id
stringlengths
33
34
title
stringlengths
6
218
authors
stringlengths
5
6.98k
categories
stringlengths
5
188
summary
stringlengths
34
3.26k
link_pdf
stringlengths
33
34
updated
timestamp[s]
published
timestamp[s]
embeddings
dict
http://arxiv.org/abs/2306.10339v1
Persian Semantic Role Labeling Using Transfer Learning and BERT-Based Models
Saeideh Niksirat Aghdam, Sayyed Ali Hossayni, Erfan Khedersolh Sadeh, Nasim Khozouei, Behrouz Minaei Bidgoli
cs.CL
Semantic role labeling (SRL) is the process of detecting the predicate-argument structure of each predicate in a sentence. SRL plays a crucial role as a pre-processing step in many NLP applications such as topic and concept extraction, question answering, summarization, machine translation, sentiment analysis, and text...
http://arxiv.org/pdf/2306.10339v1
2023-06-17T12:50:09
2023-06-17T12:50:09
{ "title": [ -0.0051116943, -0.022766113, -0.020980835, -0.06451416, -0.017120361, 0.019989014, -0.05319214, 0.011405945, -0.016479492, 0.014015198, -0.002357483, 0.016708374, -0.010612488, -0.030807495, -0.011314392, -0.0057411194, -0.030014038, ...
http://arxiv.org/abs/2306.10322v2
MO-VLN: A Multi-Task Benchmark for Open-set Zero-Shot Vision-and-Language Navigation
Xiwen Liang, Liang Ma, Shanshan Guo, Jianhua Han, Hang Xu, Shikui Ma, Xiaodan Liang
cs.CV, cs.AI, cs.CL
Given a natural language, a general robot has to comprehend the instruction and find the target object or location based on visual observations even in unexplored environments. Most agents rely on massive diverse training data to achieve better generalization, which requires expensive labor. These agents often focus on...
http://arxiv.org/pdf/2306.10322v2
2023-09-26T05:18:49
2023-06-17T11:44:04
{ "title": [ -0.01020813, -0.017074585, -0.00762558, -0.04107666, -0.026672363, -0.012626648, -0.03652954, -0.015930176, 0.018371582, 0.025802612, -0.011764526, 0.0009174347, -0.031311035, -0.018814087, -0.04534912, -0.0048980713, 0.019180298, 0....
http://arxiv.org/abs/2306.10317v1
Seen to Unseen: Exploring Compositional Generalization of Multi-Attribute Controllable Dialogue Generation
Weihao Zeng, Lulu Zhao, Keqing He, Ruotong Geng, Jingang Wang, Wei Wu, Weiran Xu
cs.CL
Existing controllable dialogue generation work focuses on the single-attribute control and lacks generalization capability to out-of-distribution multiple attribute combinations. In this paper, we explore the compositional generalization for multi-attribute controllable dialogue generation where a model can learn from ...
http://arxiv.org/pdf/2306.10317v1
2023-06-17T10:50:19
2023-06-17T10:50:19
{ "title": [ 0.00016331673, -0.0012464523, -0.021835327, -0.078430176, -0.002773285, 0.0075569153, -0.058532715, 0.011886597, 0.00041031837, 0.03375244, 0.017150879, 0.017196655, -0.014427185, -0.028381348, -0.046173096, -0.00097084045, -0.041778564,...
http://arxiv.org/abs/2306.10315v1
FutureTOD: Teaching Future Knowledge to Pre-trained Language Model for Task-Oriented Dialogue
Weihao Zeng, Keqing He, Yejie Wang, Chen Zeng, Jingang Wang, Yunsen Xian, Weiran Xu
cs.CL
Pre-trained language models based on general text enable huge success in the NLP scenario. But the intrinsical difference of linguistic patterns between general text and task-oriented dialogues makes existing pre-trained language models less useful in practice. Current dialogue pre-training methods rely on a contrastiv...
http://arxiv.org/pdf/2306.10315v1
2023-06-17T10:40:07
2023-06-17T10:40:07
{ "title": [ -0.0057868958, -0.03024292, -0.040100098, -0.091796875, -0.008544922, -0.0044021606, -0.058654785, 0.03692627, 0.004337311, 0.031173706, 0.015731812, -0.018569946, 0.006164551, -0.033813477, -0.060302734, -0.00029492378, -0.0013561249, ...
http://arxiv.org/abs/2306.10270v1
Old and New Minimalism: a Hopf algebra comparison
Matilde Marcolli, Robert C. Berwick, Noam Chomsky
cs.CL, math.QA, math.RA, 68Q70, 16T05
In this paper we compare some old formulations of Minimalism, in particular Stabler's computational minimalism, and Chomsky's new formulation of Merge and Minimalism, from the point of view of their mathematical description in terms of Hopf algebras. We show that the newer formulation has a clear advantage purely in te...
http://arxiv.org/pdf/2306.10270v1
2023-06-17T06:31:36
2023-06-17T06:31:36
{ "title": [ -0.02734375, -0.0017604828, -0.039520264, -0.072387695, 0.014678955, 0.012069702, -0.034057617, 0.0019607544, 0.0007324219, 0.046142578, 0.037902832, 0.01878357, -0.02229309, -0.023620605, -0.04864502, 0.044158936, -0.0102005005, 0.0...
http://arxiv.org/abs/2306.10249v1
Large Language Models for Telecom: The Next Big Thing?
