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/2305.10160v2
Stop Uploading Test Data in Plain Text: Practical Strategies for Mitigating Data Contamination by Evaluation Benchmarks
Alon Jacovi, Avi Caciularu, Omer Goldman, Yoav Goldberg
cs.CL, cs.AI
Data contamination has become prevalent and challenging with the rise of models pretrained on large automatically-crawled corpora. For closed models, the training data becomes a trade secret, and even for open models, it is not trivial to detect contamination. Strategies such as leaderboards with hidden answers, or usi...
http://arxiv.org/pdf/2305.10160v2
2023-10-18T13:17:13
2023-05-17T12:23:38
{ "title": [ 0.007247925, 0.023834229, -0.0143966675, -0.07421875, 0.009269714, -0.010116577, -0.043914795, -0.026168823, -0.020553589, 0.03427124, -0.005493164, 0.024490356, -0.041870117, -0.03579712, -0.042175293, -0.023132324, -0.0057907104, 0...
http://arxiv.org/abs/2305.10156v3
Personality Understanding of Fictional Characters during Book Reading
Mo Yu, Jiangnan Li, Shunyu Yao, Wenjie Pang, Xiaochen Zhou, Zhou Xiao, Fandong Meng, Jie Zhou
cs.CL, cs.AI
Comprehending characters' personalities is a crucial aspect of story reading. As readers engage with a story, their understanding of a character evolves based on new events and information; and multiple fine-grained aspects of personalities can be perceived. This leads to a natural problem of situated and fine-grained ...
http://arxiv.org/pdf/2305.10156v3
2023-10-29T08:57:47
2023-05-17T12:19:11
{ "title": [ -0.0055999756, -0.005012512, -0.028366089, -0.01436615, -0.015487671, -0.027130127, 0.008277893, -0.008300781, -0.025665283, 0.03164673, 0.0077438354, 0.0041923523, 0.028213501, -0.04196167, -0.009841919, 0.011940002, -0.009025574, 0...
http://arxiv.org/abs/2305.10154v1
Iterated learning and communication jointly explain efficient color naming systems
Emil Carlsson, Devdatt Dubhashi, Terry Regier
cs.CL
It has been argued that semantic systems reflect pressure for efficiency, and a current debate concerns the cultural evolutionary process that produces this pattern. We consider efficiency as instantiated in the Information Bottleneck (IB) principle, and a model of cultural evolution that combines iterated learning and...
http://arxiv.org/pdf/2305.10154v1
2023-05-17T12:18:44
2023-05-17T12:18:44
{ "title": [ -0.019241333, -0.0010175705, -0.053222656, -0.035888672, -0.028167725, -0.024902344, -0.01979065, 0.025466919, -0.016189575, 0.01940918, -0.013641357, 0.017532349, 0.007129669, -0.030838013, 0.02444458, -0.01537323, 0.013580322, 0.03...
http://arxiv.org/abs/2305.10149v1
Multi-Grained Knowledge Retrieval for End-to-End Task-Oriented Dialog
Fanqi Wan, Weizhou Shen, Ke Yang, Xiaojun Quan, Wei Bi
cs.CL
Retrieving proper domain knowledge from an external database lies at the heart of end-to-end task-oriented dialog systems to generate informative responses. Most existing systems blend knowledge retrieval with response generation and optimize them with direct supervision from reference responses, leading to suboptimal ...
http://arxiv.org/pdf/2305.10149v1
2023-05-17T12:12:46
2023-05-17T12:12:46
{ "title": [ -0.003004074, -0.0067100525, -0.0619812, -0.079711914, 0.019958496, -0.009529114, -0.045318604, 0.027236938, 0.02809143, 0.048339844, 0.0025939941, -0.012557983, 0.01965332, -0.01687622, -0.07702637, 0.004852295, -0.002046585, -0.003...
http://arxiv.org/abs/2305.10142v1
Improving Language Model Negotiation with Self-Play and In-Context Learning from AI Feedback
Yao Fu, Hao Peng, Tushar Khot, Mirella Lapata
cs.CL
We study whether multiple large language models (LLMs) can autonomously improve each other in a negotiation game by playing, reflecting, and criticizing. We are interested in this question because if LLMs were able to improve each other, it would imply the possibility of creating strong AI agents with minimal human int...
http://arxiv.org/pdf/2305.10142v1
2023-05-17T11:55:32
2023-05-17T11:55:32
{ "title": [ -0.02420044, -0.00052404404, -0.014030457, -0.023529053, -0.025344849, -0.0029087067, -0.03161621, -0.011177063, -0.038513184, 0.029724121, -0.012252808, 0.023223877, -0.024490356, -0.018753052, -0.029586792, -0.013595581, -0.002527237, ...
http://arxiv.org/abs/2305.10136v1
Additive manifesto decomposition: A policy domain aware method for understanding party positioning
Tanise Ceron, Dmitry Nikolaev, Sebastian Padó
cs.CL, cs.CY
Automatic extraction of party (dis)similarities from texts such as party election manifestos or parliamentary speeches plays an increasing role in computational political science. However, existing approaches are fundamentally limited to targeting only global party (dis)-similarity: they condense the relationship betwe...
http://arxiv.org/pdf/2305.10136v1
2023-05-17T11:39:31
2023-05-17T11:39:31
{ "title": [ 0.03842163, 0.033111572, -0.022277832, -0.061584473, 0.03564453, -0.009971619, -0.013717651, -0.040283203, -0.045928955, 0.01675415, 0.014518738, 0.034973145, -0.030227661, -0.027450562, -0.0025482178, -0.004295349, -0.011421204, 0.0...
http://arxiv.org/abs/2305.10122v1
Empirical Analysis of Oral and Nasal Vowels of Konkani
Swapnil Fadte, Edna Vaz, Atul Kr. Ojha, Ramdas Karmali, Jyoti D. Pawar
cs.CL
Konkani is a highly nasalised language which makes it unique among Indo-Aryan languages. This work investigates the acoustic-phonetic properties of Konkani oral and nasal vowels. For this study, speech samples from six speakers (3 male and 3 female) were collected. A total of 74 unique sentences were used as a part of ...
http://arxiv.org/pdf/2305.10122v1
2023-05-17T11:01:38
2023-05-17T11:01:38
{ "title": [ -0.052886963, -0.013946533, -0.052612305, 0.012702942, -0.05392456, -0.0093307495, -0.04425049, 0.012413025, -0.041625977, 0.039245605, -0.033447266, -0.00484848, 0.05593872, -0.03729248, 0.016342163, -0.0085372925, 0.014434814, 0.06...
http://arxiv.org/abs/2305.10096v1
Use of a Taxonomy of Empathetic Response Intents to Control and Interpret Empathy in Neural Chatbots
Anuradha Welivita, Pearl Pu
cs.CL, cs.AI
A recent trend in the domain of open-domain conversational agents is enabling them to converse empathetically to emotional prompts. Current approaches either follow an end-to-end approach or condition the responses on similar emotion labels to generate empathetic responses. But empathy is a broad concept that refers to...
http://arxiv.org/pdf/2305.10096v1
2023-05-17T10:03:03
2023-05-17T10:03:03
{ "title": [ -0.005340576, -0.013832092, -0.019958496, -0.0038909912, -0.011383057, -0.02130127, 0.03161621, -0.003288269, 0.00017309189, 0.02746582, 0.00258255, 0.0022563934, 0.03463745, -0.03967285, -0.024032593, 0.010917664, -0.012908936, 0.02...
http://arxiv.org/abs/2305.10046v1
Probing the Role of Positional Information in Vision-Language Models
Philipp J. Rösch, Jindřich Libovický
cs.CL, cs.CV, I.4; I.7
In most Vision-Language models (VL), the understanding of the image structure is enabled by injecting the position information (PI) about objects in the image. In our case study of LXMERT, a state-of-the-art VL model, we probe the use of the PI in the representation and study its effect on Visual Question Answering. We...
http://arxiv.org/pdf/2305.10046v1
2023-05-17T08:38:59
2023-05-17T08:38:59
{ "title": [ -0.026855469, -0.0061683655, -0.036010742, -0.032409668, -0.015106201, -0.019012451, -0.031173706, -0.019821167, 0.01914978, 0.040405273, -0.016113281, -0.018417358, 0.018234253, -0.016052246, -0.057037354, 0.0075531006, 0.012557983, ...
http://arxiv.org/abs/2305.10037v2
Can Language Models Solve Graph Problems in Natural Language?
Heng Wang, Shangbin Feng, Tianxing He, Zhaoxuan Tan, Xiaochuang Han, Yulia Tsvetkov
cs.CL, cs.AI
Large language models (LLMs) are increasingly adopted for a variety of tasks with implicit graphical structures, such as planning in robotics, multi-hop question answering or knowledge probing, structured commonsense reasoning, and more. While LLMs have advanced the state-of-the-art on these tasks with structure implic...
http://arxiv.org/pdf/2305.10037v2
2023-10-30T07:00:00
2023-05-17T08:29:21
{ "title": [ -0.045806885, -0.014434814, -0.017532349, -0.032440186, -0.05014038, 0.011558533, -0.077941895, -0.0017261505, 0.0028877258, 0.0055160522, -0.025863647, -0.019714355, -0.0039596558, -0.038208008, -0.012207031, -0.009536743, -0.012580872,...
http://arxiv.org/abs/2305.10036v3
Are You Copying My Model? Protecting the Copyright of Large Language Models for EaaS via Backdoor Watermark
Wenjun Peng, Jingwei Yi, Fangzhao Wu, Shangxi Wu, Bin Zhu, Lingjuan Lyu, Binxing Jiao, Tong Xu, Guangzhong Sun, Xing Xie
cs.CL, cs.CY
Large language models (LLMs) have demonstrated powerful capabilities in both text understanding and generation. Companies have begun to offer Embedding as a Service (EaaS) based on these LLMs, which can benefit various natural language processing (NLP) tasks for customers. However, previous studies have shown that EaaS...
http://arxiv.org/pdf/2305.10036v3
2023-06-02T06:56:29
2023-05-17T08:28:54
{ "title": [ -0.03186035, 0.0074386597, -0.035095215, -0.010520935, -0.013389587, -0.012001038, -0.042785645, -0.044647217, -0.04244995, 0.03189087, -0.015991211, 0.022415161, 0.017242432, -0.03878784, 0.006538391, -0.0154953, 0.004928589, 0.0294...
http://arxiv.org/abs/2305.10013v1
When Gradient Descent Meets Derivative-Free Optimization: A Match Made in Black-Box Scenario
Chengcheng Han, Liqing Cui, Renyu Zhu, Jianing Wang, Nuo Chen, Qiushi Sun, Xiang Li, Ming Gao
cs.CL, cs.AI
Large pre-trained language models (PLMs) have garnered significant attention for their versatility and potential for solving a wide spectrum of natural language processing (NLP) tasks. However, the cost of running these PLMs may be prohibitive. Furthermore, PLMs may not be open-sourced due to commercial considerations ...
http://arxiv.org/pdf/2305.10013v1
2023-05-17T07:48:28
2023-05-17T07:48:28
{ "title": [ 0.005302429, -0.004470825, -0.024230957, -0.05908203, 0.021057129, -0.018722534, -0.0030460358, -0.025039673, -0.008430481, 0.030471802, 0.008369446, 0.021316528, -0.02986145, -0.016174316, -0.02658081, 0.008605957, 0.010864258, 0.03...
http://arxiv.org/abs/2305.10010v1
AD-KD: Attribution-Driven Knowledge Distillation for Language Model Compression
Siyue Wu, Hongzhan Chen, Xiaojun Quan, Qifan Wang, Rui Wang
cs.CL
Knowledge distillation has attracted a great deal of interest recently to compress pre-trained language models. However, existing knowledge distillation methods suffer from two limitations. First, the student model simply imitates the teacher's behavior while ignoring the underlying reasoning. Second, these methods usu...
