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/2302.12246v3
Active Prompting with Chain-of-Thought for Large Language Models
Shizhe Diao, Pengcheng Wang, Yong Lin, Tong Zhang
cs.CL
The increasing scale of large language models (LLMs) brings emergent abilities to various complex tasks requiring reasoning, such as arithmetic and commonsense reasoning. It is known that the effective design of task-specific prompts is critical for LLMs' ability to produce high-quality answers. In particular, an effec...
http://arxiv.org/pdf/2302.12246v3
2023-05-23T15:43:28
2023-02-23T18:58:59
{ "title": [ -0.00554657, -0.013343811, 0.004085541, -0.064697266, -0.01701355, 0.010757446, -0.05050659, -0.0038166046, 0.028259277, 0.009498596, -0.027069092, -0.0017166138, -0.0019817352, -0.027786255, -0.025527954, 0.010894775, 0.009429932, -...
http://arxiv.org/abs/2302.12239v2
What Makes a Language Easy to Deep-Learn?
Lukas Galke, Yoav Ram, Limor Raviv
cs.CL, I.2.7
Neural networks drive the success of natural language processing. A fundamental property of language is its compositional structure, allowing humans to produce forms for new meanings systematically. However, unlike humans, neural networks notoriously struggle with systematic generalization, and do not necessarily benef...
http://arxiv.org/pdf/2302.12239v2
2023-09-22T15:02:07
2023-02-23T18:57:34
{ "title": [ 0.029083252, 0.013046265, -0.02268982, -0.04034424, -0.0119018555, -0.017181396, -0.04345703, -0.012123108, 0.030258179, -0.019515991, -0.034851074, -0.012863159, -0.048095703, -0.039642334, -0.005634308, -0.02734375, 0.013130188, 0....
http://arxiv.org/abs/2302.12211v1
Federated Nearest Neighbor Machine Translation
Yichao Du, Zhirui Zhang, Bingzhe Wu, Lemao Liu, Tong Xu, Enhong Chen
cs.CL, cs.AI, cs.LG
To protect user privacy and meet legal regulations, federated learning (FL) is attracting significant attention. Training neural machine translation (NMT) models with traditional FL algorithm (e.g., FedAvg) typically relies on multi-round model-based interactions. However, it is impractical and inefficient for machine ...
http://arxiv.org/pdf/2302.12211v1
2023-02-23T18:04:07
2023-02-23T18:04:07
{ "title": [ -0.043151855, -0.011245728, -0.013450623, -0.060791016, -0.0057792664, 0.024398804, -0.045166016, -0.0047683716, -0.00028061867, 0.064819336, -0.011734009, -0.009544373, 0.04119873, -0.014770508, -0.03616333, 0.02230835, 0.0109939575, ...
http://arxiv.org/abs/2302.12200v2
A Neural Span-Based Continual Named Entity Recognition Model
Yunan Zhang, Qingcai Chen
cs.CL, cs.AI
Named Entity Recognition (NER) models capable of Continual Learning (CL) are realistically valuable in areas where entity types continuously increase (e.g., personal assistants). Meanwhile the learning paradigm of NER advances to new patterns such as the span-based methods. However, its potential to CL has not been ful...
http://arxiv.org/pdf/2302.12200v2
2023-07-17T08:09:34
2023-02-23T17:51:29
{ "title": [ -0.007736206, 0.022415161, -0.049468994, -0.055541992, 0.022445679, -0.011444092, -0.021911621, 0.01399231, 0.0069847107, 0.04574585, -0.033966064, -0.0052833557, 0.010192871, -0.00069999695, -0.061828613, -0.0031585693, -0.006626129, ...
http://arxiv.org/abs/2302.12190v1
MCWDST: a Minimum-Cost Weighted Directed Spanning Tree Algorithm for Real-Time Fake News Mitigation in Social Media
Ciprian-Octavian Truică, Elena-Simona Apostol, Radu-Cătălin Nicolescu, Panagiotis Karras
cs.SI, cs.AI, cs.CL, cs.NE
The widespread availability of internet access and handheld devices confers to social media a power similar to the one newspapers used to have. People seek affordable information on social media and can reach it within seconds. Yet this convenience comes with dangers; any user may freely post whatever they please and t...
http://arxiv.org/pdf/2302.12190v1
2023-02-23T17:31:40
2023-02-23T17:31:40
{ "title": [ 0.0047302246, -0.008956909, 0.008583069, -0.04547119, -0.015159607, 0.00086164474, -0.0135650635, -0.014877319, -0.062438965, 0.025131226, -0.0037345886, 0.034698486, -0.009857178, -0.026672363, 0.0035495758, -0.032958984, -0.031555176, ...
http://arxiv.org/abs/2302.12189v3
HL Dataset: Visually-grounded Description of Scenes, Actions and Rationales
Michele Cafagna, Kees van Deemter, Albert Gatt
cs.CL, cs.CV
Current captioning datasets focus on object-centric captions, describing the visible objects in the image, e.g. "people eating food in a park". Although these datasets are useful to evaluate the ability of Vision & Language models to recognize and describe visual content, they do not support controlled experiments invo...
http://arxiv.org/pdf/2302.12189v3
2023-09-25T07:37:20
2023-02-23T17:30:18
{ "title": [ -0.0035858154, 0.009674072, -0.08087158, -0.048309326, -0.010566711, -0.049560547, -0.033233643, 0.0099105835, -0.0016307831, 0.03201294, -0.024154663, -0.007083893, 0.027664185, -0.041503906, -0.0045661926, -0.021514893, 0.017196655, ...
http://arxiv.org/abs/2302.12188v1
Simple and Scalable Nearest Neighbor Machine Translation
Yuhan Dai, Zhirui Zhang, Qiuzhi Liu, Qu Cui, Weihua Li, Yichao Du, Tong Xu
cs.CL
$k$NN-MT is a straightforward yet powerful approach for fast domain adaptation, which directly plugs pre-trained neural machine translation (NMT) models with domain-specific token-level $k$-nearest-neighbor ($k$NN) retrieval to achieve domain adaptation without retraining. Despite being conceptually attractive, $k$NN-M...
http://arxiv.org/pdf/2302.12188v1
2023-02-23T17:28:29
2023-02-23T17:28:29
{ "title": [ -0.041992188, -0.024810791, 0.0034809113, -0.057159424, 0.002210617, 0.033691406, -0.048675537, -0.011177063, 0.014915466, 0.050872803, -0.009933472, -0.002746582, 0.022064209, -0.008628845, -0.05227661, 0.01725769, 0.007091522, 0.01...
http://arxiv.org/abs/2302.12173v2
Not what you've signed up for: Compromising Real-World LLM-Integrated Applications with Indirect Prompt Injection
Kai Greshake, Sahar Abdelnabi, Shailesh Mishra, Christoph Endres, Thorsten Holz, Mario Fritz
cs.CR, cs.AI, cs.CL, cs.CY
Large Language Models (LLMs) are increasingly being integrated into various applications. The functionalities of recent LLMs can be flexibly modulated via natural language prompts. This renders them susceptible to targeted adversarial prompting, e.g., Prompt Injection (PI) attacks enable attackers to override original ...
http://arxiv.org/pdf/2302.12173v2
2023-05-05T14:26:17
2023-02-23T17:14:38
{ "title": [ -0.025558472, -0.010116577, -0.014892578, -0.013305664, -0.040649414, -0.0049324036, -0.041412354, -0.017807007, -0.047729492, -0.010559082, -0.025924683, 0.031082153, -0.054992676, -0.05291748, 0.03262329, -0.05050659, -0.009719849, ...
http://arxiv.org/abs/2302.12165v1
Prosodic features improve sentence segmentation and parsing
Elizabeth Nielsen, Sharon Goldwater, Mark Steedman
cs.CL
Parsing spoken dialogue presents challenges that parsing text does not, including a lack of clear sentence boundaries. We know from previous work that prosody helps in parsing single sentences (Tran et al. 2018), but we want to show the effect of prosody on parsing speech that isn't segmented into sentences. In experim...
http://arxiv.org/pdf/2302.12165v1
2023-02-23T17:03:36
2023-02-23T17:03:36
{ "title": [ 0.0034179688, 0.0075569153, -0.02394104, -0.055725098, -0.02079773, -0.019134521, -0.056732178, -0.0026988983, -0.020370483, 0.03845215, -0.002916336, -0.011619568, 0.020904541, -0.028701782, -0.0287323, 0.008033752, -0.002023697, 0....
http://arxiv.org/abs/2302.12128v1
On the Generalization Ability of Retrieval-Enhanced Transformers
Tobias Norlund, Ehsan Doostmohammadi, Richard Johansson, Marco Kuhlmann
cs.CL
Recent work on the Retrieval-Enhanced Transformer (RETRO) model has shown that off-loading memory from trainable weights to a retrieval database can significantly improve language modeling and match the performance of non-retrieval models that are an order of magnitude larger in size. It has been suggested that at leas...
http://arxiv.org/pdf/2302.12128v1
2023-02-23T16:11:04
2023-02-23T16:11:04
{ "title": [ 0.044891357, 0.028289795, 0.004852295, -0.05593872, 0.025634766, 0.004261017, 0.011695862, -0.037597656, 0.018127441, 0.05343628, -0.027069092, 0.0057258606, 0.0059661865, -0.0034561157, -0.06652832, 0.019897461, -0.0052604675, 0.040...
http://arxiv.org/abs/2302.12126v3
KHAN: Knowledge-Aware Hierarchical Attention Networks for Accurate Political Stance Prediction
Yunyong Ko, Seongeun Ryu, Soeun Han, Youngseung Jeon, Jaehoon Kim, Sohyun Park, Kyungsik Han, Hanghang Tong, Sang-Wook Kim
cs.CL, cs.AI, cs.LG
The political stance prediction for news articles has been widely studied to mitigate the echo chamber effect -- people fall into their thoughts and reinforce their pre-existing beliefs. The previous works for the political stance problem focus on (1) identifying political factors that could reflect the political stanc...
http://arxiv.org/pdf/2302.12126v3
2023-04-04T20:03:36
2023-02-23T16:09:42
{ "title": [ 0.024017334, 0.00712204, -0.025939941, -0.07635498, 0.004219055, -0.011886597, -0.02859497, -0.004360199, 0.015701294, 0.03652954, -0.025100708, -0.0015525818, -0.0019741058, -0.0024757385, -0.06488037, -0.004924774, 0.001036644, 0.0...
http://arxiv.org/abs/2302.13814v2
An Independent Evaluation of ChatGPT on Mathematical Word Problems (MWP)
Paulo Shakarian, Abhinav Koyyalamudi, Noel Ngu, Lakshmivihari Mareedu
cs.CL, cs.AI, cs.LG
We study the performance of a commercially available large language model (LLM) known as ChatGPT on math word problems (MWPs) from the dataset DRAW-1K. To our knowledge, this is the first independent evaluation of ChatGPT. We found that ChatGPT's performance changes dramatically based on the requirement to show its wor...
http://arxiv.org/pdf/2302.13814v2
2023-02-28T02:06:53
2023-02-23T16:06:16
{ "title": [ -0.0067749023, -0.0070381165, -0.0028324127, 0.0034770966, -0.008140564, -0.015151978, 0.018188477, -0.02507019, 0.0036792755, 0.01789856, -0.015327454, 0.022323608, -0.003118515, -0.023284912, -0.0071372986, 0.017410278, -0.034484863, ...
http://arxiv.org/abs/2302.12069v1
Deep learning model for Mongolian Citizens Feedback Analysis using Word Vector Embeddings
Zolzaya Dashdorj, Tsetsentsengel Munkhbayar, Stanislav Grigorev
cs.CL, cs.AI
A large amount of feedback was collected over the years. Many feedback analysis models have been developed focusing on the English language. Recognizing the concept of feedback is challenging and crucial in languages which do not have applicable corpus and tools employed in Natural Language Processing (i.e., vocabulary...