Lina Bariah, Qiyang Zhao, Hang Zou, Yu Tian, Faouzi Bader, Merouane Debbah
cs.CL, cs.AI
The evolution of generative artificial intelligence (GenAI) constitutes a turning point in reshaping the future of technology in different aspects. Wireless networks in particular, with the blooming of self-evolving networks, represent a rich field for exploiting GenAI and reaping several benefits that can fundamentall...
http://arxiv.org/pdf/2306.10249v1
2023-06-17T03:45:00
2023-06-17T03:45:00
{ "title": [ -0.051971436, -0.015007019, -0.05303955, -0.06384277, 0.005836487, 0.0098724365, -0.06750488, 0.028167725, -0.016952515, 0.04724121, -0.016815186, -0.009307861, 0.010986328, -0.011413574, -0.039733887, 0.011802673, -0.004169464, 0.02...
http://arxiv.org/abs/2306.10241v1
Snowman: A Million-scale Chinese Commonsense Knowledge Graph Distilled from Foundation Model
Jiaan Wang, Jianfeng Qu, Yunlong Liang, Zhixu Li, An Liu, Guanfeng Liu, Xin Zheng
cs.CL, cs.AI
Constructing commonsense knowledge graphs (CKGs) has attracted wide research attention due to its significant importance in cognitive intelligence. Nevertheless, existing CKGs are typically oriented to English, limiting the research in non-English languages. Meanwhile, the emergence of foundation models like ChatGPT an...
http://arxiv.org/pdf/2306.10241v1
2023-06-17T02:51:33
2023-06-17T02:51:33
{ "title": [ -0.01876831, -0.027450562, -0.062347412, -0.046966553, -0.013000488, -0.024520874, -0.02923584, 0.016998291, 0.015060425, 0.013298035, -0.039245605, 0.024246216, -0.0158844, -0.018325806, -0.018218994, -0.0038089752, 0.005458832, 0.0...
http://arxiv.org/abs/2306.10231v1
GLIMMER: generalized late-interaction memory reranker
Michiel de Jong, Yury Zemlyanskiy, Nicholas FitzGerald, Sumit Sanghai, William W. Cohen, Joshua Ainslie
cs.CL, cs.AI, cs.LG
Memory-augmentation is a powerful approach for efficiently incorporating external information into language models, but leads to reduced performance relative to retrieving text. Recent work introduced LUMEN, a memory-retrieval hybrid that partially pre-computes memory and updates memory representations on the fly with ...
http://arxiv.org/pdf/2306.10231v1
2023-06-17T01:54:25
2023-06-17T01:54:25
{ "title": [ 0.0141067505, 0.010932922, -0.02079773, -0.078430176, 0.024551392, 0.012023926, -0.02619934, -0.026672363, -0.00844574, 0.027877808, -0.0014600754, 0.008506775, -0.015533447, -0.02885437, -0.028045654, 0.0044555664, -0.0076675415, 0....
http://arxiv.org/abs/2307.05492v1
GPT4 is Slightly Helpful for Peer-Review Assistance: A Pilot Study
Zachary Robertson
cs.HC, cs.AI, cs.CL
In this pilot study, we investigate the use of GPT4 to assist in the peer-review process. Our key hypothesis was that GPT-generated reviews could achieve comparable helpfulness to human reviewers. By comparing reviews generated by both human reviewers and GPT models for academic papers submitted to a major machine lear...
http://arxiv.org/pdf/2307.05492v1
2023-06-16T23:11:06
2023-06-16T23:11:06
{ "title": [ -0.010902405, 0.0042762756, -0.043151855, 0.0095825195, -0.042541504, -0.037902832, -0.028045654, -0.03427124, -0.043701172, 0.033477783, -0.025146484, 0.0027484894, 0.029785156, -0.04083252, 0.008224487, -0.0010156631, -0.009590149, ...
http://arxiv.org/abs/2306.10196v1
Structured Thoughts Automaton: First Formalized Execution Model for Auto-Regressive Language Models
Tristan Vanderbruggen, Chunhua Liao, Peter Pirkelbauer, Pei-Hung Lin
cs.CL, cs.AI, cs.FL, cs.LG
In recent months, Language Models (LMs) have become a part of daily discourse, with focus on OpenAI and the potential of Artificial General Intelligence (AGI). Furthermore, the leaking of LLama's weights to the public has led to an influx of innovations demonstrating the impressive capabilities of generative LMs. While...
http://arxiv.org/pdf/2306.10196v1
2023-06-16T22:04:50
2023-06-16T22:04:50
{ "title": [ -0.05731201, -0.008834839, -0.0289917, -0.04034424, -0.050567627, 0.017120361, -0.10266113, -0.00005686283, 0.0067634583, 0.037322998, -0.01763916, -0.0055656433, 0.03414917, -0.020858765, -0.027740479, 0.010169983, 0.0020160675, 0.0...
http://arxiv.org/abs/2306.10193v1
Conformal Language Modeling
Victor Quach, Adam Fisch, Tal Schuster, Adam Yala, Jae Ho Sohn, Tommi S. Jaakkola, Regina Barzilay
cs.CL, cs.LG
We propose a novel approach to conformal prediction for generative language models (LMs). Standard conformal prediction produces prediction sets -- in place of single predictions -- that have rigorous, statistical performance guarantees. LM responses are typically sampled from the model's predicted distribution over th...
http://arxiv.org/pdf/2306.10193v1
2023-06-16T21:55:08
2023-06-16T21:55:08
{ "title": [ -0.040740967, 0.0046653748, -0.062683105, -0.05847168, -0.032165527, -0.0031967163, -0.09991455, 0.00756073, -0.0007314682, 0.06335449, -0.008293152, -0.02709961, 0.032928467, -0.04498291, -0.038238525, 0.030395508, -0.008110046, 0.0...
http://arxiv.org/abs/2306.10169v1
Meta-Personalizing Vision-Language Models to Find Named Instances in Video
Chun-Hsiao Yeh, Bryan Russell, Josef Sivic, Fabian Caba Heilbron, Simon Jenni
cs.CV, cs.CL, cs.LG
Large-scale vision-language models (VLM) have shown impressive results for language-guided search applications. While these models allow category-level queries, they currently struggle with personalized searches for moments in a video where a specific object instance such as ``My dog Biscuit'' appears. We present the f...
