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/2310.14587v1
Large Search Model: Redefining Search Stack in the Era of LLMs
Liang Wang, Nan Yang, Xiaolong Huang, Linjun Yang, Rangan Majumder, Furu Wei
cs.IR, cs.CL
Modern search engines are built on a stack of different components, including query understanding, retrieval, multi-stage ranking, and question answering, among others. These components are often optimized and deployed independently. In this paper, we introduce a novel conceptual framework called large search model, wh...
http://arxiv.org/pdf/2310.14587v1
2023-10-23T05:52:09
2023-10-23T05:52:09
{ "title": [ -0.038848877, -0.013511658, -0.040130615, -0.037994385, -0.004524231, -0.009880066, -0.02571106, -0.016235352, 0.0021419525, 0.014297485, -0.030593872, -0.015853882, -0.01020813, -0.03125, -0.038238525, -0.013214111, -0.00374794, 0.0...
http://arxiv.org/abs/2310.14583v1
JointMatch: A Unified Approach for Diverse and Collaborative Pseudo-Labeling to Semi-Supervised Text Classification
Henry Peng Zou, Cornelia Caragea
cs.CL, cs.LG
Semi-supervised text classification (SSTC) has gained increasing attention due to its ability to leverage unlabeled data. However, existing approaches based on pseudo-labeling suffer from the issues of pseudo-label bias and error accumulation. In this paper, we propose JointMatch, a holistic approach for SSTC that addr...
http://arxiv.org/pdf/2310.14583v1
2023-10-23T05:43:35
2023-10-23T05:43:35
{ "title": [ -0.020828247, 0.011749268, -0.049591064, -0.0680542, 0.02519226, -0.013153076, -0.031219482, -0.0014896393, -0.016448975, 0.015464783, 0.0061416626, 0.027374268, -0.026763916, -0.025161743, -0.027908325, -0.014076233, -0.02444458, 0....
http://arxiv.org/abs/2310.14577v1
DeCrisisMB: Debiased Semi-Supervised Learning for Crisis Tweet Classification via Memory Bank
Henry Peng Zou, Yue Zhou, Weizhi Zhang, Cornelia Caragea
cs.CL, cs.LG
During crisis events, people often use social media platforms such as Twitter to disseminate information about the situation, warnings, advice, and support. Emergency relief organizations leverage such information to acquire timely crisis circumstances and expedite rescue operations. While existing works utilize such i...
http://arxiv.org/pdf/2310.14577v1
2023-10-23T05:25:51
2023-10-23T05:25:51
{ "title": [ 0.00894165, 0.020614624, -0.022247314, -0.053497314, -0.006526947, 0.0029678345, -0.061767578, -0.02468872, 0.01675415, 0.017913818, -0.055877686, 0.014640808, 0.0030574799, -0.01651001, -0.04168701, -0.028335571, -0.009925842, 0.009...
http://arxiv.org/abs/2310.14573v1
Exploring the Boundaries of GPT-4 in Radiology
Qianchu Liu, Stephanie Hyland, Shruthi Bannur, Kenza Bouzid, Daniel C. Castro, Maria Teodora Wetscherek, Robert Tinn, Harshita Sharma, Fernando Pérez-García, Anton Schwaighofer, Pranav Rajpurkar, Sameer Tajdin Khanna, Hoifung Poon, Naoto Usuyama, Anja Thieme, Aditya V. Nori, Matthew P. Lungren, Ozan Oktay, Javier Alvar...
cs.CL
The recent success of general-domain large language models (LLMs) has significantly changed the natural language processing paradigm towards a unified foundation model across domains and applications. In this paper, we focus on assessing the performance of GPT-4, the most capable LLM so far, on the text-based applicati...
http://arxiv.org/pdf/2310.14573v1
2023-10-23T05:13:03
2023-10-23T05:13:03
{ "title": [ 0.030670166, 0.030136108, -0.014144897, -0.041992188, 0.02079773, -0.020996094, 0.00012874603, -0.022888184, -0.015419006, 0.003862381, -0.003408432, -0.010826111, -0.054901123, -0.044677734, -0.022827148, 0.0012149811, 0.0146865845, ...
http://arxiv.org/abs/2310.14572v1
Unveiling the Multi-Annotation Process: Examining the Influence of Annotation Quantity and Instance Difficulty on Model Performance
Pritam Kadasi, Mayank Singh
cs.CL, cs.AI
The NLP community has long advocated for the construction of multi-annotator datasets to better capture the nuances of language interpretation, subjectivity, and ambiguity. This paper conducts a retrospective study to show how performance scores can vary when a dataset expands from a single annotation per instance to m...
http://arxiv.org/pdf/2310.14572v1
2023-10-23T05:12:41
2023-10-23T05:12:41
{ "title": [ 0.012908936, -0.0071525574, -0.024261475, -0.06744385, -0.01133728, -0.018157959, -0.05065918, -0.0317688, 0.002916336, 0.03186035, -0.004085541, 0.007507324, -0.016586304, -0.02494812, -0.045959473, 0.0028762817, -0.0006289482, 0.01...
http://arxiv.org/abs/2310.18349v1
A Boundary Offset Prediction Network for Named Entity Recognition
Minghao Tang, Yongquan He, Yongxiu Xu, Hongbo Xu, Wenyuan Zhang, Yang Lin
cs.CL
Named entity recognition (NER) is a fundamental task in natural language processing that aims to identify and classify named entities in text. However, span-based methods for NER typically assign entity types to text spans, resulting in an imbalanced sample space and neglecting the connections between non-entity and en...
http://arxiv.org/pdf/2310.18349v1
2023-10-23T05:04:07
2023-10-23T05:04:07
{ "title": [ -0.0044898987, 0.007205963, -0.034942627, -0.089416504, 0.014640808, 0.00957489, -0.05493164, 0.009552002, 0.0005517006, 0.035186768, -0.02748108, 0.0013256073, 0.009124756, -0.0012750626, -0.045074463, 0.0035934448, -0.0055503845, 0...
http://arxiv.org/abs/2310.14566v1
HallusionBench: You See What You Think? Or You Think What You See? An Image-Context Reasoning Benchmark Challenging for GPT-4V(ision), LLaVA-1.5, and Other Multi-modality Models
Fuxiao Liu, Tianrui Guan, Zongxia Li, Lichang Chen, Yaser Yacoob, Dinesh Manocha, Tianyi Zhou
cs.CV, cs.CL
Large language models (LLMs), after being aligned with vision models and integrated into vision-language models (VLMs), can bring impressive improvement in image reasoning tasks. This was shown by the recently released GPT-4V(ison), LLaVA-1.5, etc. However, the strong language prior in these SOTA LVLMs can be a double-...
http://arxiv.org/pdf/2310.14566v1
2023-10-23T04:49:09
2023-10-23T04:49:09
{ "title": [ 0.004673004, 0.003030777, 0.0046844482, 0.014099121, -0.015563965, -0.004547119, -0.012908936, -0.042297363, 0.013175964, 0.04827881, -0.03781128, 0.039367676, 0.0008635521, 0.00484848, -0.024734497, -0.0211792, 0.010765076, -0.00377...
http://arxiv.org/abs/2310.14564v1
Language Models Hallucinate, but May Excel at Fact Verification
Jian Guan, Jesse Dodge, David Wadden, Minlie Huang, Hao Peng
cs.CL
Recent progress in natural language processing (NLP) owes much to remarkable advances in large language models (LLMs). Nevertheless, LLMs frequently "hallucinate," resulting in non-factual outputs. Our carefully designed human evaluation substantiates the serious hallucination issue, revealing that even GPT-3.5 produce...
http://arxiv.org/pdf/2310.14564v1
2023-10-23T04:39:01
2023-10-23T04:39:01
{ "title": [ 0.04046631, -0.011459351, -0.018218994, -0.05911255, -0.022384644, -0.020248413, -0.014167786, -0.0032482147, -0.02909851, 0.047546387, -0.013504028, 0.019256592, 0.01902771, -0.027038574, -0.011161804, -0.008735657, -0.03894043, 0.0...
http://arxiv.org/abs/2310.14563v2
NormDial: A Comparable Bilingual Synthetic Dialog Dataset for Modeling Social Norm Adherence and Violation
Oliver Li, Mallika Subramanian, Arkadiy Saakyan, Sky CH-Wang, Smaranda Muresan
cs.CL, cs.CY
Social norms fundamentally shape interpersonal communication. We present NormDial, a high-quality dyadic dialogue dataset with turn-by-turn annotations of social norm adherences and violations for Chinese and American cultures. Introducing the task of social norm observance detection, our dataset is synthetically gener...
http://arxiv.org/pdf/2310.14563v2
2023-10-25T02:00:19
2023-10-23T04:38:34
{ "title": [ -0.00969696, -0.0031166077, -0.039794922, -0.07611084, -0.002117157, -0.002565384, -0.06323242, 0.006034851, 0.011833191, 0.032592773, -0.017318726, 0.0072669983, 0.0067977905, -0.020065308, -0.029434204, 0.0084991455, -0.040496826, ...
http://arxiv.org/abs/2310.18348v2
Meaning Representations from Trajectories in Autoregressive Models
Tian Yu Liu, Matthew Trager, Alessandro Achille, Pramuditha Perera, Luca Zancato, Stefano Soatto
cs.CL, cs.AI, cs.CV, cs.LG
We propose to extract meaning representations from autoregressive language models by considering the distribution of all possible trajectories extending an input text. This strategy is prompt-free, does not require fine-tuning, and is applicable to any pre-trained autoregressive model. Moreover, unlike vector-based rep...
http://arxiv.org/pdf/2310.18348v2
2023-11-02T03:33:34
2023-10-23T04:35:58
{ "title": [ -0.008125305, 0.00014197826, -0.03765869, -0.028945923, 0.006580353, -0.0074806213, -0.022872925, -0.041656494, 0.0017166138, 0.04284668, 0.0054512024, 0.007904053, -0.01826477, -0.01852417, -0.054748535, 0.012359619, 0.0018863678, 0...
http://arxiv.org/abs/2310.14558v1
AlpaCare:Instruction-tuned Large Language Models for Medical Application
Xinlu Zhang, Chenxin Tian, Xianjun Yang, Lichang Chen, Zekun Li, Linda Ruth Petzold
cs.CL, cs.AI
Large Language Models (LLMs) have demonstrated significant enhancements in instruction-following abilities through instruction tuning, achieving notable performances across various tasks. Previous research has focused on fine-tuning medical domain-specific LLMs using an extensive array of medical-specific data, incorpo...
http://arxiv.org/pdf/2310.14558v1
2023-10-23T04:22:50
2023-10-23T04:22:50
{ "title": [ -0.063964844, -0.02986145, -0.046966553, -0.06536865, -0.008621216, 0.01789856, -0.08660889, 0.0067710876, -0.008544922, 0.037322998, 0.006389618, 0.0010356903, 0.0041160583, -0.031311035, -0.03552246, 0.010406494, 0.0019369125, 0.01...
http://arxiv.org/abs/2310.14557v1
The Skipped Beat: A Study of Sociopragmatic Understanding in LLMs for 64 Languages
Chiyu Zhang, Khai Duy Doan, Qisheng Liao, Muhammad Abdul-Mageed
cs.CL
Instruction tuned large language models (LLMs), such as ChatGPT, demonstrate remarkable performance in a wide range of tasks. Despite numerous recent studies that examine the performance of instruction-tuned LLMs on various NLP benchmarks, there remains a lack of comprehensive investigation into their ability to unders...
http://arxiv.org/pdf/2310.14557v1
2023-10-23T04:22:44
2023-10-23T04:22:44
{ "title": [ -0.037322998, -0.0055770874, -0.047729492, -0.02960205, -0.02331543, -0.014862061, -0.07287598, -0.009613037, -0.0041389465, -0.005405426, -0.071777344, 0.026138306, -0.013076782, -0.021835327, 0.007820129, -0.02420044, -0.020889282, ...
http://arxiv.org/abs/2310.14545v2
Harnessing ChatGPT for thematic analysis: Are we ready?
