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/2305.16944v3
Learning to Imagine: Visually-Augmented Natural Language Generation
Tianyi Tang, Yushuo Chen, Yifan Du, Junyi Li, Wayne Xin Zhao, Ji-Rong Wen
cs.CL
People often imagine relevant scenes to aid in the writing process. In this work, we aim to utilize visual information for composition in the same manner as humans. We propose a method, LIVE, that makes pre-trained language models (PLMs) Learn to Imagine for Visuallyaugmented natural language gEneration. First, we imag...
http://arxiv.org/pdf/2305.16944v3
2023-06-15T06:25:34
2023-05-26T13:59:45
{ "title": [ -0.031173706, -0.024337769, -0.019821167, -0.04345703, -0.026901245, 0.009735107, -0.028564453, 0.0039253235, 0.0024547577, 0.025543213, -0.027114868, 0.014587402, -0.014968872, -0.00881958, -0.055419922, -0.008598328, -0.018478394, ...
http://arxiv.org/abs/2305.16938v2
Few-shot Fine-tuning vs. In-context Learning: A Fair Comparison and Evaluation
Marius Mosbach, Tiago Pimentel, Shauli Ravfogel, Dietrich Klakow, Yanai Elazar
cs.CL
Few-shot fine-tuning and in-context learning are two alternative strategies for task adaptation of pre-trained language models. Recently, in-context learning has gained popularity over fine-tuning due to its simplicity and improved out-of-domain generalization, and because extensive evidence shows that fine-tuned model...
http://arxiv.org/pdf/2305.16938v2
2023-05-30T08:34:49
2023-05-26T13:55:17
{ "title": [ 0.02243042, -0.02658081, -0.011489868, -0.0446167, 0.0049324036, -0.012130737, 0.012283325, 0.0009870529, -0.017593384, 0.028701782, -0.037506104, 0.022018433, -0.045806885, -0.00806427, -0.04727173, -0.030349731, 0.008743286, 0.0080...
http://arxiv.org/abs/2305.16937v1
Finspector: A Human-Centered Visual Inspection Tool for Exploring and Comparing Biases among Foundation Models
Bum Chul Kwon, Nandana Mihindukulasooriya
cs.CL, cs.AI, cs.HC
Pre-trained transformer-based language models are becoming increasingly popular due to their exceptional performance on various benchmarks. However, concerns persist regarding the presence of hidden biases within these models, which can lead to discriminatory outcomes and reinforce harmful stereotypes. To address this ...
http://arxiv.org/pdf/2305.16937v1
2023-05-26T13:53:15
2023-05-26T13:53:15
{ "title": [ -0.012748718, 0.011360168, -0.06021118, -0.0112838745, -0.0051956177, -0.037750244, -0.0126571655, -0.039001465, 0.03427124, -0.004081726, -0.04458618, 0.003824234, -0.021743774, -0.03387451, -0.019210815, -0.021972656, -0.031036377, ...
http://arxiv.org/abs/2305.16935v1
Gender Lost In Translation: How Bridging The Gap Between Languages Affects Gender Bias in Zero-Shot Multilingual Translation
Lena Cabrera, Jan Niehues
cs.CL
Neural machine translation (NMT) models often suffer from gender biases that harm users and society at large. In this work, we explore how bridging the gap between languages for which parallel data is not available affects gender bias in multilingual NMT, specifically for zero-shot directions. We evaluate translation b...
http://arxiv.org/pdf/2305.16935v1
2023-05-26T13:51:50
2023-05-26T13:51:50
{ "title": [ -0.05630493, -0.010917664, -0.036132812, -0.0024795532, -0.028549194, -0.023742676, -0.022445679, -0.006137848, -0.0030212402, 0.068237305, -0.013046265, 0.0051193237, 0.049987793, -0.018615723, -0.04385376, -0.006816864, -0.022247314, ...
http://arxiv.org/abs/2305.16934v2
On Evaluating Adversarial Robustness of Large Vision-Language Models
Yunqing Zhao, Tianyu Pang, Chao Du, Xiao Yang, Chongxuan Li, Ngai-Man Cheung, Min Lin
cs.CV, cs.CL, cs.CR, cs.LG, cs.MM
Large vision-language models (VLMs) such as GPT-4 have achieved unprecedented performance in response generation, especially with visual inputs, enabling more creative and adaptable interaction than large language models such as ChatGPT. Nonetheless, multimodal generation exacerbates safety concerns, since adversaries ...
http://arxiv.org/pdf/2305.16934v2
2023-10-29T12:32:19
2023-05-26T13:49:44
{ "title": [ -0.016067505, 0.012489319, -0.019042969, -0.025482178, 0.01335907, -0.009254456, -0.027297974, -0.029724121, -0.0029888153, 0.02432251, -0.018051147, 0.0035705566, -0.02166748, -0.0017786026, -0.053619385, -0.017211914, 0.008811951, ...
http://arxiv.org/abs/2305.16932v1
A Neural State-Space Model Approach to Efficient Speech Separation
Chen Chen, Chao-Han Huck Yang, Kai Li, Yuchen Hu, Pin-Jui Ku, Eng Siong Chng
cs.SD, cs.CL, eess.AS
In this work, we introduce S4M, a new efficient speech separation framework based on neural state-space models (SSM). Motivated by linear time-invariant systems for sequence modeling, our SSM-based approach can efficiently model input signals into a format of linear ordinary differential equations (ODEs) for representa...
http://arxiv.org/pdf/2305.16932v1
2023-05-26T13:47:11
2023-05-26T13:47:11
{ "title": [ -0.0053138733, 0.0211792, -0.04724121, -0.048034668, 0.027709961, -0.005168915, -0.016021729, -0.006198883, -0.00856781, 0.04333496, -0.0055885315, 0.022705078, -0.0036029816, -0.029525757, -0.029464722, 0.011116028, 0.0046043396, 0....
http://arxiv.org/abs/2305.16917v1
Large Language Models Are Partially Primed in Pronoun Interpretation
Suet-Ying Lam, Qingcheng Zeng, Kexun Zhang, Chenyu You, Rob Voigt
cs.CL
While a large body of literature suggests that large language models (LLMs) acquire rich linguistic representations, little is known about whether they adapt to linguistic biases in a human-like way. The present study probes this question by asking whether LLMs display human-like referential biases using stimuli and pr...
http://arxiv.org/pdf/2305.16917v1
2023-05-26T13:30:48
2023-05-26T13:30:48
{ "title": [ 0.006603241, 0.0013303757, -0.027313232, -0.0770874, -0.045166016, 0.00069618225, -0.0814209, 0.008979797, -0.02330017, 0.025619507, -0.032348633, -0.030548096, 0.045135498, -0.056549072, 0.0289917, -0.00699234, 0.009544373, 0.045379...
http://arxiv.org/abs/2305.18239v1
A Study on Knowledge Distillation from Weak Teacher for Scaling Up Pre-trained Language Models
Hayeon Lee, Rui Hou, Jongpil Kim, Davis Liang, Sung Ju Hwang, Alexander Min
cs.CL, cs.AI
Distillation from Weak Teacher (DWT) is a method of transferring knowledge from a smaller, weaker teacher model to a larger student model to improve its performance. Previous studies have shown that DWT can be effective in the vision domain and natural language processing (NLP) pre-training stage. Specifically, DWT sho...
http://arxiv.org/pdf/2305.18239v1
2023-05-26T13:24:49
2023-05-26T13:24:49
{ "title": [ -0.012619019, -0.018661499, -0.046722412, -0.0748291, -0.015548706, -0.009422302, -0.074279785, 0.018188477, -0.007484436, 0.029342651, -0.032104492, 0.00013577938, 0.017944336, -0.021331787, -0.020095825, -0.025161743, 0.0067634583, ...
http://arxiv.org/abs/2305.16897v1
Inter-connection: Effective Connection between Pre-trained Encoder and Decoder for Speech Translation
Yuta Nishikawa, Satoshi Nakamura
cs.CL, cs.SD, eess.AS
In end-to-end speech translation, speech and text pre-trained models improve translation quality. Recently proposed models simply connect the pre-trained models of speech and text as encoder and decoder. Therefore, only the information from the final layer of encoders is input to the decoder. Since it is clear that the...
http://arxiv.org/pdf/2305.16897v1
2023-05-26T13:01:29
2023-05-26T13:01:29
{ "title": [ 0.012199402, 0.008926392, -0.02142334, -0.042236328, -0.011955261, -0.0146865845, -0.028411865, -0.026321411, -0.01411438, 0.050872803, -0.01676941, -0.001701355, 0.05456543, -0.04067993, 0.009017944, -0.0064888, 0.0063934326, 0.0365...
http://arxiv.org/abs/2305.16896v1
MultiTool-CoT: GPT-3 Can Use Multiple External Tools with Chain of Thought Prompting
Tatsuro Inaba, Hirokazu Kiyomaru, Fei Cheng, Sadao Kurohashi
cs.CL, cs.AI, cs.LG
Large language models (LLMs) have achieved impressive performance on various reasoning tasks. To further improve the performance, we propose MultiTool-CoT, a novel framework that leverages chain-of-thought (CoT) prompting to incorporate multiple external tools, such as a calculator and a knowledge retriever, during the...
http://arxiv.org/pdf/2305.16896v1
2023-05-26T13:00:58
2023-05-26T13:00:58
{ "title": [ 0.009101868, 0.00724411, -0.019943237, -0.020217896, -0.01965332, -0.0024375916, -0.044921875, -0.017105103, -0.014633179, 0.0029640198, -0.058746338, 0.021484375, -0.016113281, -0.014274597, 0.025054932, -0.033233643, 0.01939392, 0....
http://arxiv.org/abs/2305.16895v1
UMSE: Unified Multi-scenario Summarization Evaluation
Shen Gao, Zhitao Yao, Chongyang Tao, Xiuying Chen, Pengjie Ren, Zhaochun Ren, Zhumin Chen
cs.CL, cs.AI
Summarization quality evaluation is a non-trivial task in text summarization. Contemporary methods can be mainly categorized into two scenarios: (1) reference-based: evaluating with human-labeled reference summary; (2) reference-free: evaluating the summary consistency of the document. Recent studies mainly focus on on...
http://arxiv.org/pdf/2305.16895v1
2023-05-26T12:54:44
2023-05-26T12:54:44
{ "title": [ -0.0022296906, -0.013519287, -0.040222168, -0.10986328, 0.016159058, 0.01159668, -0.053497314, 0.00055885315, -0.013137817, 0.045288086, 0.012428284, 0.02633667, -0.005580902, -0.01524353, -0.040161133, 0.014923096, -0.016494751, 0.0...
http://arxiv.org/abs/2305.16894v1
Robustness of Multi-Source MT to Transcription Errors
Dominik Macháček, Peter Polák, Ondřej Bojar, Raj Dabre
cs.CL
Automatic speech translation is sensitive to speech recognition errors, but in a multilingual scenario, the same content may be available in various languages via simultaneous interpreting, dubbing or subtitling. In this paper, we hypothesize that leveraging multiple sources will improve translation quality if the sour...
http://arxiv.org/pdf/2305.16894v1
2023-05-26T12:54:16
2023-05-26T12:54:16
{ "title": [ -0.00712204, 0.00749588, -0.036102295, -0.103271484, 0.011566162, 0.007724762, -0.087402344, 0.011886597, 0.024520874, 0.05340576, -0.029159546, 0.0009174347, 0.013504028, -0.012023926, -0.06536865, 0.0105896, -0.017974854, 0.0115127...
http://arxiv.org/abs/2305.16885v1
Hierarchical Verbalizer for Few-Shot Hierarchical Text Classification
Ke Ji, Yixin Lian, Jingsheng Gao, Baoyuan Wang
cs.CL
Due to the complex label hierarchy and intensive labeling cost in practice, the hierarchical text classification (HTC) suffers a poor performance especially when low-resource or few-shot settings are considered. Recently, there is a growing trend of applying prompts on pre-trained language models (PLMs), which has exhi...