http://arxiv.org/pdf/2302.12069v1
2023-02-23T14:49:31
2023-02-23T14:49:31
{ "title": [ 0.027648926, 0.03237915, -0.021591187, -0.040893555, 0.026107788, -0.020889282, -0.025985718, -0.045196533, 0.017333984, 0.025650024, -0.051696777, 0.0011396408, 0.019515991, -0.014839172, -0.038391113, -0.016357422, -0.010139465, 0....
http://arxiv.org/abs/2302.12057v3
ProsAudit, a prosodic benchmark for self-supervised speech models
Maureen de Seyssel, Marvin Lavechin, Hadrien Titeux, Arthur Thomas, Gwendal Virlet, Andrea Santos Revilla, Guillaume Wisniewski, Bogdan Ludusan, Emmanuel Dupoux
cs.CL, cs.SD, eess.AS
We present ProsAudit, a benchmark in English to assess structural prosodic knowledge in self-supervised learning (SSL) speech models. It consists of two subtasks, their corresponding metrics, and an evaluation dataset. In the protosyntax task, the model must correctly identify strong versus weak prosodic boundaries. In...
http://arxiv.org/pdf/2302.12057v3
2023-06-01T08:11:15
2023-02-23T14:30:23
{ "title": [ -0.0118637085, 0.0044784546, -0.01991272, -0.037750244, -0.042510986, -0.010894775, -0.05026245, -0.0048065186, -0.0015506744, 0.04736328, 0.0033454895, -0.003917694, 0.03475952, -0.05130005, -0.01033783, 0.012145996, 0.005256653, 0....
http://arxiv.org/abs/2302.12050v1
SPINDLE: Spinning Raw Text into Lambda Terms with Graph Attention
Konstantinos Kogkalidis, Michael Moortgat, Richard Moot
cs.CL, cs.AI, cs.LG, cs.LO
This paper describes SPINDLE - an open source Python module implementing an efficient and accurate parser for written Dutch that transforms raw text input to programs for meaning composition, expressed as {\lambda} terms. The parser integrates a number of breakthrough advances made in recent years. Its output consists ...
http://arxiv.org/pdf/2302.12050v1
2023-02-23T14:22:45
2023-02-23T14:22:45
{ "title": [ -0.027450562, -0.022125244, -0.04067993, -0.04019165, -0.029708862, -0.0005259514, -0.04598999, -0.008384705, 0.008865356, 0.03250122, -0.026611328, 0.020812988, 0.021972656, -0.015319824, -0.034423828, 0.0032520294, 0.0030212402, 0....
http://arxiv.org/abs/2302.12049v1
Evaluating Automatic Speech Recognition in an Incremental Setting
Ryan Whetten, Mir Tahsin Imtiaz, Casey Kennington
cs.CL, cs.SD, eess.AS
The increasing reliability of automatic speech recognition has proliferated its everyday use. However, for research purposes, it is often unclear which model one should choose for a task, particularly if there is a requirement for speed as well as accuracy. In this paper, we systematically evaluate six speech recognize...
http://arxiv.org/pdf/2302.12049v1
2023-02-23T14:22:40
2023-02-23T14:22:40
{ "title": [ -0.0071907043, 0.005130768, -0.053222656, -0.06311035, 0.019241333, -0.017410278, -0.026947021, 0.013221741, -0.005126953, 0.026504517, -0.0635376, 0.009925842, -0.030075073, -0.002828598, -0.05239868, -0.015945435, -0.0016889572, 0....
http://arxiv.org/abs/2302.12045v1
Generative Sentiment Transfer via Adaptive Masking
Yingze Xie, Jie Xu, LiQiang Qiao, Yun Liu, Feiren Huang, Chaozhuo Li
cs.CL
Sentiment transfer aims at revising the input text to satisfy a given sentiment polarity while retaining the original semantic content. The nucleus of sentiment transfer lies in precisely separating the sentiment information from the content information. Existing explicit approaches generally identify and mask sentimen...
http://arxiv.org/pdf/2302.12045v1
2023-02-23T14:17:34
2023-02-23T14:17:34
{ "title": [ -0.005771637, 0.0013475418, -0.026611328, -0.043518066, 0.0027217865, -0.0009622574, -0.026763916, -0.01626587, 0.03729248, 0.040893555, -0.017532349, -0.0011310577, 0.014343262, -0.0129776, -0.059295654, 0.018493652, -0.006214142, 0...
http://arxiv.org/abs/2302.12044v2
Exploring Social Media for Early Detection of Depression in COVID-19 Patients
Jiageng Wu, Xian Wu, Yining Hua, Shixu Lin, Yefeng Zheng, Jie Yang
cs.CL
The COVID-19 pandemic has caused substantial damage to global health. Even though three years have passed, the world continues to struggle with the virus. Concerns are growing about the impact of COVID-19 on the mental health of infected individuals, who are more likely to experience depression, which can have long-las...
http://arxiv.org/pdf/2302.12044v2
2023-05-03T06:43:09
2023-02-23T14:13:52
{ "title": [ 0.0119018555, -0.0052871704, -0.026611328, -0.046905518, 0.012588501, -0.019210815, -0.01020813, -0.010047913, -0.038604736, 0.02897644, -0.021942139, 0.0018482208, 0.013374329, -0.03564453, -0.010536194, -0.015853882, 0.006614685, 0...
http://arxiv.org/abs/2302.12043v1
Conversational Agents and Children: Let Children Learn
Casey Kennington, Jerry Alan Fails, Katherine Landau Wright, Maria Soledad Pera
cs.CL
Using online information discovery as a case study, in this position paper we discuss the need to design, develop, and deploy (conversational) agents that can -- non-intrusively -- guide children in their quest for online resources rather than simply finding resources for them. We argue that agents should "let children...
http://arxiv.org/pdf/2302.12043v1
2023-02-23T14:12:03
2023-02-23T14:12:03
{ "title": [ 0.003583908, -0.000008404255, -0.028152466, -0.036254883, -0.017318726, -0.024734497, -0.022384644, -0.016342163, -0.021606445, 0.035491943, -0.026153564, -0.0064849854, -0.004016876, -0.024368286, -0.034210205, -0.022277832, -0.00801086...
http://arxiv.org/abs/2302.12039v1
Natural Language Processing in the Legal Domain
Daniel Martin Katz, Dirk Hartung, Lauritz Gerlach, Abhik Jana, Michael J. Bommarito II
cs.CL, cs.AI
In this paper, we summarize the current state of the field of NLP & Law with a specific focus on recent technical and substantive developments. To support our analysis, we construct and analyze a nearly complete corpus of more than six hundred NLP & Law related papers published over the past decade. Our analysis highli...
http://arxiv.org/pdf/2302.12039v1
2023-02-23T14:02:47
2023-02-23T14:02:47
{ "title": [ -0.024215698, -0.01600647, -0.027893066, -0.057891846, -0.008964539, 0.0055618286, -0.039886475, 0.0067749023, -0.002199173, 0.022216797, -0.042022705, 0.0033092499, -0.0022296906, -0.009735107, -0.03668213, 0.0015048981, -0.019699097, ...
http://arxiv.org/abs/2302.11989v1
Metric-oriented Speech Enhancement using Diffusion Probabilistic Model
Chen Chen, Yuchen Hu, Weiwei Weng, Eng Siong Chng
cs.SD, cs.CL, eess.AS
Deep neural network based speech enhancement technique focuses on learning a noisy-to-clean transformation supervised by paired training data. However, the task-specific evaluation metric (e.g., PESQ) is usually non-differentiable and can not be directly constructed in the training criteria. This mismatch between the t...
http://arxiv.org/pdf/2302.11989v1
2023-02-23T13:12:35
2023-02-23T13:12:35
{ "title": [ -0.017242432, 0.01838684, -0.07183838, -0.03488159, 0.029006958, -0.008583069, -0.005947113, 0.005355835, -0.0010528564, 0.05432129, 0.019088745, -0.016036987, 0.05630493, -0.04638672, -0.033813477, 0.0124435425, -0.01499939, 0.04565...
http://arxiv.org/abs/2302.11978v1
Does Deep Learning Learn to Abstract? A Systematic Probing Framework
Shengnan An, Zeqi Lin, Bei Chen, Qiang Fu, Nanning Zheng, Jian-Guang Lou
cs.LG, cs.CL
Abstraction is a desirable capability for deep learning models, which means to induce abstract concepts from concrete instances and flexibly apply them beyond the learning context. At the same time, there is a lack of clear understanding about both the presence and further characteristics of this capability in deep lea...
http://arxiv.org/pdf/2302.11978v1
2023-02-23T12:50:02
2023-02-23T12:50:02
{ "title": [ -0.0030403137, 0.020477295, -0.04031372, -0.053009033, 0.028320312, -0.02003479, -0.018508911, -0.023162842, 0.0023384094, 0.018432617, -0.006526947, -0.0050582886, -0.01348114, -0.036834717, -0.036743164, 0.0042533875, -0.011550903, ...
http://arxiv.org/abs/2302.11957v1
Sentence Simplification via Large Language Models
Yutao Feng, Jipeng Qiang, Yun Li, Yunhao Yuan, Yi Zhu
cs.CL, cs.AI
Sentence Simplification aims to rephrase complex sentences into simpler sentences while retaining original meaning. Large Language models (LLMs) have demonstrated the ability to perform a variety of natural language processing tasks. However, it is not yet known whether LLMs can be served as a high-quality sentence sim...
http://arxiv.org/pdf/2302.11957v1
2023-02-23T12:11:58
2023-02-23T12:11:58
{ "title": [ -0.025100708, -0.0040359497, -0.037139893, -0.07745361, 0.0098724365, 0.022125244, -0.07495117, 0.0036792755, 0.010383606, 0.044952393, 0.0020484924, 0.0029964447, 0.00356102, -0.012878418, -0.057159424, 0.016448975, -0.0020275116, 0...
http://arxiv.org/abs/2302.13782v1
Neural networks for learning personality traits from natural language
Giorgia Adorni
cs.CL, cs.AI, cs.LG
Personality is considered one of the most influential research topics in psychology, as it predicts many consequential outcomes such as mental and physical health and explains human behaviour. With the widespread use of social networks as a means of communication, it is becoming increasingly important to develop models...
http://arxiv.org/pdf/2302.13782v1
2023-02-23T10:33:40
2023-02-23T10:33:40
{ "title": [ -0.032989502, -0.005832672, -0.02746582, -0.02267456, 0.0034217834, -0.0025100708, -0.0046463013, -0.006175995, -0.007194519, 0.031051636, -0.0074005127, -0.004512787, -0.0038223267, -0.025619507, -0.044799805, 0.008850098, -0.012062073,...
http://arxiv.org/abs/2302.12258v1
Data leakage in cross-modal retrieval training: A case study
Benno Weck, Xavier Serra
cs.SD, cs.CL, cs.IR, cs.LG, eess.AS
The recent progress in text-based audio retrieval was largely propelled by the release of suitable datasets. Since the manual creation of such datasets is a laborious task, obtaining data from online resources can be a cheap solution to create large-scale datasets. We study the recently proposed SoundDesc benchmark dat...
http://arxiv.org/pdf/2302.12258v1
2023-02-23T09:51:03
2023-02-23T09:51:03
{ "title": [ 0.0101623535, 0.023651123, -0.020462036, -0.020584106, 0.023010254, -0.024108887, 0.009712219, -0.049316406, -0.021102905, 0.04421997, -0.02180481, 0.016159058, 0.007583618, -0.012245178, -0.028549194, -0.0113220215, -0.02407837, 0.0...
http://arxiv.org/abs/2302.11875v1
Improved Training of Mixture-of-Experts Language GANs
Yekun Chai, Qiyue Yin, Junge Zhang
cs.CL
Despite the dramatic success in image generation, Generative Adversarial Networks (GANs) still face great challenges in synthesizing sequences of discrete elements, in particular human language. The difficulty in generator training arises from the limited representation capacity and uninformative learning signals obtai...