http://arxiv.org/pdf/2306.10169v1
2023-06-16T20:12:11
2023-06-16T20:12:11
{ "title": [ 0.026748657, 0.008148193, -0.030181885, -0.04534912, -0.013671875, -0.027053833, -0.02796936, -0.03704834, -0.004360199, 0.047851562, -0.018875122, 0.007896423, 0.019195557, -0.011878967, -0.028198242, -0.0027389526, 0.01689148, 0.03...
http://arxiv.org/abs/2306.10165v1
Data Selection for Fine-tuning Large Language Models Using Transferred Shapley Values
Stephanie Schoch, Ritwick Mishra, Yangfeng Ji
cs.CL
Although Shapley values have been shown to be highly effective for identifying harmful training instances, dataset size and model complexity constraints limit the ability to apply Shapley-based data valuation to fine-tuning large pre-trained language models. To address this, we propose TS-DShapley, an algorithm that re...
http://arxiv.org/pdf/2306.10165v1
2023-06-16T20:07:38
2023-06-16T20:07:38
{ "title": [ -0.047088623, -0.019500732, -0.06719971, -0.042266846, -0.010292053, -0.008979797, -0.05090332, 0.017318726, 0.013320923, 0.010910034, -0.030456543, 0.0057525635, -0.011962891, -0.013786316, -0.036987305, -0.006034851, -0.027297974, ...
http://arxiv.org/abs/2306.10153v1
Semi-supervised Relation Extraction via Data Augmentation and Consistency-training
Komal K. Teru
cs.CL, cs.IR
Due to the semantic complexity of the Relation extraction (RE) task, obtaining high-quality human labelled data is an expensive and noisy process. To improve the sample efficiency of the models, semi-supervised learning (SSL) methods aim to leverage unlabelled data in addition to learning from limited labelled data poi...
http://arxiv.org/pdf/2306.10153v1
2023-06-16T19:45:42
2023-06-16T19:45:42
{ "title": [ 0.0024795532, -0.0010566711, -0.07543945, -0.08514404, 0.020629883, -0.00894928, -0.04534912, 0.019683838, 0.017333984, 0.0056419373, -0.022720337, 0.0037765503, -0.011459351, -0.009117126, -0.039154053, -0.011940002, -0.018661499, 0...
http://arxiv.org/abs/2306.10015v1
Just One Byte (per gradient): A Note on Low-Bandwidth Decentralized Language Model Finetuning Using Shared Randomness
Eric Zelikman, Qian Huang, Percy Liang, Nick Haber, Noah D. Goodman
cs.LG, cs.CL, cs.DC
Language model training in distributed settings is limited by the communication cost of gradient exchanges. In this short note, we extend recent work from Malladi et al. (2023), using shared randomness to perform distributed fine-tuning with low bandwidth. The method is a natural decentralized extension of memory-effic...
http://arxiv.org/pdf/2306.10015v1
2023-06-16T17:59:51
2023-06-16T17:59:51
{ "title": [ -0.00447464, -0.020446777, -0.023345947, -0.056396484, -0.051879883, -0.0024604797, -0.082458496, 0.01222229, -0.016998291, 0.054779053, -0.0287323, 0.023452759, 0.042022705, -0.017410278, -0.0151901245, 0.009384155, 0.005092621, 0.0...
http://arxiv.org/abs/2306.10012v1
MagicBrush: A Manually Annotated Dataset for Instruction-Guided Image Editing
Kai Zhang, Lingbo Mo, Wenhu Chen, Huan Sun, Yu Su
cs.CV, cs.AI, cs.CL
Text-guided image editing is widely needed in daily life, ranging from personal use to professional applications such as Photoshop. However, existing methods are either zero-shot or trained on an automatically synthesized dataset, which contains a high volume of noise. Thus, they still require lots of manual tuning to ...
http://arxiv.org/pdf/2306.10012v1
2023-06-16T17:58:58
2023-06-16T17:58:58
{ "title": [ -0.023834229, 0.010414124, -0.03643799, -0.030807495, 0.0067710876, -0.0115737915, -0.019424438, -0.03173828, 0.0110321045, 0.03314209, -0.018127441, 0.018753052, 0.004459381, -0.024108887, -0.026885986, 0.0053634644, 0.0076942444, 0...
http://arxiv.org/abs/2306.10098v1
Differentiable Instruction Optimization for Cross-Task Generalization
Masaru Isonuma, Junichiro Mori, Ichiro Sakata
cs.CL, cs.AI, cs.LG
Instruction tuning has been attracting much attention to achieve generalization ability across a wide variety of tasks. Although various types of instructions have been manually created for instruction tuning, it is still unclear what kind of instruction is optimal to obtain cross-task generalization ability. This work...
http://arxiv.org/pdf/2306.10098v1
2023-06-16T17:49:34
2023-06-16T17:49:34
{ "title": [ -0.055236816, -0.0024585724, -0.013519287, -0.027359009, -0.017791748, 0.007133484, -0.033721924, -0.01876831, -0.0047683716, 0.03579712, -0.050079346, 0.043884277, -0.020080566, 0.011238098, -0.00907135, -0.008522034, -0.006084442, ...
http://arxiv.org/abs/2306.09996v1
Investigating Prompting Techniques for Zero- and Few-Shot Visual Question Answering
Rabiul Awal, Le Zhang, Aishwarya Agrawal
cs.CV, cs.CL
Visual question answering (VQA) is a challenging task that requires the ability to comprehend and reason with visual information. While recent vision-language models have made strides, they continue to struggle with zero-shot VQA, particularly in handling complex compositional questions and adapting to new domains i.e....
http://arxiv.org/pdf/2306.09996v1
2023-06-16T17:47:57
2023-06-16T17:47:57
{ "title": [ 0.011749268, -0.01525116, -0.003452301, -0.03161621, -0.022125244, -0.009338379, -0.015098572, -0.0051078796, 0.03024292, 0.008651733, -0.04724121, -0.0014514923, -0.016693115, -0.0093307495, -0.06384277, -0.017440796, 0.031341553, -...