http://arxiv.org/pdf/2305.16885v1
2023-05-26T12:41:49
2023-05-26T12:41:49
{ "title": [ 0.004787445, 0.013908386, -0.040161133, -0.06933594, 0.002313614, 0.0025691986, -0.03314209, -0.0024929047, -0.0016555786, 0.0440979, -0.04397583, 0.032318115, 0.0053482056, -0.0006928444, -0.045928955, -0.018463135, -0.013511658, 0....
http://arxiv.org/abs/2305.16867v1
Playing repeated games with Large Language Models
Elif Akata, Lion Schulz, Julian Coda-Forno, Seong Joon Oh, Matthias Bethge, Eric Schulz
cs.CL
Large Language Models (LLMs) are transforming society and permeating into diverse applications. As a result, LLMs will frequently interact with us and other agents. It is, therefore, of great societal value to understand how LLMs behave in interactive social settings. Here, we propose to use behavioral game theory to s...
http://arxiv.org/pdf/2305.16867v1
2023-05-26T12:17:59
2023-05-26T12:17:59
{ "title": [ -0.014549255, 0.0021343231, -0.01574707, -0.042877197, 0.007843018, -0.0046424866, -0.03881836, -0.02444458, -0.0019273758, 0.030471802, -0.02532959, 0.016311646, -0.023269653, 0.0027751923, -0.039215088, 0.002998352, -0.016647339, 0...
http://arxiv.org/abs/2305.16863v2
Controlling Learned Effects to Reduce Spurious Correlations in Text Classifiers
Parikshit Bansal, Amit Sharma
cs.LG, cs.CL
To address the problem of NLP classifiers learning spurious correlations between training features and target labels, a common approach is to make the model's predictions invariant to these features. However, this can be counter-productive when the features have a non-zero causal effect on the target label and thus are...
http://arxiv.org/pdf/2305.16863v2
2023-06-21T07:06:15
2023-05-26T12:15:54
{ "title": [ -0.004360199, 0.009925842, -0.037078857, -0.011306763, 0.017868042, -0.026916504, 0.012275696, -0.027770996, 0.003742218, 0.033081055, -0.029510498, 0.028839111, -0.018676758, 0.0045814514, -0.054626465, -0.016021729, -0.02281189, 0....
http://arxiv.org/abs/2305.16859v1
Scientific Fact-Checking: A Survey of Resources and Approaches
Juraj Vladika, Florian Matthes
cs.CL
The task of fact-checking deals with assessing the veracity of factual claims based on credible evidence and background knowledge. In particular, scientific fact-checking is the variation of the task concerned with verifying claims rooted in scientific knowledge. This task has received significant attention due to the ...
http://arxiv.org/pdf/2305.16859v1
2023-05-26T12:12:15
2023-05-26T12:12:15
{ "title": [ -0.016464233, -0.0038661957, -0.036254883, -0.059509277, 0.013801575, 0.004436493, -0.028915405, -0.017227173, -0.04937744, 0.032073975, 0.020080566, 0.015960693, -0.018295288, -0.03918457, -0.034088135, 0.005672455, -0.02394104, 0.0...
http://arxiv.org/abs/2305.16852v1
Model-Based Simulation for Optimising Smart Reply
Benjamin Towle, Ke Zhou
cs.CL, cs.AI, cs.LG
Smart Reply (SR) systems present a user with a set of replies, of which one can be selected in place of having to type out a response. To perform well at this task, a system should be able to effectively present the user with a diverse set of options, to maximise the chance that at least one of them conveys the user's ...
http://arxiv.org/pdf/2305.16852v1
2023-05-26T12:04:33
2023-05-26T12:04:33
{ "title": [ -0.0023937225, -0.012573242, -0.010917664, -0.056854248, -0.024246216, 0.010948181, -0.04598999, -0.008598328, -0.014144897, 0.06915283, 0.0446167, -0.008712769, -0.0012073517, -0.02508545, -0.06304932, 0.016296387, 0.012527466, 0.00...
http://arxiv.org/abs/2305.18350v1
Towards Open-World Product Attribute Mining: A Lightly-Supervised Approach
Liyan Xu, Chenwei Zhang, Xian Li, Jingbo Shang, Jinho D. Choi
cs.LG, cs.CL, cs.IR
We present a new task setting for attribute mining on e-commerce products, serving as a practical solution to extract open-world attributes without extensive human intervention. Our supervision comes from a high-quality seed attribute set bootstrapped from existing resources, and we aim to expand the attribute vocabula...
http://arxiv.org/pdf/2305.18350v1
2023-05-26T11:51:31
2023-05-26T11:51:31
{ "title": [ -0.036621094, -0.009475708, -0.037506104, -0.028244019, 0.0064201355, 0.0014982224, -0.026901245, -0.017913818, -0.03353882, 0.014282227, -0.027191162, 0.0064735413, -0.032165527, -0.011474609, -0.033081055, -0.03152466, -0.018341064, ...
http://arxiv.org/abs/2305.16843v1
Randomized Positional Encodings Boost Length Generalization of Transformers
Anian Ruoss, Grégoire Delétang, Tim Genewein, Jordi Grau-Moya, Róbert Csordás, Mehdi Bennani, Shane Legg, Joel Veness
cs.LG, cs.AI, cs.CL, stat.ML
Transformers have impressive generalization capabilities on tasks with a fixed context length. However, they fail to generalize to sequences of arbitrary length, even for seemingly simple tasks such as duplicating a string. Moreover, simply training on longer sequences is inefficient due to the quadratic computation co...
http://arxiv.org/pdf/2305.16843v1
2023-05-26T11:47:52
2023-05-26T11:47:52
{ "title": [ -0.0018043518, 0.020767212, -0.03062439, -0.018630981, -0.016815186, -0.00737381, -0.047973633, -0.023910522, -0.033966064, 0.06201172, -0.020584106, 0.028640747, 0.022872925, -0.017700195, 0.011314392, 0.0045204163, 0.020187378, 0.0...
http://arxiv.org/abs/2305.16834v1
Free Lunch: Robust Cross-Lingual Transfer via Model Checkpoint Averaging
Fabian David Schmidt, Ivan Vulić, Goran Glavaš
cs.CL
Massively multilingual language models have displayed strong performance in zero-shot (ZS-XLT) and few-shot (FS-XLT) cross-lingual transfer setups, where models fine-tuned on task data in a source language are transferred without any or with only a few annotated instances to the target language(s). However, current wor...
http://arxiv.org/pdf/2305.16834v1
2023-05-26T11:24:32
2023-05-26T11:24:32
{ "title": [ -0.0061569214, -0.007583618, -0.02986145, -0.051208496, 0.011291504, 0.000415802, -0.04586792, -0.022583008, 0.020736694, 0.040924072, -0.014381409, 0.004650116, 0.04989624, -0.027420044, -0.026885986, 0.026992798, -0.0052375793, 0.0...
http://arxiv.org/abs/2305.16833v1
KNSE: A Knowledge-aware Natural Language Inference Framework for Dialogue Symptom Status Recognition
Wei Chen, Shiqi Wei, Zhongyu Wei, Xuanjing Huang
cs.CL
Symptom diagnosis in medical conversations aims to correctly extract both symptom entities and their status from the doctor-patient dialogue. In this paper, we propose a novel framework called KNSE for symptom status recognition (SSR), where the SSR is formulated as a natural language inference (NLI) task. For each men...
http://arxiv.org/pdf/2305.16833v1
2023-05-26T11:23:26
2023-05-26T11:23:26
{ "title": [ 0.010948181, -0.008300781, -0.029891968, -0.10455322, -0.007965088, 0.012504578, -0.08691406, 0.0101623535, 0.022384644, 0.009590149, -0.016311646, -0.020065308, 0.0023040771, -0.04373169, -0.028579712, -0.019454956, -0.004085541, -0...
http://arxiv.org/abs/2305.16826v1
Prompt- and Trait Relation-aware Cross-prompt Essay Trait Scoring
Heejin Do, Yunsu Kim, Gary Geunbae Lee
cs.CL, cs.AI
Automated essay scoring (AES) aims to score essays written for a given prompt, which defines the writing topic. Most existing AES systems assume to grade essays of the same prompt as used in training and assign only a holistic score. However, such settings conflict with real-education situations; pre-graded essays for ...
http://arxiv.org/pdf/2305.16826v1
2023-05-26T11:11:19
2023-05-26T11:11:19
{ "title": [ 0.001704216, 0.0067481995, -0.016403198, 0.0038909912, -0.026870728, -0.014228821, 0.0010910034, -0.031463623, 0.021392822, -0.0008044243, -0.056915283, -0.0043945312, 0.0058631897, -0.021408081, -0.013427734, -0.012779236, -0.021972656,...
http://arxiv.org/abs/2305.18226v2
HowkGPT: Investigating the Detection of ChatGPT-generated University Student Homework through Context-Aware Perplexity Analysis
Christoforos Vasilatos, Manaar Alam, Talal Rahwan, Yasir Zaki, Michail Maniatakos
cs.CL, cs.AI
As the use of Large Language Models (LLMs) in text generation tasks proliferates, concerns arise over their potential to compromise academic integrity. The education sector currently tussles with distinguishing student-authored homework assignments from AI-generated ones. This paper addresses the challenge by introduci...
http://arxiv.org/pdf/2305.18226v2
2023-06-07T11:43:44
2023-05-26T11:07:25
{ "title": [ 0.018753052, 0.009887695, 0.022598267, 0.03012085, -0.0057029724, -0.008666992, 0.025268555, -0.06109619, -0.013420105, 0.026275635, -0.018707275, 0.050567627, 0.0053367615, -0.022277832, -0.0010814667, -0.0035743713, -0.029510498, 0...
http://arxiv.org/abs/2305.16820v2
Domain Aligned Prefix Averaging for Domain Generalization in Abstractive Summarization
Pranav Ajit Nair, Sukomal Pal, Pradeepika Verma
cs.CL, cs.AI
Domain generalization is hitherto an underexplored area applied in abstractive summarization. Moreover, most existing works on domain generalization have sophisticated training algorithms. In this paper, we propose a lightweight, weight averaging based, Domain Aligned Prefix Averaging approach to domain generalization ...
http://arxiv.org/pdf/2305.16820v2
2023-05-29T06:19:48
2023-05-26T11:02:38
{ "title": [ -0.013885498, 0.008232117, -0.047546387, -0.072143555, -0.018966675, 0.0030612946, -0.06921387, -0.012298584, 0.025253296, 0.059509277, -0.0024433136, 0.018066406, 0.008056641, -0.013320923, -0.055541992, 0.01852417, -0.01612854, 0.0...
http://arxiv.org/abs/2305.16819v1
With a Little Push, NLI Models can Robustly and Efficiently Predict Faithfulness
Julius Steen, Juri Opitz, Anette Frank, Katja Markert
cs.CL
Conditional language models still generate unfaithful output that is not supported by their input. These unfaithful generations jeopardize trust in real-world applications such as summarization or human-machine interaction, motivating a need for automatic faithfulness metrics. To implement such metrics, NLI models seem...
http://arxiv.org/pdf/2305.16819v1
2023-05-26T11:00:04
2023-05-26T11:00:04
{ "title": [ 0.043182373, -0.01209259, -0.001502037, -0.05718994, -0.011894226, 0.0012893677, -0.049865723, -0.030578613, -0.020431519, 0.016357422, -0.013404846, 0.0047187805, -0.0034999847, -0.04425049, 0.01335907, -0.03074646, 0.006866455, 0.0...
http://arxiv.org/abs/2305.16816v1
Songs Across Borders: Singable and Controllable Neural Lyric Translation
Longshen Ou, Xichu Ma, Min-Yen Kan, Ye Wang
cs.CL, 68T50
The development of general-domain neural machine translation (NMT) methods has advanced significantly in recent years, but the lack of naturalness and musical constraints in the outputs makes them unable to produce singable lyric translations. This paper bridges the singability quality gap by formalizing lyric translat...