http://arxiv.org/pdf/2302.11875v1
2023-02-23T09:25:46
2023-02-23T09:25:46
{ "title": [ -0.02609253, -0.0023059845, -0.03942871, -0.050598145, 0.009178162, -0.008972168, -0.035217285, 0.013076782, 0.0055732727, 0.025650024, 0.01512146, 0.008728027, 0.005859375, -0.02885437, -0.023666382, 0.026885986, -0.0032634735, 0.02...
http://arxiv.org/abs/2302.11849v1
Coarse-to-Fine Knowledge Selection for Document Grounded Dialogs
Yeqin Zhang, Haomin Fu, Cheng Fu, Haiyang Yu, Yongbin Li, Cam-Tu Nguyen
cs.CL
Multi-document grounded dialogue systems (DGDS) belong to a class of conversational agents that answer users' requests by finding supporting knowledge from a collection of documents. Most previous studies aim to improve the knowledge retrieval model or propose more effective ways to incorporate external knowledge into ...
http://arxiv.org/pdf/2302.11849v1
2023-02-23T08:28:29
2023-02-23T08:28:29
{ "title": [ -0.012451172, -0.0021533966, -0.061065674, -0.06652832, 0.007358551, -0.023864746, -0.034362793, 0.020065308, 0.015136719, 0.040527344, -0.0435791, 0.016555786, 0.0051574707, -0.001452446, -0.058135986, -0.018310547, -0.0041046143, 0...
http://arxiv.org/abs/2302.11812v1
Teacher Intervention: Improving Convergence of Quantization Aware Training for Ultra-Low Precision Transformers
Minsoo Kim, Kyuhong Shim, Seongmin Park, Wonyong Sung, Jungwook Choi
cs.CL, cs.AI
Pre-trained Transformer models such as BERT have shown great success in a wide range of applications, but at the cost of substantial increases in model complexity. Quantization-aware training (QAT) is a promising method to lower the implementation cost and energy consumption. However, aggressive quantization below 2-bi...
http://arxiv.org/pdf/2302.11812v1
2023-02-23T06:48:24
2023-02-23T06:48:24
{ "title": [ 0.014541626, 0.018966675, -0.02003479, -0.025268555, 0.008979797, -0.032348633, 0.00598526, -0.037261963, -0.028671265, 0.05517578, -0.009422302, 0.010482788, 0.026824951, -0.024154663, -0.026428223, -0.007797241, 0.0049057007, 0.040...
http://arxiv.org/abs/2302.11799v2
FiTs: Fine-grained Two-stage Training for Knowledge-aware Question Answering
Qichen Ye, Bowen Cao, Nuo Chen, Weiyuan Xu, Yuexian Zou
cs.CL, cs.LG
Knowledge-aware question answering (KAQA) requires the model to answer questions over a knowledge base, which is essential for both open-domain QA and domain-specific QA, especially when language models alone cannot provide all the knowledge needed. Despite the promising result of recent KAQA systems which tend to inte...
http://arxiv.org/pdf/2302.11799v2
2023-03-15T14:31:56
2023-02-23T06:25:51
{ "title": [ -0.010307312, -0.0141067505, -0.027755737, -0.06524658, -0.0048828125, 0.0003118515, -0.025634766, 0.023834229, 0.0049324036, 0.025222778, -0.024414062, 0.0077781677, -0.008995056, -0.00032114983, -0.05404663, -0.0030517578, 0.0047721863...
http://arxiv.org/abs/2302.11787v1
Empathetic Response Generation via Emotion Cause Transition Graph
Yushan Qian, Bo Wang, Ting-En Lin, Yinhe Zheng, Ying Zhu, Dongming Zhao, Yuexian Hou, Yuchuan Wu, Yongbin Li
cs.CL
Empathetic dialogue is a human-like behavior that requires the perception of both affective factors (e.g., emotion status) and cognitive factors (e.g., cause of the emotion). Besides concerning emotion status in early work, the latest approaches study emotion causes in empathetic dialogue. These approaches focus on und...
http://arxiv.org/pdf/2302.11787v1
2023-02-23T05:51:17
2023-02-23T05:51:17
{ "title": [ -0.00018465519, -0.008018494, -0.041534424, -0.043029785, -0.0049438477, -0.016021729, -0.017822266, 0.00070238113, -0.000957489, 0.012893677, -0.00032687187, -0.017211914, -0.0038414001, -0.04019165, -0.028625488, -0.009147644, 0.000894...
http://arxiv.org/abs/2302.11766v1
MUTANT: A Multi-sentential Code-mixed Hinglish Dataset
Rahul Gupta, Vivek Srivastava, Mayank Singh
cs.CL, cs.LG
The multi-sentential long sequence textual data unfolds several interesting research directions pertaining to natural language processing and generation. Though we observe several high-quality long-sequence datasets for English and other monolingual languages, there is no significant effort in building such resources f...
http://arxiv.org/pdf/2302.11766v1
2023-02-23T04:04:18
2023-02-23T04:04:18
{ "title": [ -0.0020923615, -0.016433716, -0.014152527, -0.022247314, -0.031066895, -0.011314392, -0.03125, -0.0023765564, 0.001871109, 0.059020996, -0.030853271, 0.014381409, 0.042236328, -0.027893066, -0.029129028, 0.018096924, -0.011436462, 0....
http://arxiv.org/abs/2302.11752v4
VLSP2022-EVJVQA Challenge: Multilingual Visual Question Answering
Ngan Luu-Thuy Nguyen, Nghia Hieu Nguyen, Duong T. D Vo, Khanh Quoc Tran, Kiet Van Nguyen
cs.CL
Visual Question Answering (VQA) is a challenging task of natural language processing (NLP) and computer vision (CV), attracting significant attention from researchers. English is a resource-rich language that has witnessed various developments in datasets and models for visual question answering. Visual question answer...
http://arxiv.org/pdf/2302.11752v4
2023-04-12T00:44:29
2023-02-23T02:38:39
{ "title": [ -0.014381409, -0.017227173, -0.02128601, -0.045806885, -0.016479492, 0.005836487, -0.040283203, -0.003768921, 0.031173706, 0.022323608, -0.018920898, -0.005290985, 0.00187397, -0.012588501, -0.03982544, 0.016448975, 0.011779785, 0.02...
http://arxiv.org/abs/2302.11713v5
Can Pre-trained Vision and Language Models Answer Visual Information-Seeking Questions?
Yang Chen, Hexiang Hu, Yi Luan, Haitian Sun, Soravit Changpinyo, Alan Ritter, Ming-Wei Chang
cs.CV, cs.AI, cs.CL
Pre-trained vision and language models have demonstrated state-of-the-art capabilities over existing tasks involving images and texts, including visual question answering. However, it remains unclear whether these models possess the capability to answer questions that are not only querying visual content but knowledge-...
http://arxiv.org/pdf/2302.11713v5
2023-10-17T14:19:13
2023-02-23T00:33:54
{ "title": [ 0.00831604, -0.014656067, 0.018463135, 0.007247925, -0.049682617, -0.024627686, 0.010719299, -0.04046631, 0.010093689, 0.012130737, -0.045166016, -0.0054893494, -0.012687683, -0.025360107, -0.020965576, -0.032592773, 0.03451538, 0.01...
http://arxiv.org/abs/2303.00923v1
On the Role of Reviewer Expertise in Temporal Review Helpfulness Prediction
Mir Tafseer Nayeem, Davood Rafiei
cs.CL, cs.IR
Helpful reviews have been essential for the success of e-commerce services, as they help customers make quick purchase decisions and benefit the merchants in their sales. While many reviews are informative, others provide little value and may contain spam, excessive appraisal, or unexpected biases. With the large volum...
http://arxiv.org/pdf/2303.00923v1
2023-02-22T23:41:22
2023-02-22T23:41:22
{ "title": [ 0.023422241, -0.00541687, -0.023910522, -0.06439209, -0.009185791, -0.024093628, -0.020385742, -0.009849548, 0.0038280487, 0.0446167, -0.041015625, 0.019165039, 0.018493652, 0.009605408, -0.064575195, 0.0038547516, 0.0021095276, 0.04...
http://arxiv.org/abs/2302.11649v2
Grounding Complex Natural Language Commands for Temporal Tasks in Unseen Environments
Jason Xinyu Liu, Ziyi Yang, Ifrah Idrees, Sam Liang, Benjamin Schornstein, Stefanie Tellex, Ankit Shah
cs.RO, cs.AI, cs.CL, cs.FL
Grounding navigational commands to linear temporal logic (LTL) leverages its unambiguous semantics for reasoning about long-horizon tasks and verifying the satisfaction of temporal constraints. Existing approaches require training data from the specific environment and landmarks that will be used in natural language to...
http://arxiv.org/pdf/2302.11649v2
2023-10-17T18:21:27
2023-02-22T20:56:40
{ "title": [ -0.009132385, -0.0057373047, -0.039123535, -0.0670166, 0.016174316, -0.019821167, -0.021133423, 0.009063721, 0.013511658, 0.02458191, -0.020477295, -0.017929077, 0.009490967, -0.020721436, -0.046905518, -0.009780884, -0.009803772, 0....
http://arxiv.org/abs/2302.13784v1
Solution for the EPO CodeFest on Green Plastics: Hierarchical multi-label classification of patents relating to green plastics using deep learning
Tingting Qiao, Gonzalo Moro Perez
cs.CL, cs.LG
This work aims at hierarchical multi-label patents classification for patents disclosing technologies related to green plastics. This is an emerging field for which there is currently no classification scheme, and hence, no labeled data is available, making this task particularly challenging. We first propose a classif...
http://arxiv.org/pdf/2302.13784v1
2023-02-22T19:06:58
2023-02-22T19:06:58
{ "title": [ 0.008934021, 0.009880066, -0.037109375, -0.049041748, 0.012886047, -0.039031982, 0.0012302399, 0.01109314, 0.0039634705, 0.030563354, -0.036010742, 0.019073486, 0.0099105835, -0.02229309, -0.0025920868, -0.024597168, -0.018676758, 0....
http://arxiv.org/abs/2302.11550v1
Scaling Robot Learning with Semantically Imagined Experience
Tianhe Yu, Ted Xiao, Austin Stone, Jonathan Tompson, Anthony Brohan, Su Wang, Jaspiar Singh, Clayton Tan, Dee M, Jodilyn Peralta, Brian Ichter, Karol Hausman, Fei Xia
cs.RO, cs.AI, cs.CL, cs.CV, cs.LG
Recent advances in robot learning have shown promise in enabling robots to perform a variety of manipulation tasks and generalize to novel scenarios. One of the key contributing factors to this progress is the scale of robot data used to train the models. To obtain large-scale datasets, prior approaches have relied on ...
http://arxiv.org/pdf/2302.11550v1
2023-02-22T18:47:51
2023-02-22T18:47:51
{ "title": [ -0.0020828247, -0.01386261, -0.01272583, -0.00041294098, -0.01033783, -0.02960205, 0.035095215, -0.017211914, -0.007881165, 0.0231781, -0.05117798, 0.011711121, -0.014892578, -0.0044898987, -0.025756836, -0.018066406, 0.0046310425, 0...
http://arxiv.org/abs/2302.11521v1
How Does In-Context Learning Help Prompt Tuning?