V Vien Lee, Stephanie C. C. van der Lubbe, Lay Hoon Goh, Jose M. Valderas
cs.CL
ChatGPT is an advanced natural language processing tool with growing applications across various disciplines in medical research. Thematic analysis, a qualitative research method to identify and interpret patterns in data, is one application that stands to benefit from this technology. This viewpoint explores the utili...
http://arxiv.org/pdf/2310.14545v2
2023-10-24T01:56:05
2023-10-23T03:55:13
{ "title": [ 0.011108398, -0.013908386, -0.0004310608, -0.02003479, 0.009048462, 0.0052223206, -0.012069702, -0.039642334, -0.0032539368, 0.02746582, 0.0023555756, 0.034301758, -0.0211792, -0.0126571655, -0.03869629, 0.025238037, -0.028167725, 0....
http://arxiv.org/abs/2310.14542v1
Evaluating Large Language Models on Controlled Generation Tasks
Jiao Sun, Yufei Tian, Wangchunshu Zhou, Nan Xu, Qian Hu, Rahul Gupta, John Frederick Wieting, Nanyun Peng, Xuezhe Ma
cs.CL
While recent studies have looked into the abilities of large language models in various benchmark tasks, including question generation, reading comprehension, multilingual and etc, there have been few studies looking into the controllability of large language models on generation tasks. We present an extensive analysis...
http://arxiv.org/pdf/2310.14542v1
2023-10-23T03:48:24
2023-10-23T03:48:24
{ "title": [ -0.039093018, -0.012161255, -0.043060303, -0.04949951, -0.0028381348, 0.002937317, -0.060333252, 0.00856781, 0.009887695, 0.022659302, -0.009468079, -0.0061035156, -0.0029945374, -0.026321411, -0.03756714, 0.011329651, -0.021316528, ...
http://arxiv.org/abs/2310.14541v1
Continual Named Entity Recognition without Catastrophic Forgetting
Duzhen Zhang, Wei Cong, Jiahua Dong, Yahan Yu, Xiuyi Chen, Yonggang Zhang, Zhen Fang
cs.CL
Continual Named Entity Recognition (CNER) is a burgeoning area, which involves updating an existing model by incorporating new entity types sequentially. Nevertheless, continual learning approaches are often severely afflicted by catastrophic forgetting. This issue is intensified in CNER due to the consolidation of old...
http://arxiv.org/pdf/2310.14541v1
2023-10-23T03:45:30
2023-10-23T03:45:30
{ "title": [ -0.0069885254, 0.010765076, -0.0524292, -0.09655762, 0.03253174, -0.006290436, -0.037322998, 0.017578125, 0.019592285, 0.034851074, -0.026412964, 0.010620117, 0.003929138, 0.0042495728, -0.050231934, 0.008705139, -0.0041999817, 0.040...
http://arxiv.org/abs/2310.14540v1
Evaluating Spatial Understanding of Large Language Models
Yutaro Yamada, Yihan Bao, Andrew K. Lampinen, Jungo Kasai, Ilker Yildirim
cs.CL, cs.AI
Large language models (LLMs) show remarkable capabilities across a variety of tasks. Despite the models only seeing text in training, several recent studies suggest that LLM representations implicitly capture aspects of the underlying grounded concepts. Here, we explore LLM representations of a particularly salient kin...
http://arxiv.org/pdf/2310.14540v1
2023-10-23T03:44:40
2023-10-23T03:44:40
{ "title": [ -0.017837524, -0.0234375, -0.012039185, -0.046905518, -0.015327454, -0.005153656, -0.038635254, -0.01209259, 0.025985718, 0.020401001, -0.03817749, -0.032104492, -0.0022563934, -0.008491516, -0.05871582, -0.010322571, -0.0079193115, ...
http://arxiv.org/abs/2310.14534v1
Improving Seq2Seq Grammatical Error Correction via Decoding Interventions
Houquan Zhou, Yumeng Liu, Zhenghua Li, Min Zhang, Bo Zhang, Chen Li, Ji Zhang, Fei Huang
cs.CL
The sequence-to-sequence (Seq2Seq) approach has recently been widely used in grammatical error correction (GEC) and shows promising performance. However, the Seq2Seq GEC approach still suffers from two issues. First, a Seq2Seq GEC model can only be trained on parallel data, which, in GEC task, is often noisy and limite...
http://arxiv.org/pdf/2310.14534v1
2023-10-23T03:36:37
2023-10-23T03:36:37
{ "title": [ -0.011955261, -0.0046463013, -0.030227661, -0.037231445, -0.008369446, -0.01084137, -0.044677734, -0.014701843, -0.0017585754, 0.03652954, -0.012550354, 0.008682251, 0.022323608, -0.023773193, -0.036865234, 0.015220642, -0.030197144, ...
http://arxiv.org/abs/2310.14528v1
Dual-Feedback Knowledge Retrieval for Task-Oriented Dialogue Systems
Tianyuan Shi, Liangzhi Li, Zijian Lin, Tao Yang, Xiaojun Quan, Qifan Wang
cs.CL
Efficient knowledge retrieval plays a pivotal role in ensuring the success of end-to-end task-oriented dialogue systems by facilitating the selection of relevant information necessary to fulfill user requests. However, current approaches generally integrate knowledge retrieval and response generation, which poses scala...
http://arxiv.org/pdf/2310.14528v1
2023-10-23T03:21:11
2023-10-23T03:21:11
{ "title": [ -0.022232056, -0.004611969, -0.04458618, -0.068603516, 0.009803772, -0.002538681, -0.042053223, 0.017410278, 0.011833191, 0.053009033, -0.035369873, -0.0056419373, 0.016082764, -0.00078725815, -0.072509766, -0.0127334595, 0.0056877136, ...
http://arxiv.org/abs/2310.18347v1
PRCA: Fitting Black-Box Large Language Models for Retrieval Question Answering via Pluggable Reward-Driven Contextual Adapter
Haoyan Yang, Zhitao Li, Yong Zhang, Jianzong Wang, Ning Cheng, Ming Li, Jing Xiao
cs.CL
The Retrieval Question Answering (ReQA) task employs the retrieval-augmented framework, composed of a retriever and generator. The generator formulates the answer based on the documents retrieved by the retriever. Incorporating Large Language Models (LLMs) as generators is beneficial due to their advanced QA capabiliti...
http://arxiv.org/pdf/2310.18347v1
2023-10-23T03:12:00
2023-10-23T03:12:00
{ "title": [ 0.025009155, 0.0065574646, -0.037475586, -0.061553955, 0.012519836, -0.012252808, -0.030197144, -0.015716553, 0.018173218, 0.04421997, -0.025512695, 0.0036525726, 0.00258255, 0.00055360794, -0.06088257, 0.0010728836, -0.00083875656, ...
http://arxiv.org/abs/2310.14523v2
Rethinking Word-Level Auto-Completion in Computer-Aided Translation
Xingyu Chen, Lemao Liu, Guoping Huang, Zhirui Zhang, Mingming Yang, Shuming Shi, Rui Wang
cs.CL
Word-Level Auto-Completion (WLAC) plays a crucial role in Computer-Assisted Translation. It aims at providing word-level auto-completion suggestions for human translators. While previous studies have primarily focused on designing complex model architectures, this paper takes a different perspective by rethinking the f...
http://arxiv.org/pdf/2310.14523v2
2023-10-24T06:48:07
2023-10-23T03:11:46
{ "title": [ -0.04055786, -0.014762878, -0.0042877197, -0.032104492, -0.029541016, 0.016952515, -0.032470703, -0.007774353, 0.00008201599, 0.05734253, -0.03161621, 0.02494812, 0.03704834, -0.0035743713, -0.05303955, 0.022964478, -0.0069084167, 0....
http://arxiv.org/abs/2310.14520v2
QUDEVAL: The Evaluation of Questions Under Discussion Discourse Parsing
Yating Wu, Ritika Mangla, Greg Durrett, Junyi Jessy Li
cs.CL
Questions Under Discussion (QUD) is a versatile linguistic framework in which discourse progresses as continuously asking questions and answering them. Automatic parsing of a discourse to produce a QUD structure thus entails a complex question generation task: given a document and an answer sentence, generate a questio...
http://arxiv.org/pdf/2310.14520v2
2023-11-01T20:30:41
2023-10-23T03:03:58
{ "title": [ -0.006668091, 0.012016296, -0.0090408325, -0.03967285, 0.0014762878, 0.0055885315, -0.04034424, -0.03765869, -0.012626648, 0.012969971, -0.0069465637, 0.009407043, -0.022415161, -0.018127441, -0.042388916, -0.002281189, -0.01927185, ...
http://arxiv.org/abs/2310.14513v1
Turn-Level Active Learning for Dialogue State Tracking
Zihan Zhang, Meng Fang, Fanghua Ye, Ling Chen, Mohammad-Reza Namazi-Rad
cs.CL
Dialogue state tracking (DST) plays an important role in task-oriented dialogue systems. However, collecting a large amount of turn-by-turn annotated dialogue data is costly and inefficient. In this paper, we propose a novel turn-level active learning framework for DST to actively select turns in dialogues to annotate....
http://arxiv.org/pdf/2310.14513v1
2023-10-23T02:53:46
2023-10-23T02:53:46
{ "title": [ 0.006916046, -0.008094788, -0.009521484, -0.07385254, 0.0009212494, -0.0033855438, -0.030029297, 0.0027809143, 0.012420654, 0.013923645, -0.0024738312, 0.005466461, -0.03970337, -0.03717041, -0.024932861, -0.0011816025, -0.0070533752, ...
http://arxiv.org/abs/2310.14512v2
CorefPrompt: Prompt-based Event Coreference Resolution by Measuring Event Type and Argument Compatibilities
Sheng Xu, Peifeng Li, Qiaoming Zhu
cs.CL, cs.AI, cs.IR
Event coreference resolution (ECR) aims to group event mentions referring to the same real-world event into clusters. Most previous studies adopt the "encoding first, then scoring" framework, making the coreference judgment rely on event encoding. Furthermore, current methods struggle to leverage human-summarized ECR r...
http://arxiv.org/pdf/2310.14512v2
2023-10-24T02:45:55
2023-10-23T02:47:27
{ "title": [ 0.030532837, -0.0033187866, -0.035064697, -0.11236572, 0.030471802, -0.007282257, -0.04714966, -0.0038280487, -0.0047035217, 0.026031494, 0.02204895, 0.020645142, 0.026062012, -0.039642334, -0.013801575, 0.036468506, -0.011894226, 0....
http://arxiv.org/abs/2310.14510v1
CITB: A Benchmark for Continual Instruction Tuning
Zihan Zhang, Meng Fang, Ling Chen, Mohammad-Reza Namazi-Rad
cs.CL
Continual learning (CL) is a paradigm that aims to replicate the human ability to learn and accumulate knowledge continually without forgetting previous knowledge and transferring it to new tasks. Recent instruction tuning (IT) involves fine-tuning models to make them more adaptable to solving NLP tasks in general. How...
http://arxiv.org/pdf/2310.14510v1
2023-10-23T02:42:32
2023-10-23T02:42:32
{ "title": [ -0.047576904, -0.00041341782, -0.023483276, -0.058441162, -0.004764557, 0.0019664764, -0.038635254, 0.003370285, 0.017059326, 0.042266846, -0.04107666, 0.025680542, -0.00021076202, 0.0003244877, -0.051635742, 0.014060974, 0.01071167, ...
http://arxiv.org/abs/2310.14508v1
EXPLAIN, EDIT, GENERATE: Rationale-Sensitive Counterfactual Data Augmentation for Multi-hop Fact Verification
Yingjie Zhu, Jiasheng Si, Yibo Zhao, Haiyang Zhu, Deyu Zhou, Yulan He
cs.CL
Automatic multi-hop fact verification task has gained significant attention in recent years. Despite impressive results, these well-designed models perform poorly on out-of-domain data. One possible solution is to augment the training data with counterfactuals, which are generated by minimally altering the causal featu...