http://arxiv.org/abs/2306.09992v1
Rewriting the Script: Adapting Text Instructions for Voice Interaction
Alyssa Hwang, Natasha Oza, Chris Callison-Burch, Andrew Head
cs.HC, cs.CL
Voice assistants have sharply risen in popularity in recent years, but their use has been limited mostly to simple applications like music, hands-free search, or control of internet-of-things devices. What would it take for voice assistants to guide people through more complex tasks? In our work, we study the limitatio...
http://arxiv.org/pdf/2306.09992v1
2023-06-16T17:43:00
2023-06-16T17:43:00
{ "title": [ -0.026824951, -0.032958984, -0.029693604, -0.06414795, -0.020355225, 0.0075683594, -0.05557251, 0.009651184, 0.00793457, 0.037261963, -0.03161621, 0.0065689087, -0.00093364716, -0.0037174225, -0.060516357, -0.009063721, -0.0010585785, ...
http://arxiv.org/abs/2306.10097v1
CML-TTS A Multilingual Dataset for Speech Synthesis in Low-Resource Languages
Frederico S. Oliveira, Edresson Casanova, Arnaldo Cândido Júnior, Anderson S. Soares, Arlindo R. Galvão Filho
eess.AS, cs.AI, cs.CL
In this paper, we present CML-TTS, a recursive acronym for CML-Multi-Lingual-TTS, a new Text-to-Speech (TTS) dataset developed at the Center of Excellence in Artificial Intelligence (CEIA) of the Federal University of Goias (UFG). CML-TTS is based on Multilingual LibriSpeech (MLS) and adapted for training TTS models, c...
http://arxiv.org/pdf/2306.10097v1
2023-06-16T17:17:06
2023-06-16T17:17:06
{ "title": [ -0.022964478, -0.015563965, -0.04598999, -0.06274414, -0.027999878, -0.0032806396, -0.08874512, 0.01902771, -0.026809692, 0.04248047, -0.022232056, 0.005821228, 0.017059326, -0.029953003, -0.013183594, -0.0060691833, -0.008102417, 0....
http://arxiv.org/abs/2306.09968v1
ClinicalGPT: Large Language Models Finetuned with Diverse Medical Data and Comprehensive Evaluation
Guangyu Wang, Guoxing Yang, Zongxin Du, Longjun Fan, Xiaohu Li
cs.CL
Large language models have exhibited exceptional performance on various Natural Language Processing (NLP) tasks, leveraging techniques such as the pre-training, and instruction fine-tuning. Despite these advances, their effectiveness in medical applications is limited, due to challenges such as factual inaccuracies, re...
http://arxiv.org/pdf/2306.09968v1
2023-06-16T16:56:32
2023-06-16T16:56:32
{ "title": [ -0.018051147, -0.015060425, -0.056610107, -0.081970215, 0.018997192, -0.0004272461, -0.06274414, -0.013343811, -0.013946533, 0.025970459, 0.0066337585, 0.0026435852, 0.00064086914, -0.038879395, -0.022155762, 0.01486969, -0.0047302246, ...
http://arxiv.org/abs/2307.00008v1
Investigating Masking-based Data Generation in Language Models
Ed S. Ma
cs.CL, cs.AI
The current era of natural language processing (NLP) has been defined by the prominence of pre-trained language models since the advent of BERT. A feature of BERT and models with similar architecture is the objective of masked language modeling, in which part of the input is intentionally masked and the model is traine...
http://arxiv.org/pdf/2307.00008v1
2023-06-16T16:48:27
2023-06-16T16:48:27
{ "title": [ -0.0569458, -0.001074791, -0.07525635, -0.05328369, -0.008468628, 0.0038204193, -0.08404541, 0.01713562, 0.0051345825, 0.028289795, -0.010787964, -0.00010830164, 0.006500244, -0.018417358, -0.04989624, 0.013092041, -0.030853271, 0.06...
http://arxiv.org/abs/2306.10095v1
AD-AutoGPT: An Autonomous GPT for Alzheimer's Disease Infodemiology
Haixing Dai, Yiwei Li, Zhengliang Liu, Lin Zhao, Zihao Wu, Suhang Song, Ye Shen, Dajiang Zhu, Xiang Li, Sheng Li, Xiaobai Yao, Lu Shi, Quanzheng Li, Zhuo Chen, Donglan Zhang, Gengchen Mai, Tianming Liu
cs.CL, cs.AI, cs.IR, 68T01, 68T50, 92C50, I.2.7; I.2.1; J.3
In this pioneering study, inspired by AutoGPT, the state-of-the-art open-source application based on the GPT-4 large language model, we develop a novel tool called AD-AutoGPT which can conduct data collection, processing, and analysis about complex health narratives of Alzheimer's Disease in an autonomous manner via us...
http://arxiv.org/pdf/2306.10095v1
2023-06-16T16:35:59
2023-06-16T16:35:59
{ "title": [ 0.016967773, -0.0035495758, -0.030380249, -0.062408447, 0.0017690659, -0.013145447, -0.024658203, -0.013206482, -0.024734497, 0.013648987, -0.031143188, 0.021972656, -0.0029335022, -0.034423828, 0.011428833, 0.0006351471, 0.0093688965, ...
http://arxiv.org/abs/2306.09927v3
Trained Transformers Learn Linear Models In-Context
Ruiqi Zhang, Spencer Frei, Peter L. Bartlett
stat.ML, cs.AI, cs.CL, cs.LG
Attention-based neural networks such as transformers have demonstrated a remarkable ability to exhibit in-context learning (ICL): Given a short prompt sequence of tokens from an unseen task, they can formulate relevant per-token and next-token predictions without any parameter updates. By embedding a sequence of labele...