Simeng Sun, Yang Liu, Dan Iter, Chenguang Zhu, Mohit Iyyer
cs.CL
Fine-tuning large language models is becoming ever more impractical due to their rapidly-growing scale. This motivates the use of parameter-efficient adaptation methods such as prompt tuning (PT), which adds a small number of tunable embeddings to an otherwise frozen model, and in-context learning (ICL), in which demon...
http://arxiv.org/pdf/2302.11521v1
2023-02-22T17:45:12
2023-02-22T17:45:12
{ "title": [ 0.043182373, -0.0029449463, -0.020507812, -0.043304443, 0.004852295, -0.021408081, -0.016799927, -0.0071105957, -0.03164673, 0.016281128, -0.029129028, 0.029922485, -0.019058228, -0.011390686, 0.03363037, -0.036346436, 0.019241333, 0...
http://arxiv.org/abs/2302.11520v4
Guiding Large Language Models via Directional Stimulus Prompting
Zekun Li, Baolin Peng, Pengcheng He, Michel Galley, Jianfeng Gao, Xifeng Yan
cs.CL
We introduce Directional Stimulus Prompting, a novel framework for guiding black-box large language models (LLMs) toward specific desired outputs. Instead of directly adjusting LLMs, our method employs a small tunable policy model (e.g., T5) to generate an auxiliary directional stimulus prompt for each input instance. ...
http://arxiv.org/pdf/2302.11520v4
2023-10-09T21:01:22
2023-02-22T17:44:15
{ "title": [ -0.02217102, -0.005077362, -0.033355713, -0.0769043, -0.011634827, 0.0075950623, -0.0770874, 0.007369995, 0.007171631, 0.012588501, -0.030456543, -0.010269165, 0.010795593, -0.018829346, -0.029296875, 0.002937317, 0.00869751, 0.00489...
http://arxiv.org/abs/2302.11466v2
Advancements in Federated Learning: Models, Methods, and Privacy
Huiming Chen, Huandong Wang, Qingyue Long, Depeng Jin, Yong Li
cs.AI, cs.CL
Federated learning (FL) is a promising technique for addressing the rising privacy and security issues. Its main ingredient is to cooperatively learn the model among the distributed clients without uploading any sensitive data. In this paper, we conducted a thorough review of the related works, following the developmen...
http://arxiv.org/pdf/2302.11466v2
2023-03-05T11:29:10
2023-02-22T16:00:27
{ "title": [ -0.009918213, 0.02078247, -0.06689453, -0.070739746, 0.04067993, -0.019180298, -0.011924744, 0.0033664703, -0.053588867, 0.029876709, -0.03616333, 0.027908325, -0.015533447, -0.01033783, -0.006641388, -0.015716553, -0.0018482208, 0.0...
http://arxiv.org/abs/2302.11412v1
Data Augmentation for Neural NLP
Domagoj Pluščec, Jan Šnajder
cs.CL, cs.LG
Data scarcity is a problem that occurs in languages and tasks where we do not have large amounts of labeled data but want to use state-of-the-art models. Such models are often deep learning models that require a significant amount of data to train. Acquiring data for various machine learning problems is accompanied by ...
http://arxiv.org/pdf/2302.11412v1
2023-02-22T14:47:15
2023-02-22T14:47:15
{ "title": [ 0.0016040802, -0.0017757416, -0.048339844, -0.060577393, 0.002357483, -0.0135650635, -0.021392822, 0.011444092, 0.0048217773, 0.027496338, -0.037506104, 0.008003235, 0.0019397736, -0.0034885406, -0.035217285, 0.013595581, -0.007156372, ...
http://arxiv.org/abs/2302.11365v2
Impact of Subword Pooling Strategy on Cross-lingual Event Detection
Shantanu Agarwal, Steven Fincke, Chris Jenkins, Scott Miller, Elizabeth Boschee
cs.CL, cs.LG
Pre-trained multilingual language models (e.g., mBERT, XLM-RoBERTa) have significantly advanced the state-of-the-art for zero-shot cross-lingual information extraction. These language models ubiquitously rely on word segmentation techniques that break a word into smaller constituent subwords. Therefore, all word labeli...
http://arxiv.org/pdf/2302.11365v2
2023-02-23T02:04:27
2023-02-22T13:33:21
{ "title": [ -0.00090551376, 0.0024108887, -0.04043579, -0.050872803, 0.02519226, -0.018966675, -0.026275635, -0.019592285, -0.008834839, 0.049682617, -0.0015945435, 0.018737793, 0.04486084, -0.017105103, -0.02633667, 0.025802612, -0.014892578, 0...
http://arxiv.org/abs/2304.11058v1
Novel Intent Detection and Active Learning Based Classification (Student Abstract)
Ankan Mullick
cs.CL, cs.IR
Novel intent class detection is an important problem in real world scenario for conversational agents for continuous interaction. Several research works have been done to detect novel intents in a mono-lingual (primarily English) texts and images. But, current systems lack an end-to-end universal framework to detect no...
http://arxiv.org/pdf/2304.11058v1
2023-02-22T13:04:17
2023-02-22T13:04:17
{ "title": [ 0.0088272095, 0.026000977, -0.028869629, -0.028549194, 0.0413208, -0.04095459, 0.019439697, -0.026245117, -0.005382538, 0.030822754, -0.016082764, 0.0013713837, -0.024673462, -0.034423828, -0.021636963, 0.0011320114, -0.0002052784, 0...
http://arxiv.org/abs/2302.12095v5
On the Robustness of ChatGPT: An Adversarial and Out-of-distribution Perspective
Jindong Wang, Xixu Hu, Wenxin Hou, Hao Chen, Runkai Zheng, Yidong Wang, Linyi Yang, Haojun Huang, Wei Ye, Xiubo Geng, Binxin Jiao, Yue Zhang, Xing Xie
cs.AI, cs.CL, cs.LG
ChatGPT is a recent chatbot service released by OpenAI and is receiving increasing attention over the past few months. While evaluations of various aspects of ChatGPT have been done, its robustness, i.e., the performance to unexpected inputs, is still unclear to the public. Robustness is of particular concern in respon...
http://arxiv.org/pdf/2302.12095v5
2023-08-29T05:34:25
2023-02-22T11:01:20
{ "title": [ 0.002243042, 0.00069761276, 0.0014257431, -0.020202637, 0.019744873, 0.012535095, -0.009986877, -0.033966064, -0.011985779, 0.01902771, -0.010978699, 0.03616333, -0.031433105, -0.010131836, -0.03427124, -0.0023498535, -0.030212402, 0...
http://arxiv.org/abs/2302.11280v1
Topic-switch adapted Japanese Dialogue System based on PLATO-2
Donghuo Zeng, Jianming Wu, Yanan Wang, Kazunori Matsumoto, Gen Hattori, Kazushi Ikeda
cs.CL, cs.MM
Large-scale open-domain dialogue systems such as PLATO-2 have achieved state-of-the-art scores in both English and Chinese. However, little work explores whether such dialogue systems also work well in the Japanese language. In this work, we create a large-scale Japanese dialogue dataset, Dialogue-Graph, which contains...
http://arxiv.org/pdf/2302.11280v1
2023-02-22T10:57:59
2023-02-22T10:57:59
{ "title": [ -0.040161133, 0.032073975, -0.053131104, -0.00020861626, 0.008651733, -0.014892578, -0.027511597, -0.028945923, 0.022354126, 0.03390503, -0.0029411316, -0.02116394, 0.019363403, -0.037384033, -0.048736572, 0.0035324097, -0.008934021, ...
http://arxiv.org/abs/2302.11269v1
Learning from Multiple Sources for Data-to-Text and Text-to-Data
Song Duong, Alberto Lumbreras, Mike Gartrell, Patrick Gallinari
cs.LG, cs.AI, cs.CL, stat.ML
Data-to-text (D2T) and text-to-data (T2D) are dual tasks that convert structured data, such as graphs or tables into fluent text, and vice versa. These tasks are usually handled separately and use corpora extracted from a single source. Current systems leverage pre-trained language models fine-tuned on D2T or T2D tasks...
http://arxiv.org/pdf/2302.11269v1
2023-02-22T10:39:33
2023-02-22T10:39:33
{ "title": [ -0.018493652, -0.0069770813, -0.044677734, -0.07824707, 0.0050201416, -0.0045280457, -0.046447754, 0.0073776245, 0.00021088123, 0.030670166, -0.04498291, 0.02822876, -0.019470215, -0.011451721, -0.008964539, -0.0051956177, -0.00088071823...
http://arxiv.org/abs/2302.11224v1
MADI: Inter-domain Matching and Intra-domain Discrimination for Cross-domain Speech Recognition
Jiaming Zhou, Shiwan Zhao, Ning Jiang, Guoqing Zhao, Yong Qin
cs.CL, cs.SD, eess.AS
End-to-end automatic speech recognition (ASR) usually suffers from performance degradation when applied to a new domain due to domain shift. Unsupervised domain adaptation (UDA) aims to improve the performance on the unlabeled target domain by transferring knowledge from the source to the target domain. To improve tran...
http://arxiv.org/pdf/2302.11224v1
2023-02-22T09:11:06
2023-02-22T09:11:06
{ "title": [ -0.022399902, 0.0069847107, -0.068115234, -0.06781006, 0.03540039, -0.015220642, -0.029251099, 0.0107421875, -0.018615723, 0.03744507, -0.011390686, 0.012550354, 0.016983032, -0.024368286, -0.03048706, 0.0023555756, -0.027267456, 0.0...
http://arxiv.org/abs/2302.11199v1
Few-Shot Structured Policy Learning for Multi-Domain and Multi-Task Dialogues
Thibault Cordier, Tanguy Urvoy, Fabrice Lefevre, Lina M. Rojas-Barahona
cs.CL
Reinforcement learning has been widely adopted to model dialogue managers in task-oriented dialogues. However, the user simulator provided by state-of-the-art dialogue frameworks are only rough approximations of human behaviour. The ability to learn from a small number of human interactions is hence crucial, especially...
http://arxiv.org/pdf/2302.11199v1
2023-02-22T08:18:49
2023-02-22T08:18:49
{ "title": [ 0.01386261, -0.014579773, -0.039031982, -0.08270264, -0.0002360344, -0.0036144257, -0.043548584, 0.034942627, -0.020187378, -0.0055236816, -0.0010614395, 0.008255005, -0.041778564, -0.035125732, -0.007911682, -0.039001465, -0.009140015, ...
http://arxiv.org/abs/2302.11192v1
Improving Contextual Spelling Correction by External Acoustics Attention and Semantic Aware Data Augmentation
Xiaoqiang Wang, Yanqing Liu, Jinyu Li, Sheng Zhao
cs.SD, cs.CL, cs.LG, eess.AS
We previously proposed contextual spelling correction (CSC) to correct the output of end-to-end (E2E) automatic speech recognition (ASR) models with contextual information such as name, place, etc. Although CSC has achieved reasonable improvement in the biasing problem, there are still two drawbacks for further accurac...
http://arxiv.org/pdf/2302.11192v1
2023-02-22T08:00:08
2023-02-22T08:00:08
{ "title": [ 0.023529053, 0.014251709, -0.043121338, -0.02911377, 0.0056037903, -0.01939392, -0.0034103394, -0.005016327, 0.0093688965, 0.070617676, -0.049865723, 0.00447464, 0.05596924, 0.017120361, -0.055023193, -0.0064086914, -0.032928467, 0.0...
http://arxiv.org/abs/2302.11186v1
UML: A Universal Monolingual Output Layer for Multilingual ASR
Chao Zhang, Bo Li, Tara N. Sainath, Trevor Strohman, Shuo-yiin Chang
eess.AS, cs.CL, cs.SD
Word-piece models (WPMs) are commonly used subword units in state-of-the-art end-to-end automatic speech recognition (ASR) systems. For multilingual ASR, due to the differences in written scripts across languages, multilingual WPMs bring the challenges of having overly large output layers and scaling to more languages....
http://arxiv.org/pdf/2302.11186v1
2023-02-22T07:40:01
2023-02-22T07:40:01
{ "title": [ -0.018417358, -0.02645874, -0.045410156, -0.09466553, 0.00058603287, 0.009376526, -0.08227539, 0.024459839, -0.02520752, 0.05569458, 0.03781128, -0.014732361, 0.042388916, -0.043426514, -0.027359009, 0.026412964, 0.00018286705, 0.043...
http://arxiv.org/abs/2302.11157v1
FiNER: Financial Named Entity Recognition Dataset and Weak-Supervision Model
Agam Shah, Ruchit Vithani, Abhinav Gullapalli, Sudheer Chava
cs.CL, cs.IR
The development of annotated datasets over the 21st century has helped us truly realize the power of deep learning. Most of the datasets created for the named-entity-recognition (NER) task are not domain specific. Finance domain presents specific challenges to the NER task and a domain specific dataset would help push ...