http://arxiv.org/pdf/2305.16816v1
2023-05-26T10:50:17
2023-05-26T10:50:17
{ "title": [ -0.02645874, -0.02067566, -0.023162842, -0.04397583, -0.0053901672, 0.00566864, -0.029800415, 0.015319824, -0.017303467, 0.04067993, -0.022979736, -0.004032135, 0.0002360344, -0.0084991455, -0.04598999, -0.0060195923, -0.0141067505, ...
http://arxiv.org/abs/2305.16811v1
Improved Visual Story Generation with Adaptive Context Modeling
Zhangyin Feng, Yuchen Ren, Xinmiao Yu, Xiaocheng Feng, Duyu Tang, Shuming Shi, Bing Qin
cs.CV, cs.CL
Diffusion models developed on top of powerful text-to-image generation models like Stable Diffusion achieve remarkable success in visual story generation. However, the best-performing approach considers historically generated results as flattened memory cells, ignoring the fact that not all preceding images contribute ...
http://arxiv.org/pdf/2305.16811v1
2023-05-26T10:43:42
2023-05-26T10:43:42
{ "title": [ -0.027557373, -0.007621765, -0.04623413, -0.055786133, -0.005142212, 0.010276794, -0.046142578, -0.0016880035, -0.0022392273, 0.04458618, -0.022247314, 0.009559631, -0.010169983, -0.0063667297, -0.04006958, 0.0011730194, -0.013305664, ...
http://arxiv.org/abs/2305.16809v3
GenQ: Automated Question Generation to Support Caregivers While Reading Stories with Children
Arun Balajiee Lekshmi Narayanan, Ligia E. Gomez, Martha Michelle Soto Fernandez, Tri Nguyen, Chris Blais, M. Adelaida Restrepo, Art Glenberg
cs.CL, cs.AI, cs.HC
When caregivers ask open--ended questions to motivate dialogue with children, it facilitates the child's reading comprehension skills.Although there is scope for use of technological tools, referred here as "intelligent tutoring systems", to scaffold this process, it is currently unclear whether existing intelligent sy...
http://arxiv.org/pdf/2305.16809v3
2023-09-25T17:37:37
2023-05-26T10:42:21
{ "title": [ -0.05038452, -0.033081055, -0.044952393, -0.023162842, -0.045013428, -0.006122589, -0.03111267, 0.01235199, 0.0013427734, 0.0134887695, -0.026046753, -0.004547119, 0.003921509, -0.03060913, -0.02330017, -0.026855469, -0.0032520294, 0...
http://arxiv.org/abs/2305.16806v4
Do GPTs Produce Less Literal Translations?
Vikas Raunak, Arul Menezes, Matt Post, Hany Hassan Awadalla
cs.CL, cs.AI
Large Language Models (LLMs) such as GPT-3 have emerged as general-purpose language models capable of addressing many natural language generation or understanding tasks. On the task of Machine Translation (MT), multiple works have investigated few-shot prompting mechanisms to elicit better translations from LLMs. Howev...
http://arxiv.org/pdf/2305.16806v4
2023-06-06T03:15:43
2023-05-26T10:38:31
{ "title": [ -0.02041626, 0.002784729, -0.01525116, 0.027160645, -0.03366089, 0.003993988, -0.028900146, -0.04800415, -0.020950317, 0.04425049, -0.008003235, -0.004875183, 0.020858765, -0.017181396, -0.00831604, 0.0055007935, 0.014709473, 0.00544...
http://arxiv.org/abs/2305.16801v2
Motion-Based Sign Language Video Summarization using Curvature and Torsion
Evangelos G. Sartinas, Emmanouil Z. Psarakis, Dimitrios I. Kosmopoulos
cs.CV, cs.CL, 68T45, 68U10, I.4.9; I.5.4; I.2.7
An interesting problem in many video-based applications is the generation of short synopses by selecting the most informative frames, a procedure which is known as video summarization. For sign language videos the benefits of using the $t$-parameterized counterpart of the curvature of the 2-D signer's wrist trajectory ...
http://arxiv.org/pdf/2305.16801v2
2023-06-02T12:02:09
2023-05-26T10:30:23
{ "title": [ -0.02420044, -0.0010490417, -0.010940552, -0.05029297, -0.02168274, 0.020339966, -0.068359375, -0.017791748, 0.015342712, 0.05996704, -0.025787354, 0.012252808, 0.00032830238, -0.009353638, -0.06585693, -0.007858276, -0.002128601, 0....
http://arxiv.org/abs/2305.16799v1
To Revise or Not to Revise: Learning to Detect Improvable Claims for Argumentative Writing Support
Gabriella Skitalinskaya, Henning Wachsmuth
cs.CL
Optimizing the phrasing of argumentative text is crucial in higher education and professional development. However, assessing whether and how the different claims in a text should be revised is a hard task, especially for novice writers. In this work, we explore the main challenges to identifying argumentative claims i...
http://arxiv.org/pdf/2305.16799v1
2023-05-26T10:19:54
2023-05-26T10:19:54
{ "title": [ -0.0019416809, 0.0063323975, -0.010032654, -0.02658081, -0.021743774, 0.0016841888, -0.03262329, -0.04147339, -0.016906738, 0.002609253, -0.031951904, 0.02281189, -0.042388916, 0.0033111572, -0.04788208, -0.04385376, -0.031829834, 0....
http://arxiv.org/abs/2305.16798v1
Schema-Guided User Satisfaction Modeling for Task-Oriented Dialogues
Yue Feng, Yunlong Jiao, Animesh Prasad, Nikolaos Aletras, Emine Yilmaz, Gabriella Kazai
cs.CL, cs.AI
User Satisfaction Modeling (USM) is one of the popular choices for task-oriented dialogue systems evaluation, where user satisfaction typically depends on whether the user's task goals were fulfilled by the system. Task-oriented dialogue systems use task schema, which is a set of task attributes, to encode the user's t...
http://arxiv.org/pdf/2305.16798v1
2023-05-26T10:19:30
2023-05-26T10:19:30
{ "title": [ -0.003446579, -0.009757996, -0.041656494, -0.05895996, 0.0002901554, -0.009963989, -0.030929565, 0.011169434, 0.024749756, 0.033935547, -0.0017604828, -0.03741455, 0.02279663, -0.03805542, -0.06958008, -0.0040245056, -0.022384644, 0....
http://arxiv.org/abs/2305.16797v2
Calibration of Transformer-based Models for Identifying Stress and Depression in Social Media
Loukas Ilias, Spiros Mouzakitis, Dimitris Askounis
cs.CL
In today's fast-paced world, the rates of stress and depression present a surge. Social media provide assistance for the early detection of mental health conditions. Existing methods mainly introduce feature extraction approaches and train shallow machine learning classifiers. Other researches use deep neural networks ...
http://arxiv.org/pdf/2305.16797v2
2023-07-05T21:45:29
2023-05-26T10:19:04
{ "title": [ 0.020401001, 0.029754639, -0.0066566467, -0.034729004, 0.02027893, -0.003522873, -0.027282715, -0.044158936, -0.009681702, 0.031082153, -0.03555298, 0.015419006, 0.027755737, -0.03074646, -0.020645142, -0.008155823, 0.00642395, 0.012...
http://arxiv.org/abs/2305.16784v1
Incorporating Distributions of Discourse Structure for Long Document Abstractive Summarization
Dongqi Pu, Yifan Wang, Vera Demberg
cs.CL, cs.AI, cs.LG
For text summarization, the role of discourse structure is pivotal in discerning the core content of a text. Regrettably, prior studies on incorporating Rhetorical Structure Theory (RST) into transformer-based summarization models only consider the nuclearity annotation, thereby overlooking the variety of discourse rel...
http://arxiv.org/pdf/2305.16784v1
2023-05-26T09:51:47
2023-05-26T09:51:47
{ "title": [ -0.0053863525, 0.014640808, -0.025970459, -0.08312988, -0.0026378632, 0.014320374, -0.059448242, -0.010940552, 0.01033783, 0.036499023, -0.000667572, 0.004722595, -0.020202637, -0.008811951, -0.064331055, -0.0036029816, -0.017700195, ...
http://arxiv.org/abs/2305.16768v1
Towards a Common Understanding of Contributing Factors for Cross-Lingual Transfer in Multilingual Language Models: A Review
Fred Philippy, Siwen Guo, Shohreh Haddadan
cs.CL, cs.AI
In recent years, pre-trained Multilingual Language Models (MLLMs) have shown a strong ability to transfer knowledge across different languages. However, given that the aspiration for such an ability has not been explicitly incorporated in the design of the majority of MLLMs, it is challenging to obtain a unique and str...
http://arxiv.org/pdf/2305.16768v1
2023-05-26T09:31:12
2023-05-26T09:31:12
{ "title": [ -0.025466919, -0.008033752, -0.030715942, -0.04849243, 0.00027036667, 0.003232956, -0.049346924, -0.009307861, 0.0004003048, 0.04336548, 0.014175415, 0.01864624, 0.04562378, -0.033477783, -0.027114868, 0.037841797, -0.028945923, 0.05...
http://arxiv.org/abs/2306.01761v1
Distinguishing Human Generated Text From ChatGPT Generated Text Using Machine Learning
Niful Islam, Debopom Sutradhar, Humaira Noor, Jarin Tasnim Raya, Monowara Tabassum Maisha, Dewan Md Farid
cs.CL, cs.AI, cs.LG
ChatGPT is a conversational artificial intelligence that is a member of the generative pre-trained transformer of the large language model family. This text generative model was fine-tuned by both supervised learning and reinforcement learning so that it can produce text documents that seem to be written by natural int...
http://arxiv.org/pdf/2306.01761v1
2023-05-26T09:27:43
2023-05-26T09:27:43
{ "title": [ -0.0059547424, 0.0002040863, -0.01499176, -0.012458801, -0.023849487, -0.0047340393, -0.017608643, -0.030212402, -0.042236328, 0.047912598, -0.002855301, 0.024810791, -0.013969421, -0.017211914, -0.028396606, -0.0015420914, -0.030426025,...
http://arxiv.org/abs/2305.16765v1
Backpack Language Models
John Hewitt, John Thickstun, Christopher D. Manning, Percy Liang
cs.CL
We present Backpacks: a new neural architecture that marries strong modeling performance with an interface for interpretability and control. Backpacks learn multiple non-contextual sense vectors for each word in a vocabulary, and represent a word in a sequence as a context-dependent, non-negative linear combination of ...
http://arxiv.org/pdf/2305.16765v1
2023-05-26T09:26:23
2023-05-26T09:26:23
{ "title": [ -0.059783936, -0.016326904, -0.03189087, -0.015213013, -0.09649658, 0.008033752, -0.109680176, 0.011795044, -0.030303955, 0.04333496, -0.020858765, -0.043304443, -0.014526367, -0.03479004, 0.00068330765, -0.0368042, 0.030349731, -0.0...
http://arxiv.org/abs/2305.16756v2
Leveraging Domain Knowledge for Inclusive and Bias-aware Humanitarian Response Entry Classification
Nicolò Tamagnone, Selim Fekih, Ximena Contla, Nayid Orozco, Navid Rekabsaz
cs.CL, cs.LG
Accurate and rapid situation analysis during humanitarian crises is critical to delivering humanitarian aid efficiently and is fundamental to humanitarian imperatives and the Leave No One Behind (LNOB) principle. This data analysis can highly benefit from language processing systems, e.g., by classifying the text data ...
http://arxiv.org/pdf/2305.16756v2
2023-05-30T13:16:33
2023-05-26T09:15:05
{ "title": [ 0.005542755, 0.020523071, -0.06112671, -0.062408447, 0.032348633, -0.03125, -0.023330688, -0.005748749, -0.026809692, 0.008262634, -0.011444092, 0.0040779114, -0.019561768, -0.034423828, -0.019714355, -0.024642944, -0.018463135, 0.02...
http://arxiv.org/abs/2305.16755v2
Can large language models generate salient negative statements?