http://arxiv.org/pdf/2305.10010v1
2023-05-17T07:40:12
2023-05-17T07:40:12
{ "title": [ -0.04360962, -0.032806396, -0.048980713, -0.075683594, -0.018066406, 0.023925781, -0.09613037, 0.012779236, 0.016220093, 0.015838623, -0.005596161, -0.009162903, -0.0018501282, -0.02507019, -0.032562256, 0.000045657158, 0.009918213, ...
http://arxiv.org/abs/2305.10006v2
EfficientSCI: Densely Connected Network with Space-time Factorization for Large-scale Video Snapshot Compressive Imaging
Lishun Wang, Miao Cao, Xin Yuan
cs.CV, cs.CL, eess.IV
Video snapshot compressive imaging (SCI) uses a two-dimensional detector to capture consecutive video frames during a single exposure time. Following this, an efficient reconstruction algorithm needs to be designed to reconstruct the desired video frames. Although recent deep learning-based state-of-the-art (SOTA) reco...
http://arxiv.org/pdf/2305.10006v2
2023-05-18T05:13:09
2023-05-17T07:28:46
{ "title": [ -0.029449463, -0.0019207001, -0.019317627, -0.039520264, 0.0045433044, 0.009780884, -0.026473999, 0.004169464, 0.008483887, 0.01638794, 0.006248474, 0.018707275, -0.032073975, -0.040649414, -0.04031372, -0.012062073, 0.02897644, -0.0...
http://arxiv.org/abs/2305.10005v1
DinoSR: Self-Distillation and Online Clustering for Self-supervised Speech Representation Learning
Alexander H. Liu, Heng-Jui Chang, Michael Auli, Wei-Ning Hsu, James R. Glass
cs.CL
In this paper, we introduce self-distillation and online clustering for self-supervised speech representation learning (DinoSR) which combines masked language modeling, self-distillation, and online clustering. We show that these concepts complement each other and result in a strong representation learning model for sp...
http://arxiv.org/pdf/2305.10005v1
2023-05-17T07:23:46
2023-05-17T07:23:46
{ "title": [ 0.010787964, 0.01727295, -0.05014038, -0.07904053, 0.019577026, -0.0026226044, -0.057128906, 0.006298065, -0.011962891, 0.036346436, -0.008522034, 0.018951416, -0.014312744, -0.024429321, -0.027038574, 0.008552551, 0.0018873215, 0.04...
http://arxiv.org/abs/2305.09993v1
Reprompting: Automated Chain-of-Thought Prompt Inference Through Gibbs Sampling
Weijia Xu, Andrzej Banburski-Fahey, Nebojsa Jojic
cs.LG, cs.AI, cs.CL
We introduce Reprompting, an iterative sampling algorithm that searches for the Chain-of-Thought (CoT) recipes for a given task without human intervention. Through Gibbs sampling, we infer CoT recipes that work consistently well for a set of training samples. Our method iteratively samples new recipes using previously ...
http://arxiv.org/pdf/2305.09993v1
2023-05-17T06:35:43
2023-05-17T06:35:43
{ "title": [ -0.027801514, -0.023086548, -0.020080566, -0.042938232, -0.01701355, 0.0118637085, -0.032592773, -0.016952515, 0.013290405, -0.009315491, -0.046081543, 0.02218628, -0.03250122, -0.01184845, -0.034240723, -0.00491333, -0.0058021545, 0...
http://arxiv.org/abs/2305.09990v1
Dual Semantic Knowledge Composed Multimodal Dialog Systems
Xiaolin Chen, Xuemeng Song, Yinwei Wei, Liqiang Nie, Tat-Seng Chua
cs.CL, cs.MM
Textual response generation is an essential task for multimodal task-oriented dialog systems.Although existing studies have achieved fruitful progress, they still suffer from two critical limitations: 1) focusing on the attribute knowledge but ignoring the relation knowledge that can reveal the correlations between dif...
http://arxiv.org/pdf/2305.09990v1
2023-05-17T06:33:26
2023-05-17T06:33:26
{ "title": [ -0.029907227, -0.011360168, -0.057250977, -0.0848999, -0.011070251, -0.0070877075, -0.055999756, 0.029937744, 0.0030021667, 0.053497314, -0.0030536652, 0.0030231476, -0.008506775, -0.012649536, -0.05493164, -0.011680603, 0.00071525574, ...
http://arxiv.org/abs/2305.09975v1
Smart Word Suggestions for Writing Assistance
Chenshuo Wang, Shaoguang Mao, Tao Ge, Wenshan Wu, Xun Wang, Yan Xia, Jonathan Tien, Dongyan Zhao
cs.CL
Enhancing word usage is a desired feature for writing assistance. To further advance research in this area, this paper introduces "Smart Word Suggestions" (SWS) task and benchmark. Unlike other works, SWS emphasizes end-to-end evaluation and presents a more realistic writing assistance scenario. This task involves iden...
http://arxiv.org/pdf/2305.09975v1
2023-05-17T06:15:41
2023-05-17T06:15:41
{ "title": [ 0.0062408447, -0.013397217, -0.00054740906, 0.00623703, -0.06677246, -0.009735107, -0.0071754456, -0.001534462, -0.02017212, 0.058258057, -0.01259613, 0.008987427, 0.031463623, -0.014801025, -0.018630981, 0.015136719, 0.02281189, 0.0...
http://arxiv.org/abs/2305.09955v2
Knowledge Card: Filling LLMs' Knowledge Gaps with Plug-in Specialized Language Models
Shangbin Feng, Weijia Shi, Yuyang Bai, Vidhisha Balachandran, Tianxing He, Yulia Tsvetkov
cs.CL
By design, large language models (LLMs) are static general-purpose models, expensive to retrain or update frequently. As they are increasingly adopted for knowledge-intensive tasks, it becomes evident that these design choices lead to failures to generate factual, relevant, and up-to-date knowledge. To this end, we pro...
http://arxiv.org/pdf/2305.09955v2
2023-10-19T03:37:39
2023-05-17T05:25:27
{ "title": [ -0.081970215, -0.0335083, -0.055511475, -0.033355713, -0.0597229, 0.0069007874, -0.090270996, 0.014297485, -0.014823914, 0.028137207, -0.054595947, 0.034362793, -0.00017797947, -0.018615723, -0.020767212, -0.016067505, -0.005279541, ...
http://arxiv.org/abs/2305.09941v4
"I'm fully who I am": Towards Centering Transgender and Non-Binary Voices to Measure Biases in Open Language Generation
Anaelia Ovalle, Palash Goyal, Jwala Dhamala, Zachary Jaggers, Kai-Wei Chang, Aram Galstyan, Richard Zemel, Rahul Gupta
cs.CL, cs.AI, cs.CY, cs.LG, I.2; I.7; K.4
Transgender and non-binary (TGNB) individuals disproportionately experience discrimination and exclusion from daily life. Given the recent popularity and adoption of language generation technologies, the potential to further marginalize this population only grows. Although a multitude of NLP fairness literature focuses...
http://arxiv.org/pdf/2305.09941v4
2023-06-01T20:42:13
2023-05-17T04:21:45
{ "title": [ -0.0340271, -0.00793457, -0.061065674, -0.04425049, -0.0109939575, -0.015113831, -0.03552246, 0.02142334, -0.0038642883, 0.045959473, -0.011955261, 0.004814148, 0.034729004, -0.02645874, -0.03152466, -0.0071907043, -0.035461426, 0.02...
http://arxiv.org/abs/2305.09900v2
Efficient Equivariant Transfer Learning from Pretrained Models
Sourya Basu, Pulkit Katdare, Prasanna Sattigeri, Vijil Chenthamarakshan, Katherine Driggs-Campbell, Payel Das, Lav R. Varshney
cs.LG, cs.AI, cs.CL, cs.CV
Efficient transfer learning algorithms are key to the success of foundation models on diverse downstream tasks even with limited data. Recent works of Basu et al. (2023) and Kaba et al. (2022) propose group averaging (equitune) and optimization-based methods, respectively, over features from group-transformed inputs to...
http://arxiv.org/pdf/2305.09900v2
2023-10-10T19:01:41
2023-05-17T02:20:34
{ "title": [ 0.0005941391, -0.0047569275, -0.048858643, -0.062072754, 0.013214111, -0.022598267, -0.029449463, -0.012634277, 0.0019683838, 0.023742676, -0.0093688965, 0.02960205, 0.00023436546, -0.02281189, -0.0051727295, 0.014755249, 0.0073051453, ...
http://arxiv.org/abs/2305.09898v1
Balancing Lexical and Semantic Quality in Abstractive Summarization
Jeewoo Sul, Yong Suk Choi
cs.CL
An important problem of the sequence-to-sequence neural models widely used in abstractive summarization is exposure bias. To alleviate this problem, re-ranking systems have been applied in recent years. Despite some performance improvements, this approach remains underexplored. Previous works have mostly specified the ...
http://arxiv.org/pdf/2305.09898v1
2023-05-17T02:18:31
2023-05-17T02:18:31
{ "title": [ -0.009132385, 0.008491516, -0.036315918, -0.07763672, -0.014190674, 0.00894165, -0.07208252, -0.016189575, 0.008262634, 0.05895996, -0.021347046, 0.027374268, -0.01626587, 0.015823364, -0.067993164, 0.012046814, -0.012229919, 0.04434...
http://arxiv.org/abs/2305.09892v1
Clustering-Aware Negative Sampling for Unsupervised Sentence Representation
Jinghao Deng, Fanqi Wan, Tao Yang, Xiaojun Quan, Rui Wang
cs.CL, cs.AI
Contrastive learning has been widely studied in sentence representation learning. However, earlier works mainly focus on the construction of positive examples, while in-batch samples are often simply treated as negative examples. This approach overlooks the importance of selecting appropriate negative examples, potenti...
http://arxiv.org/pdf/2305.09892v1
2023-05-17T02:06:47
2023-05-17T02:06:47
{ "title": [ 0.029159546, 0.0058555603, 0.0020809174, -0.04724121, 0.009986877, -0.005622864, -0.016281128, -0.02355957, 0.03302002, 0.04034424, -0.03555298, 0.027420044, 0.021972656, 0.0046043396, -0.057769775, 0.022918701, -0.007293701, 0.02896...
http://arxiv.org/abs/2305.09877v1
Semantic Similarity Measure of Natural Language Text through Machine Learning and a Keyword-Aware Cross-Encoder-Ranking Summarizer -- A Case Study Using UCGIS GIS&T Body of Knowledge
Yuanyuan Tian, Wenwen Li, Sizhe Wang, Zhining Gu
cs.CL, cs.AI, cs.IR
Initiated by the University Consortium of Geographic Information Science (UCGIS), GIS&T Body of Knowledge (BoK) is a community-driven endeavor to define, develop, and document geospatial topics related to geographic information science and technologies (GIS&T). In recent years, GIS&T BoK has undergone rigorous developm...
http://arxiv.org/pdf/2305.09877v1
2023-05-17T01:17:57
2023-05-17T01:17:57
{ "title": [ -0.051361084, 0.001162529, -0.011276245, 0.022659302, -0.02154541, 0.0027256012, -0.003660202, -0.032470703, 0.009307861, 0.052947998, -0.03564453, 0.025039673, 0.008552551, -0.006286621, -0.036102295, -0.015701294, -0.0041275024, 0....
http://arxiv.org/abs/2305.09864v1
The Jaseci Programming Paradigm and Runtime Stack: Building Scale-out Production Applications Easy and Fast
Jason Mars, Yiping Kang, Roland Daynauth, Baichuan Li, Ashish Mahendra, Krisztian Flautner, Lingjia Tang
cs.CL, cs.DC, cs.PL, cs.SE
Today's production scale-out applications include many sub-application components, such as storage backends, logging infrastructure and AI models. These components have drastically different characteristics, are required to work in collaboration, and interface with each other as microservices. This leads to increasingl...
http://arxiv.org/pdf/2305.09864v1
2023-05-17T00:34:36
2023-05-17T00:34:36
{ "title": [ -0.055511475, -0.04055786, -0.082458496, -0.044403076, -0.0110321045, -0.022735596, -0.045532227, 0.051971436, -0.066589355, 0.033843994, -0.0073509216, -0.01235199, -0.017349243, -0.036468506, -0.0007491112, -0.04067993, -0.011154175, ...
