00087451935, -0.021316528, 0.03050232, -0.054840088, 0.0045051575, 0.024215698, -0.021453857, 0.012992859, -0.03540039, -0.017028809, ...
http://arxiv.org/abs/2305.19402v2
Contextual Vision Transformers for Robust Representation Learning
Yujia Bao, Theofanis Karaletsos
cs.CV, cs.AI, cs.CL
We introduce Contextual Vision Transformers (ContextViT), a method designed to generate robust image representations for datasets experiencing shifts in latent factors across various groups. Derived from the concept of in-context learning, ContextViT incorporates an additional context token to encapsulate group-specifi...
http://arxiv.org/pdf/2305.19402v2
2023-09-28T20:01:05
2023-05-30T20:31:26
{ "title": [ 0.03366089, 0.029846191, -0.034301758, -0.07043457, 0.04397583, -0.012573242, -0.023880005, -0.038482666, 0.015380859, 0.047180176, -0.054016113, 0.014450073, 0.0095825195, 0.0027313232, -0.051361084, -0.0013227463, 0.009468079, 0.03...
http://arxiv.org/abs/2305.19396v1
Resource-Efficient Fine-Tuning Strategies for Automatic MOS Prediction in Text-to-Speech for Low-Resource Languages
Phat Do, Matt Coler, Jelske Dijkstra, Esther Klabbers
eess.AS, cs.CL
We train a MOS prediction model based on wav2vec 2.0 using the open-access data sets BVCC and SOMOS. Our test with neural TTS data in the low-resource language (LRL) West Frisian shows that pre-training on BVCC before fine-tuning on SOMOS leads to the best accuracy for both fine-tuned and zero-shot prediction. Further ...
http://arxiv.org/pdf/2305.19396v1
2023-05-30T20:19:56
2023-05-30T20:19:56
{ "title": [ -0.013847351, -0.021194458, -0.014724731, -0.09674072, -0.016815186, 0.020904541, -0.07647705, 0.035827637, -0.022338867, 0.062683105, -0.01789856, 0.012245178, 0.03479004, -0.019195557, -0.049194336, 0.012245178, -0.014060974, 0.058...
http://arxiv.org/abs/2305.19395v2
DyGen: Learning from Noisy Labels via Dynamics-Enhanced Generative Modeling
Yuchen Zhuang, Yue Yu, Lingkai Kong, Xiang Chen, Chao Zhang
cs.CL, cs.LG
Learning from noisy labels is a challenge that arises in many real-world applications where training data can contain incorrect or corrupted labels. When fine-tuning language models with noisy labels, models can easily overfit the label noise, leading to decreased performance. Most existing methods for learning from no...
http://arxiv.org/pdf/2305.19395v2
2023-06-13T05:14:45
2023-05-30T20:19:41
{ "title": [ -0.011009216, 0.002292633, -0.030197144, -0.05050659, 0.009544373, -0.009223938, -0.010421753, -0.024597168, 0.007980347, 0.016143799, -0.025421143, 0.030319214, -0.017166138, -0.02053833, -0.019119263, 0.006828308, 0.01146698, 0.047...
http://arxiv.org/abs/2305.19383v1
Quantum Natural Language Processing based Sentiment Analysis using lambeq Toolkit
Srinjoy Ganguly, Sai Nandan Morapakula, Luis Miguel Pozo Coronado
quant-ph, cs.CL
Sentiment classification is one the best use case of classical natural language processing (NLP) where we can witness its power in various daily life domains such as banking, business and marketing industry. We already know how classical AI and machine learning can change and improve technology. Quantum natural languag...
http://arxiv.org/pdf/2305.19383v1
2023-05-30T19:54:02
2023-05-30T19:54:02
{ "title": [ -0.01802063, -0.014289856, -0.037719727, -0.009155273, -0.0061073303, -0.022033691, 0.01600647, -0.009635925, -0.010627747, 0.06463623, 0.010017395, -0.024230957, 0.02319336, -0.025177002, -0.067993164, -0.00021398067, -0.011695862, ...
http://arxiv.org/abs/2306.00021v1
Explaining Hate Speech Classification with Model Agnostic Methods
Durgesh Nandini, Ute Schmid
cs.CL, cs.AI, cs.HC, cs.SI
There have been remarkable breakthroughs in Machine Learning and Artificial Intelligence, notably in the areas of Natural Language Processing and Deep Learning. Additionally, hate speech detection in dialogues has been gaining popularity among Natural Language Processing researchers with the increased use of social med...