Hiba Arnaout, Simon Razniewski
cs.CL, cs.AI
We examine the ability of large language models (LLMs) to generate salient (interesting) negative statements about real-world entities; an emerging research topic of the last few years. We probe the LLMs using zero- and k-shot unconstrained probes, and compare with traditional methods for negation generation, i.e., pat...
http://arxiv.org/pdf/2305.16755v2
2023-09-21T13:36:03
2023-05-26T09:13:59
{ "title": [ 0.01676941, -0.014381409, 0.0013694763, -0.03375244, -0.04348755, -0.007850647, -0.020126343, -0.0059890747, 0.015609741, 0.022155762, -0.056427002, -0.026168823, 0.00894928, -0.014755249, -0.034973145, -0.026855469, -0.0018720627, 0...
http://arxiv.org/abs/2305.16750v1
Automating the Analysis of Institutional Design in International Agreements
Anna Wróblewska, Bartosz Pieliński, Karolina Seweryn, Sylwia Sysko-Romańczuk, Karol Saputa, Aleksandra Wichrowska, Hanna Schreiber
cs.CL, cs.IR
This paper explores the automatic knowledge extraction of formal institutional design - norms, rules, and actors - from international agreements. The focus was to analyze the relationship between the visibility and centrality of actors in the formal institutional design in regulating critical aspects of cultural herita...
http://arxiv.org/pdf/2305.16750v1
2023-05-26T08:57:11
2023-05-26T08:57:11
{ "title": [ -0.052642822, 0.0072364807, -0.037750244, -0.0068893433, -0.028396606, -0.017288208, -0.015930176, -0.013679504, -0.019500732, 0.022949219, -0.048217773, 0.014953613, -0.019439697, -0.005180359, -0.030654907, -0.02268982, -0.0014095306, ...
http://arxiv.org/abs/2305.16742v1
Parameter-Efficient Fine-Tuning without Introducing New Latency
Baohao Liao, Yan Meng, Christof Monz
cs.CL, cs.AI, cs.LG
Parameter-efficient fine-tuning (PEFT) of pre-trained language models has recently demonstrated remarkable achievements, effectively matching the performance of full fine-tuning while utilizing significantly fewer trainable parameters, and consequently addressing the storage and communication constraints. Nonetheless, ...
http://arxiv.org/pdf/2305.16742v1
2023-05-26T08:44:42
2023-05-26T08:44:42
{ "title": [ 0.019454956, -0.019363403, -0.002922058, -0.10095215, 0.011016846, 0.0032577515, -0.02041626, -0.00050878525, -0.012489319, 0.03741455, 0.018249512, -0.004333496, -0.0059051514, -0.02885437, -0.025039673, 0.014656067, 0.0138549805, 0...
http://arxiv.org/abs/2305.16740v1
Conjunct Resolution in the Face of Verbal Omissions
Royi Rassin, Yoav Goldberg, Reut Tsarfaty
cs.CL
Verbal omissions are complex syntactic phenomena in VP coordination structures. They occur when verbs and (some of) their arguments are omitted from subsequent clauses after being explicitly stated in an initial clause. Recovering these omitted elements is necessary for accurate interpretation of the sentence, and whil...
http://arxiv.org/pdf/2305.16740v1
2023-05-26T08:44:02
2023-05-26T08:44:02
{ "title": [ 0.0077781677, -0.013473511, -0.04916382, -0.08087158, -0.047424316, -0.017089844, -0.06274414, 0.022994995, -0.006793976, 0.02961731, -0.022369385, -0.02722168, 0.034851074, -0.047668457, 0.0012207031, -0.022537231, -0.0016460419, 0....
http://arxiv.org/abs/2305.18212v1
Multimodal Recommendation Dialog with Subjective Preference: A New Challenge and Benchmark
Yuxing Long, Binyuan Hui, Caixia Yuan1, Fei Huang, Yongbin Li, Xiaojie Wang
cs.IR, cs.AI, cs.CL, cs.CV, cs.LG, cs.MM
Existing multimodal task-oriented dialog data fails to demonstrate the diverse expressions of user subjective preferences and recommendation acts in the real-life shopping scenario. This paper introduces a new dataset SURE (Multimodal Recommendation Dialog with SUbjective PREference), which contains 12K shopping dialog...
http://arxiv.org/pdf/2305.18212v1
2023-05-26T08:43:46
2023-05-26T08:43:46
{ "title": [ 0.01386261, 0.0006814003, -0.034423828, -0.05822754, 0.012229919, -0.024978638, -0.020584106, 0.0010023117, 0.008033752, 0.038085938, -0.0071487427, 0.015342712, -0.009117126, -0.012001038, -0.04260254, 0.005519867, -0.025650024, 0.0...
http://arxiv.org/abs/2305.16739v1
AlignScore: Evaluating Factual Consistency with a Unified Alignment Function
Yuheng Zha, Yichi Yang, Ruichen Li, Zhiting Hu
cs.CL
Many text generation applications require the generated text to be factually consistent with input information. Automatic evaluation of factual consistency is challenging. Previous work has developed various metrics that often depend on specific functions, such as natural language inference (NLI) or question answering ...
http://arxiv.org/pdf/2305.16739v1
2023-05-26T08:41:59
2023-05-26T08:41:59
{ "title": [ -0.02911377, -0.0060653687, -0.05441284, -0.05670166, 0.019714355, 0.0026798248, -0.029907227, -0.023483276, -0.0061836243, 0.032104492, -0.0054397583, -0.0005378723, -0.004760742, -0.023101807, -0.05255127, 0.00004249811, -0.015625, ...
http://arxiv.org/abs/2305.16734v1
AMPERE: AMR-Aware Prefix for Generation-Based Event Argument Extraction Model
I-Hung Hsu, Zhiyu Xie, Kuan-Hao Huang, Prem Natarajan, Nanyun Peng
cs.CL, cs.AI
Event argument extraction (EAE) identifies event arguments and their specific roles for a given event. Recent advancement in generation-based EAE models has shown great performance and generalizability over classification-based models. However, existing generation-based EAE models mostly focus on problem re-formulation...
http://arxiv.org/pdf/2305.16734v1
2023-05-26T08:38:25
2023-05-26T08:38:25
{ "title": [ 0.002494812, -0.018966675, -0.042755127, -0.114990234, -0.0004401207, 0.019470215, -0.08508301, 0.0013303757, -0.02696228, 0.022445679, 0.071899414, -0.009391785, 0.013221741, -0.08105469, -0.00806427, 0.03100586, -0.02204895, 0.0174...
http://arxiv.org/abs/2305.16731v4
Automatic Emotion Experiencer Recognition
Maximilian Wegge, Roman Klinger
cs.CL
The most prominent subtask in emotion analysis is emotion classification; to assign a category to a textual unit, for instance a social media post. Many research questions from the social sciences do, however, not only require the detection of the emotion of an author of a post but to understand who is ascribed an emot...
http://arxiv.org/pdf/2305.16731v4
2023-07-27T15:09:38
2023-05-26T08:33:28
{ "title": [ -0.004142761, 0.020950317, -0.02973938, -0.036315918, 0.004219055, -0.01184082, -0.013771057, -0.028320312, 0.028015137, 0.041748047, -0.031921387, -0.01928711, 0.022857666, -0.033996582, -0.044036865, 0.022323608, 0.018875122, 0.023...
http://arxiv.org/abs/2305.16726v1
RankCSE: Unsupervised Sentence Representations Learning via Learning to Rank
Jiduan Liu, Jiahao Liu, Qifan Wang, Jingang Wang, Wei Wu, Yunsen Xian, Dongyan Zhao, Kai Chen, Rui Yan
cs.CL, cs.AI
Unsupervised sentence representation learning is one of the fundamental problems in natural language processing with various downstream applications. Recently, contrastive learning has been widely adopted which derives high-quality sentence representations by pulling similar semantics closer and pushing dissimilar ones...
http://arxiv.org/pdf/2305.16726v1
2023-05-26T08:27:07
2023-05-26T08:27:07
{ "title": [ -0.0047073364, -0.0050201416, -0.034820557, -0.052215576, 0.0128479, 0.0015735626, -0.031066895, -0.016036987, 0.0127334595, 0.025817871, -0.009613037, 0.011352539, -0.0079956055, -0.016937256, -0.03527832, 0.0077667236, -0.014656067, ...
http://arxiv.org/abs/2305.16724v2
Code-Switched Text Synthesis in Unseen Language Pairs
I-Hung Hsu, Avik Ray, Shubham Garg, Nanyun Peng, Jing Huang
cs.CL, cs.AI
Existing efforts on text synthesis for code-switching mostly require training on code-switched texts in the target language pairs, limiting the deployment of the models to cases lacking code-switched data. In this work, we study the problem of synthesizing code-switched texts for language pairs absent from the training...
http://arxiv.org/pdf/2305.16724v2
2023-07-07T07:51:38
2023-05-26T08:22:35
{ "title": [ -0.053833008, -0.01449585, -0.007587433, -0.010658264, -0.04244995, 0.011627197, -0.046447754, -0.007423401, -0.011924744, 0.026901245, -0.014656067, 0.0013999939, -0.010314941, -0.018875122, -0.03488159, -0.008583069, -0.030548096, ...
http://arxiv.org/abs/2305.16718v2
People and Places of Historical Europe: Bootstrapping Annotation Pipeline and a New Corpus of Named Entities in Late Medieval Texts
Vít Novotný, Kristýna Luger, Michal Štefánik, Tereza Vrabcová, Aleš Horák
cs.CL, 68T50, I.2.7; I.7.0
Although pre-trained named entity recognition (NER) models are highly accurate on modern corpora, they underperform on historical texts due to differences in language OCR errors. In this work, we develop a new NER corpus of 3.6M sentences from late medieval charters written mainly in Czech, Latin, and German. We show...
http://arxiv.org/pdf/2305.16718v2
2023-06-06T20:42:10
2023-05-26T08:05:01
{ "title": [ 0.018814087, 0.017913818, -0.029037476, -0.06378174, 0.030822754, -0.008552551, -0.0068511963, -0.008377075, 0.0070381165, 0.05630493, 0.0031814575, 0.026855469, 0.039093018, -0.0149383545, -0.036224365, 0.021255493, -0.00032925606, ...
http://arxiv.org/abs/2305.16701v1
PIP: Parse-Instructed Prefix for Syntactically Controlled Paraphrase Generation
Yixin Wan, Kuan-Hao Huang, Kai-Wei Chang
cs.CL, cs.AI
Syntactically controlled paraphrase generation requires language models to generate paraphrases for sentences according to specific syntactic structures. Existing fine-tuning methods for this task are costly as all the parameters of the model need to be updated during the training process. Inspired by recent studies on...
http://arxiv.org/pdf/2305.16701v1
2023-05-26T07:42:38
2023-05-26T07:42:38
{ "title": [ -0.0206604, -0.02796936, -0.03427124, -0.060302734, -0.04034424, 0.012451172, -0.06378174, -0.0012769699, -0.044830322, 0.044281006, 0.01751709, 0.012390137, 0.03753662, -0.042816162, -0.003686905, 0.009475708, -0.039398193, 0.054077...
http://arxiv.org/abs/2305.16697v1
DKAF: KB Arbitration for Learning Task-Oriented Dialog Systems with Dialog-KB Inconsistencies
Vishal Vivek Saley, Rocktim Jyoti Das, Dinesh Raghu, Mausam
cs.CL
Task-oriented dialog (TOD) agents often ground their responses on external knowledge bases (KBs). These KBs can be dynamic and may be updated frequently. Existing approaches for learning TOD agents assume the KB snapshot contemporary to each individual dialog is available during training. However, in real-world scenari...