http://arxiv.org/abs/2305.09863v2
Explaining black box text modules in natural language with language models
Chandan Singh, Aliyah R. Hsu, Richard Antonello, Shailee Jain, Alexander G. Huth, Bin Yu, Jianfeng Gao
cs.AI, cs.CL, cs.LG, q-bio.NC
Large language models (LLMs) have demonstrated remarkable prediction performance for a growing array of tasks. However, their rapid proliferation and increasing opaqueness have created a growing need for interpretability. Here, we ask whether we can automatically obtain natural language explanations for black box text ...
http://arxiv.org/pdf/2305.09863v2
2023-11-15T17:19:10
2023-05-17T00:29:18
{ "title": [ -0.044128418, -0.013626099, -0.019683838, -0.07312012, -0.026473999, 0.025985718, -0.08081055, -0.0013237, -0.011940002, 0.01687622, -0.027191162, 0.011543274, -0.019989014, -0.0026054382, -0.033966064, -0.015258789, -0.032928467, 0....
http://arxiv.org/abs/2305.09860v2
Epsilon Sampling Rocks: Investigating Sampling Strategies for Minimum Bayes Risk Decoding for Machine Translation
Markus Freitag, Behrooz Ghorbani, Patrick Fernandes
cs.CL, cs.AI, cs.LG
Recent advances in machine translation (MT) have shown that Minimum Bayes Risk (MBR) decoding can be a powerful alternative to beam search decoding, especially when combined with neural-based utility functions. However, the performance of MBR decoding depends heavily on how and how many candidates are sampled from the ...
http://arxiv.org/pdf/2305.09860v2
2023-05-18T02:24:56
2023-05-17T00:11:38
{ "title": [ -0.051971436, -0.037963867, -0.003320694, -0.008659363, -0.057281494, 0.003232956, -0.032318115, -0.0010719299, -0.0042877197, 0.061798096, -0.0579834, 0.038726807, 0.046661377, 0.003124237, -0.045684814, 0.01586914, -0.0066490173, 0...
http://arxiv.org/abs/2305.09859v1
Smaller Language Models are Better Black-box Machine-Generated Text Detectors
Fatemehsadat Mireshghallah, Justus Mattern, Sicun Gao, Reza Shokri, Taylor Berg-Kirkpatrick
cs.CL, cs.LG
With the advent of fluent generative language models that can produce convincing utterances very similar to those written by humans, distinguishing whether a piece of text is machine-generated or human-written becomes more challenging and more important, as such models could be used to spread misinformation, fake news,...
http://arxiv.org/pdf/2305.09859v1
2023-05-17T00:09:08
2023-05-17T00:09:08
{ "title": [ -0.01574707, -0.008094788, -0.03753662, -0.02909851, -0.03189087, -0.029403687, -0.038269043, -0.019165039, -0.025604248, 0.043273926, -0.046081543, 0.013549805, 0.0036697388, -0.0070877075, -0.030319214, -0.012008667, -0.0012769699, ...
http://arxiv.org/abs/2305.09858v1
Knowledge Graph Completion Models are Few-shot Learners: An Empirical Study of Relation Labeling in E-commerce with LLMs
Jiao Chen, Luyi Ma, Xiaohan Li, Nikhil Thakurdesai, Jianpeng Xu, Jason H. D. Cho, Kaushiki Nag, Evren Korpeoglu, Sushant Kumar, Kannan Achan
cs.IR, cs.AI, cs.CL, cs.LG
Knowledge Graphs (KGs) play a crucial role in enhancing e-commerce system performance by providing structured information about entities and their relationships, such as complementary or substitutable relations between products or product types, which can be utilized in recommender systems. However, relation labeling i...
http://arxiv.org/pdf/2305.09858v1
2023-05-17T00:08:36
2023-05-17T00:08:36
{ "title": [ -0.030899048, -0.02532959, -0.061798096, -0.048828125, -0.048553467, -0.0025081635, -0.07043457, 0.02470398, 0.0021629333, 0.0077056885, -0.048583984, 0.029510498, -0.01159668, -0.010360718, -0.017837524, -0.026535034, 0.00041389465, ...
http://arxiv.org/abs/2305.09857v2
CoEdIT: Text Editing by Task-Specific Instruction Tuning
Vipul Raheja, Dhruv Kumar, Ryan Koo, Dongyeop Kang
cs.CL, cs.AI, I.2.7
We introduce CoEdIT, a state-of-the-art text editing system for writing assistance. CoEdIT takes instructions from the user specifying the attributes of the desired text, such as "Make the sentence simpler" or "Write it in a more neutral style," and outputs the edited text. We present a large language model fine-tuned ...
http://arxiv.org/pdf/2305.09857v2
2023-10-23T23:17:13
2023-05-17T00:05:24
{ "title": [ -0.060791016, -0.023208618, -0.010856628, 0.0027866364, -0.054229736, 0.011260986, -0.038208008, -0.01852417, -0.025497437, 0.06488037, -0.02520752, 0.021987915, 0.058380127, -0.010231018, -0.020019531, 0.0005044937, -0.026931763, 0....
http://arxiv.org/abs/2305.09846v2
CPL-NoViD: Context-Aware Prompt-based Learning for Norm Violation Detection in Online Communities
Zihao He, Jonathan May, Kristina Lerman
cs.CL, cs.SI
Detecting norm violations in online communities is critical to maintaining healthy and safe spaces for online discussions. Existing machine learning approaches often struggle to adapt to the diverse rules and interpretations across different communities due to the inherent challenges of fine-tuning models for such cont...
http://arxiv.org/pdf/2305.09846v2
2023-05-18T18:33:01
2023-05-16T23:27:59
{ "title": [ 0.044708252, -0.0045318604, -0.022079468, -0.044830322, 0.030014038, -0.02406311, 0.022705078, -0.004146576, 0.020935059, 0.020889282, -0.06774902, 0.030197144, 0.023361206, -0.007003784, -0.015464783, 0.0008497238, -0.01965332, 0.05...
http://arxiv.org/abs/2305.09807v2
On Dataset Transferability in Active Learning for Transformers
Fran Jelenić, Josip Jukić, Nina Drobac, Jan Šnajder
cs.LG, cs.CL
Active learning (AL) aims to reduce labeling costs by querying the examples most beneficial for model learning. While the effectiveness of AL for fine-tuning transformer-based pre-trained language models (PLMs) has been demonstrated, it is less clear to what extent the AL gains obtained with one model transfer to other...
http://arxiv.org/pdf/2305.09807v2
2023-09-29T09:05:00
2023-05-16T21:10:54
{ "title": [ 0.0027160645, 0.025100708, -0.016494751, -0.04888916, 0.03793335, -0.004131317, -0.000310421, -0.030059814, 0.008621216, 0.015777588, -0.029434204, 0.015365601, -0.036621094, -0.028244019, -0.01977539, 0.0044288635, 0.020248413, 0.03...
http://arxiv.org/abs/2305.09800v2
Mirages: On Anthropomorphism in Dialogue Systems
Gavin Abercrombie, Amanda Cercas Curry, Tanvi Dinkar, Verena Rieser, Zeerak Talat
cs.CL
Automated dialogue or conversational systems are anthropomorphised by developers and personified by users. While a degree of anthropomorphism may be inevitable due to the choice of medium, conscious and unconscious design choices can guide users to personify such systems to varying degrees. Encouraging users to relate ...
http://arxiv.org/pdf/2305.09800v2
2023-10-23T09:26:22
2023-05-16T20:50:46
{ "title": [ -0.030258179, 0.0014677048, -0.02709961, -0.026412964, 0.0045700073, -0.0061187744, -0.011001587, -0.026916504, -0.011558533, 0.040283203, -0.016403198, 0.0021762848, 0.0059127808, -0.009170532, -0.05206299, -0.012161255, -0.021957397, ...
http://arxiv.org/abs/2305.09798v1
The Ways of Words: The Impact of Word Choice on Information Engagement and Decision Making
Nimrod Dvir, Elaine Friedman, Suraj Commuri, Fan Yang, Jennifer Romano
cs.CL, cs.HC, cs.SY, eess.SY, stat.AP, 28-08, H.5.2; H.1.2
Little research has explored how information engagement (IE), the degree to which individuals interact with and use information in a manner that manifests cognitively, behaviorally, and affectively. This study explored the impact of phrasing, specifically word choice, on IE and decision making. Synthesizing two theoret...
http://arxiv.org/pdf/2305.09798v1
2023-05-16T20:46:36
2023-05-16T20:46:36
{ "title": [ -0.019485474, -0.0062294006, -0.05090332, -0.019104004, 0.012367249, -0.029296875, 0.0015707016, -0.01965332, -0.04031372, 0.034851074, -0.01890564, 0.030685425, -0.018310547, -0.004878998, -0.033172607, -0.0057144165, -0.010139465, ...
http://arxiv.org/abs/2305.09785v1
Distilling Semantic Concept Embeddings from Contrastively Fine-Tuned Language Models
Na Li, Hanane Kteich, Zied Bouraoui, Steven Schockaert
cs.CL, cs.AI
Learning vectors that capture the meaning of concepts remains a fundamental challenge. Somewhat surprisingly, perhaps, pre-trained language models have thus far only enabled modest improvements to the quality of such concept embeddings. Current strategies for using language models typically represent a concept by avera...
http://arxiv.org/pdf/2305.09785v1
2023-05-16T20:17:02
2023-05-16T20:17:02
{ "title": [ 0.004547119, -0.021438599, -0.019332886, -0.07244873, -0.001326561, -0.00062036514, -0.03918457, -0.0003745556, 0.044006348, 0.028289795, -0.03149414, 0.020858765, -0.00049448013, 0.002784729, -0.057861328, 0.0096206665, -0.012802124, ...
http://arxiv.org/abs/2305.09781v2
SpecInfer: Accelerating Generative Large Language Model Serving with Speculative Inference and Token Tree Verification
Xupeng Miao, Gabriele Oliaro, Zhihao Zhang, Xinhao Cheng, Zeyu Wang, Rae Ying Yee Wong, Alan Zhu, Lijie Yang, Xiaoxiang Shi, Chunan Shi, Zhuoming Chen, Daiyaan Arfeen, Reyna Abhyankar, Zhihao Jia
cs.CL, cs.DC, cs.LG
The high computational and memory requirements of generative large language models (LLMs) make it challenging to serve them quickly and cheaply. This paper introduces SpecInfer, an LLM serving system that accelerates generative LLM inference with speculative inference and token tree verification. A key insight behind S...
http://arxiv.org/pdf/2305.09781v2
2023-08-16T13:33:06
2023-05-16T20:12:59
{ "title": [ -0.02407837, -0.032043457, -0.053894043, -0.097595215, -0.0069389343, 0.009437561, -0.07727051, 0.034973145, 0.0065574646, 0.021194458, -0.023361206, -0.0041885376, 0.014587402, -0.027755737, -0.033233643, 0.005207062, -0.019973755, ...
http://arxiv.org/abs/2305.10450v1
Understanding of Normal and Abnormal Hearts by Phase Space Analysis and Convolutional Neural Networks
Bekir Yavuz Koc, Taner Arsan, Onder Pekcan
eess.IV, cs.CL, cs.CV, cs.LG, cs.NE, eess.SP, 68, I.5.1
Cardiac diseases are one of the leading mortality factors in modern, industrialized societies, which cause high expenses in public health systems. Due to high costs, developing analytical methods to improve cardiac diagnostics is essential. The heart's electric activity was first modeled using a set of nonlinear differ...
http://arxiv.org/pdf/2305.10450v1
2023-05-16T19:52:40
2023-05-16T19:52:40
{ "title": [ 0.017700195, 0.0026893616, -0.027786255, -0.0128479, 0.010101318, -0.038970947, 0.01838684, -0.020904541, -0.023269653, 0.036224365, -0.027542114, 0.0211792, 0.020111084, -0.041809082, -0.0059928894, 0.00075531006, 0.028808594, 0.032...