http://arxiv.org/pdf/2306.09927v3
2023-10-19T20:31:32
2023-06-16T15:50:03
{ "title": [ 0.037994385, 0.014854431, -0.005367279, -0.029449463, -0.025619507, -0.028076172, -0.03161621, -0.029037476, -0.05331421, 0.026016235, -0.055023193, 0.004737854, -0.014877319, -0.028808594, 0.031219482, -0.03439331, 0.021255493, 0.02...
http://arxiv.org/abs/2306.09922v1
Learning to Summarize and Answer Questions about a Virtual Robot's Past Actions
Chad DeChant, Iretiayo Akinola, Daniel Bauer
cs.RO, cs.AI, cs.CL, cs.LG
When robots perform long action sequences, users will want to easily and reliably find out what they have done. We therefore demonstrate the task of learning to summarize and answer questions about a robot agent's past actions using natural language alone. A single system with a large language model at its core is trai...
http://arxiv.org/pdf/2306.09922v1
2023-06-16T15:47:24
2023-06-16T15:47:24
{ "title": [ 0.01550293, 0.0013494492, -0.023590088, -0.034851074, -0.017303467, -0.025741577, -0.019378662, -0.0077781677, -0.011810303, 0.030883789, -0.038024902, 0.022644043, -0.019897461, -0.0043563843, -0.025817871, -0.028656006, 0.025238037, ...
http://arxiv.org/abs/2306.09918v1
No Strong Feelings One Way or Another: Re-operationalizing Neutrality in Natural Language Inference
Animesh Nighojkar, Antonio Laverghetta Jr., John Licato
cs.CL, cs.AI
Natural Language Inference (NLI) has been a cornerstone task in evaluating language models' inferential reasoning capabilities. However, the standard three-way classification scheme used in NLI has well-known shortcomings in evaluating models' ability to capture the nuances of natural human reasoning. In this paper, we...
http://arxiv.org/pdf/2306.09918v1
2023-06-16T15:45:08
2023-06-16T15:45:08
{ "title": [ -0.011062622, -0.0058670044, -0.019943237, -0.061553955, 0.002653122, 0.00059747696, -0.034973145, -0.011505127, 0.005809784, 0.022964478, -0.032836914, -0.0012159348, 0.0020484924, -0.0031642914, -0.05706787, -0.010894775, -0.017166138,...
http://arxiv.org/abs/2306.09896v4
Is Self-Repair a Silver Bullet for Code Generation?
Theo X. Olausson, Jeevana Priya Inala, Chenglong Wang, Jianfeng Gao, Armando Solar-Lezama
cs.CL, cs.AI, cs.PL, cs.SE
Large language models have shown remarkable aptitude in code generation, but still struggle on challenging tasks. Self-repair -- in which the model debugs and fixes mistakes in its own code -- has recently become a popular way to boost performance in these settings. However, only very limited studies on how and when se...
http://arxiv.org/pdf/2306.09896v4
2023-10-17T17:51:27
2023-06-16T15:13:17
{ "title": [ 0.008651733, -0.017105103, -0.0362854, -0.06945801, -0.041503906, -0.017990112, -0.05343628, 0.027664185, -0.0569458, 0.026367188, 0.011726379, -0.005558014, -0.018447876, -0.05532837, 0.009979248, -0.029953003, 0.024398804, 0.010017...
http://arxiv.org/abs/2306.09877v1
Revealing the impact of social circumstances on the selection of cancer therapy through natural language processing of social work notes
Shenghuan Sun, Travis Zack, Christopher Y. K. Williams, Atul J. Butte, Madhumita Sushil
cs.CL
We aimed to investigate the impact of social circumstances on cancer therapy selection using natural language processing to derive insights from social worker documentation. We developed and employed a Bidirectional Encoder Representations from Transformers (BERT) based approach, using a hierarchical multi-step BERT mo...
http://arxiv.org/pdf/2306.09877v1
2023-06-16T14:40:39
2023-06-16T14:40:39
{ "title": [ 0.02609253, -0.016845703, 0.007030487, -0.06390381, -0.0178833, -0.0022182465, -0.026290894, 0.009750366, -0.022079468, 0.0061187744, -0.042938232, 0.022125244, -0.025558472, -0.010925293, 0.0017766953, -0.039245605, -0.009147644, 0....
http://arxiv.org/abs/2306.09869v3
Energy-Based Cross Attention for Bayesian Context Update in Text-to-Image Diffusion Models
Geon Yeong Park, Jeongsol Kim, Beomsu Kim, Sang Wan Lee, Jong Chul Ye
cs.CV, cs.AI, cs.CL, cs.LG
Despite the remarkable performance of text-to-image diffusion models in image generation tasks, recent studies have raised the issue that generated images sometimes cannot capture the intended semantic contents of the text prompts, which phenomenon is often called semantic misalignment. To address this, here we present...
http://arxiv.org/pdf/2306.09869v3
2023-11-04T18:18:10
2023-06-16T14:30:41
{ "title": [ 0.011512756, 0.02406311, -0.020599365, -0.026367188, 0.021240234, -0.015975952, 0.015975952, -0.010810852, 0.04071045, 0.053741455, -0.037017822, 0.014724731, 0.05316162, -0.008201599, -0.026138306, 0.024963379, 0.0025177002, 0.02903...
http://arxiv.org/abs/2306.09841v3
Are Large Language Models Really Good Logical Reasoners? A Comprehensive Evaluation and Beyond
Fangzhi Xu, Qika Lin, Jiawei Han, Tianzhe Zhao, Jun Liu, Erik Cambria
cs.CL, cs.AI
Logical reasoning consistently plays a fundamental and significant role in the domains of knowledge engineering and artificial intelligence. Recently, Large Language Models (LLMs) have emerged as a noteworthy innovation in natural language processing (NLP), exhibiting impressive achievements across various classic NLP ...