http://arxiv.org/pdf/2310.14508v1
2023-10-23T02:39:14
2023-10-23T02:39:14
{ "title": [ 0.01449585, -0.002626419, -0.04510498, -0.04345703, -0.0041236877, -0.030410767, -0.0036411285, -0.0016727448, -0.012542725, 0.029144287, -0.023651123, 0.017456055, 0.0074424744, -0.017486572, -0.031585693, -0.009284973, -0.029769897, ...
http://arxiv.org/abs/2310.14505v1
Sentiment analysis with adaptive multi-head attention in Transformer
Fanfei Meng, David Demeter
cs.CL
We propose a novel framework based on the attention mechanism to identify the sentiment of a movie review document. Previous efforts on deep neural networks with attention mechanisms focus on encoder and decoder with fixed numbers of multi-head attention. Therefore, we need a mechanism to stop the attention process aut...
http://arxiv.org/pdf/2310.14505v1
2023-10-23T02:32:30
2023-10-23T02:32:30
{ "title": [ 0.0016174316, 0.014472961, -0.042022705, -0.034240723, 0.021911621, -0.012611389, -0.020080566, -0.015823364, 0.016784668, 0.050201416, -0.035003662, -0.00020849705, 0.03164673, -0.001288414, -0.043670654, 0.01348114, -0.0044670105, ...
http://arxiv.org/abs/2310.14503v1
Diversify Question Generation with Retrieval-Augmented Style Transfer
Qi Gou, Zehua Xia, Bowen Yu, Haiyang Yu, Fei Huang, Yongbin Li, Nguyen Cam-Tu
cs.CL
Given a textual passage and an answer, humans are able to ask questions with various expressions, but this ability is still challenging for most question generation (QG) systems. Existing solutions mainly focus on the internal knowledge within the given passage or the semantic word space for diverse content planning. T...
http://arxiv.org/pdf/2310.14503v1
2023-10-23T02:27:31
2023-10-23T02:27:31
{ "title": [ -0.013298035, 0.00057554245, -0.04949951, -0.033111572, -0.001531601, -0.008651733, -0.01927185, -0.0043258667, 0.013786316, 0.008575439, -0.0030956268, 0.032470703, -0.026779175, -0.019256592, -0.025024414, 0.012573242, 0.010314941, ...
http://arxiv.org/abs/2310.14495v1
InstructExcel: A Benchmark for Natural Language Instruction in Excel
Justin Payan, Swaroop Mishra, Mukul Singh, Carina Negreanu, Christian Poelitz, Chitta Baral, Subhro Roy, Rasika Chakravarthy, Benjamin Van Durme, Elnaz Nouri
cs.CL, cs.AI
With the evolution of Large Language Models (LLMs) we can solve increasingly more complex NLP tasks across various domains, including spreadsheets. This work investigates whether LLMs can generate code (Excel OfficeScripts, a TypeScript API for executing many tasks in Excel) that solves Excel specific tasks provided vi...
http://arxiv.org/pdf/2310.14495v1
2023-10-23T02:00:55
2023-10-23T02:00:55
{ "title": [ -0.04272461, -0.0335083, -0.04711914, -0.03744507, -0.023040771, -0.0079956055, -0.03137207, 0.006790161, -0.009643555, 0.02859497, -0.036346436, 0.019805908, 0.0030174255, -0.011512756, -0.029632568, -0.014457703, -0.018295288, 0.04...
http://arxiv.org/abs/2310.14491v1
Towards a Mechanistic Interpretation of Multi-Step Reasoning Capabilities of Language Models
Yifan Hou, Jiaoda Li, Yu Fei, Alessandro Stolfo, Wangchunshu Zhou, Guangtao Zeng, Antoine Bosselut, Mrinmaya Sachan
cs.CL
Recent work has shown that language models (LMs) have strong multi-step (i.e., procedural) reasoning capabilities. However, it is unclear whether LMs perform these tasks by cheating with answers memorized from pretraining corpus, or, via a multi-step reasoning mechanism. In this paper, we try to answer this question by...
http://arxiv.org/pdf/2310.14491v1
2023-10-23T01:47:29
2023-10-23T01:47:29
{ "title": [ -0.0345459, -0.031463623, -0.031951904, -0.08734131, -0.030517578, 0.017105103, -0.09100342, 0.029342651, 0.0075912476, 0.022094727, 0.0061187744, -0.0026493073, -0.00944519, -0.033325195, -0.045959473, 0.02330017, -0.006290436, 0.01...
http://arxiv.org/abs/2310.14486v1
Text Fact Transfer
Nishant Balepur, Jie Huang, Kevin Chen-Chuan Chang
cs.CL
Text style transfer is a prominent task that aims to control the style of text without inherently changing its factual content. To cover more text modification applications, such as adapting past news for current events and repurposing educational materials, we propose the task of text fact transfer, which seeks to tra...
http://arxiv.org/pdf/2310.14486v1
2023-10-23T01:36:44
2023-10-23T01:36:44
{ "title": [ 0.006504059, -0.019058228, -0.006324768, -0.048095703, -0.05810547, -0.013580322, -0.04272461, 0.0022258759, -0.008522034, 0.0597229, -0.008842468, -0.00024461746, 0.041137695, -0.016448975, 0.00566864, 0.0059165955, 0.02027893, 0.01...
http://arxiv.org/abs/2310.14483v1
"Why Should I Review This Paper?" Unifying Semantic, Topic, and Citation Factors for Paper-Reviewer Matching
Yu Zhang, Yanzhen Shen, Xiusi Chen, Bowen Jin, Jiawei Han
cs.IR, cs.CL, cs.DL, cs.LG
As many academic conferences are overwhelmed by a rapidly increasing number of paper submissions, automatically finding appropriate reviewers for each submission becomes a more urgent need than ever. Various factors have been considered by previous attempts on this task to measure the expertise relevance between a pape...
http://arxiv.org/pdf/2310.14483v1
2023-10-23T01:29:18
2023-10-23T01:29:18
{ "title": [ -0.0340271, 0.0043258667, -0.043640137, -0.013755798, -0.004802704, -0.021148682, -0.020706177, -0.032989502, -0.010414124, 0.031707764, -0.028762817, 0.042419434, 0.012863159, 0.0014219284, -0.012962341, 0.007007599, 0.00070762634, ...
http://arxiv.org/abs/2310.14479v1
DetectGPT-SC: Improving Detection of Text Generated by Large Language Models through Self-Consistency with Masked Predictions
Rongsheng Wang, Qi Li, Sihong Xie
cs.CL
General large language models (LLMs) such as ChatGPT have shown remarkable success, but it has also raised concerns among people about the misuse of AI-generated texts. Therefore, an important question is how to detect whether the texts are generated by ChatGPT or by humans. Existing detectors are built on the assumpti...
http://arxiv.org/pdf/2310.14479v1
2023-10-23T01:23:10
2023-10-23T01:23:10
{ "title": [ -0.014762878, 0.020370483, -0.04067993, -0.03970337, 0.009841919, -0.0041770935, -0.05517578, -0.032043457, -0.00868988, 0.02279663, -0.007610321, 0.011352539, 0.016113281, -0.026321411, -0.021499634, 0.013145447, -0.025741577, 0.055...
http://arxiv.org/abs/2310.14478v1
GeoLM: Empowering Language Models for Geospatially Grounded Language Understanding
Zekun Li, Wenxuan Zhou, Yao-Yi Chiang, Muhao Chen
cs.CL
Humans subconsciously engage in geospatial reasoning when reading articles. We recognize place names and their spatial relations in text and mentally associate them with their physical locations on Earth. Although pretrained language models can mimic this cognitive process using linguistic context, they do not utilize ...
http://arxiv.org/pdf/2310.14478v1
2023-10-23T01:20:01
2023-10-23T01:20:01
{ "title": [ -0.026260376, -0.013496399, -0.027526855, -0.06390381, -0.0014038086, 0.0060195923, -0.054351807, -0.00025105476, 0.009284973, 0.018432617, -0.019348145, -0.015319824, 0.013000488, -0.024169922, -0.04031372, 0.005203247, -0.01625061, ...
http://arxiv.org/abs/2310.14451v1
Domain Terminology Integration into Machine Translation: Leveraging Large Language Models
Yasmin Moslem, Gianfranco Romani, Mahdi Molaei, Rejwanul Haque, John D. Kelleher, Andy Way
cs.CL
This paper discusses the methods that we used for our submissions to the WMT 2023 Terminology Shared Task for German-to-English (DE-EN), English-to-Czech (EN-CS), and Chinese-to-English (ZH-EN) language pairs. The task aims to advance machine translation (MT) by challenging participants to develop systems that accurate...
http://arxiv.org/pdf/2310.14451v1
2023-10-22T23:25:28
2023-10-22T23:25:28
{ "title": [ -0.02758789, -0.0102005005, -0.038360596, -0.06970215, 0.0027561188, 0.00059890747, -0.046722412, 0.013114929, 0.004840851, 0.053497314, 0.005290985, 0.002910614, 0.06365967, -0.01890564, -0.035064697, 0.03451538, -0.015686035, 0.037...
http://arxiv.org/abs/2310.14450v2
TATA: Stance Detection via Topic-Agnostic and Topic-Aware Embeddings
Hans W. A. Hanley, Zakir Durumeric
cs.CL, cs.CY, cs.LG
Stance detection is important for understanding different attitudes and beliefs on the Internet. However, given that a passage's stance toward a given topic is often highly dependent on that topic, building a stance detection model that generalizes to unseen topics is difficult. In this work, we propose using contrasti...
http://arxiv.org/pdf/2310.14450v2
2023-11-13T03:22:32
2023-10-22T23:23:44
{ "title": [ 0.0029354095, 0.010910034, -0.03201294, -0.05709839, 0.0032482147, -0.0107421875, -0.036193848, -0.029083252, 0.034301758, 0.01965332, -0.046020508, 0.01399231, -0.0021076202, -0.006137848, -0.03488159, -0.0048065186, -0.011581421, 0...
http://arxiv.org/abs/2310.14435v1
Retrieval-Augmented Chain-of-Thought in Semi-structured Domains
Vaibhav Mavi, Abulhair Saparov, Chen Zhao
cs.CL, cs.AI
Applying existing question answering (QA) systems to specialized domains like law and finance presents challenges that necessitate domain expertise. Although large language models (LLMs) have shown impressive language comprehension and in-context learning capabilities, their inability to handle very long inputs/context...
http://arxiv.org/pdf/2310.14435v1
2023-10-22T22:45:14
2023-10-22T22:45:14
{ "title": [ 0.027862549, -0.006702423, -0.013656616, -0.049468994, 0.0023479462, -0.022338867, -0.00023770332, -0.015899658, 0.0016422272, 0.02532959, -0.006504059, -0.013687134, 0.012580872, -0.034820557, -0.033050537, 0.0025291443, -0.0058403015, ...
http://arxiv.org/abs/2310.14429v1
Text generation for dataset augmentation in security classification tasks
Alexander P. Welsh, Matthew Edwards
cs.CR, cs.CL
Security classifiers, designed to detect malicious content in computer systems and communications, can underperform when provided with insufficient training data. In the security domain, it is often easy to find samples of the negative (benign) class, and challenging to find enough samples of the positive (malicious) c...
http://arxiv.org/pdf/2310.14429v1
2023-10-22T22:25:14
2023-10-22T22:25:14
{ "title": [ -0.015457153, -0.00009739399, -0.060516357, -0.034851074, 0.011993408, -0.018096924, -0.02067566, 0.0026760101, -0.022216797, 0.017501831, -0.021621704, 0.019592285, -0.02859497, -0.013465881, -0.035980225, -0.022384644, -0.024749756, ...