http://arxiv.org/abs/2305.09770v6
ConvXAI: Delivering Heterogeneous AI Explanations via Conversations to Support Human-AI Scientific Writing
Hua Shen, Chieh-Yang Huang, Tongshuang Wu, Ting-Hao 'Kenneth' Huang
cs.HC, cs.AI, cs.CL
Despite a surge collection of XAI methods, users still struggle to obtain required AI explanations. Previous research suggests chatbots as dynamic solutions, but the effective design of conversational XAI agents for practical human needs remains under-explored. This paper focuses on Conversational XAI for AI-assisted s...
http://arxiv.org/pdf/2305.09770v6
2023-10-27T16:08:32
2023-05-16T19:48:49
{ "title": [ -0.05657959, -0.0390625, -0.03326416, -0.023788452, -0.033172607, 0.0037975311, -0.021011353, 0.004135132, 0.007820129, 0.009117126, -0.03338623, 0.025390625, -0.0053520203, -0.014190674, -0.021331787, -0.006137848, -0.024230957, 0.0...
http://arxiv.org/abs/2305.09764v1
Application-Agnostic Language Modeling for On-Device ASR
Markus Nußbaum-Thom, Lyan Verwimp, Youssef Oualil
cs.CL, cs.SD, eess.AS
On-device automatic speech recognition systems face several challenges compared to server-based systems. They have to meet stricter constraints in terms of speed, disk size and memory while maintaining the same accuracy. Often they have to serve several applications with different distributions at once, such as communi...
http://arxiv.org/pdf/2305.09764v1
2023-05-16T19:31:18
2023-05-16T19:31:18
{ "title": [ -0.03591919, -0.0104522705, -0.04611206, -0.060150146, -0.004535675, 0.0031719208, -0.06390381, 0.028915405, -0.030029297, 0.037994385, 0.0119018555, -0.01096344, 0.0033226013, -0.028503418, -0.034362793, 0.010902405, -0.007385254, 0...
http://arxiv.org/abs/2305.09758v3
A Video Is Worth 4096 Tokens: Verbalize Videos To Understand Them In Zero Shot
Aanisha Bhattacharya, Yaman K Singla, Balaji Krishnamurthy, Rajiv Ratn Shah, Changyou Chen
cs.CV, cs.CL
Multimedia content, such as advertisements and story videos, exhibit a rich blend of creativity and multiple modalities. They incorporate elements like text, visuals, audio, and storytelling techniques, employing devices like emotions, symbolism, and slogans to convey meaning. There is a dearth of large annotated train...
http://arxiv.org/pdf/2305.09758v3
2023-10-26T10:08:31
2023-05-16T19:13:11
{ "title": [ 0.03475952, -0.018371582, -0.03765869, -0.024734497, -0.019699097, -0.035217285, 0.01751709, 0.014228821, -0.017364502, 0.014976501, -0.06008911, -0.008285522, -0.031311035, -0.031082153, -0.022583008, -0.053375244, -0.009010315, 0.0...
http://arxiv.org/abs/2305.09756v1
Clinical Note Owns its Hierarchy: Multi-Level Hypergraph Neural Networks for Patient-Level Representation Learning
Nayeon Kim, Yinhua Piao, Sun Kim
cs.CL
Leveraging knowledge from electronic health records (EHRs) to predict a patient's condition is essential to the effective delivery of appropriate care. Clinical notes of patient EHRs contain valuable information from healthcare professionals, but have been underused due to their difficult contents and complex hierarchi...
http://arxiv.org/pdf/2305.09756v1
2023-05-16T19:08:18
2023-05-16T19:08:18
{ "title": [ 0.017807007, -0.00083351135, -0.04147339, -0.0635376, 0.0004441738, -0.016693115, -0.035888672, -0.012145996, -0.01663208, 0.030395508, -0.010131836, 0.014266968, -0.013000488, -0.028564453, -0.029556274, -0.014846802, 0.007083893, 0...
http://arxiv.org/abs/2305.09731v1
What In-Context Learning "Learns" In-Context: Disentangling Task Recognition and Task Learning
Jane Pan, Tianyu Gao, Howard Chen, Danqi Chen
cs.CL, cs.LG
Large language models (LLMs) exploit in-context learning (ICL) to solve tasks with only a few demonstrations, but its mechanisms are not yet well-understood. Some works suggest that LLMs only recall already learned concepts from pre-training, while others hint that ICL performs implicit learning over demonstrations. We...
http://arxiv.org/pdf/2305.09731v1
2023-05-16T18:05:19
2023-05-16T18:05:19
{ "title": [ -0.0102005005, 0.009887695, -0.06652832, -0.022506714, 0.013809204, -0.047851562, 0.01586914, -0.018081665, -0.013870239, 0.027938843, -0.047546387, 0.0070724487, -0.019180298, -0.01737976, -0.022903442, -0.023849487, 0.014877319, 0....
http://arxiv.org/abs/2305.09656v3
SatLM: Satisfiability-Aided Language Models Using Declarative Prompting
Xi Ye, Qiaochu Chen, Isil Dillig, Greg Durrett
cs.CL, cs.AI
Prior work has combined chain-of-thought prompting in large language models (LLMs) with programmatic representations to perform effective and transparent reasoning. While such an approach works well for tasks that only require forward reasoning (e.g., straightforward arithmetic), it is less effective for constraint sol...
http://arxiv.org/pdf/2305.09656v3
2023-10-11T21:38:22
2023-05-16T17:55:51
{ "title": [ -0.028152466, -0.02142334, -0.015220642, -0.09844971, -0.011802673, 0.030593872, -0.09667969, 0.014587402, 0.009025574, 0.0032215118, -0.028167725, 0.00756073, -0.007583618, -0.020645142, -0.03805542, 0.0044822693, -0.019424438, 0.02...
http://arxiv.org/abs/2305.09652v2
The Interpreter Understands Your Meaning: End-to-end Spoken Language Understanding Aided by Speech Translation
Mutian He, Philip N. Garner
cs.CL, cs.SD, eess.AS
End-to-end spoken language understanding (SLU) remains elusive even with current large pretrained language models on text and speech, especially in multilingual cases. Machine translation has been established as a powerful pretraining objective on text as it enables the model to capture high-level semantics of the inpu...
http://arxiv.org/pdf/2305.09652v2
2023-10-17T14:59:28
2023-05-16T17:53:03
{ "title": [ -0.021224976, -0.008262634, -0.027191162, -0.06317139, -0.008460999, -0.0102005005, -0.045959473, -0.0061950684, -0.0039043427, 0.016845703, -0.006034851, 0.0032444, 0.0054969788, -0.051757812, -0.027633667, -0.002494812, -0.00049829483,...
http://arxiv.org/abs/2305.09651v1
Tailoring Instructions to Student's Learning Levels Boosts Knowledge Distillation
Yuxin Ren, Zihan Zhong, Xingjian Shi, Yi Zhu, Chun Yuan, Mu Li
cs.CL, cs.LG
It has been commonly observed that a teacher model with superior performance does not necessarily result in a stronger student, highlighting a discrepancy between current teacher training practices and effective knowledge transfer. In order to enhance the guidance of the teacher training process, we introduce the conce...
http://arxiv.org/pdf/2305.09651v1
2023-05-16T17:50:09
2023-05-16T17:50:09
{ "title": [ 0.041870117, -0.006565094, -0.043762207, -0.06512451, -0.017440796, -0.02470398, -0.044036865, -0.002166748, -0.008560181, 0.02619934, -0.023727417, -0.008399963, 0.0008392334, -0.039764404, 0.0062713623, -0.034820557, 0.017120361, -...
http://arxiv.org/abs/2305.09645v2
StructGPT: A General Framework for Large Language Model to Reason over Structured Data
Jinhao Jiang, Kun Zhou, Zican Dong, Keming Ye, Wayne Xin Zhao, Ji-Rong Wen
cs.CL
In this paper, we study how to improve the zero-shot reasoning ability of large language models~(LLMs) over structured data in a unified way. Inspired by the study on tool augmentation for LLMs, we develop an \emph{Iterative Reading-then-Reasoning~(IRR)} approach for solving question answering tasks based on structured...
http://arxiv.org/pdf/2305.09645v2
2023-10-23T07:51:23
2023-05-16T17:45:23
{ "title": [ -0.029418945, -0.034240723, -0.058044434, -0.090270996, -0.0004386902, 0.00623703, -0.07556152, 0.011192322, 0.013397217, 0.012786865, -0.013366699, -0.015014648, 0.0020275116, -0.031219482, -0.026885986, -0.0014410019, -0.0072288513, ...
http://arxiv.org/abs/2305.09620v1
AI-Augmented Surveys: Leveraging Large Language Models for Opinion Prediction in Nationally Representative Surveys
Junsol Kim, Byungkyu Lee
cs.CL, cs.AI, cs.LG
How can we use large language models (LLMs) to augment surveys? This paper investigates three distinct applications of LLMs fine-tuned by nationally representative surveys for opinion prediction -- missing data imputation, retrodiction, and zero-shot prediction. We present a new methodological framework that incorporat...
http://arxiv.org/pdf/2305.09620v1
2023-05-16T17:13:07
2023-05-16T17:13:07
{ "title": [ -0.014709473, -0.009742737, -0.027801514, -0.03894043, 0.0044021606, -0.00198555, -0.015945435, -0.016067505, 0.0076522827, 0.01979065, -0.039276123, -0.0017280579, -0.016311646, -0.009254456, -0.05706787, -0.019317627, -0.008575439, ...
http://arxiv.org/abs/2305.09617v1
Towards Expert-Level Medical Question Answering with Large Language Models
Karan Singhal, Tao Tu, Juraj Gottweis, Rory Sayres, Ellery Wulczyn, Le Hou, Kevin Clark, Stephen Pfohl, Heather Cole-Lewis, Darlene Neal, Mike Schaekermann, Amy Wang, Mohamed Amin, Sami Lachgar, Philip Mansfield, Sushant Prakash, Bradley Green, Ewa Dominowska, Blaise Aguera y Arcas, Nenad Tomasev, Yun Liu, Renee Wong, ...
cs.CL, cs.AI, cs.LG
Recent artificial intelligence (AI) systems have reached milestones in "grand challenges" ranging from Go to protein-folding. The capability to retrieve medical knowledge, reason over it, and answer medical questions comparably to physicians has long been viewed as one such grand challenge. Large language models (LLM...
http://arxiv.org/pdf/2305.09617v1
2023-05-16T17:11:29
2023-05-16T17:11:29
{ "title": [ -0.008613586, -0.0038909912, -0.031066895, -0.07562256, 0.0072631836, 0.012008667, -0.04711914, 0.0037021637, 0.007843018, 0.030181885, -0.0016775131, -0.014030457, 0.005332947, -0.01411438, -0.05596924, 0.014221191, 0.010520935, 0.0...
http://arxiv.org/abs/2305.09612v1
Large Language Models are Built-in Autoregressive Search Engines
Noah Ziems, Wenhao Yu, Zhihan Zhang, Meng Jiang
cs.CL, cs.IR
Document retrieval is a key stage of standard Web search engines. Existing dual-encoder dense retrievers obtain representations for questions and documents independently, allowing for only shallow interactions between them. To overcome this limitation, recent autoregressive search engines replace the dual-encoder archi...
http://arxiv.org/pdf/2305.09612v1
2023-05-16T17:04:48
2023-05-16T17:04:48
{ "title": [ 0.025039673, -0.014289856, -0.019073486, -0.039764404, -0.039123535, -0.021011353, -0.024719238, 0.017837524, -0.005832672, 0.049041748, -0.028961182, -0.02482605, 0.021133423, -0.028381348, -0.018234253, -0.01890564, 0.017120361, 0....
http://arxiv.org/abs/2305.09598v1
Boosting Event Extraction with Denoised Structure-to-Text Augmentation
bo wang, Heyan Huang, Xiaochi Wei, Ge Shi, Xiao Liu, Chong Feng, Tong Zhou, Shuaiqiang Wang, Dawei Yin
cs.CL
Event extraction aims to recognize pre-defined event triggers and arguments from texts, which suffer from the lack of high-quality annotations. In most NLP applications, involving a large scale of synthetic training data is a practical and effective approach to alleviate the problem of data scarcity. However, when appl...