http://arxiv.org/pdf/2306.09841v3
2023-08-08T12:57:18
2023-06-16T13:39:35
{ "title": [ -0.020446777, -0.013069153, -0.035888672, -0.087646484, 0.00491333, 0.010276794, -0.0713501, 0.005973816, -0.0048103333, 0.021224976, -0.020141602, -0.006416321, 0.007598877, -0.018447876, -0.038909912, 0.001824379, -0.024658203, 0.0...
http://arxiv.org/abs/2306.10090v1
Improving Audio Caption Fluency with Automatic Error Correction
Hanxue Zhang, Zeyu Xie, Xuenan Xu, Mengyue Wu, Kai Yu
eess.AS, cs.AI, cs.CL, cs.SD
Automated audio captioning (AAC) is an important cross-modality translation task, aiming at generating descriptions for audio clips. However, captions generated by previous AAC models have faced ``false-repetition'' errors due to the training objective. In such scenarios, we propose a new task of AAC error correction a...
http://arxiv.org/pdf/2306.10090v1
2023-06-16T13:37:01
2023-06-16T13:37:01
{ "title": [ -0.000019431114, 0.008140564, -0.034484863, -0.035583496, 0.004055023, -0.019058228, -0.0056762695, -0.0040893555, -0.0034637451, 0.0703125, -0.023208618, 0.0019779205, 0.0236969, 0.0007753372, -0.06304932, 0.005622864, -0.013290405, ...
http://arxiv.org/abs/2306.09830v1
Sheffield's Submission to the AmericasNLP Shared Task on Machine Translation into Indigenous Languages
Edward Gow-Smith, Danae Sánchez Villegas
cs.CL
In this paper we describe the University of Sheffield's submission to the AmericasNLP 2023 Shared Task on Machine Translation into Indigenous Languages which comprises the translation from Spanish to eleven indigenous languages. Our approach consists of extending, training, and ensembling different variations of NLLB-2...
http://arxiv.org/pdf/2306.09830v1
2023-06-16T13:15:26
2023-06-16T13:15:26
{ "title": [ -0.024169922, -0.013160706, 0.0036506653, 0.011108398, -0.043792725, 0.0124435425, -0.012283325, -0.012496948, -0.037078857, 0.062316895, -0.011108398, 0.00630188, 0.044799805, -0.027038574, 0.00049495697, -0.011154175, -0.008872986, ...
http://arxiv.org/abs/2306.09824v1
Process Knowledge-infused Learning for Clinician-friendly Explanations
Kaushik Roy, Yuxin Zi, Manas Gaur, Jinendra Malekar, Qi Zhang, Vignesh Narayanan, Amit Sheth
cs.CL, cs.AI
Language models have the potential to assess mental health using social media data. By analyzing online posts and conversations, these models can detect patterns indicating mental health conditions like depression, anxiety, or suicidal thoughts. They examine keywords, language markers, and sentiment to gain insights in...
http://arxiv.org/pdf/2306.09824v1
2023-06-16T13:08:17
2023-06-16T13:08:17
{ "title": [ 0.0069770813, -0.026916504, -0.026504517, -0.08807373, -0.014564514, -0.011390686, -0.05307007, 0.004611969, -0.026443481, 0.0054855347, 0.009429932, -0.0036201477, -0.018844604, -0.052490234, -0.0022220612, -0.0008864403, 0.022903442, ...
http://arxiv.org/abs/2306.09821v2
Unlocking the Potential of User Feedback: Leveraging Large Language Model as User Simulator to Enhance Dialogue System
Zhiyuan Hu, Yue Feng, Anh Tuan Luu, Bryan Hooi, Aldo Lipani
cs.CL
Dialogue systems and large language models (LLMs) have gained considerable attention. However, the direct utilization of LLMs as task-oriented dialogue (TOD) models has been found to underperform compared to smaller task-specific models. Nonetheless, it is crucial to acknowledge the significant potential of LLMs and ex...
http://arxiv.org/pdf/2306.09821v2
2023-10-19T16:51:06
2023-06-16T13:04:56
{ "title": [ -0.016494751, -0.012397766, -0.02684021, -0.04534912, -0.014854431, -0.009498596, -0.055236816, -0.013175964, -0.0001026392, 0.033050537, -0.0020103455, -0.0035037994, 0.008171082, -0.023269653, -0.048706055, -0.0016889572, -0.012138367,...
http://arxiv.org/abs/2306.09814v1
Investigating the Utility of Surprisal from Large Language Models for Speech Synthesis Prosody
Sofoklis Kakouros, Juraj Šimko, Martti Vainio, Antti Suni
eess.AS, cs.CL
This paper investigates the use of word surprisal, a measure of the predictability of a word in a given context, as a feature to aid speech synthesis prosody. We explore how word surprisal extracted from large language models (LLMs) correlates with word prominence, a signal-based measure of the salience of a word in a ...
http://arxiv.org/pdf/2306.09814v1
2023-06-16T12:49:44
2023-06-16T12:49:44
{ "title": [ -0.022979736, -0.004108429, -0.035491943, -0.06903076, -0.017150879, -0.0027370453, -0.07196045, 0.02128601, -0.020690918, 0.029922485, 0.0023422241, -0.01361084, 0.01991272, -0.0496521, -0.0028800964, -0.0033931732, -0.013587952, 0....
http://arxiv.org/abs/2306.09802v2
RED$^{\rm FM}$: a Filtered and Multilingual Relation Extraction Dataset
Pere-Lluís Huguet Cabot, Simone Tedeschi, Axel-Cyrille Ngonga Ngomo, Roberto Navigli
cs.CL
Relation Extraction (RE) is a task that identifies relationships between entities in a text, enabling the acquisition of relational facts and bridging the gap between natural language and structured knowledge. However, current RE models often rely on small datasets with low coverage of relation types, particularly when...