http://arxiv.org/pdf/2302.11157v1
2023-02-22T05:41:27
2023-02-22T05:41:27
{ "title": [ -0.012763977, 0.0113220215, -0.061035156, -0.060943604, 0.024536133, -0.014198303, -0.038909912, -0.008453369, -0.012893677, 0.03378296, -0.053375244, 0.017623901, 0.022979736, -0.0006508827, -0.021224976, -0.0068740845, -0.013999939, ...
http://arxiv.org/abs/2302.11154v2
Open-domain Visual Entity Recognition: Towards Recognizing Millions of Wikipedia Entities
Hexiang Hu, Yi Luan, Yang Chen, Urvashi Khandelwal, Mandar Joshi, Kenton Lee, Kristina Toutanova, Ming-Wei Chang
cs.CV, cs.AI, cs.CL
Large-scale multi-modal pre-training models such as CLIP and PaLI exhibit strong generalization on various visual domains and tasks. However, existing image classification benchmarks often evaluate recognition on a specific domain (e.g., outdoor images) or a specific task (e.g., classifying plant species), which falls ...
http://arxiv.org/pdf/2302.11154v2
2023-02-24T00:50:25
2023-02-22T05:31:26
{ "title": [ 0.00806427, 0.034698486, -0.048339844, -0.056640625, 0.043029785, -0.011253357, -0.01637268, -0.027404785, 0.013839722, 0.02368164, -0.03942871, 0.016998291, -0.0013198853, 0.0014076233, -0.03994751, 0.0013494492, 0.0055656433, 0.030...
http://arxiv.org/abs/2302.11091v1
GTRL: An Entity Group-Aware Temporal Knowledge Graph Representation Learning Method
Xing Tang, Ling Chen
cs.LG, cs.AI, cs.CL
Temporal Knowledge Graph (TKG) representation learning embeds entities and event types into a continuous low-dimensional vector space by integrating the temporal information, which is essential for downstream tasks, e.g., event prediction and question answering. Existing methods stack multiple graph convolution layers ...
http://arxiv.org/pdf/2302.11091v1
2023-02-22T01:57:42
2023-02-22T01:57:42
{ "title": [ -0.0038604736, -0.0018310547, -0.09991455, -0.09350586, 0.032714844, -0.025634766, -0.038726807, 0.043273926, 0.0109939575, 0.03125, -0.010147095, 0.0084991455, 0.008171082, -0.015670776, -0.045776367, 0.0020599365, -0.0026607513, 0....
http://arxiv.org/abs/2302.11084v2
Test-Time Distribution Normalization for Contrastively Learned Vision-language Models
Yifei Zhou, Juntao Ren, Fengyu Li, Ramin Zabih, Ser-Nam Lim
cs.LG, cs.CL, cs.CV
Advances in the field of vision-language contrastive learning have made it possible for many downstream applications to be carried out efficiently and accurately by simply taking the dot product between image and text representations. One of the most representative approaches proposed recently known as CLIP has garnere...
http://arxiv.org/pdf/2302.11084v2
2023-10-18T23:06:06
2023-02-22T01:14:30
{ "title": [ -0.043151855, 0.0037250519, -0.045776367, -0.0284729, -0.012191772, -0.013748169, -0.020355225, -0.001490593, 0.016311646, 0.035308838, -0.04550171, 0.020019531, -0.0058898926, 0.0036582947, -0.057891846, -0.01486969, 0.010429382, 0....
http://arxiv.org/abs/2302.11074v1
Preventing Catastrophic Forgetting in Continual Learning of New Natural Language Tasks
Sudipta Kar, Giuseppe Castellucci, Simone Filice, Shervin Malmasi, Oleg Rokhlenko
cs.CL, cs.AI, cs.LG
Multi-Task Learning (MTL) is widely-accepted in Natural Language Processing as a standard technique for learning multiple related tasks in one model. Training an MTL model requires having the training data for all tasks available at the same time. As systems usually evolve over time, (e.g., to support new functionaliti...
http://arxiv.org/pdf/2302.11074v1
2023-02-22T00:18:25
2023-02-22T00:18:25
{ "title": [ -0.00749588, 0.0068855286, -0.040252686, -0.06677246, 0.009971619, -0.015487671, -0.02180481, 0.00096559525, -0.013420105, 0.03463745, -0.031402588, 0.02444458, -0.008903503, 0.0005750656, -0.042114258, -0.003068924, 0.0059814453, 0....
http://arxiv.org/abs/2302.11054v1
Conversational Text-to-SQL: An Odyssey into State-of-the-Art and Challenges Ahead
Sree Hari Krishnan Parthasarathi, Lu Zeng, Dilek Hakkani-Tur
cs.CL
Conversational, multi-turn, text-to-SQL (CoSQL) tasks map natural language utterances in a dialogue to SQL queries. State-of-the-art (SOTA) systems use large, pre-trained and finetuned language models, such as the T5-family, in conjunction with constrained decoding. With multi-tasking (MT) over coherent tasks with disc...
http://arxiv.org/pdf/2302.11054v1
2023-02-21T23:15:33
2023-02-21T23:15:33
{ "title": [ -0.009490967, -0.006881714, -0.047302246, -0.0713501, 0.021377563, -0.0007548332, -0.034332275, 0.018936157, 0.02444458, 0.048309326, -0.031951904, 0.0028686523, 0.01713562, -0.012954712, -0.06530762, -0.0027561188, 0.0007419586, 0.0...
http://arxiv.org/abs/2302.11050v1
Edgeformers: Graph-Empowered Transformers for Representation Learning on Textual-Edge Networks
Bowen Jin, Yu Zhang, Yu Meng, Jiawei Han
cs.LG, cs.CL, cs.SI
Edges in many real-world social/information networks are associated with rich text information (e.g., user-user communications or user-product reviews). However, mainstream network representation learning models focus on propagating and aggregating node attributes, lacking specific designs to utilize text semantics on ...
http://arxiv.org/pdf/2302.11050v1
2023-02-21T23:09:17
2023-02-21T23:09:17
{ "title": [ -0.012184143, 0.003856659, -0.036743164, -0.054016113, 0.009803772, 0.01259613, -0.037231445, -0.0054359436, 0.011993408, 0.02394104, -0.027862549, 0.021392822, 0.0070343018, -0.013442993, -0.029220581, 0.0060539246, -0.0053977966, 0...
http://arxiv.org/abs/2302.11042v2
In-context Example Selection with Influences
Tai Nguyen, Eric Wong
cs.CL, cs.LG
In-context learning (ICL) is a powerful paradigm emerged from large language models (LLMs). Despite its promises, ICL performance is known to be highly sensitive to input examples. In this work, we use $\textit{in-context influences}$ to analyze few-shot ICL performance directly from the in-context examples. Our propos...
http://arxiv.org/pdf/2302.11042v2
2023-06-05T17:49:58
2023-02-21T22:47:45
{ "title": [ 0.06512451, -0.0011558533, -0.024658203, -0.052642822, -0.007347107, -0.029708862, -0.017745972, -0.028549194, -0.0015745163, 0.018051147, -0.038604736, 0.014778137, 0.0015287399, -0.045043945, 0.02128601, -0.011054993, 0.015640259, ...
http://arxiv.org/abs/2302.10978v1
Learning to Retrieve Engaging Follow-Up Queries
Christopher Richardson, Sudipta Kar, Anjishnu Kumar, Anand Ramachandran, Omar Zia Khan, Zeynab Raeesy, Abhinav Sethy
cs.CL, cs.AI, cs.IR, cs.LG
Open domain conversational agents can answer a broad range of targeted queries. However, the sequential nature of interaction with these systems makes knowledge exploration a lengthy task which burdens the user with asking a chain of well phrased questions. In this paper, we present a retrieval based system and associa...
http://arxiv.org/pdf/2302.10978v1
2023-02-21T20:26:23
2023-02-21T20:26:23
{ "title": [ 0.047424316, 0.010299683, -0.022079468, -0.0703125, 0.016143799, -0.02809143, -0.01927185, -0.010864258, 0.007369995, 0.030334473, -0.024124146, 0.0009713173, -0.026550293, -0.0051193237, -0.0501709, -0.006958008, 0.011047363, 0.0179...
http://arxiv.org/abs/2302.10879v1
$k$NN-Adapter: Efficient Domain Adaptation for Black-Box Language Models
Yangsibo Huang, Daogao Liu, Zexuan Zhong, Weijia Shi, Yin Tat Lee
cs.CL
Fine-tuning a language model on a new domain is standard practice for domain adaptation. However, it can be infeasible when it comes to modern large-scale language models such as GPT-3, which can only be accessed through APIs, making it difficult to access the internal parameters of the model. In this paper, we propose...
http://arxiv.org/pdf/2302.10879v1
2023-02-21T18:54:21
2023-02-21T18:54:21
{ "title": [ -0.02519226, -0.01335907, -0.04321289, -0.0692749, -0.0059165955, -0.0039100647, -0.06439209, 0.00390625, 0.004432678, 0.023086548, -0.0013570786, 0.006954193, 0.01247406, -0.011566162, -0.03656006, 0.009414673, -0.017868042, 0.04956...
http://arxiv.org/abs/2302.10871v1
Efficient CTC Regularization via Coarse Labels for End-to-End Speech Translation
Biao Zhang, Barry Haddow, Rico Sennrich
cs.CL, cs.SD, eess.AS
For end-to-end speech translation, regularizing the encoder with the Connectionist Temporal Classification (CTC) objective using the source transcript or target translation as labels can greatly improve quality metrics. However, CTC demands an extra prediction layer over the vocabulary space, bringing in nonnegligible ...
http://arxiv.org/pdf/2302.10871v1
2023-02-21T18:38:41
2023-02-21T18:38:41
{ "title": [ -0.01977539, -0.008674622, -0.015808105, -0.07421875, -0.007156372, 0.019485474, -0.055419922, 0.02319336, -0.012886047, 0.04144287, -0.014144897, 0.015281677, 0.01398468, -0.027664185, -0.027908325, 0.012451172, -0.0023708344, 0.028...
http://arxiv.org/abs/2302.10866v3
Hyena Hierarchy: Towards Larger Convolutional Language Models
Michael Poli, Stefano Massaroli, Eric Nguyen, Daniel Y. Fu, Tri Dao, Stephen Baccus, Yoshua Bengio, Stefano Ermon, Christopher Ré
cs.LG, cs.CL
Recent advances in deep learning have relied heavily on the use of large Transformers due to their ability to learn at scale. However, the core building block of Transformers, the attention operator, exhibits quadratic cost in sequence length, limiting the amount of context accessible. Existing subquadratic methods bas...
http://arxiv.org/pdf/2302.10866v3
2023-04-19T20:08:39
2023-02-21T18:29:25
{ "title": [ -0.021850586, -0.010253906, -0.048980713, -0.06750488, 0.008377075, 0.0006008148, -0.05480957, 0.017044067, 0.0076408386, 0.029067993, -0.01763916, 0.0044288635, 0.020385742, -0.008026123, -0.032196045, 0.010795593, -0.0137786865, 0....
http://arxiv.org/abs/2302.10850v2
Offline Reinforcement Learning for Mixture-of-Expert Dialogue Management
Dhawal Gupta, Yinlam Chow, Aza Tulepbergenov, Mohammad Ghavamzadeh, Craig Boutilier
cs.LG, cs.AI, cs.CL
Reinforcement learning (RL) has shown great promise for developing dialogue management (DM) agents that are non-myopic, conduct rich conversations, and maximize overall user satisfaction. Despite recent developments in RL and language models (LMs), using RL to power conversational chatbots remains challenging, in part ...
http://arxiv.org/pdf/2302.10850v2
2023-10-29T13:05:52
2023-02-21T18:02:20
{ "title": [ -0.00024962425, -0.0073661804, -0.041656494, -0.0635376, 0.009521484, -0.013954163, -0.021408081, 0.017623901, -0.034729004, 0.015930176, 0.020599365, -0.016311646, -0.024383545, -0.042022705, -0.008171082, -0.031555176, -0.022888184, ...