http://arxiv.org/pdf/2305.09598v1
2023-05-16T16:52:07
2023-05-16T16:52:07
{ "title": [ 0.018035889, 0.018295288, -0.04736328, -0.1015625, 0.027938843, 0.0051498413, -0.054779053, -0.018859863, 0.0357666, 0.022842407, 0.0022945404, 0.008781433, -0.0026893616, -0.019500732, -0.05444336, 0.014053345, -0.0102005005, -0.007...
http://arxiv.org/abs/2305.09574v1
UOR: Universal Backdoor Attacks on Pre-trained Language Models
Wei Du, Peixuan Li, Boqun Li, Haodong Zhao, Gongshen Liu
cs.CL, cs.AI, cs.CR
Backdoors implanted in pre-trained language models (PLMs) can be transferred to various downstream tasks, which exposes a severe security threat. However, most existing backdoor attacks against PLMs are un-targeted and task-specific. Few targeted and task-agnostic methods use manually pre-defined triggers and output re...
http://arxiv.org/pdf/2305.09574v1
2023-05-16T16:11:48
2023-05-16T16:11:48
{ "title": [ -0.01876831, -0.013511658, -0.028213501, -0.061920166, 0.0058784485, -0.0042877197, -0.035827637, -0.000746727, -0.046203613, 0.031341553, 0.017089844, 0.018859863, -0.0015163422, -0.03302002, -0.018676758, 0.00819397, -0.015808105, ...
http://arxiv.org/abs/2305.09556v1
Adapting Sentence Transformers for the Aviation Domain
Liya Wang, Jason Chou, Dave Rouck, Alex Tien, Diane M Baumgartner
cs.CL
Learning effective sentence representations is crucial for many Natural Language Processing (NLP) tasks, including semantic search, semantic textual similarity (STS), and clustering. While multiple transformer models have been developed for sentence embedding learning, these models may not perform optimally when dealin...
http://arxiv.org/pdf/2305.09556v1
2023-05-16T15:53:24
2023-05-16T15:53:24
{ "title": [ 0.0051612854, -0.010597229, -0.005180359, -0.061309814, -0.0051460266, 0.010307312, -0.05102539, -0.012084961, -0.02835083, 0.030975342, -0.037750244, 0.0005865097, 0.0062713623, -0.020599365, -0.011253357, -0.00737381, -0.0073013306, ...
http://arxiv.org/abs/2305.09550v2
Life of PII -- A PII Obfuscation Transformer
Ajinkya Deshmukh, Saumya Banthia, Anantha Sharma
cs.CL, cs.IR, cs.IT, math.IT
Protecting sensitive information is crucial in today's world of Large Language Models (LLMs) and data-driven services. One common method used to preserve privacy is by using data perturbation techniques to reduce overreaching utility of (sensitive) Personal Identifiable Information (PII) data while maintaining its stat...
http://arxiv.org/pdf/2305.09550v2
2023-05-17T12:32:26
2023-05-16T15:48:36
{ "title": [ 0.01121521, 0.003528595, -0.032409668, -0.014183044, -0.042633057, -0.011260986, -0.011451721, -0.016479492, -0.04446411, 0.038269043, -0.035827637, 0.017044067, 0.037506104, -0.03552246, 0.01890564, -0.03640747, 0.0007133484, 0.0292...
http://arxiv.org/abs/2305.09548v2
Measuring Social Dimensions of Self-Presentation in Social Media Biographies with an Identity-based Approach
Navid Madani, Rabiraj Bandyopadhyay, Briony Swire-Thompson, Michael Miller Yoder, Kenneth Joseph
cs.CL
Social media users on sites like Twitter, Instagram, and Tiktok use the profile description, or bio, field of user profiles to present themselves to the world. In contrast to the ``offline'' world, where social context often encourages us to adopt a single identity, the profile description is a free-text field in which...
http://arxiv.org/pdf/2305.09548v2
2023-10-11T19:57:43
2023-05-16T15:45:59
{ "title": [ -0.0098724365, 0.0068855286, -0.039916992, -0.05117798, 0.02658081, -0.012207031, -0.012069702, 0.0015926361, -0.044769287, 0.03567505, -0.008872986, 0.04812622, -0.008842468, -0.025238037, -0.0071868896, -0.00085926056, -0.017608643, ...
http://arxiv.org/abs/2305.09534v1
MetaSRL++: A Uniform Scheme for Modelling Deeper Semantics
Fritz Hohl, Nianheng Wu, Martina Galetti, Remi van Trijp
cs.CL
Despite enormous progress in Natural Language Processing (NLP), our field is still lacking a common deep semantic representation scheme. As a result, the problem of meaning and understanding is typically sidestepped through more simple, approximative methods. This paper argues that in order to arrive at such a scheme, ...
http://arxiv.org/pdf/2305.09534v1
2023-05-16T15:26:52
2023-05-16T15:26:52
{ "title": [ -0.009666443, -0.015640259, -0.04067993, -0.10064697, 0.015487671, 0.018234253, -0.070373535, -0.012413025, 0.02053833, 0.023742676, 0.0033512115, -0.013885498, -0.002565384, -0.027038574, -0.04727173, 0.017730713, -0.0062332153, 0.0...
http://arxiv.org/abs/2305.10448v1
Sequence-to-Sequence Pre-training with Unified Modality Masking for Visual Document Understanding
Shuwei Feng, Tianyang Zhan, Zhanming Jie, Trung Quoc Luong, Xiaoran Jin
cs.CL, cs.AI
This paper presents GenDoc, a general sequence-to-sequence document understanding model pre-trained with unified masking across three modalities: text, image, and layout. The proposed model utilizes an encoder-decoder architecture, which allows for increased adaptability to a wide range of downstream tasks with diverse...
http://arxiv.org/pdf/2305.10448v1
2023-05-16T15:25:19
2023-05-16T15:25:19
{ "title": [ -0.0029258728, 0.006286621, -0.03012085, -0.08331299, -0.0025634766, -0.009666443, -0.055511475, 0.0072746277, 0.022079468, 0.027938843, -0.04348755, 0.015449524, 0.034820557, -0.0118255615, -0.022537231, -0.0033988953, 0.010154724, ...
http://arxiv.org/abs/2305.09520v1
DLUE: Benchmarking Document Language Understanding
Ruoxi Xu, Hongyu Lin, Xinyan Guan, Xianpei Han, Yingfei Sun, Le Sun
cs.CL
Understanding documents is central to many real-world tasks but remains a challenging topic. Unfortunately, there is no well-established consensus on how to comprehensively evaluate document understanding abilities, which significantly hinders the fair comparison and measuring the progress of the field. To benchmark do...
http://arxiv.org/pdf/2305.09520v1
2023-05-16T15:16:24
2023-05-16T15:16:24
{ "title": [ -0.0491333, -0.026275635, -0.030593872, -0.07183838, -0.008659363, 0.01701355, -0.06518555, 0.0028877258, 0.007286072, 0.022735596, -0.023284912, 0.008369446, -0.009597778, -0.007133484, -0.03918457, -0.002828598, 0.0013103485, 0.024...
http://arxiv.org/abs/2305.09515v2
AR-Diffusion: Auto-Regressive Diffusion Model for Text Generation
Tong Wu, Zhihao Fan, Xiao Liu, Yeyun Gong, Yelong Shen, Jian Jiao, Hai-Tao Zheng, Juntao Li, Zhongyu Wei, Jian Guo, Nan Duan, Weizhu Chen
cs.CL
Diffusion models have gained significant attention in the realm of image generation due to their exceptional performance. Their success has been recently expanded to text generation via generating all tokens within a sequence concurrently. However, natural language exhibits a far more pronounced sequential dependency i...
http://arxiv.org/pdf/2305.09515v2
2023-05-19T02:08:36
2023-05-16T15:10:22
{ "title": [ -0.02961731, -0.010108948, -0.039245605, -0.0289917, 0.0014591217, 0.011558533, -0.027862549, -0.017364502, 0.016418457, 0.028442383, -0.011222839, 0.029159546, -0.005130768, -0.01033783, -0.046966553, 0.013114929, -0.03152466, 0.037...
http://arxiv.org/abs/2305.09509v1
Bidirectional Generative Framework for Cross-domain Aspect-based Sentiment Analysis
Yue Deng, Wenxuan Zhang, Sinno Jialin Pan, Lidong Bing
cs.CL
Cross-domain aspect-based sentiment analysis (ABSA) aims to perform various fine-grained sentiment analysis tasks on a target domain by transferring knowledge from a source domain. Since labeled data only exists in the source domain, a model is expected to bridge the domain gap for tackling cross-domain ABSA. Though do...
http://arxiv.org/pdf/2305.09509v1
2023-05-16T15:02:23
2023-05-16T15:02:23
{ "title": [ 0.020965576, -0.0014381409, -0.033935547, -0.089660645, 0.013496399, -0.006336212, -0.034210205, 0.007835388, 0.03201294, 0.04724121, -0.0026721954, 0.00060224533, 0.010757446, -0.014343262, -0.06604004, 0.01965332, -0.032287598, 0.0...
http://arxiv.org/abs/2305.09506v1
Fuzzy Temporal Protoforms for the Quantitative Description of Processes in Natural Language
Yago Fontenla-Seco, Alberto Bugarín-Diz, Manuel Lama
cs.CL
In this paper, we propose a series of fuzzy temporal protoforms in the framework of the automatic generation of quantitative and qualitative natural language descriptions of processes. The model includes temporal and causal information from processes and attributes, quantifies attributes in time during the process life...
http://arxiv.org/pdf/2305.09506v1
2023-05-16T14:59:38
2023-05-16T14:59:38
{ "title": [ -0.0021400452, 0.002822876, -0.033843994, -0.069885254, 0.017562866, -0.006652832, -0.038360596, -0.010284424, 0.012062073, 0.050872803, -0.001376152, -0.0015287399, 0.02279663, -0.017333984, -0.06124878, 0.025985718, -0.013870239, 0...
http://arxiv.org/abs/2305.09438v3
MPI-rical: Data-Driven MPI Distributed Parallelism Assistance with Transformers
Nadav Schneider, Tal Kadosh, Niranjan Hasabnis, Timothy Mattson, Yuval Pinter, Gal Oren
cs.DC, cs.CL, cs.LG
Message Passing Interface (MPI) plays a crucial role in distributed memory parallelization across multiple nodes. However, parallelizing MPI code manually, and specifically, performing domain decomposition, is a challenging, error-prone task. In this paper, we address this problem by developing MPI-RICAL, a novel data-...
http://arxiv.org/pdf/2305.09438v3
2023-08-30T14:56:16
2023-05-16T13:50:24
{ "title": [ -0.035949707, 0.029159546, -0.045043945, -0.051635742, 0.008628845, 0.0048446655, -0.041168213, -0.015731812, -0.015007019, 0.0385437, -0.023727417, 0.023880005, 0.0143966675, -0.032989502, -0.015388489, 0.013137817, 0.0018157959, 0....
http://arxiv.org/abs/2305.09410v1
About Evaluation of F1 Score for RECENT Relation Extraction System
Michał Olek
cs.CL, cs.AI
This document contains a discussion of the F1 score evaluation used in the article 'Relation Classification with Entity Type Restriction' by Shengfei Lyu, Huanhuan Chen published on Findings of the Association for Computational Linguistics: ACL-IJCNLP 2021. The authors created a system named RECENT and claim it achieve...
http://arxiv.org/pdf/2305.09410v1
2023-05-16T12:55:43
2023-05-16T12:55:43
{ "title": [ 0.02067566, -0.0026760101, -0.06939697, -0.059631348, 0.0022354126, -0.03390503, -0.027313232, 0.02758789, -0.0066108704, 0.017654419, -0.047790527, -0.0070533752, -0.0034542084, -0.031585693, -0.024124146, -0.026260376, -0.013053894, ...