http://arxiv.org/pdf/2305.16697v1
2023-05-26T07:36:23
2023-05-26T07:36:23
{ "title": [ -0.036315918, -0.0015363693, -0.056518555, -0.050720215, -0.017227173, -0.0077056885, -0.057037354, 0.0070991516, 0.026306152, 0.0440979, -0.034942627, -0.0016021729, 0.007358551, -0.0051994324, -0.0647583, -0.014198303, -0.024932861, ...
http://arxiv.org/abs/2305.16675v1
Multiview Identifiers Enhanced Generative Retrieval
Yongqi Li, Nan Yang, Liang Wang, Furu Wei, Wenjie Li
cs.CL, cs.AI, cs.IR, cs.LG
Instead of simply matching a query to pre-existing passages, generative retrieval generates identifier strings of passages as the retrieval target. At a cost, the identifier must be distinctive enough to represent a passage. Current approaches use either a numeric ID or a text piece (such as a title or substrings) as t...
http://arxiv.org/pdf/2305.16675v1
2023-05-26T06:50:21
2023-05-26T06:50:21
{ "title": [ 0.00459671, 0.0107803345, -0.0569458, -0.06011963, -0.0079422, -0.026901245, -0.028869629, -0.0009899139, 0.015686035, 0.041748047, -0.017303467, 0.023025513, -0.0015115738, -0.025482178, -0.028076172, 0.022018433, -0.0038776398, 0.0...
http://arxiv.org/abs/2305.16664v1
Score-balanced Loss for Multi-aspect Pronunciation Assessment
Heejin Do, Yunsu Kim, Gary Geunbae Lee
cs.CL, cs.SD, eess.AS
With rapid technological growth, automatic pronunciation assessment has transitioned toward systems that evaluate pronunciation in various aspects, such as fluency and stress. However, despite the highly imbalanced score labels within each aspect, existing studies have rarely tackled the data imbalance problem. In this...
http://arxiv.org/pdf/2305.16664v1
2023-05-26T06:21:37
2023-05-26T06:21:37
{ "title": [ -0.0013389587, -0.008888245, -0.04269409, -0.051330566, -0.027267456, -0.0129852295, -0.024658203, 0.03186035, 0.0061569214, 0.05419922, -0.009635925, -0.005558014, 0.0036411285, -0.0317688, -0.061157227, -0.0054016113, -0.01687622, ...
http://arxiv.org/abs/2305.16663v2
GDA: Generative Data Augmentation Techniques for Relation Extraction Tasks
Xuming Hu, Aiwei Liu, Zeqi Tan, Xin Zhang, Chenwei Zhang, Irwin King, Philip S. Yu
cs.CL, 68T01, I.2.7
Relation extraction (RE) tasks show promising performance in extracting relations from two entities mentioned in sentences, given sufficient annotations available during training. Such annotations would be labor-intensive to obtain in practice. Existing work adopts data augmentation techniques to generate pseudo-annota...
http://arxiv.org/pdf/2305.16663v2
2023-06-15T02:43:12
2023-05-26T06:21:01
{ "title": [ 0.00017523766, -0.014305115, -0.055725098, -0.085998535, 0.019317627, 0.006702423, -0.03817749, 0.0018825531, 0.04147339, 0.01260376, -0.01876831, 0.010627747, -0.0101623535, -0.01576233, -0.04360962, 0.0103302, -0.015319824, 0.04724...
http://arxiv.org/abs/2305.16653v1
AdaPlanner: Adaptive Planning from Feedback with Language Models
Haotian Sun, Yuchen Zhuang, Lingkai Kong, Bo Dai, Chao Zhang
cs.CL, cs.AI, cs.LG
Large language models (LLMs) have recently demonstrated the potential in acting as autonomous agents for sequential decision-making tasks. However, most existing methods either take actions greedily without planning or rely on static plans that are not adaptable to environmental feedback. Consequently, the sequential d...
http://arxiv.org/pdf/2305.16653v1
2023-05-26T05:52:27
2023-05-26T05:52:27
{ "title": [ -0.04562378, -0.011169434, -0.0440979, -0.057861328, 0.0072021484, 0.012626648, -0.056274414, -0.017715454, -0.017944336, 0.015129089, -0.0368042, 0.019744873, -0.031707764, -0.011131287, -0.017318726, -0.018737793, 0.0016307831, 0.0...
http://arxiv.org/abs/2305.16651v1
TADA: Task-Agnostic Dialect Adapters for English
Will Held, Caleb Ziems, Diyi Yang
cs.CL
Large Language Models, the dominant starting point for Natural Language Processing (NLP) applications, fail at a higher rate for speakers of English dialects other than Standard American English (SAE). Prior work addresses this using task-specific data or synthetic data augmentation, both of which require intervention ...
http://arxiv.org/pdf/2305.16651v1
2023-05-26T05:45:03
2023-05-26T05:45:03
{ "title": [ -0.028366089, -0.01473999, -0.04550171, -0.05593872, 0.0058135986, 0.006462097, -0.0579834, 0.011680603, -0.00004464388, 0.057434082, 0.013870239, -0.006389618, 0.041259766, -0.019134521, -0.06188965, 0.037994385, 0.00031590462, 0.03...
http://arxiv.org/abs/2305.16648v1
Dramatic Conversation Disentanglement
Kent K. Chang, Danica Chen, David Bamman
cs.CL
We present a new dataset for studying conversation disentanglement in movies and TV series. While previous work has focused on conversation disentanglement in IRC chatroom dialogues, movies and TV shows provide a space for studying complex pragmatic patterns of floor and topic change in face-to-face multi-party interac...
http://arxiv.org/pdf/2305.16648v1
2023-05-26T05:39:49
2023-05-26T05:39:49
{ "title": [ 0.057739258, 0.016418457, -0.03656006, -0.05819702, -0.014694214, -0.033691406, -0.024917603, 0.006839752, -0.020141602, 0.031921387, 0.011734009, 0.002620697, 0.03479004, -0.052764893, 0.0418396, 0.005001068, 0.010681152, 0.02180481...
http://arxiv.org/abs/2305.16646v2
Language Models Can Improve Event Prediction by Few-Shot Abductive Reasoning
Xiaoming Shi, Siqiao Xue, Kangrui Wang, Fan Zhou, James Y. Zhang, Jun Zhou, Chenhao Tan, Hongyuan Mei
cs.CL, cs.AI, cs.LG
Large language models have shown astonishing performance on a wide range of reasoning tasks. In this paper, we investigate whether they could reason about real-world events and help improve the prediction performance of event sequence models. We design LAMP, a framework that integrates a large language model in event p...
http://arxiv.org/pdf/2305.16646v2
2023-10-08T03:27:20
2023-05-26T05:32:29
{ "title": [ 0.013404846, -0.02758789, -0.015975952, -0.074401855, -0.043395996, -0.005470276, -0.064208984, 0.0087509155, -0.016540527, 0.022018433, 0.0101623535, -0.017028809, 0.037139893, -0.04937744, 0.009750366, 0.0039367676, 0.0128479, 0.02...
http://arxiv.org/abs/2305.16641v1
Are Fairy Tales Fair? Analyzing Gender Bias in Temporal Narrative Event Chains of Children's Fairy Tales
Paulina Toro Isaza, Guangxuan Xu, Akintoye Oloko, Yufang Hou, Nanyun Peng, Dakuo Wang
cs.CL, 68T50
Social biases and stereotypes are embedded in our culture in part through their presence in our stories, as evidenced by the rich history of humanities and social science literature analyzing such biases in children stories. Because these analyses are often conducted manually and at a small scale, such investigations c...
http://arxiv.org/pdf/2305.16641v1
2023-05-26T05:29:37
2023-05-26T05:29:37
{ "title": [ -0.011138916, -0.0008997917, -0.06695557, -0.03491211, 0.012718201, -0.04159546, 0.002658844, 0.01109314, -0.026168823, 0.033966064, -0.020935059, 0.007789612, 0.014587402, -0.013809204, -0.023071289, -0.02545166, -0.036865234, 0.058...
http://arxiv.org/abs/2305.16638v1
Adversarial Multi-task Learning for End-to-end Metaphor Detection
Shenglong Zhang, Ying Liu
cs.CL
Metaphor detection (MD) suffers from limited training data. In this paper, we started with a linguistic rule called Metaphor Identification Procedure and then proposed a novel multi-task learning framework to transfer knowledge in basic sense discrimination (BSD) to MD. BSD is constructed from word sense disambiguation...
http://arxiv.org/pdf/2305.16638v1
2023-05-26T05:28:00
2023-05-26T05:28:00
{ "title": [ 0.0014686584, 0.01626587, -0.016143799, -0.01965332, 0.021438599, -0.013244629, 0.00043606758, -0.032348633, -0.0020103455, 0.048309326, 0.015098572, 0.007545471, 0.027648926, -0.017822266, -0.037384033, 0.020904541, -0.014427185, 0....
http://arxiv.org/abs/2305.16636v2
DataFinder: Scientific Dataset Recommendation from Natural Language Descriptions
Vijay Viswanathan, Luyu Gao, Tongshuang Wu, Pengfei Liu, Graham Neubig
cs.IR, cs.CL, cs.DL
Modern machine learning relies on datasets to develop and validate research ideas. Given the growth of publicly available data, finding the right dataset to use is increasingly difficult. Any research question imposes explicit and implicit constraints on how well a given dataset will enable researchers to answer this q...
http://arxiv.org/pdf/2305.16636v2
2023-06-07T03:08:27
2023-05-26T05:22:36
{ "title": [ -0.029373169, -0.015380859, -0.044921875, -0.049926758, -0.0038833618, 0.001996994, -0.027633667, -0.0062294006, 0.011108398, 0.03137207, -0.029144287, 0.017196655, -0.014808655, -0.0019025803, -0.045532227, 0.0013256073, -0.014434814, ...
http://arxiv.org/abs/2305.16635v1
Impossible Distillation: from Low-Quality Model to High-Quality Dataset & Model for Summarization and Paraphrasing
Jaehun Jung, Peter West, Liwei Jiang, Faeze Brahman, Ximing Lu, Jillian Fisher, Taylor Sorensen, Yejin Choi
cs.CL, cs.AI, cs.LG
It is commonly perceived that the strongest language models (LMs) rely on a combination of massive scale, instruction data, and human feedback to perform specialized tasks -- e.g. summarization and paraphrasing, without supervision. In this paper, we propose that language models can learn to summarize and paraphrase se...
http://arxiv.org/pdf/2305.16635v1
2023-05-26T05:19:24
2023-05-26T05:19:24
{ "title": [ 0.030914307, 0.014434814, -0.032836914, -0.0869751, 0.005554199, -0.008262634, -0.06347656, -0.033416748, 0.0071105957, 0.038360596, -0.039916992, -0.019195557, 0.03427124, -0.023666382, -0.040496826, -0.013572693, -0.009666443, 0.05...
http://arxiv.org/abs/2305.16633v1
Zero is Not Hero Yet: Benchmarking Zero-Shot Performance of LLMs for Financial Tasks
Agam Shah, Sudheer Chava
cs.CL, q-fin.GN
Recently large language models (LLMs) like ChatGPT have shown impressive performance on many natural language processing tasks with zero-shot. In this paper, we investigate the effectiveness of zero-shot LLMs in the financial domain. We compare the performance of ChatGPT along with some open-source generative LLMs in z...
http://arxiv.org/pdf/2305.16633v1
2023-05-26T05:13:01
2023-05-26T05:13:01
{ "title": [ -0.03881836, -0.015792847, -0.029922485, -0.030776978, -0.042388916, -0.009490967, -0.050323486, -0.0028934479, 0.009689331, 0.04598999, -0.025161743, 0.013633728, 0.0093307495, -0.022003174, -0.034057617, -0.0033798218, -0.0037727356, ...
http://arxiv.org/abs/2305.16626v1
Evaluation of Question Generation Needs More References
Shinhyeok Oh, Hyojun Go, Hyeongdon Moon, Yunsung Lee, Myeongho Jeong, Hyun Seung Lee, Seungtaek Choi
cs.CL, cs.AI, I.2.7
Question generation (QG) is the task of generating a valid and fluent question based on a given context and the target answer. According to various purposes, even given the same context, instructors can ask questions about different concepts, and even the same concept can be written in different ways. However, the eval...