http://arxiv.org/abs/2302.10845v1
TherapyView: Visualizing Therapy Sessions with Temporal Topic Modeling and AI-Generated Arts
Baihan Lin, Stefan Zecevic, Djallel Bouneffouf, Guillermo Cecchi
cs.CL, cs.AI, cs.HC, cs.LG
We present the TherapyView, a demonstration system to help therapists visualize the dynamic contents of past treatment sessions, enabled by the state-of-the-art neural topic modeling techniques to analyze the topical tendencies of various psychiatric conditions and deep learning-based image generation engine to provide...
http://arxiv.org/pdf/2302.10845v1
2023-02-21T17:53:45
2023-02-21T17:53:45
{ "title": [ 0.011108398, 0.0032424927, -0.030166626, -0.044036865, 0.0047454834, -0.0073547363, -0.01713562, -0.028076172, 0.01889038, 0.01386261, -0.00730896, 0.02357483, 0.0013828278, -0.038513184, -0.01473999, 0.010131836, 0.013366699, 0.0137...
http://arxiv.org/abs/2302.10786v1
Real-World Deployment and Evaluation of Kwame for Science, An AI Teaching Assistant for Science Education in West Africa
George Boateng, Samuel John, Samuel Boateng, Philemon Badu, Patrick Agyeman-Budu, Victor Kumbol
cs.CL, cs.CY, cs.HC, cs.IR
Africa has a high student-to-teacher ratio which limits students' access to teachers for learning support such as educational question answering. In this work, we extended Kwame, our previous AI teaching assistant for coding education, adapted it for science education, and deployed it as a web app. Kwame for Science pr...
http://arxiv.org/pdf/2302.10786v1
2023-02-21T16:20:17
2023-02-21T16:20:17
{ "title": [ -0.026992798, -0.04574585, 0.015525818, 0.051574707, -0.07495117, -0.002773285, 0.011299133, -0.021194458, 0.0015850067, 0.05166626, -0.025177002, 0.01802063, 0.09246826, -0.009880066, 0.015686035, -0.0013313293, -0.0023956299, 0.017...
http://arxiv.org/abs/2302.10724v4
ChatGPT: Jack of all trades, master of none
Jan Kocoń, Igor Cichecki, Oliwier Kaszyca, Mateusz Kochanek, Dominika Szydło, Joanna Baran, Julita Bielaniewicz, Marcin Gruza, Arkadiusz Janz, Kamil Kanclerz, Anna Kocoń, Bartłomiej Koptyra, Wiktoria Mieleszczenko-Kowszewicz, Piotr Miłkowski, Marcin Oleksy, Maciej Piasecki, Łukasz Radliński, Konrad Wojtasik, Stanisław ...
cs.CL, cs.AI, cs.CY, cs.LG
OpenAI has released the Chat Generative Pre-trained Transformer (ChatGPT) and revolutionized the approach in artificial intelligence to human-model interaction. Several publications on ChatGPT evaluation test its effectiveness on well-known natural language processing (NLP) tasks. However, the existing studies are most...
http://arxiv.org/pdf/2302.10724v4
2023-06-09T19:52:34
2023-02-21T15:20:37
{ "title": [ -0.020385742, -0.011505127, -0.019989014, 0.020339966, -0.033813477, -0.013145447, 0.0011034012, -0.040100098, -0.015853882, 0.026016235, -0.02494812, 0.053009033, -0.03643799, -0.02748108, 0.009994507, 0.019073486, -0.03515625, 0.03...
http://arxiv.org/abs/2302.10707v1
Parallel Sentence-Level Explanation Generation for Real-World Low-Resource Scenarios
Yan Liu, Xiaokang Chen, Qi Dai
cs.CL
In order to reveal the rationale behind model predictions, many works have exploited providing explanations in various forms. Recently, to further guarantee readability, more and more works turn to generate sentence-level human language explanations. However, current works pursuing sentence-level explanations rely heav...
http://arxiv.org/pdf/2302.10707v1
2023-02-21T14:52:21
2023-02-21T14:52:21
{ "title": [ -0.032073975, -0.02381897, -0.0154953, -0.07751465, -0.031402588, 0.031677246, -0.078430176, 0.0058135986, -0.019683838, 0.03100586, 0.014091492, 0.017532349, -0.011512756, -0.01928711, -0.03060913, 0.008369446, -0.010864258, 0.02720...
http://arxiv.org/abs/2302.10680v1
Generic Dependency Modeling for Multi-Party Conversation
Weizhou Shen, Xiaojun Quan, Ke Yang
cs.CL
To model the dependencies between utterances in multi-party conversations, we propose a simple and generic framework based on the dependency parsing results of utterances. Particularly, we present an approach to encoding the dependencies in the form of relative dependency encoding (ReDE) and illustrate how to implement...
http://arxiv.org/pdf/2302.10680v1
2023-02-21T13:58:19
2023-02-21T13:58:19
{ "title": [ -0.013671875, -0.0067977905, -0.06311035, -0.12927246, 0.014427185, 0.013183594, -0.11102295, 0.036621094, 0.009529114, 0.03173828, 0.021697998, -0.013648987, -0.0070228577, -0.035583496, -0.060058594, 0.001411438, -0.019439697, 0.02...
http://arxiv.org/abs/2302.10646v1
Playing the Werewolf game with artificial intelligence for language understanding
Hisaichi Shibata, Soichiro Miki, Yuta Nakamura
cs.AI, cs.CL
The Werewolf game is a social deduction game based on free natural language communication, in which players try to deceive others in order to survive. An important feature of this game is that a large portion of the conversations are false information, and the behavior of artificial intelligence (AI) in such a situatio...
http://arxiv.org/pdf/2302.10646v1
2023-02-21T13:03:20
2023-02-21T13:03:20
{ "title": [ -0.0211792, -0.0012512207, -0.023330688, -0.0071372986, -0.022201538, -0.016906738, 0.004714966, -0.029693604, -0.014816284, 0.030273438, -0.005760193, -0.00957489, 0.009010315, -0.041107178, -0.024673462, 0.003004074, -0.0072669983, ...
http://arxiv.org/abs/2302.10593v1
Connecting Humanities and Social Sciences: Applying Language and Speech Technology to Online Panel Surveys
Henk van den Heuvel, Martijn Bentum, Simone Wills, Judith C. Koops
cs.CL
In this paper, we explore the application of language and speech technology to open-ended questions in a Dutch panel survey. In an experimental wave respondents could choose to answer open questions via speech or keyboard. Automatic speech recognition (ASR) was used to process spoken responses. We evaluated answers fro...
http://arxiv.org/pdf/2302.10593v1
2023-02-21T10:52:15
2023-02-21T10:52:15
{ "title": [ -0.0090408325, 0.022415161, -0.04360962, -0.025054932, 0.036376953, -0.024856567, 0.0064430237, -0.023742676, 0.0048828125, 0.019805908, -0.024398804, -0.0038204193, -0.02255249, -0.022232056, -0.055999756, -0.020568848, -0.024154663, ...
http://arxiv.org/abs/2302.10570v1
Co-Driven Recognition of Semantic Consistency via the Fusion of Transformer and HowNet Sememes Knowledge
Fan Chen, Yan Huang, Xinfang Zhang, Kang Luo, Jinxuan Zhu, Ruixian He
cs.CL, cs.AI, cs.SC
Semantic consistency recognition aims to detect and judge whether the semantics of two text sentences are consistent with each other. However, the existing methods usually encounter the challenges of synonyms, polysemy and difficulty to understand long text. To solve the above problems, this paper proposes a co-driven ...
http://arxiv.org/pdf/2302.10570v1
2023-02-21T09:53:19
2023-02-21T09:53:19
{ "title": [ -0.028427124, 0.021621704, -0.08850098, -0.045898438, 0.0076446533, -0.020996094, -0.06021118, 0.009101868, -0.02041626, 0.054016113, -0.019180298, -0.0044136047, 0.053955078, -0.020736694, -0.020339966, -0.011566162, -0.012489319, 0...
http://arxiv.org/abs/2302.10479v1
Tell Model Where to Attend: Improving Interpretability of Aspect-Based Sentiment Classification via Small Explanation Annotations
Zhenxiao Cheng, Jie Zhou, Wen Wu, Qin Chen, Liang He
cs.CL
Gradient-based explanation methods play an important role in the field of interpreting complex deep neural networks for NLP models. However, the existing work has shown that the gradients of a model are unstable and easily manipulable, which impacts the model's reliability largely. According to our preliminary analyses...
http://arxiv.org/pdf/2302.10479v1
2023-02-21T06:55:08
2023-02-21T06:55:08
{ "title": [ 0.011116028, -0.0063285828, -0.0011663437, -0.076660156, -0.0016021729, 0.01512146, -0.042297363, -0.0045280457, 0.012046814, 0.028213501, -0.005962372, -0.022247314, -0.016220093, -0.010734558, -0.06878662, -0.012916565, -0.0062561035, ...
http://arxiv.org/abs/2302.10454v2
KG-ECO: Knowledge Graph Enhanced Entity Correction for Query Rewriting
Jinglun Cai, Mingda Li, Ziyan Jiang, Eunah Cho, Zheng Chen, Yang Liu, Xing Fan, Chenlei Guo
cs.CL, cs.LG
Query Rewriting (QR) plays a critical role in large-scale dialogue systems for reducing frictions. When there is an entity error, it imposes extra challenges for a dialogue system to produce satisfactory responses. In this work, we propose KG-ECO: Knowledge Graph enhanced Entity COrrection for query rewriting, an entit...
http://arxiv.org/pdf/2302.10454v2
2023-02-22T21:19:54
2023-02-21T05:42:06
{ "title": [ -0.024139404, -0.016983032, -0.051971436, -0.087646484, 0.009567261, 0.014213562, -0.050872803, 0.02154541, 0.0026187897, 0.03237915, -0.013877869, 0.010215759, 0.03930664, -0.013458252, -0.030212402, 0.003709793, -0.023727417, 0.026...
http://arxiv.org/abs/2302.10447v3
Mask-guided BERT for Few Shot Text Classification
Wenxiong Liao, Zhengliang Liu, Haixing Dai, Zihao Wu, Yiyang Zhang, Xiaoke Huang, Yuzhong Chen, Xi Jiang, Wei Liu, Dajiang Zhu, Tianming Liu, Sheng Li, Xiang Li, Hongmin Cai
cs.CL, cs.AI
Transformer-based language models have achieved significant success in various domains. However, the data-intensive nature of the transformer architecture requires much labeled data, which is challenging in low-resource scenarios (i.e., few-shot learning (FSL)). The main challenge of FSL is the difficulty of training r...
http://arxiv.org/pdf/2302.10447v3
2023-03-09T01:38:38
2023-02-21T05:24:00
{ "title": [ -0.018859863, -0.006801605, -0.02456665, -0.0259552, -0.022003174, -0.012260437, -0.011543274, 0.0050735474, 0.00034856796, 0.030639648, -0.050048828, 0.022201538, 0.015960693, -0.009124756, -0.031143188, -0.009414673, -0.009773254, ...
http://arxiv.org/abs/2302.10406v1
Time to Embrace Natural Language Processing (NLP)-based Digital Pathology: Benchmarking NLP- and Convolutional Neural Network-based Deep Learning Pipelines
Min Cen, Xingyu Li, Bangwei Guo, Jitendra Jonnagaddala, Hong Zhang, Xu Steven Xu
cs.CL, cs.CV, cs.LG, eess.IV, q-bio.QM
NLP-based computer vision models, particularly vision transformers, have been shown to outperform CNN models in many imaging tasks. However, most digital pathology artificial-intelligence models are based on CNN architectures, probably owing to a lack of data regarding NLP models for pathology images. In this study, we...
http://arxiv.org/pdf/2302.10406v1
2023-02-21T02:42:03
2023-02-21T02:42:03
{ "title": [ -0.013168335, -0.0034866333, -0.03778076, -0.046569824, -0.02468872, -0.019378662, -0.058410645, -0.0028686523, 0.005050659, 0.020599365, -0.025650024, -0.0054512024, 0.0013341904, -0.032928467, -0.02468872, -0.009780884, -0.010383606, ...