http://arxiv.org/abs/2305.09402v1
A Preliminary Analysis on the Code Generation Capabilities of GPT-3.5 and Bard AI Models for Java Functions
Giuseppe Destefanis, Silvia Bartolucci, Marco Ortu
cs.SE, cs.CL
This paper evaluates the capability of two state-of-the-art artificial intelligence (AI) models, GPT-3.5 and Bard, in generating Java code given a function description. We sourced the descriptions from CodingBat.com, a popular online platform that provides practice problems to learn programming. We compared the Java co...
http://arxiv.org/pdf/2305.09402v1
2023-05-16T12:44:39
2023-05-16T12:44:39
{ "title": [ 0.01637268, -0.01058197, 0.015220642, -0.051239014, -0.033966064, 0.004550934, -0.060333252, -0.022506714, -0.009338379, 0.02180481, -0.034942627, 0.030181885, -0.01979065, -0.018173218, -0.01889038, -0.0039634705, -0.021820068, 0.03...
http://arxiv.org/abs/2305.09400v1
Consistent Multi-Granular Rationale Extraction for Explainable Multi-hop Fact Verification
Jiasheng Si, Yingjie Zhu, Deyu Zhou
cs.CL
The success of deep learning models on multi-hop fact verification has prompted researchers to understand the behavior behind their veracity. One possible way is erasure search: obtaining the rationale by entirely removing a subset of input without compromising the veracity prediction. Although extensively explored, ex...
http://arxiv.org/pdf/2305.09400v1
2023-05-16T12:31:53
2023-05-16T12:31:53
{ "title": [ 0.00021266937, -0.006465912, -0.033172607, -0.10760498, 0.0011234283, 0.016311646, -0.062683105, 0.024719238, -0.008666992, 0.030731201, -0.003742218, 0.012626648, 0.022705078, -0.018722534, -0.017425537, -0.006164551, -0.047546387, ...
http://arxiv.org/abs/2305.09360v3
GIFT: Graph-Induced Fine-Tuning for Multi-Party Conversation Understanding
Jia-Chen Gu, Zhen-Hua Ling, Quan Liu, Cong Liu, Guoping Hu
cs.CL
Addressing the issues of who saying what to whom in multi-party conversations (MPCs) has recently attracted a lot of research attention. However, existing methods on MPC understanding typically embed interlocutors and utterances into sequential information flows, or utilize only the superficial of inherent graph struct...
http://arxiv.org/pdf/2305.09360v3
2023-07-18T02:01:14
2023-05-16T11:35:24
{ "title": [ -0.01928711, -0.017242432, -0.029067993, -0.07324219, -0.017074585, 0.008934021, -0.0769043, 0.015037537, -0.0033454895, 0.018188477, -0.011199951, 0.018066406, -0.011306763, -0.014724731, -0.039916992, -0.012374878, -0.010955811, 0....
http://arxiv.org/abs/2305.09359v2
Constructing and Interpreting Causal Knowledge Graphs from News
Fiona Anting Tan, Debdeep Paul, Sahim Yamaura, Miura Koji, See-Kiong Ng
cs.CL
Many financial jobs rely on news to learn about causal events in the past and present, to make informed decisions and predictions about the future. With the ever-increasing amount of news available online, there is a need to automate the extraction of causal events from unstructured texts. In this work, we propose a me...
http://arxiv.org/pdf/2305.09359v2
2023-07-30T06:28:42
2023-05-16T11:33:32
{ "title": [ -0.03942871, -0.013381958, -0.058135986, -0.09832764, -0.0029945374, 0.008575439, -0.06500244, 0.020904541, -0.026626587, 0.0042419434, -0.015434265, 0.01991272, -0.045532227, -0.013938904, -0.029327393, -0.023391724, -0.0040893555, ...
http://arxiv.org/abs/2305.09335v1
MsPrompt: Multi-step Prompt Learning for Debiasing Few-shot Event Detection
Siyuan Wang, Jianming Zheng, Xuejun Hu, Fei Cai, Chengyu Song, Xueshan Luo
cs.CL
Event detection (ED) is aimed to identify the key trigger words in unstructured text and predict the event types accordingly. Traditional ED models are too data-hungry to accommodate real applications with scarce labeled data. Besides, typical ED models are facing the context-bypassing and disabled generalization issue...
http://arxiv.org/pdf/2305.09335v1
2023-05-16T10:19:12
2023-05-16T10:19:12
{ "title": [ 0.037475586, 0.0104522705, -0.030319214, -0.06536865, 0.030273438, -0.018615723, -0.033996582, -0.036956787, 0.00617218, -0.0073776245, -0.027755737, 0.0043792725, -0.021697998, -0.03488159, -0.0158844, -0.019104004, 0.0104904175, 0....
http://arxiv.org/abs/2305.09333v1
Multi-modal Visual Understanding with Prompts for Semantic Information Disentanglement of Image
Yuzhou Peng
cs.CV, cs.AI, cs.CL
Multi-modal visual understanding of images with prompts involves using various visual and textual cues to enhance the semantic understanding of images. This approach combines both vision and language processing to generate more accurate predictions and recognition of images. By utilizing prompt-based techniques, models...
http://arxiv.org/pdf/2305.09333v1
2023-05-16T10:15:44
2023-05-16T10:15:44
{ "title": [ 0.001748085, 0.0030059814, -0.05239868, -0.05822754, 0.009216309, -0.02178955, -0.0362854, -0.0042762756, 0.045013428, 0.009849548, -0.05496216, 0.007949829, 0.028747559, -0.022155762, -0.018310547, 0.0008740425, 0.022003174, 0.00069...
http://arxiv.org/abs/2305.09329v2
BERTTM: Leveraging Contextualized Word Embeddings from Pre-trained Language Models for Neural Topic Modeling
Zheng Fang, Yulan He, Rob Procter
cs.CL, cs.AI
With the development of neural topic models in recent years, topic modelling is playing an increasingly important role in natural language understanding. However, most existing topic models still rely on bag-of-words (BoW) information, either as training input or training target. This limits their ability to capture wo...
http://arxiv.org/pdf/2305.09329v2
2023-05-17T09:20:35
2023-05-16T10:07:33
{ "title": [ -0.0076026917, -0.0020427704, -0.023803711, -0.04901123, 0.0039863586, 0.00178051, -0.04437256, -0.026809692, 0.01171875, 0.027938843, -0.007080078, 0.033081055, 0.0053138733, -0.01374054, -0.018081665, 0.025985718, -0.013374329, 0.0...
http://arxiv.org/abs/2305.09316v1
Enhancing Keyphrase Extraction from Long Scientific Documents using Graph Embeddings
Roberto Martínez-Cruz, Debanjan Mahata, Alvaro J. López-López, José Portela
cs.CL, cs.AI
In this study, we investigate using graph neural network (GNN) representations to enhance contextualized representations of pre-trained language models (PLMs) for keyphrase extraction from lengthy documents. We show that augmenting a PLM with graph embeddings provides a more comprehensive semantic understanding of word...
http://arxiv.org/pdf/2305.09316v1
2023-05-16T09:44:38
2023-05-16T09:44:38
{ "title": [ -0.028701782, -0.002117157, -0.025619507, -0.026489258, -0.012107849, -0.00038170815, -0.022537231, -0.0009551048, 0.017227173, 0.043792725, 0.0026340485, 0.0019664764, -0.0029315948, -0.017181396, -0.06011963, 0.009361267, -0.031829834,...
http://arxiv.org/abs/2305.09313v1
Hybrid and Collaborative Passage Reranking
Zongmeng Zhang, Wengang Zhou, Jiaxin Shi, Houqiang Li
cs.IR, cs.AI, cs.CL
In passage retrieval system, the initial passage retrieval results may be unsatisfactory, which can be refined by a reranking scheme. Existing solutions to passage reranking focus on enriching the interaction between query and each passage separately, neglecting the context among the top-ranked passages in the initial ...
http://arxiv.org/pdf/2305.09313v1
2023-05-16T09:38:52
2023-05-16T09:38:52
{ "title": [ -0.01638794, 0.005256653, -0.013320923, -0.032073975, -0.036010742, 0.0008845329, -0.048187256, -0.04244995, -0.0046691895, 0.04660034, -0.01689148, 0.020523071, -0.011482239, -0.012931824, -0.023025513, -0.010612488, -0.016189575, 0...
http://arxiv.org/abs/2305.09312v1
Exploring the Impact of Layer Normalization for Zero-shot Neural Machine Translation
Zhuoyuan Mao, Raj Dabre, Qianying Liu, Haiyue Song, Chenhui Chu, Sadao Kurohashi
cs.CL
This paper studies the impact of layer normalization (LayerNorm) on zero-shot translation (ZST). Recent efforts for ZST often utilize the Transformer architecture as the backbone, with LayerNorm at the input of layers (PreNorm) set as the default. However, Xu et al. (2019) has revealed that PreNorm carries the risk of ...
http://arxiv.org/pdf/2305.09312v1
2023-05-16T09:37:08
2023-05-16T09:37:08
{ "title": [ -0.01966858, -0.0021152496, -0.021942139, -0.04876709, -0.005695343, 0.0021877289, -0.030212402, 0.012168884, 0.00021255016, 0.061798096, 0.0236969, 0.0057258606, 0.02027893, -0.014785767, -0.05807495, 0.03967285, 0.006668091, 0.0300...
http://arxiv.org/abs/2305.09299v1
UniS-MMC: Multimodal Classification via Unimodality-supervised Multimodal Contrastive Learning
Heqing Zou, Meng Shen, Chen Chen, Yuchen Hu, Deepu Rajan, Eng Siong Chng
cs.CV, cs.CL
Multimodal learning aims to imitate human beings to acquire complementary information from multiple modalities for various downstream tasks. However, traditional aggregation-based multimodal fusion methods ignore the inter-modality relationship, treat each modality equally, suffer sensor noise, and thus reduce multimod...
http://arxiv.org/pdf/2305.09299v1
2023-05-16T09:18:38
2023-05-16T09:18:38
{ "title": [ -0.01210022, 0.015487671, -0.037750244, -0.040863037, 0.008811951, -0.013763428, -0.03286743, -0.037017822, 0.009857178, 0.025863647, -0.014732361, 0.020385742, 0.0038795471, -0.01777649, -0.02017212, 0.004501343, -0.01979065, 0.0718...
http://arxiv.org/abs/2305.09287v2
Adversarial Word Dilution as Text Data Augmentation in Low-Resource Regime
Junfan Chen, Richong Zhang, Zheyan Luo, Chunming Hu, Yongyi Mao
cs.CL
Data augmentation is widely used in text classification, especially in the low-resource regime where a few examples for each class are available during training. Despite the success, generating data augmentations as hard positive examples that may increase their effectiveness is under-explored. This paper proposes an A...
http://arxiv.org/pdf/2305.09287v2
2023-08-09T10:45:52
2023-05-16T08:46:11
{ "title": [ -0.019821167, 0.0041999817, -0.066345215, -0.04144287, 0.0075798035, -0.016098022, -0.030792236, -0.008659363, -0.02218628, 0.04058838, -0.043182373, 0.030029297, -0.011276245, 0.00554657, -0.04562378, -0.021026611, -0.019821167, 0.0...
http://arxiv.org/abs/2305.09281v1
On the Origins of Bias in NLP through the Lens of the Jim Code
Fatma Elsafoury, Gavin Abercrombie
cs.CL
In this paper, we trace the biases in current natural language processing (NLP) models back to their origins in racism, sexism, and homophobia over the last 500 years. We review literature from critical race theory, gender studies, data ethics, and digital humanities studies, and summarize the origins of bias in NLP mo...
http://arxiv.org/pdf/2305.09281v1
2023-05-16T08:37:13
2023-05-16T08:37:13
{ "title": [ -0.010902405, -0.025756836, -0.03994751, -0.027114868, -0.011817932, -0.028778076, 0.0055618286, 0.004627228, -0.029266357, 0.021774292, -0.05682373, 0.000022351742, 0.016815186, -0.013763428, -0.015625, -0.02104187, -0.040161133, 0....