http://arxiv.org/pdf/2305.16626v1
2023-05-26T04:40:56
2023-05-26T04:40:56
{ "title": [ 0.024017334, 0.015396118, -0.006717682, -0.020431519, -0.026184082, -0.0048828125, -0.0011758804, -0.025512695, -0.020217896, 0.041107178, -0.04537964, -0.02607727, -0.0025081635, -0.026672363, -0.013214111, -0.012763977, 0.01499176, ...
http://arxiv.org/abs/2305.16617v1
Efficient Detection of LLM-generated Texts with a Bayesian Surrogate Model
Zhijie Deng, Hongcheng Gao, Yibo Miao, Hao Zhang
cs.LG, cs.AI, cs.CL
The detection of machine-generated text, especially from large language models (LLMs), is crucial in preventing serious social problems resulting from their misuse. Some methods train dedicated detectors on specific datasets but fall short in generalizing to unseen test data, while other zero-shot ones often yield subo...
http://arxiv.org/pdf/2305.16617v1
2023-05-26T04:23:10
2023-05-26T04:23:10
{ "title": [ -0.027130127, 0.012214661, -0.022094727, -0.0236969, -0.01977539, 0.012962341, -0.060791016, -0.033935547, -0.038848877, 0.013290405, 0.0040779114, 0.024353027, -0.02267456, -0.03945923, -0.0008568764, -0.018417358, -0.04046631, 0.04...
http://arxiv.org/abs/2306.05537v1
AaKOS: Aspect-adaptive Knowledge-based Opinion Summarization
Guan Wang, Weihua Li, Edmund M-K. Lai, Quan Bai
cs.CL
The rapid growth of information on the Internet has led to an overwhelming amount of opinions and comments on various activities, products, and services. This makes it difficult and time-consuming for users to process all the available information when making decisions. Text summarization, a Natural Language Processing...
http://arxiv.org/pdf/2306.05537v1
2023-05-26T03:44:35
2023-05-26T03:44:35
{ "title": [ -0.017974854, -0.020355225, -0.016525269, -0.078186035, -0.019073486, 0.024276733, -0.05316162, 0.009338379, 0.028961182, 0.054718018, -0.0039100647, -0.014511108, 0.010139465, 0.000104784966, -0.09552002, 0.011734009, 0.0075416565, ...
http://arxiv.org/abs/2305.16599v1
Bridging the Domain Gaps in Context Representations for k-Nearest Neighbor Neural Machine Translation
Zhiwei Cao, Baosong Yang, Huan Lin, Suhang Wu, Xiangpeng Wei, Dayiheng Liu, Jun Xie, Min Zhang, Jinsong Su
cs.CL
$k$-Nearest neighbor machine translation ($k$NN-MT) has attracted increasing attention due to its ability to non-parametrically adapt to new translation domains. By using an upstream NMT model to traverse the downstream training corpus, it is equipped with a datastore containing vectorized key-value pairs, which are re...
http://arxiv.org/pdf/2305.16599v1
2023-05-26T03:04:42
2023-05-26T03:04:42
{ "title": [ -0.020568848, -0.0041122437, 0.008056641, -0.034973145, -0.008918762, 0.018310547, -0.026489258, -0.018432617, -0.00020051003, 0.066345215, -0.017532349, 0.005176544, 0.035949707, -0.0021343231, -0.042785645, 0.023010254, 0.0012693405, ...
http://arxiv.org/abs/2305.16598v1
NormMark: A Weakly Supervised Markov Model for Socio-cultural Norm Discovery
Farhad Moghimifar, Shilin Qu, Tongtong Wu, Yuan-Fang Li, Gholamreza Haffari
cs.CL
Norms, which are culturally accepted guidelines for behaviours, can be integrated into conversational models to generate utterances that are appropriate for the socio-cultural context. Existing methods for norm recognition tend to focus only on surface-level features of dialogues and do not take into account the intera...
http://arxiv.org/pdf/2305.16598v1
2023-05-26T03:03:37
2023-05-26T03:03:37
{ "title": [ -0.0018901825, 0.0072784424, -0.025878906, -0.04220581, 0.01007843, -0.008483887, -0.028839111, -0.023803711, -0.0061416626, 0.008041382, -0.028381348, 0.030975342, -0.01537323, -0.021591187, -0.019546509, 0.004135132, -0.019638062, ...
http://arxiv.org/abs/2305.16597v1
Neural Architecture Search for Parameter-Efficient Fine-tuning of Large Pre-trained Language Models
Neal Lawton, Anoop Kumar, Govind Thattai, Aram Galstyan, Greg Ver Steeg
cs.CL, cs.AI, cs.LG, I.2.7
Parameter-efficient tuning (PET) methods fit pre-trained language models (PLMs) to downstream tasks by either computing a small compressed update for a subset of model parameters, or appending and fine-tuning a small number of new model parameters to the pre-trained network. Hand-designed PET architectures from the lit...
http://arxiv.org/pdf/2305.16597v1
2023-05-26T03:01:07
2023-05-26T03:01:07
{ "title": [ -0.017074585, -0.011474609, -0.0073280334, -0.06890869, -0.0067825317, 0.008087158, -0.050750732, 0.0013656616, -0.009567261, 0.04260254, -0.004234314, 0.0010557175, 0.018203735, -0.016082764, -0.04272461, 0.009002686, 0.0018043518, ...
http://arxiv.org/abs/2305.17147v2
Heterogeneous Value Evaluation for Large Language Models
Zhaowei Zhang, Nian Liu, Siyuan Qi, Ceyao Zhang, Ziqi Rong, Song-Chun Zhu, Shuguang Cui, Yaodong Yang
cs.CL, cs.AI, cs.HC, cs.LG, I.2.0; K.4.1
The emergent capabilities of Large Language Models (LLMs) have made it crucial to align their values with those of humans. Current methodologies typically attempt alignment with a homogeneous human value and requires human verification, yet lack consensus on the desired aspect and depth of alignment and resulting human...
http://arxiv.org/pdf/2305.17147v2
2023-06-01T17:00:50
2023-05-26T02:34:20
{ "title": [ -0.0335083, 0.0011911392, -0.05807495, -0.066101074, 0.017044067, -0.0049934387, -0.0541687, 0.009300232, 0.012763977, 0.024353027, -0.02999878, -0.0072250366, -0.0076065063, -0.014762878, -0.03652954, -0.0062942505, -0.028427124, 0....
http://arxiv.org/abs/2305.16585v1
ParaAMR: A Large-Scale Syntactically Diverse Paraphrase Dataset by AMR Back-Translation
Kuan-Hao Huang, Varun Iyer, I-Hung Hsu, Anoop Kumar, Kai-Wei Chang, Aram Galstyan
cs.CL
Paraphrase generation is a long-standing task in natural language processing (NLP). Supervised paraphrase generation models, which rely on human-annotated paraphrase pairs, are cost-inefficient and hard to scale up. On the other hand, automatically annotated paraphrase pairs (e.g., by machine back-translation), usually...
http://arxiv.org/pdf/2305.16585v1
2023-05-26T02:27:33
2023-05-26T02:27:33
{ "title": [ -0.017654419, -0.0013580322, -0.034576416, -0.08416748, -0.0025634766, 0.012886047, -0.07299805, 0.0047836304, -0.005065918, 0.045806885, -0.0036525726, 0.016784668, 0.03515625, -0.019073486, -0.030166626, 0.027328491, -0.011131287, ...
http://arxiv.org/abs/2305.16582v1
Beyond Chain-of-Thought, Effective Graph-of-Thought Reasoning in Large Language Models
Yao Yao, Zuchao Li, Hai Zhao
cs.CL
With the widespread use of large language models (LLMs) in NLP tasks, researchers have discovered the potential of Chain-of-thought (CoT) to assist LLMs in accomplishing complex reasoning tasks by generating intermediate steps. However, human thought processes are often non-linear, rather than simply sequential chains ...
http://arxiv.org/pdf/2305.16582v1
2023-05-26T02:15:09
2023-05-26T02:15:09
{ "title": [ -0.010910034, -0.034423828, -0.009231567, -0.07183838, -0.026306152, 0.010055542, -0.047729492, 0.007209778, 0.010444641, 0.0056419373, -0.016448975, -0.0038146973, 0.016983032, -0.033477783, -0.017486572, 0.005455017, -0.02267456, 0...
http://arxiv.org/abs/2305.16581v1
An Investigation of Noise in Morphological Inflection
Adam Wiemerslage, Changbing Yang, Garrett Nicolai, Miikka Silfverberg, Katharina Kann
cs.CL
With a growing focus on morphological inflection systems for languages where high-quality data is scarce, training data noise is a serious but so far largely ignored concern. We aim at closing this gap by investigating the types of noise encountered within a pipeline for truly unsupervised morphological paradigm comple...
http://arxiv.org/pdf/2305.16581v1
2023-05-26T02:14:34
2023-05-26T02:14:34
{ "title": [ -0.0047073364, 0.008384705, -0.035491943, -0.03793335, -0.0011978149, -0.021377563, -0.014808655, 0.020248413, -0.012435913, 0.03149414, -0.03488159, 0.0031261444, -0.0010032654, -0.03274536, -0.027389526, -0.0066452026, -0.01876831, ...
http://arxiv.org/abs/2305.16579v3
NLP Reproducibility For All: Understanding Experiences of Beginners
Shane Storks, Keunwoo Peter Yu, Ziqiao Ma, Joyce Chai
cs.CL, cs.AI
As natural language processing (NLP) has recently seen an unprecedented level of excitement, and more people are eager to enter the field, it is unclear whether current research reproducibility efforts are sufficient for this group of beginners to apply the latest developments. To understand their needs, we conducted a...
http://arxiv.org/pdf/2305.16579v3
2023-06-03T14:01:24
2023-05-26T02:08:54
{ "title": [ -0.030548096, -0.014579773, -0.026351929, -0.038085938, -0.014442444, 0.0070648193, -0.04434204, -0.01637268, 0.0121154785, 0.011276245, -0.037078857, 0.0025596619, -0.0075187683, -0.02067566, -0.03010559, 0.0074539185, -0.021469116, ...
http://arxiv.org/abs/2305.16577v1
Nichelle and Nancy: The Influence of Demographic Attributes and Tokenization Length on First Name Biases
Haozhe An, Rachel Rudinger
cs.CL
Through the use of first name substitution experiments, prior research has demonstrated the tendency of social commonsense reasoning models to systematically exhibit social biases along the dimensions of race, ethnicity, and gender (An et al., 2023). Demographic attributes of first names, however, are strongly correlat...
http://arxiv.org/pdf/2305.16577v1
2023-05-26T01:57:42
2023-05-26T01:57:42
{ "title": [ -0.03289795, -0.02319336, -0.05303955, -0.006416321, -0.042877197, -0.025543213, -0.0395813, -0.0027484894, -0.025390625, 0.047851562, -0.03475952, 0.02532959, 0.0051612854, -0.014671326, -0.015159607, -0.021514893, -0.04107666, 0.06...
http://arxiv.org/abs/2305.16572v1
Counterfactual reasoning: Testing language models' understanding of hypothetical scenarios
Jiaxuan Li, Lang Yu, Allyson Ettinger
cs.CL
Current pre-trained language models have enabled remarkable improvements in downstream tasks, but it remains difficult to distinguish effects of statistical correlation from more systematic logical reasoning grounded on the understanding of real world. We tease these factors apart by leveraging counterfactual condition...
http://arxiv.org/pdf/2305.16572v1
2023-05-26T01:42:57
2023-05-26T01:42:57
{ "title": [ -0.011352539, -0.025512695, -0.009033203, -0.05508423, -0.017410278, 0.0048217773, -0.05419922, -0.016571045, 0.031051636, 0.012016296, -0.032928467, -0.0074768066, 0.002620697, -0.023666382, -0.05596924, -0.0023288727, -0.026184082, ...