http://arxiv.org/pdf/2306.00021v1
2023-05-30T19:52:56
2023-05-30T19:52:56
{ "title": [ 0.008979797, 0.0155181885, -0.03704834, -0.047912598, 0.014533997, -0.021331787, -0.009689331, -0.016708374, -0.028915405, 0.026229858, -0.012916565, 0.0041007996, -0.025177002, -0.014556885, -0.043151855, -0.020523071, -0.019943237, ...
http://arxiv.org/abs/2306.01779v1
Conceptual Design Generation Using Large Language Models
Kevin Ma, Daniele Grandi, Christopher McComb, Kosa Goucher-Lambert
cs.CL, cs.AI
Concept generation is a creative step in the conceptual design phase, where designers often turn to brainstorming, mindmapping, or crowdsourcing design ideas to complement their own knowledge of the domain. Recent advances in natural language processing (NLP) and machine learning (ML) have led to the rise of Large Lang...
http://arxiv.org/pdf/2306.01779v1
2023-05-30T19:32:39
2023-05-30T19:32:39
{ "title": [ -0.052093506, -0.029449463, -0.03326416, -0.052001953, -0.018432617, 0.020431519, -0.07891846, -0.005973816, 0.021835327, 0.023666382, -0.019927979, -0.0211792, -0.01423645, -0.017150879, -0.0418396, -0.004623413, -0.013504028, 0.026...
http://arxiv.org/abs/2305.19373v1
Mining Themes in Clinical Notes to Identify Phenotypes and to Predict Length of Stay in Patients admitted with Heart Failure
Ankita Agarwal, Tanvi Banerjee, William L. Romine, Krishnaprasad Thirunarayan, Lingwei Chen, Mia Cajita
cs.LG, cs.AI, cs.CL
Heart failure is a syndrome which occurs when the heart is not able to pump blood and oxygen to support other organs in the body. Identifying the underlying themes in the diagnostic codes and procedure reports of patients admitted for heart failure could reveal the clinical phenotypes associated with heart failure and ...
http://arxiv.org/pdf/2305.19373v1
2023-05-30T19:30:40
2023-05-30T19:30:40
{ "title": [ 0.027526855, -0.033416748, 0.014907837, -0.049224854, -0.03277588, 0.004421234, -0.029220581, -0.0076065063, -0.03994751, 0.029663086, -0.02796936, 0.025314331, -0.028045654, -0.013320923, -0.014778137, -0.046661377, -0.022476196, 0....
http://arxiv.org/abs/2305.19370v3
Blockwise Parallel Transformer for Large Context Models
Hao Liu, Pieter Abbeel
cs.CL, cs.LG
Transformers have emerged as the cornerstone of state-of-the-art natural language processing models, showcasing exceptional performance across a wide range of AI applications. However, the memory demands posed by the self-attention mechanism and the large feedforward network in Transformers limit their ability to handl...
http://arxiv.org/pdf/2305.19370v3
2023-08-28T20:13:33
2023-05-30T19:25:51
{ "title": [ -0.0110321045, 0.017288208, -0.06793213, -0.109802246, 0.024841309, 0.014015198, -0.07519531, -0.0017166138, 0.00021159649, 0.08380127, -0.019851685, -0.010391235, 0.054138184, -0.016311646, -0.033233643, 0.022201538, 0.009651184, 0....
http://arxiv.org/abs/2305.19358v2
Stable Anisotropic Regularization
William Rudman, Carsten Eickhoff
cs.CL, cs.AI, cs.LG
Given the success of Large Language Models (LLMs), there has been considerable interest in studying the properties of model activations. The literature overwhelmingly agrees that LLM representations are dominated by a few ``outlier dimensions'' with exceedingly high variance and magnitude. Several studies in Natural La...
http://arxiv.org/pdf/2305.19358v2
2023-09-29T21:23:14
2023-05-30T18:57:45
{ "title": [ -0.014411926, 0.015541077, -0.036590576, -0.053466797, 0.018829346, 0.0055122375, -0.035003662, -0.035125732, 0.020828247, 0.017822266, 0.010971069, 0.022277832, -0.014671326, -0.03955078, -0.004386902, 0.040130615, 0.023361206, 0.00...