http://arxiv.org/abs/2305.09269v1
ContrastNet: A Contrastive Learning Framework for Few-Shot Text Classification
Junfan Chen, Richong Zhang, Yongyi Mao, Jie Xu
cs.CL
Few-shot text classification has recently been promoted by the meta-learning paradigm which aims to identify target classes with knowledge transferred from source classes with sets of small tasks named episodes. Despite their success, existing works building their meta-learner based on Prototypical Networks are unsatis...
http://arxiv.org/pdf/2305.09269v1
2023-05-16T08:22:17
2023-05-16T08:22:17
{ "title": [ 0.0066375732, -0.0074386597, -0.017868042, -0.034057617, 0.006500244, -0.009750366, -0.00003284216, -0.020004272, 0.023422241, 0.021957397, -0.030471802, 0.016921997, -0.014450073, -0.0075263977, -0.050109863, -0.012428284, -0.009750366,...
http://arxiv.org/abs/2305.09258v1
HyHTM: Hyperbolic Geometry based Hierarchical Topic Models
Simra Shahid, Tanay Anand, Nikitha Srikanth, Sumit Bhatia, Balaji Krishnamurthy, Nikaash Puri
cs.IR, cs.CL
Hierarchical Topic Models (HTMs) are useful for discovering topic hierarchies in a collection of documents. However, traditional HTMs often produce hierarchies where lowerlevel topics are unrelated and not specific enough to their higher-level topics. Additionally, these methods can be computationally expensive. We pre...
http://arxiv.org/pdf/2305.09258v1
2023-05-16T08:06:11
2023-05-16T08:06:11
{ "title": [ -0.014060974, 0.015068054, -0.04815674, -0.05105591, 0.0050735474, -0.0074501038, -0.02848816, -0.039398193, 0.012435913, 0.044128418, 0.00026774406, 0.02319336, 0.00623703, -0.014060974, -0.027450562, 0.02166748, 0.0036621094, -0.00...
http://arxiv.org/abs/2305.09249v1
xPQA: Cross-Lingual Product Question Answering across 12 Languages
Xiaoyu Shen, Akari Asai, Bill Byrne, Adrià de Gispert
cs.CL
Product Question Answering (PQA) systems are key in e-commerce applications to provide responses to customers' questions as they shop for products. While existing work on PQA focuses mainly on English, in practice there is need to support multiple customer languages while leveraging product information available in Eng...
http://arxiv.org/pdf/2305.09249v1
2023-05-16T07:56:19
2023-05-16T07:56:19
{ "title": [ -0.048187256, -0.02999878, -0.047454834, -0.05807495, 0.00029420853, 0.009674072, -0.06286621, 0.0011434555, 0.004005432, 0.023880005, -0.01576233, 0.0045547485, -0.0024700165, -0.010696411, -0.04309082, 0.0031814575, 0.00084638596, ...
http://arxiv.org/abs/2305.09246v1
Maybe Only 0.5% Data is Needed: A Preliminary Exploration of Low Training Data Instruction Tuning
Hao Chen, Yiming Zhang, Qi Zhang, Hantao Yang, Xiaomeng Hu, Xuetao Ma, Yifan Yanggong, Junbo Zhao
cs.AI, cs.CL
Instruction tuning for large language models (LLMs) has gained attention from researchers due to its ability to unlock the potential of LLMs in following instructions. While instruction tuning offers advantages for facilitating the adaptation of large language models (LLMs) to downstream tasks as a fine-tuning approach...
http://arxiv.org/pdf/2305.09246v1
2023-05-16T07:52:57
2023-05-16T07:52:57
{ "title": [ -0.035064697, -0.008804321, -0.03857422, -0.019195557, -0.027420044, -0.03378296, -0.025177002, -0.009735107, -0.012336731, 0.026977539, -0.04888916, 0.010429382, -0.019363403, -0.018478394, -0.01789856, -0.02331543, 0.014678955, 0.0...
http://arxiv.org/abs/2305.09220v1
Towards Unifying Multi-Lingual and Cross-Lingual Summarization
Jiaan Wang, Fandong Meng, Duo Zheng, Yunlong Liang, Zhixu Li, Jianfeng Qu, Jie Zhou
cs.CL, cs.AI
To adapt text summarization to the multilingual world, previous work proposes multi-lingual summarization (MLS) and cross-lingual summarization (CLS). However, these two tasks have been studied separately due to the different definitions, which limits the compatible and systematic research on both of them. In this pape...
http://arxiv.org/pdf/2305.09220v1
2023-05-16T06:53:21
2023-05-16T06:53:21
{ "title": [ -0.0046195984, 0.0011167526, -0.029220581, -0.08483887, 0.009674072, 0.011329651, -0.06970215, -0.006793976, 0.0033740997, 0.0619812, -0.0019779205, 0.01701355, 0.022537231, 0.0068893433, -0.04724121, 0.024871826, 0.00054216385, 0.03...
http://arxiv.org/abs/2305.09210v2
Towards Speech Dialogue Translation Mediating Speakers of Different Languages
Shuichiro Shimizu, Chenhui Chu, Sheng Li, Sadao Kurohashi
cs.CL
We present a new task, speech dialogue translation mediating speakers of different languages. We construct the SpeechBSD dataset for the task and conduct baseline experiments. Furthermore, we consider context to be an important aspect that needs to be addressed in this task and propose two ways of utilizing context, na...
http://arxiv.org/pdf/2305.09210v2
2023-05-22T04:45:29
2023-05-16T06:37:55
{ "title": [ -0.023712158, 0.000852108, -0.04043579, -0.039916992, -0.011169434, -0.018920898, -0.041656494, 0.007080078, -0.033935547, 0.049743652, 0.016113281, -0.0057640076, 0.010955811, -0.03967285, -0.0345459, 0.008392334, 0.0018377304, 0.03...
http://arxiv.org/abs/2305.09696v1
Generative Table Pre-training Empowers Models for Tabular Prediction
Tianping Zhang, Shaowen Wang, Shuicheng Yan, Jian Li, Qian Liu
cs.LG, cs.AI, cs.CL
Recently, the topic of table pre-training has attracted considerable research interest. However, how to employ table pre-training to boost the performance of tabular prediction remains an open challenge. In this paper, we propose TapTap, the first attempt that leverages table pre-training to empower models for tabular ...
http://arxiv.org/pdf/2305.09696v1
2023-05-16T06:37:38
2023-05-16T06:37:38
{ "title": [ 0.02848816, 0.006000519, -0.06286621, -0.1116333, -0.014770508, -0.029769897, -0.088012695, 0.0010156631, -0.009597778, 0.051940918, -0.029434204, 0.0090408325, 0.035217285, -0.028671265, -0.01676941, -0.0060539246, 0.024124146, 0.01...
http://arxiv.org/abs/2305.09204v1
The Weighted Möbius Score: A Unified Framework for Feature Attribution
Yifan Jiang, Shane Steinert-Threlkeld
cs.LG, cs.AI, cs.CL
Feature attribution aims to explain the reasoning behind a black-box model's prediction by identifying the impact of each feature on the prediction. Recent work has extended feature attribution to interactions between multiple features. However, the lack of a unified framework has led to a proliferation of methods that...
http://arxiv.org/pdf/2305.09204v1
2023-05-16T06:27:27
2023-05-16T06:27:27
{ "title": [ -0.008743286, -0.023208618, -0.027175903, -0.028259277, 0.0026760101, -0.005455017, -0.020629883, -0.03665161, -0.011650085, 0.010391235, -0.001455307, -0.019973755, -0.014266968, -0.0473938, -0.027160645, -0.006088257, -0.00894928, ...
http://arxiv.org/abs/2305.09193v2
Easy-to-Hard Learning for Information Extraction
Chang Gao, Wenxuan Zhang, Wai Lam, Lidong Bing
cs.CL
Information extraction (IE) systems aim to automatically extract structured information, such as named entities, relations between entities, and events, from unstructured texts. While most existing work addresses a particular IE task, universally modeling various IE tasks with one model has achieved great success recen...
http://arxiv.org/pdf/2305.09193v2
2023-05-19T11:27:02
2023-05-16T06:04:14
{ "title": [ -0.03338623, -0.009162903, -0.039886475, -0.055480957, 0.0042152405, 0.0017118454, -0.026794434, 0.0035476685, 0.009941101, 0.025314331, -0.03842163, 0.017913818, -0.027893066, -0.0017261505, -0.042633057, -0.004070282, -0.0018558502, ...
http://arxiv.org/abs/2305.09167v1
Adversarial Speaker Disentanglement Using Unannotated External Data for Self-supervised Representation Based Voice Conversion
Xintao Zhao, Shuai Wang, Yang Chao, Zhiyong Wu, Helen Meng
cs.SD, cs.CL, eess.AS
Nowadays, recognition-synthesis-based methods have been quite popular with voice conversion (VC). By introducing linguistics features with good disentangling characters extracted from an automatic speech recognition (ASR) model, the VC performance achieved considerable breakthroughs. Recently, self-supervised learning ...
http://arxiv.org/pdf/2305.09167v1
2023-05-16T04:52:29
2023-05-16T04:52:29
{ "title": [ 0.019882202, 0.024612427, -0.04574585, -0.047088623, 0.018096924, -0.015899658, -0.019454956, -0.016815186, -0.028335571, 0.046844482, 0.001786232, 0.006931305, 0.011268616, -0.036102295, -0.017791748, 0.0008544922, 0.002986908, 0.03...
http://arxiv.org/abs/2305.09154v1
Progressive Translation: Improving Domain Robustness of Neural Machine Translation with Intermediate Sequences
Chaojun Wang, Yang Liu, Wai Lam
cs.CL
Previous studies show that intermediate supervision signals benefit various Natural Language Processing tasks. However, it is not clear whether there exist intermediate signals that benefit Neural Machine Translation (NMT). Borrowing techniques from Statistical Machine Translation, we propose intermediate signals which...
http://arxiv.org/pdf/2305.09154v1
2023-05-16T04:15:25
2023-05-16T04:15:25
{ "title": [ -0.0059661865, 0.01411438, -0.026763916, -0.07635498, 0.027816772, 0.005329132, -0.04333496, -0.004386902, -0.003068924, 0.055664062, 0.011184692, 0.012001038, 0.01940918, -0.014564514, -0.054992676, 0.025360107, -0.006580353, 0.0360...
http://arxiv.org/abs/2305.09148v1
Dual-Alignment Pre-training for Cross-lingual Sentence Embedding
Ziheng Li, Shaohan Huang, Zihan Zhang, Zhi-Hong Deng, Qiang Lou, Haizhen Huang, Jian Jiao, Furu Wei, Weiwei Deng, Qi Zhang
cs.CL, cs.AI
Recent studies have shown that dual encoder models trained with the sentence-level translation ranking task are effective methods for cross-lingual sentence embedding. However, our research indicates that token-level alignment is also crucial in multilingual scenarios, which has not been fully explored previously. Base...
http://arxiv.org/pdf/2305.09148v1
2023-05-16T03:53:30
2023-05-16T03:53:30
{ "title": [ -0.009407043, 0.00434494, -0.011489868, -0.05581665, -0.0018472672, 0.005722046, -0.049926758, -0.014862061, 0.005783081, 0.041870117, -0.015319824, 0.02279663, 0.017349243, -0.006385803, -0.028366089, 0.016326904, -0.009155273, 0.04...
http://arxiv.org/abs/2305.09144v1
Retentive or Forgetful? Diving into the Knowledge Memorizing Mechanism of Language Models
Boxi Cao, Qiaoyu Tang, Hongyu Lin, Xianpei Han, Jiawei Chen, Tianshu Wang, Le Sun
cs.CL, cs.AI
Memory is one of the most essential cognitive functions serving as a repository of world knowledge and episodes of activities. In recent years, large-scale pre-trained language models have shown remarkable memorizing ability. On the contrary, vanilla neural networks without pre-training have been long observed sufferin...
http://arxiv.org/pdf/2305.09144v1
2023-05-16T03:50:38
2023-05-16T03:50:38
{ "title": [ -0.036956787, -0.009552002, -0.04171753, -0.05041504, -0.018157959, -0.01210022, -0.045898438, 0.0060195923, 0.011978149, 0.046722412, -0.036956787, 0.0071983337, 0.023086548, 0.002565384, -0.066101074, 0.0050582886, 0.0018281937, 0....