http://arxiv.org/abs/2305.18342v2
Neural Task Synthesis for Visual Programming
Victor-Alexandru Pădurean, Georgios Tzannetos, Adish Singla
cs.LG, cs.AI, cs.CL, cs.CY, cs.PL
Generative neural models hold great promise in enhancing programming education by synthesizing new content for students. We seek to design neural models that can automatically generate programming tasks for a given specification in the context of visual programming domains. Despite the recent successes of large generat...
http://arxiv.org/pdf/2305.18342v2
2023-06-01T14:05:10
2023-05-26T01:08:18
{ "title": [ -0.034729004, -0.004135132, -0.024627686, -0.032165527, -0.0055999756, 0.0024719238, -0.03881836, -0.009803772, 0.018341064, 0.02267456, -0.03540039, -0.011138916, -0.018951416, -0.017837524, -0.059783936, -0.011886597, 0.0018224716, ...
http://arxiv.org/abs/2305.16559v1
Teamwork Is Not Always Good: An Empirical Study of Classifier Drift in Class-incremental Information Extraction
Minqian Liu, Lifu Huang
cs.CL, cs.AI, cs.LG
Class-incremental learning (CIL) aims to develop a learning system that can continually learn new classes from a data stream without forgetting previously learned classes. When learning classes incrementally, the classifier must be constantly updated to incorporate new classes, and the drift in decision boundary may le...
http://arxiv.org/pdf/2305.16559v1
2023-05-26T00:57:43
2023-05-26T00:57:43
{ "title": [ -0.018447876, 0.010803223, -0.05935669, -0.034423828, 0.014450073, -0.037902832, 0.0006489754, -0.015655518, -0.0009236336, -0.005092621, -0.06048584, 0.026123047, -0.05947876, 0.002691269, -0.026641846, -0.047546387, -0.023605347, 0...
http://arxiv.org/abs/2305.18620v1
CONA: A novel CONtext-Aware instruction paradigm for communication using large language model
Nan Zhou, Xinghui Tao, Xi Chen
cs.CL, cs.AI, cs.HC
We introduce CONA, a novel context-aware instruction paradigm for effective knowledge dissemination using generative pre-trained transformer (GPT) models. CONA is a flexible framework designed to leverage the capabilities of Large Language Models (LLMs) and incorporate DIKW (Data, Information, Knowledge, Wisdom) hierar...
http://arxiv.org/pdf/2305.18620v1
2023-05-26T00:53:18
2023-05-26T00:53:18
{ "title": [ -0.040100098, -0.014076233, -0.04724121, -0.047851562, 0.007499695, -0.0032100677, -0.04385376, 0.01159668, 0.0021438599, 0.039154053, -0.017990112, -0.003030777, 0.02508545, -0.016540527, -0.018356323, -0.0019741058, -0.02633667, 0....
http://arxiv.org/abs/2305.16548v1
Annotating and Detecting Fine-grained Factual Errors for Dialogue Summarization
Rongxin Zhu, Jianzhong Qi, Jey Han Lau
cs.CL, cs.AI
A series of datasets and models have been proposed for summaries generated for well-formatted documents such as news articles. Dialogue summaries, however, have been under explored. In this paper, we present the first dataset with fine-grained factual error annotations named DIASUMFACT. We define fine-grained factual e...
http://arxiv.org/pdf/2305.16548v1
2023-05-26T00:18:33
2023-05-26T00:18:33
{ "title": [ 0.008872986, -0.0023326874, -0.030227661, -0.07556152, 0.00466156, 0.00072717667, -0.046813965, 0.0036010742, 0.006828308, 0.044708252, 0.005710602, 0.015777588, 0.0034008026, -0.004386902, -0.06781006, 0.013214111, -0.033477783, 0.0...
http://arxiv.org/abs/2305.16521v1
Label Agnostic Pre-training for Zero-shot Text Classification
Christopher Clarke, Yuzhao Heng, Yiping Kang, Krisztian Flautner, Lingjia Tang, Jason Mars
cs.CL, cs.LG
Conventional approaches to text classification typically assume the existence of a fixed set of predefined labels to which a given text can be classified. However, in real-world applications, there exists an infinite label space for describing a given text. In addition, depending on the aspect (sentiment, topic, etc.) ...
http://arxiv.org/pdf/2305.16521v1
2023-05-25T22:55:32
2023-05-25T22:55:32
{ "title": [ -0.006542206, 0.0014028549, -0.03387451, -0.054107666, -0.0024871826, -0.0104904175, -0.010665894, 0.005214691, 0.009651184, 0.04171753, -0.022155762, 0.020950317, -0.015434265, -0.003440857, -0.047943115, -0.00030899048, -0.0015764236, ...
http://arxiv.org/abs/2305.16519v1
The Dangers of trusting Stochastic Parrots: Faithfulness and Trust in Open-domain Conversational Question Answering
Sabrina Chiesurin, Dimitris Dimakopoulos, Marco Antonio Sobrevilla Cabezudo, Arash Eshghi, Ioannis Papaioannou, Verena Rieser, Ioannis Konstas
cs.CL
Large language models are known to produce output which sounds fluent and convincing, but is also often wrong, e.g. "unfaithful" with respect to a rationale as retrieved from a knowledge base. In this paper, we show that task-based systems which exhibit certain advanced linguistic dialog behaviors, such as lexical alig...
http://arxiv.org/pdf/2305.16519v1
2023-05-25T22:54:13
2023-05-25T22:54:13
{ "title": [ 0.012336731, 0.008239746, -0.045654297, -0.06121826, 0.0178833, -0.0287323, -0.015365601, 0.0007944107, -0.015617371, 0.04006958, -0.010467529, 0.012039185, 0.011688232, -0.016479492, -0.053131104, -0.013389587, -0.006477356, 0.00799...
http://arxiv.org/abs/2305.16504v1
On the Tool Manipulation Capability of Open-source Large Language Models
Qiantong Xu, Fenglu Hong, Bo Li, Changran Hu, Zhengyu Chen, Jian Zhang
cs.CL, cs.AI, cs.LG
Recent studies on software tool manipulation with large language models (LLMs) mostly rely on closed model APIs. The industrial adoption of these models is substantially constrained due to the security and robustness risks in exposing information to closed LLM API services. In this paper, we ask can we enhance open-sou...
http://arxiv.org/pdf/2305.16504v1
2023-05-25T22:10:20
2023-05-25T22:10:20
{ "title": [ -0.039398193, -0.0034599304, -0.035491943, -0.064208984, -0.0008826256, 0.011398315, -0.0657959, -0.007408142, -0.010345459, 0.050628662, -0.0010223389, -0.022476196, 0.036376953, -0.019378662, -0.04711914, 0.014770508, -0.005695343, ...
http://arxiv.org/abs/2305.16503v1
IMBERT: Making BERT Immune to Insertion-based Backdoor Attacks
Xuanli He, Jun Wang, Benjamin Rubinstein, Trevor Cohn
cs.CL
Backdoor attacks are an insidious security threat against machine learning models. Adversaries can manipulate the predictions of compromised models by inserting triggers into the training phase. Various backdoor attacks have been devised which can achieve nearly perfect attack success without affecting model prediction...
http://arxiv.org/pdf/2305.16503v1
2023-05-25T22:08:57
2023-05-25T22:08:57
{ "title": [ -0.005870819, -0.0043678284, -0.016738892, -0.045684814, -0.011566162, 0.0027160645, -0.020843506, -0.01158905, -0.06097412, 0.048217773, 0.025527954, 0.007896423, 0.012748718, -0.040100098, -0.031799316, 0.010238647, -0.030258179, 0...
http://arxiv.org/abs/2305.16490v1
Prototype-Based Interpretability for Legal Citation Prediction
Chu Fei Luo, Rohan Bhambhoria, Samuel Dahan, Xiaodan Zhu
cs.CL
Deep learning has made significant progress in the past decade, and demonstrates potential to solve problems with extensive social impact. In high-stakes decision making areas such as law, experts often require interpretability for automatic systems to be utilized in practical settings. In this work, we attempt to addr...
http://arxiv.org/pdf/2305.16490v1
2023-05-25T21:40:58
2023-05-25T21:40:58
{ "title": [ -0.022659302, 0.00793457, -0.0209198, -0.070739746, 0.0056152344, 0.026855469, -0.0619812, -0.014663696, 0.00065279007, 0.017929077, -0.018859863, -0.009880066, 0.017868042, -0.018066406, -0.039489746, -0.0015449524, -0.02709961, 0.0...
http://arxiv.org/abs/2305.16470v1
Measuring the Effect of Influential Messages on Varying Personas
Chenkai Sun, Jinning Li, Hou Pong Chan, ChengXiang Zhai, Heng Ji
cs.CL, cs.LG
Predicting how a user responds to news events enables important applications such as allowing intelligent agents or content producers to estimate the effect on different communities and revise unreleased messages to prevent unexpected bad outcomes such as social conflict and moral injury. We present a new task, Respons...
http://arxiv.org/pdf/2305.16470v1
2023-05-25T21:01:00
2023-05-25T21:01:00
{ "title": [ 0.0013475418, 0.014816284, -0.025634766, -0.023391724, 0.0028858185, -0.030899048, 0.008781433, -0.017959595, -0.03579712, 0.02848816, -0.0047912598, 0.013893127, -0.01235199, -0.038116455, -0.0055618286, -0.00019586086, -0.0038337708, ...
http://arxiv.org/abs/2305.16444v1
Don't Retrain, Just Rewrite: Countering Adversarial Perturbations by Rewriting Text
Ashim Gupta, Carter Wood Blum, Temma Choji, Yingjie Fei, Shalin Shah, Alakananda Vempala, Vivek Srikumar
cs.CL
Can language models transform inputs to protect text classifiers against adversarial attacks? In this work, we present ATINTER, a model that intercepts and learns to rewrite adversarial inputs to make them non-adversarial for a downstream text classifier. Our experiments on four datasets and five attack mechanisms reve...
http://arxiv.org/pdf/2305.16444v1
2023-05-25T19:42:51
2023-05-25T19:42:51
{ "title": [ -0.006729126, -0.010223389, 0.00003796816, -0.042175293, -0.0009264946, 0.013137817, -0.028366089, -0.036865234, -0.03692627, 0.019973755, 0.00036764145, 0.041046143, -0.012512207, -0.024627686, -0.0051727295, -0.002073288, -0.003364563,...
http://arxiv.org/abs/2305.16433v1
Neural Machine Translation for Mathematical Formulae
Felix Petersen, Moritz Schubotz, Andre Greiner-Petter, Bela Gipp
cs.CL, cs.SC, stat.AP
We tackle the problem of neural machine translation of mathematical formulae between ambiguous presentation languages and unambiguous content languages. Compared to neural machine translation on natural language, mathematical formulae have a much smaller vocabulary and much longer sequences of symbols, while their tran...
http://arxiv.org/pdf/2305.16433v1
2023-05-25T19:15:06
2023-05-25T19:15:06
{ "title": [ -0.026687622, 0.0075263977, -0.018951416, -0.04827881, 0.0071792603, 0.0049362183, -0.021209717, 0.001499176, 0.009124756, 0.05053711, 0.0001449585, -0.012771606, 0.015686035, -0.013420105, -0.05014038, 0.019195557, 0.0021324158, 0.0...
http://arxiv.org/abs/2305.16426v2
Not wacky vs. definitely wacky: A study of scalar adverbs in pretrained language models
Isabelle Lorge, Janet Pierrehumbert
cs.CL
Vector space models of word meaning all share the assumption that words occurring in similar contexts have similar meanings. In such models, words that are similar in their topical associations but differ in their logical force tend to emerge as semantically close, creating well-known challenges for NLP applications th...