http://arxiv.org/abs/2310.14424v1
Which Prompts Make The Difference? Data Prioritization For Efficient Human LLM Evaluation
Meriem Boubdir, Edward Kim, Beyza Ermis, Marzieh Fadaee, Sara Hooker
cs.CL, cs.AI
Human evaluation is increasingly critical for assessing large language models, capturing linguistic nuances, and reflecting user preferences more accurately than traditional automated metrics. However, the resource-intensive nature of this type of annotation process poses significant challenges. The key question drivin...
http://arxiv.org/pdf/2310.14424v1
2023-10-22T21:48:51
2023-10-22T21:48:51
{ "title": [ -0.030319214, -0.0028038025, -0.056121826, -0.05545044, -0.025131226, -0.0132369995, -0.07946777, -0.009857178, -0.016677856, 0.008117676, -0.032958984, 0.010673523, -0.020370483, -0.025909424, -0.015701294, -0.023269653, -0.031555176, ...
http://arxiv.org/abs/2310.14422v1
Large Language Models are biased to overestimate profoundness
Eugenio Herrera-Berg, Tomás Vergara Browne, Pablo León-Villagrá, Marc-Lluís Vives, Cristian Buc Calderon
cs.CL
Recent advancements in natural language processing by large language models (LLMs), such as GPT-4, have been suggested to approach Artificial General Intelligence. And yet, it is still under dispute whether LLMs possess similar reasoning abilities to humans. This study evaluates GPT-4 and various other LLMs in judging ...
http://arxiv.org/pdf/2310.14422v1
2023-10-22T21:33:50
2023-10-22T21:33:50
{ "title": [ -0.013725281, -0.008552551, -0.028945923, -0.02154541, -0.056365967, -0.033111572, -0.029083252, -0.012893677, 0.013954163, 0.026184082, -0.048553467, -0.044006348, 0.02255249, -0.035095215, -0.026565552, -0.014457703, -0.006084442, ...
http://arxiv.org/abs/2310.14418v1
REFER: An End-to-end Rationale Extraction Framework for Explanation Regularization
Mohammad Reza Ghasemi Madani, Pasquale Minervini
cs.CL
Human-annotated textual explanations are becoming increasingly important in Explainable Natural Language Processing. Rationale extraction aims to provide faithful (i.e., reflective of the behavior of the model) and plausible (i.e., convincing to humans) explanations by highlighting the inputs that had the largest impac...
http://arxiv.org/pdf/2310.14418v1
2023-10-22T21:20:52
2023-10-22T21:20:52
{ "title": [ 0.03704834, -0.024978638, -0.010040283, -0.10467529, -0.038970947, 0.00472641, -0.05303955, 0.009498596, -0.008178711, 0.0074501038, 0.011154175, 0.008255005, 0.0020942688, -0.04168701, 0.025680542, -0.007019043, -0.009971619, 0.0457...
http://arxiv.org/abs/2310.14404v1
Be Selfish, But Wisely: Investigating the Impact of Agent Personality in Mixed-Motive Human-Agent Interactions
Kushal Chawla, Ian Wu, Yu Rong, Gale M. Lucas, Jonathan Gratch
cs.CL, cs.AI
A natural way to design a negotiation dialogue system is via self-play RL: train an agent that learns to maximize its performance by interacting with a simulated user that has been designed to imitate human-human dialogue data. Although this procedure has been adopted in prior work, we find that it results in a fundame...
http://arxiv.org/pdf/2310.14404v1
2023-10-22T20:31:35
2023-10-22T20:31:35
{ "title": [ 0.0027942657, 0.006690979, -0.030014038, -0.03543091, 0.022644043, -0.034454346, 0.019210815, -0.022155762, -0.043151855, 0.02067566, -0.0077171326, 0.0049362183, -0.016601562, -0.026275635, -0.022323608, -0.02229309, -0.015838623, 0...
http://arxiv.org/abs/2310.14403v1
O3D: Offline Data-driven Discovery and Distillation for Sequential Decision-Making with Large Language Models
Yuchen Xiao, Yanchao Sun, Mengda Xu, Udari Madhushani, Jared Vann, Deepeka Garg, Sumitra Ganesh
cs.AI, cs.CL
Recent advancements in large language models (LLMs) have exhibited promising performance in solving sequential decision-making problems. By imitating few-shot examples provided in the prompts (i.e., in-context learning), an LLM agent can interact with an external environment and complete given tasks without additional ...
http://arxiv.org/pdf/2310.14403v1
2023-10-22T20:28:33
2023-10-22T20:28:33
{ "title": [ -0.024597168, -0.023284912, -0.053741455, -0.07861328, -0.0007095337, 0.0015935898, -0.050933838, 0.011909485, 0.01096344, 0.0029659271, -0.03918457, -0.007965088, -0.011932373, -0.018569946, -0.01725769, -0.025421143, -0.011299133, ...
http://arxiv.org/abs/2310.14393v1
Merging Generated and Retrieved Knowledge for Open-Domain QA
Yunxiang Zhang, Muhammad Khalifa, Lajanugen Logeswaran, Moontae Lee, Honglak Lee, Lu Wang
cs.CL, cs.AI
Open-domain question answering (QA) systems are often built with retrieval modules. However, retrieving passages from a given source is known to suffer from insufficient knowledge coverage. Alternatively, prompting large language models (LLMs) to generate contextual passages based on their parametric knowledge has been...
http://arxiv.org/pdf/2310.14393v1
2023-10-22T19:37:06
2023-10-22T19:37:06
{ "title": [ -0.013671875, 0.0048713684, -0.057922363, -0.08404541, 0.011543274, -0.0011997223, -0.04006958, 0.011543274, -0.01083374, 0.036468506, -0.02557373, 0.00018119812, 0.001200676, -0.011070251, -0.026748657, -0.017181396, 0.0052452087, 0...
http://arxiv.org/abs/2310.14389v1
Evaluating Subjective Cognitive Appraisals of Emotions from Large Language Models
Hongli Zhan, Desmond C. Ong, Junyi Jessy Li
cs.CL
The emotions we experience involve complex processes; besides physiological aspects, research in psychology has studied cognitive appraisals where people assess their situations subjectively, according to their own values (Scherer, 2005). Thus, the same situation can often result in different emotional experiences. Whi...
http://arxiv.org/pdf/2310.14389v1
2023-10-22T19:12:17
2023-10-22T19:12:17
{ "title": [ -0.017593384, -0.0045776367, -0.026489258, -0.03326416, -0.0049209595, -0.013717651, -0.027252197, -0.020523071, 0.028305054, 0.023208618, -0.019744873, -0.033111572, 0.0010118484, -0.026245117, -0.070739746, -0.0064811707, -0.0087509155...
http://arxiv.org/abs/2310.14384v1
ARCOQ: Arabic Closest Opposite Questions Dataset
Sandra Rizkallah, Amir F. Atiya, Samir Shaheen
cs.CL, cs.AI
This paper presents a dataset for closest opposite questions in Arabic language. The dataset is the first of its kind for the Arabic language. It is beneficial for the assessment of systems on the aspect of antonymy detection. The structure is similar to that of the Graduate Record Examination (GRE) closest opposite qu...
http://arxiv.org/pdf/2310.14384v1
2023-10-22T18:41:26
2023-10-22T18:41:26
{ "title": [ -0.025100708, -0.017349243, -0.032714844, 0.00009256601, -0.054840088, -0.014633179, -0.028076172, -0.013534546, 0.021057129, 0.029800415, -0.013824463, 0.0027313232, 0.016952515, -0.05529785, -0.0040740967, 0.008041382, 0.011001587, ...
http://arxiv.org/abs/2310.14366v1
Bi-Encoders based Species Normalization -- Pairwise Sentence Learning to Rank
Zainab Awan, Tim Kahlke, Peter Ralph, Paul Kennedy
cs.CL, cs.LG
Motivation: Biomedical named-entity normalization involves connecting biomedical entities with distinct database identifiers in order to facilitate data integration across various fields of biology. Existing systems for biomedical named entity normalization heavily rely on dictionaries, manually created rules, and high...
http://arxiv.org/pdf/2310.14366v1
2023-10-22T17:30:16
2023-10-22T17:30:16
{ "title": [ -0.010650635, 0.004070282, -0.027374268, -0.021453857, 0.0006594658, -0.014076233, -0.010757446, -0.0069274902, 0.0041542053, 0.040740967, -0.026794434, 0.015327454, -0.00045967102, -0.013626099, -0.032440186, -0.0020313263, -0.020812988...
http://arxiv.org/abs/2310.14360v2
Is ChatGPT a game changer for geocoding -- a benchmark for geocoding address parsing techniques
Zhengcong Yin, Diya Li, Daniel W. Goldberg
cs.CL, cs.LG
The remarkable success of GPT models across various tasks, including toponymy recognition motivates us to assess the performance of the GPT-3 model in the geocoding address parsing task. To ensure that the evaluation more accurately mirrors performance in real-world scenarios with diverse user input qualities and resol...
http://arxiv.org/pdf/2310.14360v2
2023-11-09T07:29:14
2023-10-22T17:03:56
{ "title": [ 0.007987976, -0.011398315, 0.038604736, 0.01550293, -0.04888916, 0.01058197, -0.012168884, -0.032043457, -0.045043945, 0.0463562, -0.02734375, 0.020904541, 0.009109497, -0.019378662, 0.012710571, -0.024536133, -0.00440979, 0.04727173...
http://arxiv.org/abs/2310.14358v1
Right, No Matter Why: AI Fact-checking and AI Authority in Health-related Inquiry Settings
Elena Sergeeva, Anastasia Sergeeva, Huiyun Tang, Kerstin Bongard-Blanchy, Peter Szolovits
cs.CL, cs.AI, cs.HC
Previous research on expert advice-taking shows that humans exhibit two contradictory behaviors: on the one hand, people tend to overvalue their own opinions undervaluing the expert opinion, and on the other, people often defer to other people's advice even if the advice itself is rather obviously wrong. In our study, ...
http://arxiv.org/pdf/2310.14358v1
2023-10-22T16:59:31
2023-10-22T16:59:31
{ "title": [ 0.0073242188, -0.013145447, -0.031433105, -0.052886963, 0.015182495, -0.013313293, 0.0045318604, -0.010604858, -0.044433594, 0.020614624, -0.03677368, 0.034179688, -0.0030612946, -0.019317627, -0.018463135, -0.013214111, -0.025436401, ...
http://arxiv.org/abs/2310.14356v1
Cultural and Linguistic Diversity Improves Visual Representations
Andre Ye, Sebastin Santy, Jena D. Hwang, Amy X. Zhang, Ranjay Krishna
cs.CV, cs.CL, cs.CY, cs.HC
Computer vision often treats perception as objective, and this assumption gets reflected in the way that datasets are collected and models are trained. For instance, image descriptions in different languages are typically assumed to be translations of the same semantic content. However, work in cross-cultural psycholog...
http://arxiv.org/pdf/2310.14356v1
2023-10-22T16:51:42
2023-10-22T16:51:42
{ "title": [ 0.048034668, 0.031799316, -0.029785156, -0.04071045, 0.00067329407, -0.035217285, -0.029800415, -0.03881836, -0.016098022, 0.03894043, -0.00843811, -0.004840851, 0.0058670044, -0.047576904, 0.009094238, 0.0042533875, 0.039245605, 0.0...
http://arxiv.org/abs/2310.14346v1
The Law and NLP: Bridging Disciplinary Disconnects
Robert Mahari, Dominik Stammbach, Elliott Ash, Alex 'Sandy' Pentland
cs.CL
Legal practice is intrinsically rooted in the fabric of language, yet legal practitioners and scholars have been slow to adopt tools from natural language processing (NLP). At the same time, the legal system is experiencing an access to justice crisis, which could be partially alleviated with NLP. In this position pape...