http://arxiv.org/abs/2302.10346v1
Exploring the Limits of Transfer Learning with Unified Model in the Cybersecurity Domain
Kuntal Kumar Pal, Kazuaki Kashihara, Ujjwala Anantheswaran, Kirby C. Kuznia, Siddhesh Jagtap, Chitta Baral
cs.CL, cs.AI, cs.CR
With the increase in cybersecurity vulnerabilities of software systems, the ways to exploit them are also increasing. Besides these, malware threats, irregular network interactions, and discussions about exploits in public forums are also on the rise. To identify these threats faster, to detect potentially relevant ent...
http://arxiv.org/pdf/2302.10346v1
2023-02-20T22:21:26
2023-02-20T22:21:26
{ "title": [ 0.000426054, 0.022888184, -0.05001831, -0.020065308, 0.018936157, -0.031066895, 0.0021209717, -0.031204224, -0.048309326, 0.030593872, -0.014465332, 0.027740479, -0.015556335, -0.027862549, -0.0004067421, -0.0077285767, -0.015548706, ...
http://arxiv.org/abs/2302.10342v1
Fantastic Rewards and How to Tame Them: A Case Study on Reward Learning for Task-oriented Dialogue Systems
Yihao Feng, Shentao Yang, Shujian Zhang, Jianguo Zhang, Caiming Xiong, Mingyuan Zhou, Huan Wang
cs.CL
When learning task-oriented dialogue (ToD) agents, reinforcement learning (RL) techniques can naturally be utilized to train dialogue strategies to achieve user-specific goals. Prior works mainly focus on adopting advanced RL techniques to train the ToD agents, while the design of the reward function is not well studie...
http://arxiv.org/pdf/2302.10342v1
2023-02-20T22:10:04
2023-02-20T22:10:04
{ "title": [ 0.011062622, 0.0030651093, -0.031402588, -0.043945312, 0.024856567, -0.02748108, 0.013442993, 0.0010709763, -0.022445679, 0.027923584, -0.013771057, 0.010002136, -0.01953125, -0.023330688, -0.037139893, -0.02330017, -0.019943237, 0.0...
http://arxiv.org/abs/2302.10322v1
Deep Transformers without Shortcuts: Modifying Self-attention for Faithful Signal Propagation
Bobby He, James Martens, Guodong Zhang, Aleksandar Botev, Andrew Brock, Samuel L Smith, Yee Whye Teh
cs.LG, cs.AI, cs.CL, stat.ML
Skip connections and normalisation layers form two standard architectural components that are ubiquitous for the training of Deep Neural Networks (DNNs), but whose precise roles are poorly understood. Recent approaches such as Deep Kernel Shaping have made progress towards reducing our reliance on them, using insights ...
http://arxiv.org/pdf/2302.10322v1
2023-02-20T21:26:25
2023-02-20T21:26:25
{ "title": [ 0.01260376, 0.021530151, -0.010299683, -0.045288086, 0.021087646, -0.005592346, -0.005859375, -0.037017822, -0.007167816, 0.048797607, -0.002325058, 0.00623703, 0.021102905, -0.02420044, -0.042175293, 0.019500732, 0.010299683, 0.0318...
http://arxiv.org/abs/2302.10166v3
Learning Deep Semantics for Test Completion
Pengyu Nie, Rahul Banerjee, Junyi Jessy Li, Raymond J. Mooney, Milos Gligoric
cs.SE, cs.CL, cs.LG
Writing tests is a time-consuming yet essential task during software development. We propose to leverage recent advances in deep learning for text and code generation to assist developers in writing tests. We formalize the novel task of test completion to automatically complete the next statement in a test method based...
http://arxiv.org/pdf/2302.10166v3
2023-03-07T06:47:30
2023-02-20T18:53:56
{ "title": [ -0.019363403, -0.009254456, -0.044525146, -0.11413574, 0.020980835, 0.01423645, -0.07165527, 0.031066895, 0.0034942627, 0.038238525, -0.029159546, 0.025360107, 0.00349617, 0.0002297163, -0.0395813, 0.009048462, -0.018203735, 0.033538...
http://arxiv.org/abs/2302.10143v1
Hashtag-Guided Low-Resource Tweet Classification
Shizhe Diao, Sedrick Scott Keh, Liangming Pan, Zhiliang Tian, Yan Song, Tong Zhang
cs.CL, cs.CY
Social media classification tasks (e.g., tweet sentiment analysis, tweet stance detection) are challenging because social media posts are typically short, informal, and ambiguous. Thus, training on tweets is challenging and demands large-scale human-annotated labels, which are time-consuming and costly to obtain. In th...
http://arxiv.org/pdf/2302.10143v1
2023-02-20T18:21:02
2023-02-20T18:21:02
{ "title": [ 0.009750366, 0.0017614365, -0.027801514, -0.061706543, 0.015899658, 0.0016832352, -0.0178833, -0.009468079, -0.005065918, 0.04586792, -0.038482666, 0.04650879, -0.027130127, 0.012329102, -0.051452637, -0.0035171509, -0.012428284, 0.0...
http://arxiv.org/abs/2302.10086v1
A Two-Sided Discussion of Preregistration of NLP Research
Anders Søgaard, Daniel Hershcovich, Miryam de Lhoneux
cs.CL
Van Miltenburg et al. (2021) suggest NLP research should adopt preregistration to prevent fishing expeditions and to promote publication of negative results. At face value, this is a very reasonable suggestion, seemingly solving many methodological problems with NLP research. We discuss pros and cons -- some old, some ...
http://arxiv.org/pdf/2302.10086v1
2023-02-20T16:48:34
2023-02-20T16:48:34
{ "title": [ 0.026641846, -0.016036987, -0.021697998, -0.04815674, -0.028060913, -0.01776123, -0.020858765, 0.0072135925, -0.077941895, 0.027572632, -0.03100586, 0.033294678, -0.018356323, -0.03074646, 0.042907715, -0.045410156, 0.003063202, 0.05...
http://arxiv.org/abs/2302.13795v1
ChatGPT: A Meta-Analysis after 2.5 Months
Christoph Leiter, Ran Zhang, Yanran Chen, Jonas Belouadi, Daniil Larionov, Vivian Fresen, Steffen Eger
cs.CL
ChatGPT, a chatbot developed by OpenAI, has gained widespread popularity and media attention since its release in November 2022. However, little hard evidence is available regarding its perception in various sources. In this paper, we analyze over 300,000 tweets and more than 150 scientific papers to investigate how Ch...
http://arxiv.org/pdf/2302.13795v1
2023-02-20T15:43:22
2023-02-20T15:43:22
{ "title": [ -0.0053482056, -0.001830101, -0.015487671, -0.008392334, -0.025222778, 0.000600338, -0.039031982, -0.030334473, -0.016464233, 0.040802002, 0.004184723, 0.019927979, 0.01940918, -0.036499023, -0.006816864, 0.03918457, -0.017105103, 0....
http://arxiv.org/abs/2302.10025v1
DINOISER: Diffused Conditional Sequence Learning by Manipulating Noises
Jiasheng Ye, Zaixiang Zheng, Yu Bao, Lihua Qian, Mingxuan Wang
cs.CL, cs.AI, cs.LG
While diffusion models have achieved great success in generating continuous signals such as images and audio, it remains elusive for diffusion models in learning discrete sequence data like natural languages. Although recent advances circumvent this challenge of discreteness by embedding discrete tokens as continuous s...
http://arxiv.org/pdf/2302.10025v1
2023-02-20T15:14:46
2023-02-20T15:14:46
{ "title": [ -0.021408081, 0.0061798096, -0.0692749, -0.06750488, 0.005130768, -0.013153076, -0.039154053, 0.0023555756, -0.002986908, 0.011886597, -0.091552734, 0.03237915, -0.0023345947, -0.0012340546, -0.01675415, -0.026428223, 0.00894165, 0.0...
http://arxiv.org/abs/2302.10016v1
Boosting classification reliability of NLP transformer models in the long run
Zoltán Kmetty, Bence Kollányi, Krisztián Boros
cs.CL, cs.LG
Transformer-based machine learning models have become an essential tool for many natural language processing (NLP) tasks since the introduction of the method. A common objective of these projects is to classify text data. Classification models are often extended to a different topic and/or time period. In these situati...
http://arxiv.org/pdf/2302.10016v1
2023-02-20T14:46:54
2023-02-20T14:46:54
{ "title": [ 0.0066452026, 0.020233154, -0.030532837, -0.048461914, 0.0099487305, -0.015930176, -0.039215088, -0.036621094, -0.006088257, 0.016342163, -0.037719727, 0.0037937164, 0.0013170242, -0.015960693, -0.03161621, -0.008972168, -0.027267456, ...
http://arxiv.org/abs/2302.09991v1
Towards Measuring and Scoring Speaker Diarization Fairness
Yannis Tevissen, Jérôme Boudy, Gérard Chollet, Frédéric Petitpont
cs.SD, cs.CL, eess.AS
Speaker diarization, or the task of finding "who spoke and when", is now used in almost every speech processing application. Nevertheless, its fairness has not yet been evaluated because there was no protocol to study its biases one by one. In this paper we propose a protocol and a scoring method designed to evaluate s...
http://arxiv.org/pdf/2302.09991v1
2023-02-20T14:07:43
2023-02-20T14:07:43
{ "title": [ 0.0003569126, 0.008880615, -0.04763794, -0.038360596, -0.0031490326, -0.014862061, -0.029663086, -0.01789856, -0.0018119812, 0.03186035, -0.047302246, 0.011177063, 0.0038528442, -0.0082092285, -0.029434204, -0.0090408325, -0.023361206, ...
http://arxiv.org/abs/2302.10205v1
Zero-Shot Information Extraction via Chatting with ChatGPT
Xiang Wei, Xingyu Cui, Ning Cheng, Xiaobin Wang, Xin Zhang, Shen Huang, Pengjun Xie, Jinan Xu, Yufeng Chen, Meishan Zhang, Yong Jiang, Wenjuan Han
cs.CL
Zero-shot information extraction (IE) aims to build IE systems from the unannotated text. It is challenging due to involving little human intervention. Challenging but worthwhile, zero-shot IE reduces the time and effort that data labeling takes. Recent efforts on large language models (LLMs, e.g., GPT-3, ChatGPT) show...
http://arxiv.org/pdf/2302.10205v1
2023-02-20T12:57:12
2023-02-20T12:57:12
{ "title": [ -0.009742737, -0.017654419, -0.03427124, -0.037506104, -0.013587952, -0.0063323975, -0.021057129, 0.0003232956, 0.011108398, 0.028762817, -0.019760132, 0.031402588, -0.0129470825, -0.012825012, -0.036254883, 0.020462036, -0.0031585693, ...
http://arxiv.org/abs/2302.09908v2
A Sidecar Separator Can Convert a Single-Talker Speech Recognition System to a Multi-Talker One
Lingwei Meng, Jiawen Kang, Mingyu Cui, Yuejiao Wang, Xixin Wu, Helen Meng
cs.SD, cs.AI, cs.CL, cs.LG, eess.AS
Although automatic speech recognition (ASR) can perform well in common non-overlapping environments, sustaining performance in multi-talker overlapping speech recognition remains challenging. Recent research revealed that ASR model's encoder captures different levels of information with different layers -- the lower la...
http://arxiv.org/pdf/2302.09908v2
2023-03-05T23:10:24
2023-02-20T11:09:37
{ "title": [ 0.021224976, -0.007297516, -0.05392456, -0.064941406, -0.006187439, -0.011940002, -0.03164673, 0.039794922, -0.04925537, 0.023071289, -0.032928467, 0.008781433, -0.01977539, -0.034820557, 0.030227661, -0.046203613, 0.0010213852, 0.01...
http://arxiv.org/abs/2302.09887v2
90% F1 Score in Relational Triple Extraction: Is it Real ?