http://arxiv.org/pdf/2305.16426v2
2023-10-22T17:06:35
2023-05-25T18:56:26
{ "title": [ 0.00920105, 0.01373291, -0.003129959, -0.043914795, 0.00025701523, -0.0030345917, -0.04034424, -0.027191162, -0.022766113, 0.036987305, -0.0082473755, 0.018859863, 0.020004272, -0.018844604, -0.030227661, 0.0065231323, -0.014221191, ...
http://arxiv.org/abs/2305.16409v2
Counterfactual Probing for the Influence of Affect and Specificity on Intergroup Bias
Venkata S Govindarajan, Kyle Mahowald, David I. Beaver, Junyi Jessy Li
cs.CL, cs.CY
While existing work on studying bias in NLP focues on negative or pejorative language use, Govindarajan et al. (2023) offer a revised framing of bias in terms of intergroup social context, and its effects on language behavior. In this paper, we investigate if two pragmatic features (specificity and affect) systematical...
http://arxiv.org/pdf/2305.16409v2
2023-06-02T19:11:52
2023-05-25T18:19:39
{ "title": [ 0.030914307, 0.01626587, -0.027236938, -0.04156494, 0.024795532, -0.034698486, 0.014839172, -0.032592773, -0.017288208, 0.027755737, -0.04336548, -0.0014324188, -0.002483368, -0.01991272, -0.030395508, -0.0309906, -0.017593384, 0.040...
http://arxiv.org/abs/2305.16407v1
Script Normalization for Unconventional Writing of Under-Resourced Languages in Bilingual Communities
Sina Ahmadi, Antonios Anastasopoulos
cs.CL
The wide accessibility of social media has provided linguistically under-represented communities with an extraordinary opportunity to create content in their native languages. This, however, comes with certain challenges in script normalization, particularly where the speakers of a language in a bilingual community rel...
http://arxiv.org/pdf/2305.16407v1
2023-05-25T18:18:42
2023-05-25T18:18:42
{ "title": [ -0.049926758, 0.010139465, -0.0413208, -0.027893066, -0.019973755, -0.0035915375, -0.035980225, 0.0064964294, -0.049316406, 0.058898926, -0.026657104, 0.04284668, 0.006954193, -0.007709503, -0.0033855438, -0.00554657, -0.0059013367, ...
http://arxiv.org/abs/2305.16406v2
Context-aware attention layers coupled with optimal transport domain adaptation and multimodal fusion methods for recognizing dementia from spontaneous speech
Loukas Ilias, Dimitris Askounis
cs.CL
Alzheimer's disease (AD) constitutes a complex neurocognitive disease and is the main cause of dementia. Although many studies have been proposed targeting at diagnosing dementia through spontaneous speech, there are still limitations. Existing state-of-the-art approaches, which propose multimodal methods, train separa...
http://arxiv.org/pdf/2305.16406v2
2023-07-26T20:53:22
2023-05-25T18:18:09
{ "title": [ 0.047729492, 0.010002136, -0.04043579, -0.082092285, 0.015899658, -0.02142334, -0.022567749, 0.007713318, -0.008575439, 0.046203613, -0.0063552856, -0.00869751, 0.005947113, -0.029037476, -0.036071777, -0.028259277, -0.008468628, 0.0...
http://arxiv.org/abs/2305.16397v3
Are Diffusion Models Vision-And-Language Reasoners?
Benno Krojer, Elinor Poole-Dayan, Vikram Voleti, Christopher Pal, Siva Reddy
cs.CV, cs.AI, cs.CL
Text-conditioned image generation models have recently shown immense qualitative success using denoising diffusion processes. However, unlike discriminative vision-and-language models, it is a non-trivial task to subject these diffusion-based generative models to automatic fine-grained quantitative evaluation of high-l...
http://arxiv.org/pdf/2305.16397v3
2023-11-02T18:36:09
2023-05-25T18:02:22
{ "title": [ 0.0018463135, 0.0060424805, -0.020324707, -0.0010070801, -0.03149414, -0.02078247, 0.000395298, -0.019760132, 0.013092041, -0.0030517578, -0.04547119, -0.022918701, 0.00004297495, -0.034362793, 0.010192871, -0.02394104, 0.0018177032, ...
http://arxiv.org/abs/2305.17144v2
Ghost in the Minecraft: Generally Capable Agents for Open-World Environments via Large Language Models with Text-based Knowledge and Memory
Xizhou Zhu, Yuntao Chen, Hao Tian, Chenxin Tao, Weijie Su, Chenyu Yang, Gao Huang, Bin Li, Lewei Lu, Xiaogang Wang, Yu Qiao, Zhaoxiang Zhang, Jifeng Dai
cs.AI, cs.CL, cs.CV, cs.LG
The captivating realm of Minecraft has attracted substantial research interest in recent years, serving as a rich platform for developing intelligent agents capable of functioning in open-world environments. However, the current research landscape predominantly focuses on specific objectives, such as the popular "Obtai...
http://arxiv.org/pdf/2305.17144v2
2023-06-01T09:18:01
2023-05-25T17:59:49
{ "title": [ 0.010322571, -0.005592346, -0.0036182404, -0.0022449493, -0.023040771, -0.009918213, 0.0034446716, -0.028701782, 0.0041503906, 0.035339355, -0.02583313, -0.024291992, 0.04144287, -0.0357666, -0.027145386, -0.01914978, -0.029891968, 0...
http://arxiv.org/abs/2305.16307v2
IndicTrans2: Towards High-Quality and Accessible Machine Translation Models for all 22 Scheduled Indian Languages
AI4Bharat, Jay Gala, Pranjal A. Chitale, Raghavan AK, Sumanth Doddapaneni, Varun Gumma, Aswanth Kumar, Janki Nawale, Anupama Sujatha, Ratish Puduppully, Vivek Raghavan, Pratyush Kumar, Mitesh M. Khapra, Raj Dabre, Anoop Kunchukuttan
cs.CL
India has a rich linguistic landscape with languages from 4 major language families spoken by over a billion people. 22 of these languages are listed in the Constitution of India (referred to as scheduled languages) are the focus of this work. Given the linguistic diversity, high-quality and accessible Machine Translat...
http://arxiv.org/pdf/2305.16307v2
2023-06-17T04:00:19
2023-05-25T17:57:43
{ "title": [ -0.060638428, -0.012535095, -0.032684326, -0.038238525, -0.02684021, 0.01285553, -0.06890869, 0.0093307495, -0.027313232, 0.040863037, -0.0044784546, 0.008850098, 0.029129028, -0.02684021, -0.02809143, 0.008613586, -0.0059890747, 0.0...
http://arxiv.org/abs/2305.16302v1
Cross-Lingual Knowledge Distillation for Answer Sentence Selection in Low-Resource Languages
Shivanshu Gupta, Yoshitomo Matsubara, Ankit Chadha, Alessandro Moschitti
cs.CL
While impressive performance has been achieved on the task of Answer Sentence Selection (AS2) for English, the same does not hold for languages that lack large labeled datasets. In this work, we propose Cross-Lingual Knowledge Distillation (CLKD) from a strong English AS2 teacher as a method to train AS2 models for low...
http://arxiv.org/pdf/2305.16302v1
2023-05-25T17:56:04
2023-05-25T17:56:04
{ "title": [ -0.022445679, -0.016860962, -0.03579712, -0.060394287, -0.013153076, -0.009094238, -0.06488037, 0.0040779114, 0.04473877, 0.02835083, -0.020080566, 0.010147095, 0.003211975, -0.006454468, -0.06329346, 0.012191772, 0.005680084, 0.0175...
http://arxiv.org/abs/2305.16300v1
Landmark Attention: Random-Access Infinite Context Length for Transformers
Amirkeivan Mohtashami, Martin Jaggi
cs.CL, cs.LG
While transformers have shown remarkable success in natural language processing, their attention mechanism's large memory requirements have limited their ability to handle longer contexts. Prior approaches, such as recurrent memory or retrieval-based augmentation, have either compromised the random-access flexibility o...
http://arxiv.org/pdf/2305.16300v1
2023-05-25T17:53:42
2023-05-25T17:53:42
{ "title": [ -0.0002784729, 0.018981934, -0.023605347, -0.055847168, 0.021224976, -0.0052871704, 0.0007305145, -0.017501831, 0.012336731, 0.04434204, -0.05178833, 0.008850098, 0.007827759, -0.00044631958, -0.036895752, 0.000042676926, 0.00070858, ...
http://arxiv.org/abs/2306.03823v1
Transformative Effects of ChatGPT on Modern Education: Emerging Era of AI Chatbots
Sukhpal Singh Gill, Minxian Xu, Panos Patros, Huaming Wu, Rupinder Kaur, Kamalpreet Kaur, Stephanie Fuller, Manmeet Singh, Priyansh Arora, Ajith Kumar Parlikad, Vlado Stankovski, Ajith Abraham, Soumya K. Ghosh, Hanan Lutfiyya, Salil S. Kanhere, Rami Bahsoon, Omer Rana, Schahram Dustdar, Rizos Sakellariou, Steve Uhlig, ...
cs.CY, cs.AI, cs.CL
ChatGPT, an AI-based chatbot, was released to provide coherent and useful replies based on analysis of large volumes of data. In this article, leading scientists, researchers and engineers discuss the transformative effects of ChatGPT on modern education. This research seeks to improve our knowledge of ChatGPT capabili...
http://arxiv.org/pdf/2306.03823v1
2023-05-25T17:35:57
2023-05-25T17:35:57
{ "title": [ -0.007255554, -0.0076789856, -0.018371582, 0.0034122467, -0.01133728, -0.021621704, 0.030014038, -0.010681152, -0.03274536, 0.054382324, -0.018203735, 0.021697998, 0.012992859, -0.003993988, -0.023162842, 0.0023822784, -0.018447876, ...
http://arxiv.org/abs/2305.16265v3
UNITE: A Unified Benchmark for Text-to-SQL Evaluation
Wuwei Lan, Zhiguo Wang, Anuj Chauhan, Henghui Zhu, Alexander Li, Jiang Guo, Sheng Zhang, Chung-Wei Hang, Joseph Lilien, Yiqun Hu, Lin Pan, Mingwen Dong, Jun Wang, Jiarong Jiang, Stephen Ash, Vittorio Castelli, Patrick Ng, Bing Xiang
cs.CL
A practical text-to-SQL system should generalize well on a wide variety of natural language questions, unseen database schemas, and novel SQL query structures. To comprehensively evaluate text-to-SQL systems, we introduce a UNIfied benchmark for Text-to-SQL Evaluation (UNITE). It is composed of publicly available text-...
http://arxiv.org/pdf/2305.16265v3
2023-07-14T15:56:31
2023-05-25T17:19:52
{ "title": [ -0.023117065, -0.009353638, -0.033203125, -0.079833984, 0.025375366, 0.014892578, -0.02961731, 0.014953613, 0.01940918, 0.05215454, -0.031234741, 0.025436401, 0.0020103455, 0.0033950806, -0.0713501, -0.006755829, -0.013519287, 0.0296...
http://arxiv.org/abs/2305.16264v4
Scaling Data-Constrained Language Models
Niklas Muennighoff, Alexander M. Rush, Boaz Barak, Teven Le Scao, Aleksandra Piktus, Nouamane Tazi, Sampo Pyysalo, Thomas Wolf, Colin Raffel
cs.CL, cs.AI, cs.LG
The current trend of scaling language models involves increasing both parameter count and training dataset size. Extrapolating this trend suggests that training dataset size may soon be limited by the amount of text data available on the internet. Motivated by this limit, we investigate scaling language models in data-...
http://arxiv.org/pdf/2305.16264v4
2023-10-26T02:24:24
2023-05-25T17:18:55
{ "title": [ -0.058807373, -0.016052246, -0.07165527, -0.06542969, -0.013381958, 0.0071411133, -0.07635498, 0.036621094, 0.004005432, 0.026062012, -0.038513184, -0.016021729, 0.00831604, -0.015472412, -0.035736084, -0.00059986115, -0.0052604675, ...