http://arxiv.org/pdf/2310.14346v1
2023-10-22T16:34:31
2023-10-22T16:34:31
{ "title": [ -0.015014648, -0.010093689, -0.04043579, -0.0357666, -0.013450623, -0.018127441, -0.025604248, -0.008468628, -0.01675415, 0.0053253174, -0.057769775, 0.005115509, -0.018676758, -0.027923584, -0.01461792, -0.031982422, -0.033233643, 0...
http://arxiv.org/abs/2310.14340v1
Social Commonsense-Guided Search Query Generation for Open-Domain Knowledge-Powered Conversations
Revanth Gangi Reddy, Hao Bai, Wentao Yao, Sharath Chandra Etagi Suresh, Heng Ji, ChengXiang Zhai
cs.CL
Open-domain dialog involves generating search queries that help obtain relevant knowledge for holding informative conversations. However, it can be challenging to determine what information to retrieve when the user is passive and does not express a clear need or request. To tackle this issue, we present a novel approa...
http://arxiv.org/pdf/2310.14340v1
2023-10-22T16:14:56
2023-10-22T16:14:56
{ "title": [ 0.008834839, -0.011978149, -0.024978638, -0.0635376, -0.0046539307, 0.0020656586, -0.025894165, 0.0017375946, 0.017242432, 0.030258179, -0.014892578, 0.018600464, -0.0023269653, -0.012390137, -0.04208374, -0.0012578964, -0.0115356445, ...
http://arxiv.org/abs/2310.14338v2
From Chaos to Clarity: Claim Normalization to Empower Fact-Checking
Megha Sundriyal, Tanmoy Chakraborty, Preslav Nakov
cs.CL, cs.AI
With the rise of social media, users are exposed to many misleading claims. However, the pervasive noise inherent in these posts presents a challenge in identifying precise and prominent claims that require verification. Extracting the important claims from such posts is arduous and time-consuming, yet it is an underex...
http://arxiv.org/pdf/2310.14338v2
2023-11-13T13:23:44
2023-10-22T16:07:06
{ "title": [ 0.021209717, 0.011688232, -0.04660034, -0.08392334, 0.028442383, -0.013771057, -0.018859863, 0.00178051, -0.05154419, 0.028533936, -0.017181396, 0.04397583, 0.01751709, -0.022476196, -0.013214111, 0.00044322014, -0.010398865, 0.04077...
http://arxiv.org/abs/2311.02084v1
ITEm: Unsupervised Image-Text Embedding Learning for eCommerce
Baohao Liao, Michael Kozielski, Sanjika Hewavitharana, Jiangbo Yuan, Shahram Khadivi, Tomer Lancewicki
cs.CV, cs.CL, cs.IR
Product embedding serves as a cornerstone for a wide range of applications in eCommerce. The product embedding learned from multiple modalities shows significant improvement over that from a single modality, since different modalities provide complementary information. However, some modalities are more informatively do...
http://arxiv.org/pdf/2311.02084v1
2023-10-22T15:39:44
2023-10-22T15:39:44
{ "title": [ -0.0025291443, 0.00182724, -0.029937744, -0.0044937134, -0.01235199, -0.031585693, -0.005218506, -0.0309906, -0.0011177063, 0.03878784, -0.058776855, 0.012420654, 0.023147583, -0.010597229, -0.016189575, -0.016204834, 0.01701355, 0.0...
http://arxiv.org/abs/2310.14329v1
DiFair: A Benchmark for Disentangled Assessment of Gender Knowledge and Bias
Mahdi Zakizadeh, Kaveh Eskandari Miandoab, Mohammad Taher Pilehvar
cs.CL
Numerous debiasing techniques have been proposed to mitigate the gender bias that is prevalent in pretrained language models. These are often evaluated on datasets that check the extent to which the model is gender-neutral in its predictions. Importantly, this evaluation protocol overlooks the possible adverse impact o...
http://arxiv.org/pdf/2310.14329v1
2023-10-22T15:27:16
2023-10-22T15:27:16
{ "title": [ -0.031433105, 0.00944519, -0.08605957, -0.05130005, 0.027038574, -0.032989502, -0.018218994, -0.0032024384, 0.031555176, 0.014518738, -0.039367676, -0.0074043274, 0.008644104, -0.0154418945, -0.043640137, -0.010475159, -0.012504578, ...
http://arxiv.org/abs/2310.14326v1
CLMSM: A Multi-Task Learning Framework for Pre-training on Procedural Text
Abhilash Nandy, Manav Nitin Kapadnis, Pawan Goyal, Niloy Ganguly
cs.CL, cs.AI
In this paper, we propose CLMSM, a domain-specific, continual pre-training framework, that learns from a large set of procedural recipes. CLMSM uses a Multi-Task Learning Framework to optimize two objectives - a) Contrastive Learning using hard triplets to learn fine-grained differences across entities in the procedure...
http://arxiv.org/pdf/2310.14326v1
2023-10-22T15:20:11
2023-10-22T15:20:11
{ "title": [ -0.026870728, -0.011352539, -0.043884277, -0.066589355, -0.025466919, -0.0033187866, -0.07330322, 0.00005966425, -0.019515991, 0.041229248, 0.0010299683, 0.030029297, -0.014633179, -0.013938904, -0.024124146, -0.0062789917, -0.00737381, ...
http://arxiv.org/abs/2310.14325v1
Towards Harmful Erotic Content Detection through Coreference-Driven Contextual Analysis
Inez Okulska, Emilia Wiśnios
cs.CL
Adult content detection still poses a great challenge for automation. Existing classifiers primarily focus on distinguishing between erotic and non-erotic texts. However, they often need more nuance in assessing the potential harm. Unfortunately, the content of this nature falls beyond the reach of generative models du...
http://arxiv.org/pdf/2310.14325v1
2023-10-22T15:19:04
2023-10-22T15:19:04
{ "title": [ 0.045318604, 0.029724121, -0.036712646, -0.033691406, 0.06427002, -0.037231445, 0.034118652, -0.029083252, 0.017089844, 0.056427002, -0.005672455, 0.015274048, 0.040863037, -0.012763977, -0.045959473, 0.029891968, -0.0118637085, 0.05...
http://arxiv.org/abs/2310.18345v1
A Survey on Semantic Processing Techniques
Rui Mao, Kai He, Xulang Zhang, Guanyi Chen, Jinjie Ni, Zonglin Yang, Erik Cambria
cs.CL, cs.AI
Semantic processing is a fundamental research domain in computational linguistics. In the era of powerful pre-trained language models and large language models, the advancement of research in this domain appears to be decelerating. However, the study of semantics is multi-dimensional in linguistics. The research depth ...
http://arxiv.org/pdf/2310.18345v1
2023-10-22T15:09:51
2023-10-22T15:09:51
{ "title": [ -0.014144897, -0.0056877136, -0.038604736, -0.06365967, 0.028244019, 0.00674057, -0.033233643, -0.003944397, 0.012901306, 0.035583496, -0.01322937, -0.005722046, 0.00067424774, -0.008476257, -0.055633545, 0.019073486, -0.009109497, 0...
http://arxiv.org/abs/2310.18344v1
Chainpoll: A high efficacy method for LLM hallucination detection
Robert Friel, Atindriyo Sanyal
cs.CL, cs.AI, cs.LG
Large language models (LLMs) have experienced notable advancements in generating coherent and contextually relevant responses. However, hallucinations - incorrect or unfounded claims - are still prevalent, prompting the creation of automated metrics to detect these in LLM outputs. Our contributions include: introducing...
http://arxiv.org/pdf/2310.18344v1
2023-10-22T14:45:14
2023-10-22T14:45:14
{ "title": [ -0.014160156, 0.02619934, -0.04324341, -0.028244019, -0.0101623535, -0.014526367, -0.037597656, -0.035125732, -0.012123108, 0.028945923, -0.003396988, 0.024993896, -0.0042266846, -0.043426514, -0.01802063, 0.013877869, -0.026168823, ...
http://arxiv.org/abs/2310.14319v1
4 and 7-bit Labeling for Projective and Non-Projective Dependency Trees
Carlos Gómez-Rodríguez, Diego Roca, David Vilares
cs.CL, cs.FL, 68T50, I.2.7
We introduce an encoding for parsing as sequence labeling that can represent any projective dependency tree as a sequence of 4-bit labels, one per word. The bits in each word's label represent (1) whether it is a right or left dependent, (2) whether it is the outermost (left/right) dependent of its parent, (3) whether ...
http://arxiv.org/pdf/2310.14319v1
2023-10-22T14:43:53
2023-10-22T14:43:53
{ "title": [ -0.044677734, -0.01210022, -0.052368164, -0.08117676, -0.0099487305, 0.022262573, -0.07397461, 0.027008057, -0.0027446747, 0.0033798218, -0.035888672, 0.008323669, -0.02368164, -0.023895264, -0.013549805, -0.011787415, -0.0071487427, ...
http://arxiv.org/abs/2310.14312v1
Neural Text Sanitization with Privacy Risk Indicators: An Empirical Analysis
Anthi Papadopoulou, Pierre Lison, Mark Anderson, Lilja Øvrelid, Ildikó Pilán
cs.CL
Text sanitization is the task of redacting a document to mask all occurrences of (direct or indirect) personal identifiers, with the goal of concealing the identity of the individual(s) referred in it. In this paper, we consider a two-step approach to text sanitization and provide a detailed analysis of its empirical p...
http://arxiv.org/pdf/2310.14312v1
2023-10-22T14:17:27
2023-10-22T14:17:27
{ "title": [ 0.007068634, 0.02986145, -0.03543091, -0.045318604, 0.03555298, -0.02180481, -0.016708374, -0.01675415, -0.030593872, 0.042053223, -0.011993408, 0.023422241, 0.015235901, -0.008850098, -0.027313232, -0.0063323975, -0.020599365, 0.032...
http://arxiv.org/abs/2310.14303v2
Language Model Unalignment: Parametric Red-Teaming to Expose Hidden Harms and Biases
Rishabh Bhardwaj, Soujanya Poria
cs.CL
Red-teaming has been a widely adopted way to evaluate the harmfulness of Large Language Models (LLMs). It aims to jailbreak a model's safety behavior to make it act as a helpful agent disregarding the harmfulness of the query. Existing methods are primarily based on input text-based red-teaming such as adversarial prom...
http://arxiv.org/pdf/2310.14303v2
2023-11-13T05:28:47
2023-10-22T13:55:46
{ "title": [ 0.00497818, 0.030044556, -0.013595581, -0.03010559, -0.03540039, -0.005718231, -0.046325684, -0.048614502, -0.04248047, 0.023956299, -0.010749817, -0.0055236816, 0.032440186, -0.05545044, 0.015960693, -0.03201294, -0.010978699, 0.027...
http://arxiv.org/abs/2310.14282v1
NERetrieve: Dataset for Next Generation Named Entity Recognition and Retrieval
Uri Katz, Matan Vetzler, Amir DN Cohen, Yoav Goldberg
cs.CL, cs.AI, cs.IR
Recognizing entities in texts is a central need in many information-seeking scenarios, and indeed, Named Entity Recognition (NER) is arguably one of the most successful examples of a widely adopted NLP task and corresponding NLP technology. Recent advances in large language models (LLMs) appear to provide effective sol...
http://arxiv.org/pdf/2310.14282v1
2023-10-22T12:23:00
2023-10-22T12:23:00
{ "title": [ -0.009796143, -0.009208679, -0.050720215, -0.07922363, 0.027542114, 0.0037517548, -0.026260376, 0.006336212, 0.007106781, 0.02456665, -0.032043457, 0.020309448, 0.009803772, 0.004940033, -0.036743164, 0.008369446, -0.011253357, 0.044...