http://arxiv.org/pdf/2306.09802v2
2023-06-19T09:25:27
2023-06-16T12:29:59
{ "title": [ -0.018066406, -0.012878418, -0.046569824, -0.08087158, 0.015113831, 0.01083374, -0.059417725, 0.00017786026, 0.03265381, 0.011940002, 0.0053901672, -0.0013360977, -0.010093689, -0.022720337, -0.044677734, 0.01171875, -0.015716553, 0....
http://arxiv.org/abs/2306.09782v1
Full Parameter Fine-tuning for Large Language Models with Limited Resources
Kai Lv, Yuqing Yang, Tengxiao Liu, Qinghui Gao, Qipeng Guo, Xipeng Qiu
cs.CL
Large Language Models (LLMs) have revolutionized Natural Language Processing (NLP) but demand massive GPU resources for training. Lowering the threshold for LLMs training would encourage greater participation from researchers, benefiting both academia and society. While existing approaches have focused on parameter-eff...
http://arxiv.org/pdf/2306.09782v1
2023-06-16T11:37:15
2023-06-16T11:37:15
{ "title": [ -0.02720642, -0.021194458, -0.027175903, -0.09289551, -0.008773804, 0.016601562, -0.07141113, 0.023910522, -0.00084781647, 0.027664185, -0.008140564, -0.005279541, 0.0020256042, -0.016464233, -0.044952393, 0.010093689, -0.009109497, ...
http://arxiv.org/abs/2306.09752v1
Politeness Stereotypes and Attack Vectors: Gender Stereotypes in Japanese and Korean Language Models
Victor Steinborn, Antonis Maronikolakis, Hinrich Schütze
cs.CL, cs.AI, cs.CY, cs.LG
In efforts to keep up with the rapid progress and use of large language models, gender bias research is becoming more prevalent in NLP. Non-English bias research, however, is still in its infancy with most work focusing on English. In our work, we study how grammatical gender bias relating to politeness levels manifest...
http://arxiv.org/pdf/2306.09752v1
2023-06-16T10:36:18
2023-06-16T10:36:18
{ "title": [ -0.02583313, 0.000028133392, -0.032958984, -0.018722534, -0.025634766, -0.022583008, -0.019989014, -0.008895874, -0.034332275, 0.041137695, -0.021118164, 0.009346008, 0.013298035, -0.017868042, -0.04031372, -0.01852417, -0.050933838, ...
http://arxiv.org/abs/2306.09737v1
Using Natural Language Processing and Networks to Automate Structured Literature Reviews: An Application to Farmers Climate Change Adaptation
Sofia Gil-Clavel, Tatiana Filatova
cs.CL
The fast-growing number of research articles makes it problematic for scholars to keep track of the new findings related to their areas of expertise. Furthermore, linking knowledge across disciplines in rapidly developing fields becomes challenging for complex topics like climate change that demand interdisciplinary so...
http://arxiv.org/pdf/2306.09737v1
2023-06-16T10:05:47
2023-06-16T10:05:47
{ "title": [ -0.053497314, -0.009475708, -0.035095215, -0.02180481, -0.030578613, 0.003255844, -0.052856445, -0.014266968, -0.039764404, 0.031982422, -0.030319214, 0.015716553, -0.025924683, -0.003665924, -0.05291748, -0.03213501, -0.005382538, 0...
http://arxiv.org/abs/2306.09725v1
Discourse Representation Structure Parsing for Chinese
Chunliu Wang, Xiao Zhang, Johan Bos
cs.CL
Previous work has predominantly focused on monolingual English semantic parsing. We, instead, explore the feasibility of Chinese semantic parsing in the absence of labeled data for Chinese meaning representations. We describe the pipeline of automatically collecting the linearized Chinese meaning representation data fo...
http://arxiv.org/pdf/2306.09725v1
2023-06-16T09:47:45
2023-06-16T09:47:45
{ "title": [ -0.01789856, -0.0001629591, -0.009223938, -0.07165527, -0.009895325, 0.018005371, -0.070007324, 0.008163452, -0.0065231323, 0.022033691, -0.011581421, -0.0062446594, -0.0039787292, -0.025039673, -0.032226562, -0.0022945404, -0.016281128,...
http://arxiv.org/abs/2306.09719v2
Pushing the Limits of ChatGPT on NLP Tasks
Xiaofei Sun, Linfeng Dong, Xiaoya Li, Zhen Wan, Shuhe Wang, Tianwei Zhang, Jiwei Li, Fei Cheng, Lingjuan Lyu, Fei Wu, Guoyin Wang
cs.CL, cs.AI
Despite the success of ChatGPT, its performances on most NLP tasks are still well below the supervised baselines. In this work, we looked into the causes, and discovered that its subpar performance was caused by the following factors: (1) token limit in the prompt does not allow for the full utilization of the supervis...
http://arxiv.org/pdf/2306.09719v2
2023-10-09T15:48:23
2023-06-16T09:40:05
{ "title": [ -0.0065307617, -0.020828247, -0.014823914, -0.042755127, -0.012557983, 0.006526947, -0.017074585, -0.0062789917, 0.005393982, 0.023269653, -0.03074646, 0.016235352, -0.016464233, -0.019119263, -0.026885986, 0.020370483, -0.0025920868, ...
http://arxiv.org/abs/2306.09712v1
Semi-Offline Reinforcement Learning for Optimized Text Generation
Changyu Chen, Xiting Wang, Yiqiao Jin, Victor Ye Dong, Li Dong, Jie Cao, Yi Liu, Rui Yan
cs.LG, cs.AI, cs.CL
In reinforcement learning (RL), there are two major settings for interacting with the environment: online and offline. Online methods explore the environment at significant time cost, and offline methods efficiently obtain reward signals by sacrificing exploration capability. We propose semi-offline RL, a novel paradig...