http://arxiv.org/abs/2305.09137v1
Pre-Training to Learn in Context
Yuxian Gu, Li Dong, Furu Wei, Minlie Huang
cs.CL
In-context learning, where pre-trained language models learn to perform tasks from task examples and instructions in their contexts, has attracted much attention in the NLP community. However, the ability of in-context learning is not fully exploited because language models are not explicitly trained to learn in contex...
http://arxiv.org/pdf/2305.09137v1
2023-05-16T03:38:06
2023-05-16T03:38:06
{ "title": [ 0.0602417, -0.003698349, -0.026443481, -0.07897949, -0.009887695, -0.027801514, -0.023513794, 0.015823364, -0.048187256, 0.01737976, -0.025253296, 0.008087158, -0.01637268, -0.023544312, 0.05923462, -0.052490234, 0.046722412, -0.0074...
http://arxiv.org/abs/2305.09107v1
Is a Video worth $n\times n$ Images? A Highly Efficient Approach to Transformer-based Video Question Answering
Chenyang Lyu, Tianbo Ji, Yvette Graham, Jennifer Foster
cs.CV, cs.AI, cs.CL, cs.MM
Conventional Transformer-based Video Question Answering (VideoQA) approaches generally encode frames independently through one or more image encoders followed by interaction between frames and question. However, such schema would incur significant memory use and inevitably slow down the training and inference speed. In...
http://arxiv.org/pdf/2305.09107v1
2023-05-16T02:12:57
2023-05-16T02:12:57
{ "title": [ 0.009811401, 0.013328552, -0.035461426, -0.044403076, 0.025146484, -0.024490356, -0.013183594, -0.014259338, 0.0026950836, 0.026763916, -0.060791016, 0.008811951, 0.016906738, -0.0040893555, -0.030075073, -0.025512695, 0.0096206665, ...
http://arxiv.org/abs/2305.09098v1
Weight-Inherited Distillation for Task-Agnostic BERT Compression
Taiqiang Wu, Cheng Hou, Zhe Zhao, Shanshan Lao, Jiayi Li, Ngai Wong, Yujiu Yang
cs.CL, cs.LG
Knowledge Distillation (KD) is a predominant approach for BERT compression. Previous KD-based methods focus on designing extra alignment losses for the student model to mimic the behavior of the teacher model. These methods transfer the knowledge in an indirect way. In this paper, we propose a novel Weight-Inherited Di...
http://arxiv.org/pdf/2305.09098v1
2023-05-16T01:51:22
2023-05-16T01:51:22
{ "title": [ -0.0435791, -0.032409668, -0.05291748, -0.10821533, -0.02557373, 0.03173828, -0.089538574, 0.06213379, 0.023361206, 0.060180664, -0.0005197525, -0.015640259, 0.03062439, -0.022781372, -0.05328369, 0.019683838, 0.017425537, -0.0366210...
http://arxiv.org/abs/2305.09067v1
SGP-TOD: Building Task Bots Effortlessly via Schema-Guided LLM Prompting
Xiaoying Zhang, Baolin Peng, Kun Li, Jingyan Zhou, Helen Meng
cs.CL
Building end-to-end task bots and maintaining their integration with new functionalities using minimal human efforts is a long-standing challenge in dialog research. Recently large language models (LLMs) have demonstrated exceptional proficiency in conversational engagement and adherence to instructions across various ...
http://arxiv.org/pdf/2305.09067v1
2023-05-15T23:29:56
2023-05-15T23:29:56
{ "title": [ -0.019210815, -0.033813477, -0.054656982, -0.06335449, -0.017913818, -0.012435913, -0.039611816, 0.02607727, -0.0011014938, 0.015090942, -0.02798462, 0.010063171, 0.009307861, -0.029693604, -0.011413574, -0.010971069, -0.0075950623, ...
http://arxiv.org/abs/2305.09025v1
Soft Prompt Decoding for Multilingual Dense Retrieval
Zhiqi Huang, Hansi Zeng, Hamed Zamani, James Allan
cs.IR, cs.CL
In this work, we explore a Multilingual Information Retrieval (MLIR) task, where the collection includes documents in multiple languages. We demonstrate that applying state-of-the-art approaches developed for cross-lingual information retrieval to MLIR tasks leads to sub-optimal performance. This is due to the heteroge...
http://arxiv.org/pdf/2305.09025v1
2023-05-15T21:17:17
2023-05-15T21:17:17
{ "title": [ -0.029876709, -0.020065308, -0.01777649, -0.06506348, -0.012359619, 0.017532349, -0.059570312, 0.008735657, 0.02947998, 0.056793213, -0.013885498, -0.013008118, 0.043884277, -0.013053894, -0.04446411, 0.029342651, 0.01776123, 0.01310...
http://arxiv.org/abs/2305.09022v1
It Takes Two to Tango: Navigating Conceptualizations of NLP Tasks and Measurements of Performance
Arjun Subramonian, Xingdi Yuan, Hal Daumé III, Su Lin Blodgett
cs.CL
Progress in NLP is increasingly measured through benchmarks; hence, contextualizing progress requires understanding when and why practitioners may disagree about the validity of benchmarks. We develop a taxonomy of disagreement, drawing on tools from measurement modeling, and distinguish between two types of disagreeme...
http://arxiv.org/pdf/2305.09022v1
2023-05-15T21:12:07
2023-05-15T21:12:07
{ "title": [ -0.029205322, -0.010055542, -0.037384033, -0.02331543, -0.0058059692, -0.016357422, -0.011581421, -0.012519836, -0.010101318, 0.013206482, -0.030776978, 0.01927185, -0.02784729, -0.008094788, -0.04727173, -0.020614624, -0.017105103, ...
http://arxiv.org/abs/2305.08982v1
Helping the Helper: Supporting Peer Counselors via AI-Empowered Practice and Feedback
Shang-Ling Hsu, Raj Sanjay Shah, Prathik Senthil, Zahra Ashktorab, Casey Dugan, Werner Geyer, Diyi Yang
cs.HC, cs.CL
Millions of users come to online peer counseling platforms to seek support on diverse topics ranging from relationship stress to anxiety. However, studies show that online peer support groups are not always as effective as expected largely due to users' negative experiences with unhelpful counselors. Peer counselors ar...
http://arxiv.org/pdf/2305.08982v1
2023-05-15T19:48:59
2023-05-15T19:48:59
{ "title": [ -0.020080566, -0.017562866, -0.02722168, -0.016357422, -0.027648926, -0.010940552, -0.008956909, -0.015823364, 0.012275696, 0.037902832, -0.040008545, 0.029846191, 0.024765015, -0.018417358, -0.014213562, 0.012298584, -0.007774353, 0...
http://arxiv.org/abs/2305.08978v1
An assessment of measuring local levels of homelessness through proxy social media signals
Yoshi Meke Bird, Sarah E. Grobe, Michael V. Arnold, Sean P. Rogers, Mikaela I. Fudolig, Julia Witte Zimmerman, Christopher M. Danforth, Peter Sheridan Dodds
cs.SI, cs.CL, cs.CY
Recent studies suggest social media activity can function as a proxy for measures of state-level public health, detectable through natural language processing. We present results of our efforts to apply this approach to estimate homelessness at the state level throughout the US during the period 2010-2019 and 2022 usin...
http://arxiv.org/pdf/2305.08978v1
2023-05-15T19:40:28
2023-05-15T19:40:28
{ "title": [ 0.057525635, 0.0345459, -0.017211914, -0.025909424, 0.042236328, -0.039093018, 0.0463562, -0.03262329, -0.038726807, 0.026626587, -0.038085938, -0.0025749207, -0.008003235, -0.03326416, 0.0034313202, -0.04284668, 0.0184021, 0.0064620...
http://arxiv.org/abs/2305.09688v1
OOD-Speech: A Large Bengali Speech Recognition Dataset for Out-of-Distribution Benchmarking
Fazle Rabbi Rakib, Souhardya Saha Dip, Samiul Alam, Nazia Tasnim, Md. Istiak Hossain Shihab, Md. Nazmuddoha Ansary, Syed Mobassir Hossen, Marsia Haque Meghla, Mamunur Mamun, Farig Sadeque, Sayma Sultana Chowdhury, Tahsin Reasat, Asif Sushmit, Ahmed Imtiaz Humayun
eess.AS, cs.CL, cs.LG
We present OOD-Speech, the first out-of-distribution (OOD) benchmarking dataset for Bengali automatic speech recognition (ASR). Being one of the most spoken languages globally, Bengali portrays large diversity in dialects and prosodic features, which demands ASR frameworks to be robust towards distribution shifts. For ...
http://arxiv.org/pdf/2305.09688v1
2023-05-15T18:00:39
2023-05-15T18:00:39
{ "title": [ -0.009185791, 0.00014340878, -0.026107788, -0.064086914, 0.010536194, 0.006954193, -0.0256958, 0.009979248, -0.007331848, 0.03286743, -0.044525146, 0.020309448, -0.01512146, -0.013366699, -0.04525757, -0.020584106, -0.0064888, 0.0645...
http://arxiv.org/abs/2305.08853v1
CQE: A Comprehensive Quantity Extractor
Satya Almasian, Vivian Kazakova, Philip Göldner, Michael Gertz
cs.CL, I.7; I.7.1; I.7.5
Quantities are essential in documents to describe factual information. They are ubiquitous in application domains such as finance, business, medicine, and science in general. Compared to other information extraction approaches, interestingly only a few works exist that describe methods for a proper extraction and repre...
http://arxiv.org/pdf/2305.08853v1
2023-05-15T17:59:41
2023-05-15T17:59:41
{ "title": [ -0.03970337, -0.0026416779, -0.07434082, -0.062072754, 0.019821167, -0.014312744, -0.029785156, -0.014778137, 0.0017023087, 0.030776978, -0.007896423, 0.008682251, -0.021514893, -0.0289917, -0.032165527, 0.019851685, 0.0032653809, 0....
http://arxiv.org/abs/2305.08848v1
Small Models are Valuable Plug-ins for Large Language Models
Canwen Xu, Yichong Xu, Shuohang Wang, Yang Liu, Chenguang Zhu, Julian McAuley
cs.CL, cs.AI, cs.LG
Large language models (LLMs) such as GPT-3 and GPT-4 are powerful but their weights are often publicly unavailable and their immense sizes make the models difficult to be tuned with common hardware. As a result, effectively tuning these models with large-scale supervised data can be challenging. As an alternative, In-C...
http://arxiv.org/pdf/2305.08848v1
2023-05-15T17:59:01
2023-05-15T17:59:01
{ "title": [ 0.0018720627, -0.021911621, -0.03564453, -0.06414795, -0.060333252, -0.008171082, -0.080444336, 0.016479492, -0.052215576, 0.04336548, 0.0003142357, -0.020004272, 0.0066719055, -0.058898926, 0.005756378, -0.01928711, 0.01612854, 0.00...
http://arxiv.org/abs/2305.08845v1
Large Language Models are Zero-Shot Rankers for Recommender Systems
Yupeng Hou, Junjie Zhang, Zihan Lin, Hongyu Lu, Ruobing Xie, Julian McAuley, Wayne Xin Zhao
cs.IR, cs.CL
Recently, large language models (LLMs) (e.g. GPT-4) have demonstrated impressive general-purpose task-solving abilities, including the potential to approach recommendation tasks. Along this line of research, this work aims to investigate the capacity of LLMs that act as the ranking model for recommender systems. To con...
http://arxiv.org/pdf/2305.08845v1
2023-05-15T17:57:39
2023-05-15T17:57:39
{ "title": [ 0.021774292, -0.014572144, -0.025268555, -0.026412964, -0.038024902, -0.02796936, -0.016418457, 0.010139465, 0.0013246536, 0.037475586, -0.026290894, -0.010055542, 0.014785767, -0.035827637, -0.006755829, -0.0113220215, 0.0018558502, ...