http://arxiv.org/abs/2310.14278v1
Conversational Speech Recognition by Learning Audio-textual Cross-modal Contextual Representation
Kun Wei, Bei Li, Hang Lv, Quan Lu, Ning Jiang, Lei Xie
cs.SD, cs.CL, eess.AS
Automatic Speech Recognition (ASR) in conversational settings presents unique challenges, including extracting relevant contextual information from previous conversational turns. Due to irrelevant content, error propagation, and redundancy, existing methods struggle to extract longer and more effective contexts. To add...
http://arxiv.org/pdf/2310.14278v1
2023-10-22T11:57:33
2023-10-22T11:57:33
{ "title": [ -0.010498047, 0.021057129, -0.054718018, -0.061920166, 0.012451172, -0.0034980774, -0.052581787, 0.0031337738, 0.0044174194, 0.046020508, -0.038238525, 0.013427734, 0.02229309, -0.0050811768, -0.03640747, -0.0015106201, -0.013969421, ...
http://arxiv.org/abs/2310.14265v2
CT-GAT: Cross-Task Generative Adversarial Attack based on Transferability
Minxuan Lv, Chengwei Dai, Kun Li, Wei Zhou, Songlin Hu
cs.CL
Neural network models are vulnerable to adversarial examples, and adversarial transferability further increases the risk of adversarial attacks. Current methods based on transferability often rely on substitute models, which can be impractical and costly in real-world scenarios due to the unavailability of training dat...
http://arxiv.org/pdf/2310.14265v2
2023-11-05T14:07:04
2023-10-22T11:00:04
{ "title": [ -0.009262085, -0.0041007996, -0.023880005, -0.033355713, 0.016296387, -0.013198853, 0.0013561249, -0.026275635, -0.02633667, 0.023605347, 0.008407593, 0.032104492, -0.020446777, -0.020462036, -0.018981934, 0.008033752, -0.009170532, ...
http://arxiv.org/abs/2310.14262v1
Boosting Unsupervised Machine Translation with Pseudo-Parallel Data
Ivana Kvapilíková, Ondřej Bojar
cs.CL
Even with the latest developments in deep learning and large-scale language modeling, the task of machine translation (MT) of low-resource languages remains a challenge. Neural MT systems can be trained in an unsupervised way without any translation resources but the quality lags behind, especially in truly low-resourc...
http://arxiv.org/pdf/2310.14262v1
2023-10-22T10:57:12
2023-10-22T10:57:12
{ "title": [ -0.04559326, 0.012748718, -0.031707764, -0.060058594, 0.0032920837, 0.016159058, -0.06274414, -0.0074310303, -0.015274048, 0.050689697, -0.0011053085, 0.014122009, 0.020050049, -0.018615723, -0.036621094, 0.01260376, -0.0079956055, 0...
http://arxiv.org/abs/2310.14261v1
RSM-NLP at BLP-2023 Task 2: Bangla Sentiment Analysis using Weighted and Majority Voted Fine-Tuned Transformers
Pratinav Seth, Rashi Goel, Komal Mathur, Swetha Vemulapalli
cs.CL, cs.AI, cs.LG
This paper describes our approach to submissions made at Shared Task 2 at BLP Workshop - Sentiment Analysis of Bangla Social Media Posts. Sentiment Analysis is an action research area in the digital age. With the rapid and constant growth of online social media sites and services and the increasing amount of textual da...
http://arxiv.org/pdf/2310.14261v1
2023-10-22T10:55:56
2023-10-22T10:55:56
{ "title": [ 0.012588501, -0.022872925, 0.0001193285, -0.028381348, -0.016052246, 0.006702423, -0.007926941, -0.008895874, 0.010429382, 0.028930664, -0.004951477, -0.019210815, -0.010009766, -0.029800415, -0.047821045, -0.020492554, -0.020202637, ...
http://arxiv.org/abs/2310.14248v1
From Static to Dynamic: A Continual Learning Framework for Large Language Models
Mingzhe Du, Anh Tuan Luu, Bin Ji, See-kiong Ng
cs.CL
The vast number of parameters in large language models (LLMs) endows them with remarkable capabilities, allowing them to excel in a variety of natural language processing tasks. However, this complexity also presents challenges, making LLMs difficult to train and inhibiting their ability to continuously assimilate new ...
http://arxiv.org/pdf/2310.14248v1
2023-10-22T10:18:53
2023-10-22T10:18:53
{ "title": [ -0.015808105, -0.0068855286, -0.039916992, -0.09240723, 0.016967773, 0.0050582886, -0.06555176, 0.013961792, 0.015625, 0.03363037, -0.011619568, 0.0061912537, -0.0052986145, -0.013809204, -0.054138184, 0.019622803, -0.004760742, 0.04...
http://arxiv.org/abs/2310.18343v2
PHD: Pixel-Based Language Modeling of Historical Documents
Nadav Borenstein, Phillip Rust, Desmond Elliott, Isabelle Augenstein
cs.CL
The digitisation of historical documents has provided historians with unprecedented research opportunities. Yet, the conventional approach to analysing historical documents involves converting them from images to text using OCR, a process that overlooks the potential benefits of treating them as images and introduces h...
http://arxiv.org/pdf/2310.18343v2
2023-11-04T11:30:01
2023-10-22T08:45:48
{ "title": [ -0.03274536, 0.004322052, -0.03390503, 0.0031147003, -0.04498291, -0.019210815, -0.029541016, -0.0001322031, 0.036590576, 0.040008545, -0.034118652, -0.01361084, 0.05999756, -0.029205322, -0.027435303, -0.002729416, 0.0009045601, 0.0...
http://arxiv.org/abs/2310.18342v1
MIRACLE: Towards Personalized Dialogue Generation with Latent-Space Multiple Personal Attribute Control
Zhenyi Lu, Wei Wei, Xiaoye Qu, XianLing Mao, Dangyang Chen, Jixiong Chen
cs.CL, cs.AI
Personalized dialogue systems aim to endow the chatbot agent with more anthropomorphic traits for human-like interactions. Previous approaches have explored explicitly user profile modeling using text descriptions, implicit derivation of user embeddings, or utilizing handicraft prompts for ChatGPT-like models. However,...
http://arxiv.org/pdf/2310.18342v1
2023-10-22T08:44:26
2023-10-22T08:44:26
{ "title": [ -0.00944519, -0.011581421, -0.051757812, -0.064941406, 0.002248764, 0.00093507767, -0.0340271, 0.013664246, -0.01184082, 0.036102295, 0.023422241, 0.022399902, -0.009681702, -0.031799316, -0.036956787, 0.011192322, -0.01966858, 0.030...
http://arxiv.org/abs/2310.14225v1
Customising General Large Language Models for Specialised Emotion Recognition Tasks
Liyizhe Peng, Zixing Zhang, Tao Pang, Jing Han, Huan Zhao, Hao Chen, Björn W. Schuller
cs.CL
The advent of large language models (LLMs) has gained tremendous attention over the past year. Previous studies have shown the astonishing performance of LLMs not only in other tasks but also in emotion recognition in terms of accuracy, universality, explanation, robustness, few/zero-shot learning, and others. Leveragi...
http://arxiv.org/pdf/2310.14225v1
2023-10-22T08:09:13
2023-10-22T08:09:13
{ "title": [ -0.021102905, -0.010002136, -0.031829834, -0.058685303, -0.015991211, -0.0033168793, -0.04510498, 0.014297485, 0.025314331, 0.033355713, -0.02130127, -0.031463623, 0.030136108, -0.026428223, -0.05783081, 0.008934021, -0.0010128021, 0...
http://arxiv.org/abs/2310.14211v1
LUNA: A Model-Based Universal Analysis Framework for Large Language Models
Da Song, Xuan Xie, Jiayang Song, Derui Zhu, Yuheng Huang, Felix Juefei-Xu, Lei Ma
cs.LG, cs.AI, cs.CL, cs.CR, cs.SE
Over the past decade, Artificial Intelligence (AI) has had great success recently and is being used in a wide range of academic and industrial fields. More recently, LLMs have made rapid advancements that have propelled AI to a new level, enabling even more diverse applications and industrial domains with intelligence,...
http://arxiv.org/pdf/2310.14211v1
2023-10-22T07:26:21
2023-10-22T07:26:21
{ "title": [ -0.02330017, -0.012557983, -0.048583984, -0.09625244, 0.011390686, 0.010864258, -0.09051514, 0.01260376, 0.023895264, 0.0435791, -0.000113248825, 0.009422302, 0.034240723, -0.015396118, -0.04748535, 0.037872314, -0.017089844, 0.04214...
http://arxiv.org/abs/2310.14206v1
Manifold-Preserving Transformers are Effective for Short-Long Range Encoding
Ayan Sengupta, Md Shad Akhtar, Tanmoy Chakraborty
cs.CL, cs.LG
Multi-head self-attention-based Transformers have shown promise in different learning tasks. Albeit these models exhibit significant improvement in understanding short-term and long-term contexts from sequences, encoders of Transformers and their variants fail to preserve layer-wise contextual information. Transformers...
http://arxiv.org/pdf/2310.14206v1
2023-10-22T06:58:28
2023-10-22T06:58:28
{ "title": [ 0.03704834, -0.006454468, 0.029724121, -0.029434204, -0.035186768, -0.011894226, 0.0010375977, -0.034362793, -0.03729248, 0.06439209, -0.0011863708, -0.014175415, 0.06225586, -0.041534424, 0.00541687, 0.009750366, 0.020614624, 0.0144...
http://arxiv.org/abs/2310.14198v1
QA-NatVer: Question Answering for Natural Logic-based Fact Verification
Rami Aly, Marek Strong, Andreas Vlachos
cs.CL
Fact verification systems assess a claim's veracity based on evidence. An important consideration in designing them is faithfulness, i.e. generating explanations that accurately reflect the reasoning of the model. Recent works have focused on natural logic, which operates directly on natural language by capturing the s...
http://arxiv.org/pdf/2310.14198v1
2023-10-22T06:27:31
2023-10-22T06:27:31
{ "title": [ -0.022094727, -0.018051147, -0.04711914, -0.08203125, 0.007434845, 0.012199402, -0.04397583, 0.017944336, -0.0012636185, 0.028945923, -0.013015747, 0.012565613, -0.0033454895, -0.006855011, -0.05126953, 0.00053167343, -0.0345459, 0.0...
http://arxiv.org/abs/2310.18341v2
CXR-LLaVA: Multimodal Large Language Model for Interpreting Chest X-ray Images
Seowoo Lee, Jiwon Youn, Mansu Kim, Soon Ho Yoon
cs.CL, cs.AI
Purpose: Recent advancements in large language models (LLMs) have expanded their capabilities in a multimodal fashion, potentially replicating the image interpretation of human radiologists. This study aimed to develop open-source multimodal large language model for interpreting chest X-ray images (CXR-LLaVA). We also ...
http://arxiv.org/pdf/2310.18341v2
2023-10-31T21:08:19
2023-10-22T06:22:37
{ "title": [ -0.04888916, 0.00074243546, -0.049865723, -0.037475586, -0.011199951, -0.0017232895, -0.06323242, -0.015411377, 0.0018911362, 0.034332275, -0.0184021, -0.005970001, 0.011817932, -0.023376465, -0.03829956, 0.0044441223, 0.005176544, 0...
http://arxiv.org/abs/2310.14192v1
PromptMix: A Class Boundary Augmentation Method for Large Language Model Distillation
Gaurav Sahu, Olga Vechtomova, Dzmitry Bahdanau, Issam H. Laradji
cs.CL, cs.AI
Data augmentation is a widely used technique to address the problem of text classification when there is a limited amount of training data. Recent work often tackles this problem using large language models (LLMs) like GPT3 that can generate new examples given already available ones. In this work, we propose a method t...
http://arxiv.org/pdf/2310.14192v1
2023-10-22T05:43:23
2023-10-22T05:43:23
{ "title": [ -0.0020866394, -0.007598877, -0.05718994, -0.09893799, 0.0154800415, 0.005935669, -0.0871582, 0.00995636, 0.017959595, 0.01802063, -0.026855469, 0.0021743774, 0.016601562, -0.018875122, -0.026763916, 0.0017280579, -0.021942139, 0.036...
http://arxiv.org/abs/2310.14174v1
An In-Context Schema Understanding Method for Knowledge Base Question Answering
Yantao Liu, Zixuan Li, Xiaolong Jin, Long Bai, Saiping Guan, Jiafeng Guo, Xueqi Cheng
cs.CL
The Knowledge Base Question Answering (KBQA) task aims to answer natural language questions based on a given knowledge base. As a kind of common method for this task, semantic parsing-based ones first convert natural language questions to logical forms (e.g., SPARQL queries) and then execute them on knowledge bases to ...
http://arxiv.org/pdf/2310.14174v1
2023-10-22T04:19:17
2023-10-22T04:19:17
{ "title": [ -0.015113831, -0.007045746, -0.035339355, -0.07470703, 0.00843811, 0.00573349, -0.02909851, 0.019500732, 0.040252686, 0.014335632, -0.0413208, -0.027679443, 0.011367798, -0.014465332, -0.04714966, -0.010955811, 0.0013217926, 0.009902...
http://arxiv.org/abs/2310.14159v2
Can Language Models Laugh at YouTube Short-form Videos?