Pratik Saini, Samiran Pal, Tapas Nayak, Indrajit Bhattacharya
cs.CL
Extracting relational triples from text is a crucial task for constructing knowledge bases. Recent advancements in joint entity and relation extraction models have demonstrated remarkable F1 scores ($\ge 90\%$) in accurately extracting relational triples from free text. However, these models have been evaluated under r...
http://arxiv.org/pdf/2302.09887v2
2023-10-27T05:14:46
2023-02-20T10:30:16
{ "title": [ -0.018447876, 0.008163452, -0.07397461, -0.07495117, 0.012184143, -0.024505615, -0.03829956, 0.0030345917, -0.026275635, 0.04360962, -0.020324707, -0.00084495544, -0.028182983, -0.013412476, -0.062683105, -0.023895264, -0.009803772, ...
http://arxiv.org/abs/2302.09885v1
Mental Health Coping Stories on Social Media: A Causal-Inference Study of Papageno Effect
Yunhao Yuan, Koustuv Saha, Barbara Keller, Erkki Tapio Isometsä, Talayeh Aledavood
cs.CY, cs.CL, cs.SI
The Papageno effect concerns how media can play a positive role in preventing and mitigating suicidal ideation and behaviors. With the increasing ubiquity and widespread use of social media, individuals often express and share lived experiences and struggles with mental health. However, there is a gap in our understand...
http://arxiv.org/pdf/2302.09885v1
2023-02-20T10:25:28
2023-02-20T10:25:28
{ "title": [ 0.043182373, 0.015945435, 0.008506775, -0.060333252, 0.027297974, -0.0027580261, -0.009048462, -0.02168274, -0.04586792, 0.05581665, -0.013191223, 0.029953003, 0.012710571, -0.019363403, -0.028640747, -0.02229309, -0.025680542, 0.008...
http://arxiv.org/abs/2302.09865v2
Can discrete information extraction prompts generalize across language models?
Nathanaël Carraz Rakotonirina, Roberto Dessì, Fabio Petroni, Sebastian Riedel, Marco Baroni
cs.CL, cs.AI, cs.LG
We study whether automatically-induced prompts that effectively extract information from a language model can also be used, out-of-the-box, to probe other language models for the same information. After confirming that discrete prompts induced with the AutoPrompt algorithm outperform manual and semi-manual prompts on t...
http://arxiv.org/pdf/2302.09865v2
2023-03-07T07:23:38
2023-02-20T09:56:51
{ "title": [ -0.019927979, 0.008201599, -0.03866577, -0.047058105, -0.02482605, -0.019210815, -0.05908203, 0.0059776306, -0.010513306, 0.032165527, -0.04776001, 0.01637268, 0.0046691895, -0.005054474, -0.03488159, -0.0069007874, -0.010543823, 0.0...
http://arxiv.org/abs/2302.09860v1
Awkward to RDataFrame and back
Ianna Osborne, Jim Pivarski
hep-ex, astro-ph.IM, cs.CL, physics.data-an
Awkward Arrays and RDataFrame provide two very different ways of performing calculations at scale. By adding the ability to zero-copy convert between them, users get the best of both. It gives users a better flexibility in mixing different packages and languages in their analysis. In Awkward Array version 2, the ak.to_...
http://arxiv.org/pdf/2302.09860v1
2023-02-20T09:41:02
2023-02-20T09:41:02
{ "title": [ 0.002380371, -0.046722412, 0.013687134, -0.025405884, -0.043060303, 0.022628784, -0.017929077, 0.02027893, 0.013191223, 0.04272461, -0.032348633, -0.011062622, -0.041534424, -0.009559631, -0.021148682, -0.024398804, 0.044769287, -0.0...
http://arxiv.org/abs/2302.09856v3
Knowledge-aware Bayesian Co-attention for Multimodal Emotion Recognition
Zihan Zhao, Yu Wang, Yanfeng Wang
cs.CL, cs.SD, eess.AS
Multimodal emotion recognition is a challenging research area that aims to fuse different modalities to predict human emotion. However, most existing models that are based on attention mechanisms have difficulty in learning emotionally relevant parts on their own. To solve this problem, we propose to incorporate extern...
http://arxiv.org/pdf/2302.09856v3
2023-03-07T07:53:35
2023-02-20T09:38:11
{ "title": [ 0.003025055, 0.016235352, -0.039916992, -0.060943604, 0.008361816, -0.0158844, -0.03414917, -0.004901886, 0.036346436, 0.033843994, -0.022018433, -0.019683838, 0.043823242, -0.023971558, -0.04272461, 0.009719849, -0.0024089813, 0.006...
http://arxiv.org/abs/2302.09852v2
Unsupervised Layer-wise Score Aggregation for Textual OOD Detection
Maxime Darrin, Guillaume Staerman, Eduardo Dadalto Câmara Gomes, Jackie CK Cheung, Pablo Piantanida, Pierre Colombo
cs.CL, cs.AI
Out-of-distribution (OOD) detection is a rapidly growing field due to new robustness and security requirements driven by an increased number of AI-based systems. Existing OOD textual detectors often rely on an anomaly score (e.g., Mahalanobis distance) computed on the embedding output of the last layer of the encoder. ...
http://arxiv.org/pdf/2302.09852v2
2023-05-29T19:31:57
2023-02-20T09:26:11
{ "title": [ -0.0030651093, 0.025909424, -0.021362305, -0.04598999, 0.033935547, -0.0065689087, -0.0019702911, -0.02482605, -0.011550903, 0.033477783, -0.0413208, 0.018753052, 0.017608643, -0.0045318604, -0.021148682, -0.012252808, -0.009506226, ...
http://arxiv.org/abs/2302.09820v1
Improving User Controlled Table-To-Text Generation Robustness
Hanxu Hu, Yunqing Liu, Zhongyi Yu, Laura Perez-Beltrachini
cs.CL
In this work we study user controlled table-to-text generation where users explore the content in a table by selecting cells and reading a natural language description thereof automatically produce by a natural language generator. Such generation models usually learn from carefully selected cell combinations (clean cel...
http://arxiv.org/pdf/2302.09820v1
2023-02-20T07:51:15
2023-02-20T07:51:15
{ "title": [ -0.0060157776, 0.010658264, -0.040893555, -0.0496521, 0.015380859, -0.0063552856, -0.04333496, -0.026397705, 0.0009098053, 0.06585693, -0.025817871, 0.022140503, 0.03591919, 0.008552551, -0.059631348, 0.0066947937, -0.00055122375, 0....
http://arxiv.org/abs/2306.17582v2
ChatGPT for Robotics: Design Principles and Model Abilities
Sai Vemprala, Rogerio Bonatti, Arthur Bucker, Ashish Kapoor
cs.AI, cs.CL, cs.HC, cs.LG, cs.RO
This paper presents an experimental study regarding the use of OpenAI's ChatGPT for robotics applications. We outline a strategy that combines design principles for prompt engineering and the creation of a high-level function library which allows ChatGPT to adapt to different robotics tasks, simulators, and form factor...
http://arxiv.org/pdf/2306.17582v2
2023-07-19T19:30:28
2023-02-20T06:39:06
{ "title": [ -0.014305115, -0.017074585, -0.011802673, -0.010864258, -0.0045204163, -0.0009880066, -0.01210022, -0.025558472, -0.020309448, 0.02784729, -0.022857666, 0.031829834, -0.009262085, -0.017410278, -0.011161804, 0.006084442, -0.0096206665, ...
http://arxiv.org/abs/2302.09775v2
Persian topic detection based on Human Word association and graph embedding
Mehrdad Ranjbar-Khadivi, Shahin Akbarpour, Mohammad-Reza Feizi-Derakhshi, Babak Anari
cs.CL, cs.AI, cs.SI
In this paper, we propose a framework to detect topics in social media based on Human Word Association. Identifying topics discussed in these media has become a critical and significant challenge. Most of the work done in this area is in English, but much has been done in the Persian language, especially microblogs wri...
http://arxiv.org/pdf/2302.09775v2
2023-07-18T10:19:50
2023-02-20T05:46:47
{ "title": [ -0.009681702, 0.01184082, -0.024932861, -0.01184845, -0.0052108765, -0.013900757, -0.025375366, -0.0435791, 0.024475098, 0.012390137, -0.017547607, 0.014167786, -0.015319824, -0.030700684, -0.016799927, -0.0011329651, -0.02017212, 0....
http://arxiv.org/abs/2302.09736v2
STOA-VLP: Spatial-Temporal Modeling of Object and Action for Video-Language Pre-training
Weihong Zhong, Mao Zheng, Duyu Tang, Xuan Luo, Heng Gong, Xiaocheng Feng, Bing Qin
cs.CV, cs.CL
Although large-scale video-language pre-training models, which usually build a global alignment between the video and the text, have achieved remarkable progress on various downstream tasks, the idea of adopting fine-grained information during the pre-training stage is not well explored. In this work, we propose STOA-V...
http://arxiv.org/pdf/2302.09736v2
2023-05-24T01:03:09
2023-02-20T03:13:45
{ "title": [ -0.0076408386, -0.0038146973, -0.048614502, -0.0814209, -0.0045433044, -0.015029907, -0.045837402, -0.0067977905, -0.007675171, 0.021591187, -0.016296387, 0.011108398, 0.001036644, -0.023345947, -0.024673462, -0.021118164, -0.0062408447,...
http://arxiv.org/abs/2302.09723v2
Emphasizing Unseen Words: New Vocabulary Acquisition for End-to-End Speech Recognition
Leyuan Qu, Cornelius Weber, Stefan Wermter
cs.CL, cs.SD, eess.AS
Due to the dynamic nature of human language, automatic speech recognition (ASR) systems need to continuously acquire new vocabulary. Out-Of-Vocabulary (OOV) words, such as trending words and new named entities, pose problems to modern ASR systems that require long training times to adapt their large numbers of paramete...
http://arxiv.org/pdf/2302.09723v2
2023-02-21T09:44:33
2023-02-20T02:21:30
{ "title": [ -0.0158844, -0.0012493134, -0.046020508, -0.062805176, 0.015411377, -0.010848999, -0.027954102, 0.019989014, -0.018814087, 0.025100708, -0.035247803, 0.027557373, -0.022384644, -0.0065193176, -0.03793335, -0.010826111, -0.01234436, 0...
http://arxiv.org/abs/2302.09715v2
What happens before and after: Multi-Event Commonsense in Event Coreference Resolution
Sahithya Ravi, Chris Tanner, Raymond Ng, Vered Shwartz
cs.CL
Event coreference models cluster event mentions pertaining to the same real-world event. Recent models rely on contextualized representations to recognize coreference among lexically or contextually similar mentions. However, models typically fail to leverage commonsense inferences, which is particularly limiting for r...
http://arxiv.org/pdf/2302.09715v2
2023-02-21T22:44:34
2023-02-20T01:51:01
{ "title": [ 0.035339355, -0.005004883, -0.026611328, -0.113708496, 0.013725281, -0.028945923, -0.04196167, 0.011985779, -0.006549835, 0.04724121, -0.0019454956, 0.026367188, 0.053222656, -0.02394104, -0.012374878, 0.030044556, -0.00026226044, 0....
http://arxiv.org/abs/2302.09685v1
Intent Identification and Entity Extraction for Healthcare Queries in Indic Languages
Ankan Mullick, Ishani Mondal, Sourjyadip Ray, R Raghav, G Sai Chaitanya, Pawan Goyal
cs.IR, cs.CL
Scarcity of data and technological limitations for resource-poor languages in developing countries like India poses a threat to the development of sophisticated NLU systems for healthcare. To assess the current status of various state-of-the-art language models in healthcare, this paper studies the problem by initially...
http://arxiv.org/pdf/2302.09685v1
2023-02-19T22:53:03
2023-02-19T22:53:03
{ "title": [ -0.0055885315, 0.013313293, -0.06640625, -0.061462402, 0.041290283, -0.019470215, -0.029510498, 0.011497498, -0.028305054, 0.030792236, -0.005458832, 0.023162842, 0.0021152496, -0.023590088, -0.017807007, -0.0025863647, 0.007297516, ...