Dayoon Ko, Sangho Lee, Gunhee Kim
cs.CL, cs.CV
As short-form funny videos on social networks are gaining popularity, it becomes demanding for AI models to understand them for better communication with humans. Unfortunately, previous video humor datasets target specific domains, such as speeches or sitcoms, and mostly focus on verbal cues. We curate a user-generated...
http://arxiv.org/pdf/2310.14159v2
2023-10-26T09:27:13
2023-10-22T03:01:38
{ "title": [ -0.00040888786, 0.012275696, -0.03475952, -0.014411926, -0.02645874, -0.008201599, -0.035888672, -0.0079193115, 0.016403198, 0.024932861, -0.03717041, -0.023254395, 0.00919342, -0.023712158, -0.028274536, -0.010063171, -0.010986328, ...
http://arxiv.org/abs/2310.18340v1
When Urban Region Profiling Meets Large Language Models
Yibo Yan, Haomin Wen, Siru Zhong, Wei Chen, Haodong Chen, Qingsong Wen, Roger Zimmermann, Yuxuan Liang
cs.CL, cs.AI
Urban region profiling from web-sourced data is of utmost importance for urban planning and sustainable development. We are witnessing a rising trend of LLMs for various fields, especially dealing with multi-modal data research such as vision-language learning, where the text modality serves as a supplement information...
http://arxiv.org/pdf/2310.18340v1
2023-10-22T02:32:53
2023-10-22T02:32:53
{ "title": [ -0.0044898987, 0.012756348, -0.03326416, -0.05517578, 0.003578186, -0.0129776, -0.050079346, -0.018859863, -0.022659302, 0.03765869, -0.026992798, -0.0064926147, -0.027542114, -0.022583008, -0.043029785, -0.013244629, 0.0020523071, 0...
http://arxiv.org/abs/2310.14152v1
Orthogonal Subspace Learning for Language Model Continual Learning
Xiao Wang, Tianze Chen, Qiming Ge, Han Xia, Rong Bao, Rui Zheng, Qi Zhang, Tao Gui, Xuanjing Huang
cs.CL, cs.LG
Benefiting from massive corpora and advanced hardware, large language models (LLMs) exhibit remarkable capabilities in language understanding and generation. However, their performance degrades in scenarios where multiple tasks are encountered sequentially, also known as catastrophic forgetting. In this paper, we propo...
http://arxiv.org/pdf/2310.14152v1
2023-10-22T02:23:44
2023-10-22T02:23:44
{ "title": [ -0.01876831, 0.010314941, -0.04736328, -0.028457642, -0.0028266907, -0.02104187, -0.03414917, -0.006439209, -0.013916016, 0.03793335, -0.03164673, 0.0016736984, 0.013633728, -0.029190063, -0.025665283, -0.0028915405, -0.00051784515, ...
http://arxiv.org/abs/2310.14151v1
PromptCBLUE: A Chinese Prompt Tuning Benchmark for the Medical Domain
Wei Zhu, Xiaoling Wang, Huanran Zheng, Mosha Chen, Buzhou Tang
cs.CL
Biomedical language understanding benchmarks are the driving forces for artificial intelligence applications with large language model (LLM) back-ends. However, most current benchmarks: (a) are limited to English which makes it challenging to replicate many of the successes in English for other languages, or (b) focus ...
http://arxiv.org/pdf/2310.14151v1
2023-10-22T02:20:38
2023-10-22T02:20:38
{ "title": [ -0.03152466, -0.023162842, -0.016921997, -0.050689697, -0.01361084, 0.004711151, -0.049591064, -0.00982666, -0.0037708282, 0.037475586, 0.01260376, 0.00021982193, 0.0061302185, -0.041748047, -0.03878784, 0.003967285, -0.0022735596, 0...
http://arxiv.org/abs/2310.14126v1
Ask To The Point: Open-Domain Entity-Centric Question Generation
Yuxiang Liu, Jie Huang, Kevin Chen-Chuan Chang
cs.CL, cs.AI
We introduce a new task called *entity-centric question generation* (ECQG), motivated by real-world applications such as topic-specific learning, assisted reading, and fact-checking. The task aims to generate questions from an entity perspective. To solve ECQG, we propose a coherent PLM-based framework GenCONE with two...
http://arxiv.org/pdf/2310.14126v1
2023-10-21T22:19:19
2023-10-21T22:19:19
{ "title": [ -0.006816864, 0.0018434525, -0.029678345, -0.048187256, 0.011528015, 0.004714966, -0.007686615, -0.005268097, 0.00077056885, 0.023757935, -0.014289856, 0.033691406, -0.0058937073, -0.013389587, -0.025375366, 0.0115356445, -0.011909485, ...
http://arxiv.org/abs/2310.14124v1
Structural generalization in COGS: Supertagging is (almost) all you need
Alban Petit, Caio Corro, François Yvon
cs.CL
In many Natural Language Processing applications, neural networks have been found to fail to generalize on out-of-distribution examples. In particular, several recent semantic parsing datasets have put forward important limitations of neural networks in cases where compositional generalization is required. In this work...
http://arxiv.org/pdf/2310.14124v1
2023-10-21T21:51:25
2023-10-21T21:51:25
{ "title": [ -0.0016222, -0.0017309189, -0.023986816, -0.09051514, 0.0075569153, 0.0064582825, -0.04336548, -0.014533997, -0.025802612, 0.02456665, -0.040649414, 0.016326904, -0.0002515316, -0.025299072, -0.011070251, -0.020004272, 0.015365601, 0...
http://arxiv.org/abs/2310.14120v1
Sentiment Analysis Across Multiple African Languages: A Current Benchmark
Saurav K. Aryal, Howard Prioleau, Surakshya Aryal
cs.CL, cs.AI
Sentiment analysis is a fundamental and valuable task in NLP. However, due to limitations in data and technological availability, research into sentiment analysis of African languages has been fragmented and lacking. With the recent release of the AfriSenti-SemEval Shared Task 12, hosted as a part of The 17th Internati...
http://arxiv.org/pdf/2310.14120v1
2023-10-21T21:38:06
2023-10-21T21:38:06
{ "title": [ -0.019012451, -0.02268982, 0.003780365, -0.035369873, -0.018814087, 0.012626648, -0.055725098, -0.015174866, 0.014884949, 0.046081543, -0.0052986145, 0.0030498505, 0.021942139, -0.012023926, -0.05908203, 0.024002075, -0.00095796585, ...
http://arxiv.org/abs/2310.14110v1
Finite-context Indexing of Restricted Output Space for NLP Models Facing Noisy Input
Minh Nguyen, Nancy F. Chen
cs.CL
NLP models excel on tasks with clean inputs, but are less accurate with noisy inputs. In particular, character-level noise such as human-written typos and adversarially-engineered realistic-looking misspellings often appears in text and can easily trip up NLP models. Prior solutions to address character-level noise oft...
http://arxiv.org/pdf/2310.14110v1
2023-10-21T20:28:26
2023-10-21T20:28:26
{ "title": [ -0.008094788, 0.004875183, -0.012298584, -0.033233643, -0.015975952, -0.0039634705, -0.01876831, -0.02519226, 0.019821167, 0.015731812, -0.085632324, 0.019699097, -0.010955811, 0.0021400452, -0.038970947, -0.0236969, -0.009155273, 0....
http://arxiv.org/abs/2310.14107v1
On the Transferability of Visually Grounded PCFGs
Yanpeng Zhao, Ivan Titov
cs.CL, cs.AI
There has been a significant surge of interest in visually grounded grammar induction in recent times. While a variety of models have been developed for the task and have demonstrated impressive performance, they have not been evaluated on text domains that are different from the training domain, so it is unclear if th...
http://arxiv.org/pdf/2310.14107v1
2023-10-21T20:19:51
2023-10-21T20:19:51
{ "title": [ 0.031585693, -0.0077552795, -0.00242424, -0.05291748, -0.010223389, 0.004333496, -0.024612427, -0.020080566, 0.027359009, 0.047668457, -0.018157959, -0.0045928955, 0.01991272, -0.016708374, -0.07019043, -0.0016813278, 0.013473511, 0....
http://arxiv.org/abs/2310.14103v1
Revisiting Instruction Fine-tuned Model Evaluation to Guide Industrial Applications
Manuel Faysse, Gautier Viaud, Céline Hudelot, Pierre Colombo
cs.LG, cs.AI, cs.CL
Instruction Fine-Tuning (IFT) is a powerful paradigm that strengthens the zero-shot capabilities of Large Language Models (LLMs), but in doing so induces new evaluation metric requirements. We show LLM-based metrics to be well adapted to these requirements, and leverage them to conduct an investigation of task-speciali...
http://arxiv.org/pdf/2310.14103v1
2023-10-21T20:04:55
2023-10-21T20:04:55
{ "title": [ -0.035095215, -0.020126343, -0.014877319, -0.036499023, -0.022033691, -0.008018494, -0.036743164, 0.008285522, -0.018829346, 0.026428223, -0.037231445, 0.007949829, -0.022842407, -0.0032291412, -0.03515625, -0.010040283, -0.0093688965, ...
http://arxiv.org/abs/2310.14093v1
Leveraging Knowledge Graphs for Orphan Entity Allocation in Resume Processing
Aagam Bakliwal, Shubham Manish Gandhi, Yashodhara Haribhakta
cs.CL
Significant challenges are posed in talent acquisition and recruitment by processing and analyzing unstructured data, particularly resumes. This research presents a novel approach for orphan entity allocation in resume processing using knowledge graphs. Techniques of association mining, concept extraction, external kno...
http://arxiv.org/pdf/2310.14093v1
2023-10-21T19:10:30
2023-10-21T19:10:30
{ "title": [ 0.00806427, -0.014541626, -0.04067993, -0.074279785, -0.011810303, -0.0017004013, -0.037353516, 0.011993408, 0.006587982, 0.04623413, -0.047027588, -0.001698494, 0.049041748, -0.0017852783, -0.02418518, -0.014175415, -0.014343262, 0....
http://arxiv.org/abs/2310.14088v3
MedEval: A Multi-Level, Multi-Task, and Multi-Domain Medical Benchmark for Language Model Evaluation
Zexue He, Yu Wang, An Yan, Yao Liu, Eric Y. Chang, Amilcare Gentili, Julian McAuley, Chun-Nan Hsu
cs.CL
Curated datasets for healthcare are often limited due to the need of human annotations from experts. In this paper, we present MedEval, a multi-level, multi-task, and multi-domain medical benchmark to facilitate the development of language models for healthcare. MedEval is comprehensive and consists of data from severa...
http://arxiv.org/pdf/2310.14088v3
2023-11-14T21:59:56
2023-10-21T18:59:41
{ "title": [ -0.036193848, -0.005279541, -0.03543091, -0.059509277, 0.0014629364, 0.0068092346, -0.07098389, -0.019302368, -0.010787964, 0.038604736, 0.029083252, 0.006164551, 0.012969971, -0.040802002, -0.042297363, 0.021469116, -0.011627197, 0....
http://arxiv.org/abs/2310.18339v1
MOELoRA: An MOE-based Parameter Efficient Fine-Tuning Method for Multi-task Medical Applications
Qidong Liu, Xian Wu, Xiangyu Zhao, Yuanshao Zhu, Derong Xu, Feng Tian, Yefeng Zheng
cs.CL, cs.AI
The recent surge in the field of Large Language Models (LLMs) has gained significant attention in numerous domains. In order to tailor an LLM to a specific domain such as a web-based healthcare system, fine-tuning with domain knowledge is necessary. However, two issues arise during fine-tuning LLMs for medical applicat...
http://arxiv.org/pdf/2310.18339v1
2023-10-21T17:18:09
2023-10-21T17:18:09
{ "title": [ -0.030822754, -0.01335907, -0.02293396, -0.07232666, 0.009468079, 0.016845703, -0.037139893, 0.0068511963, -0.016662598, 0.047058105, 0.030960083, -0.008880615, 0.019622803, -0.035095215, -0.045532227, 0.004085541, -0.009063721, 0.01...
http://arxiv.org/abs/2310.14053v1
Beyond Accuracy: Evaluating Self-Consistency of Code Large Language Models with IdentityChain
Marcus J. Min, Yangruibo Ding, Luca Buratti, Saurabh Pujar, Gail Kaiser, Suman Jana, Baishakhi Ray
cs.LG, cs.CL, cs.SE, 68, I.2; D.2
Code Large Language Models (Code LLMs) are being increasingly employed in real-life applications, so evaluating them is critical. While the general accuracy of Code LLMs on individual tasks has been extensively evaluated, their self-consistency across different tasks is overlooked. Intuitively, a trustworthy model shou...
http://arxiv.org/pdf/2310.14053v1
2023-10-21T16:14:56
2023-10-21T16:14:56
{ "title": [ -0.053253174, -0.0052604675, -0.06976318, -0.0692749, 0.013168335, -0.0018177032, -0.052093506, 0.039611816, -0.018005371, 0.054718018, -0.0090408325, 0.0020141602, 0.012390137, -0.013549805, -0.052368164, -0.000044882298, -0.016647339, ...
http://arxiv.org/abs/2310.14050v1
Code-Switching with Word Senses for Pretraining in Neural Machine Translation
Vivek Iyer, Edoardo Barba, Alexandra Birch, Jeff Z. Pan, Roberto Navigli
cs.CL
Lexical ambiguity is a significant and pervasive challenge in Neural Machine Translation (NMT), with many state-of-the-art (SOTA) NMT systems struggling to handle polysemous words (Campolungo et al., 2022). The same holds for the NMT pretraining paradigm of denoising synthetic "code-switched" text (Pan et al., 2021; Iy...
http://arxiv.org/pdf/2310.14050v1
2023-10-21T16:13:01
2023-10-21T16:13:01
{ "title": [ -0.021469116, -0.0077323914, 0.00031113625, -0.013473511, -0.018981934, 0.0011405945, -0.009300232, -0.015350342, -0.01625061, 0.04663086, -0.007232666, 0.00623703, 0.022583008, -0.023849487, -0.029541016, 0.009147644, -0.0033683777, ...
http://arxiv.org/abs/2310.14047v1
MeaeQ: Mount Model Extraction Attacks with Efficient Queries
Chengwei Dai, Minxuan Lv, Kun Li, Wei Zhou
cs.CL
We study model extraction attacks in natural language processing (NLP) where attackers aim to steal victim models by repeatedly querying the open Application Programming Interfaces (APIs). Recent works focus on limited-query budget settings and adopt random sampling or active learning-based sampling strategies on publi...
http://arxiv.org/pdf/2310.14047v1
2023-10-21T16:07:16
2023-10-21T16:07:16
{ "title": [ -0.0058937073, 0.0033664703, -0.053527832, -0.02571106, 0.01637268, -0.026550293, -0.0009698868, -0.024353027, -0.019302368, 0.03427124, 0.014701843, 0.0076141357, -0.006401062, -0.0435791, -0.034423828, 0.019699097, 0.002286911, -0....
http://arxiv.org/abs/2310.18338v1
Small Language Models Fine-tuned to Coordinate Larger Language Models improve Complex Reasoning
Gurusha Juneja, Subhabrata Dutta, Soumen Chakrabarti, Sunny Manchanda, Tanmoy Chakraborty
cs.CL, cs.AI
Large Language Models (LLMs) prompted to generate chain-of-thought (CoT) exhibit impressive reasoning capabilities. Recent attempts at prompt decomposition toward solving complex, multi-step reasoning problems depend on the ability of the LLM to simultaneously decompose and solve the problem. A significant disadvantage...
http://arxiv.org/pdf/2310.18338v1
2023-10-21T15:23:20
2023-10-21T15:23:20
{ "title": [ 0.0040397644, -0.03164673, -0.028656006, -0.07696533, -0.047424316, -0.012001038, -0.074645996, 0.030059814, -0.05807495, 0.057678223, -0.0039634705, 0.009994507, 0.032836914, -0.03137207, 0.007610321, -0.0015096664, 0.0022506714, 0....
http://arxiv.org/abs/2310.14034v1
Tree Prompting: Efficient Task Adaptation without Fine-Tuning
John X. Morris, Chandan Singh, Alexander M. Rush, Jianfeng Gao, Yuntian Deng
cs.CL, cs.LG
Prompting language models (LMs) is the main interface for applying them to new tasks. However, for smaller LMs, prompting provides low accuracy compared to gradient-based finetuning. Tree Prompting is an approach to prompting which builds a decision tree of prompts, linking multiple LM calls together to solve a task. A...
http://arxiv.org/pdf/2310.14034v1
2023-10-21T15:18:22
2023-10-21T15:18:22
{ "title": [ -0.020614624, -0.024719238, -0.030654907, -0.06188965, -0.011779785, 0.002281189, -0.042175293, -0.0027046204, 0.00907135, 0.0098724365, -0.022994995, -0.009597778, -0.029937744, -0.023086548, -0.049346924, -0.016326904, 0.017440796, ...
http://arxiv.org/abs/2310.14032v1
Analysing State-Backed Propaganda Websites: a New Dataset and Linguistic Study
Freddy Heppell, Kalina Bontcheva, Carolina Scarton
cs.CL
This paper analyses two hitherto unstudied sites sharing state-backed disinformation, Reliable Recent News (rrn.world) and WarOnFakes (waronfakes.com), which publish content in Arabic, Chinese, English, French, German, and Spanish. We describe our content acquisition methodology and perform cross-site unsupervised topi...
http://arxiv.org/pdf/2310.14032v1
2023-10-21T15:00:27
2023-10-21T15:00:27
{ "title": [ 0.0018043518, -0.00315094, -0.019256592, -0.02960205, 0.02746582, -0.024917603, 0.0154800415, -0.012550354, -0.0435791, 0.017730713, -0.004371643, 0.025817871, -0.009124756, -0.014640808, -0.0051727295, -0.007713318, -0.023117065, 0....
http://arxiv.org/abs/2310.14029v1
LLM-Prop: Predicting Physical And Electronic Properties Of Crystalline Solids From Their Text Descriptions
Andre Niyongabo Rubungo, Craig Arnold, Barry P. Rand, Adji Bousso Dieng
cs.CL, cond-mat.mtrl-sci
The prediction of crystal properties plays a crucial role in the crystal design process. Current methods for predicting crystal properties focus on modeling crystal structures using graph neural networks (GNNs). Although GNNs are powerful, accurately modeling the complex interactions between atoms and molecules within ...
http://arxiv.org/pdf/2310.14029v1
2023-10-21T14:49:58
2023-10-21T14:49:58
{ "title": [ -0.037109375, -0.0038928986, -0.040740967, -0.033081055, -0.014549255, -0.013160706, -0.030838013, -0.005214691, -0.014251709, 0.019439697, -0.0009455681, 0.031463623, 0.012199402, -0.03918457, 0.014846802, 0.014503479, -0.024505615, ...
http://arxiv.org/abs/2310.14028v1
GASCOM: Graph-based Attentive Semantic Context Modeling for Online Conversation Understanding
Vibhor Agarwal, Yu Chen, Nishanth Sastry
cs.CL
Online conversation understanding is an important yet challenging NLP problem which has many useful applications (e.g., hate speech detection). However, online conversations typically unfold over a series of posts and replies to those posts, forming a tree structure within which individual posts may refer to semantic c...
http://arxiv.org/pdf/2310.14028v1
2023-10-21T14:45:26
2023-10-21T14:45:26
{ "title": [ -0.015808105, -0.010475159, -0.024139404, -0.07305908, -0.0033950806, 0.016677856, -0.070251465, 0.010444641, 0.025726318, 0.0435791, -0.030090332, 0.0070648193, 0.02658081, -0.0018072128, -0.05001831, 0.0021209717, -0.016281128, 0.0...
http://arxiv.org/abs/2310.14025v1
Large Language Models and Multimodal Retrieval for Visual Word Sense Disambiguation
Anastasia Kritharoula, Maria Lymperaiou, Giorgos Stamou
cs.CL
Visual Word Sense Disambiguation (VWSD) is a novel challenging task with the goal of retrieving an image among a set of candidates, which better represents the meaning of an ambiguous word within a given context. In this paper, we make a substantial step towards unveiling this interesting task by applying a varying set...
http://arxiv.org/pdf/2310.14025v1
2023-10-21T14:35:42
2023-10-21T14:35:42
{ "title": [ -0.010795593, 0.008514404, -0.03225708, -0.037353516, -0.006767273, -0.0032348633, -0.033050537, -0.006210327, 0.048187256, 0.047973633, -0.044677734, -0.005542755, 0.022155762, 0.0062561035, -0.0748291, -0.005176544, 0.005092621, 0....
http://arxiv.org/abs/2310.14001v1
Toward Stronger Textual Attack Detectors
Pierre Colombo, Marine Picot, Nathan Noiry, Guillaume Staerman, Pablo Piantanida
cs.CL
The landscape of available textual adversarial attacks keeps growing, posing severe threats and raising concerns regarding the deep NLP system's integrity. However, the crucial problem of defending against malicious attacks has only drawn the attention of the NLP community. The latter is nonetheless instrumental in dev...
http://arxiv.org/pdf/2310.14001v1
2023-10-21T13:01:29
2023-10-21T13:01:29
{ "title": [ -0.004901886, -0.0016145706, -0.023956299, -0.05508423, 0.021469116, -0.003730774, -0.013572693, -0.0023498535, -0.019638062, 0.03479004, -0.007949829, 0.016159058, -0.0034561157, -0.0047035217, -0.03805542, 0.005458832, -0.007980347, ...
http://arxiv.org/abs/2310.13998v1
Transductive Learning for Textual Few-Shot Classification in API-based Embedding Models
Pierre Colombo, Victor Pellegrain, Malik Boudiaf, Victor Storchan, Myriam Tami, Ismail Ben Ayed, Celine Hudelot, Pablo Piantanida
cs.CL
Proprietary and closed APIs are becoming increasingly common to process natural language, and are impacting the practical applications of natural language processing, including few-shot classification. Few-shot classification involves training a model to perform a new classification task with a handful of labeled data....
http://arxiv.org/pdf/2310.13998v1
2023-10-21T12:47:10
2023-10-21T12:47:10
{ "title": [ 0.009399414, -0.025741577, -0.019180298, -0.030029297, -0.021911621, -0.015899658, -0.015670776, 0.006378174, 0.026016235, 0.014801025, -0.06112671, 0.029281616, -0.011207581, -0.009765625, -0.03314209, -0.02243042, -0.0015850067, 0....