http://arxiv.org/pdf/2306.09712v1
2023-06-16T09:24:29
2023-06-16T09:24:29
{ "title": [ -0.028808594, -0.010887146, -0.02458191, -0.032470703, -0.0044822693, 0.0012626648, -0.03491211, -0.00012040138, -0.019882202, 0.02746582, -0.002916336, 0.0063972473, -0.013809204, -0.01625061, -0.027862549, -0.016296387, -0.023544312, ...
http://arxiv.org/abs/2306.09705v1
Reducing Computational Costs in Sentiment Analysis: Tensorized Recurrent Networks vs. Recurrent Networks
Gabriel Lopez, Anna Nguyen, Joe Kaul
cs.LG, cs.CL
Anticipating audience reaction towards a certain text is integral to several facets of society ranging from politics, research, and commercial industries. Sentiment analysis (SA) is a useful natural language processing (NLP) technique that utilizes lexical/statistical and deep learning methods to determine whether diff...
http://arxiv.org/pdf/2306.09705v1
2023-06-16T09:18:08
2023-06-16T09:18:08
{ "title": [ -0.004196167, -0.009422302, -0.011459351, -0.0390625, -0.0052948, -0.004337311, -0.019638062, -0.014785767, 0.000020384789, 0.047973633, -0.015541077, 0.008224487, 0.0032787323, -0.007381439, -0.054504395, 0.0124435425, -0.006477356, ...
http://arxiv.org/abs/2306.09704v2
Cross-corpus Readability Compatibility Assessment for English Texts
Zhenzhen Li, Han Ding, Shaohong Zhang
cs.CL
Text readability assessment has gained significant attention from researchers in various domains. However, the lack of exploration into corpus compatibility poses a challenge as different research groups utilize different corpora. In this study, we propose a novel evaluation framework, Cross-corpus text Readability Com...
http://arxiv.org/pdf/2306.09704v2
2023-09-13T09:43:28
2023-06-16T09:15:39
{ "title": [ -0.02999878, 0.015991211, -0.025009155, -0.024978638, 0.008926392, 0.0068626404, -0.031234741, -0.03286743, 0.0284729, 0.036743164, -0.040252686, -0.012939453, 0.029754639, -0.004878998, -0.058746338, 0.0057754517, -0.0234375, 0.0537...
http://arxiv.org/abs/2306.09697v1
Class-Adaptive Self-Training for Relation Extraction with Incompletely Annotated Training Data
Qingyu Tan, Lu Xu, Lidong Bing, Hwee Tou Ng
cs.CL
Relation extraction (RE) aims to extract relations from sentences and documents. Existing relation extraction models typically rely on supervised machine learning. However, recent studies showed that many RE datasets are incompletely annotated. This is known as the false negative problem in which valid relations are fa...
http://arxiv.org/pdf/2306.09697v1
2023-06-16T09:01:45
2023-06-16T09:01:45
{ "title": [ -0.009651184, 0.007133484, -0.055999756, -0.057037354, 0.00793457, -0.0039024353, -0.02960205, -0.0054512024, 0.018798828, 0.009414673, -0.0039634705, 0.0026855469, -0.03942871, -0.012832642, -0.054229736, -0.0042152405, 0.0070266724, ...
http://arxiv.org/abs/2306.09657v1
Online Distillation for Pseudo-Relevance Feedback
Sean MacAvaney, Xi Wang
cs.IR, cs.CL
Model distillation has emerged as a prominent technique to improve neural search models. To date, distillation taken an offline approach, wherein a new neural model is trained to predict relevance scores between arbitrary queries and documents. In this paper, we explore a departure from this offline distillation strate...
http://arxiv.org/pdf/2306.09657v1
2023-06-16T07:26:33
2023-06-16T07:26:33
{ "title": [ 0.0065345764, 0.0021400452, -0.023620605, -0.07897949, 0.0138549805, 0.0025939941, -0.03677368, -0.010688782, 0.050628662, 0.03918457, -0.026931763, -0.038635254, 0.014785767, -0.018432617, -0.06549072, -0.0054092407, 0.013633728, 0....
http://arxiv.org/abs/2306.09649v1
ReactGenie: An Object-Oriented State Abstraction for Complex Multimodal Interactions Using Large Language Models
Jackie, Yang, Karina Li, Daniel Wan Rosli, Shuning Zhang, Yuhan Zhang, Monica S. Lam, James A. Landay
cs.HC, cs.CL
Multimodal interactions have been shown to be more flexible, efficient, and adaptable for diverse users and tasks than traditional graphical interfaces. However, existing multimodal development frameworks either do not handle the complexity and compositionality of multimodal commands well or require developers to write...
http://arxiv.org/pdf/2306.09649v1
2023-06-16T06:53:26
2023-06-16T06:53:26
{ "title": [ -0.013961792, -0.010902405, -0.049865723, -0.09692383, 0.012489319, 0.00056552887, -0.077819824, 0.015609741, 0.0033435822, 0.036499023, 0.004699707, 0.008003235, 0.006149292, -0.028625488, -0.036590576, -0.00063943863, -0.017227173, ...
http://arxiv.org/abs/2306.09642v1
Cross-Domain Toxic Spans Detection
Stefan F. Schouten, Baran Barbarestani, Wondimagegnhue Tufa, Piek Vossen, Ilia Markov
cs.CL, cs.LG
Given the dynamic nature of toxic language use, automated methods for detecting toxic spans are likely to encounter distributional shift. To explore this phenomenon, we evaluate three approaches for detecting toxic spans under cross-domain conditions: lexicon-based, rationale extraction, and fine-tuned language models....
http://arxiv.org/pdf/2306.09642v1
2023-06-16T06:10:00
2023-06-16T06:10:00
{ "title": [ 0.03186035, 0.01689148, -0.025558472, -0.07141113, 0.020065308, -0.027191162, -0.019989014, -0.012504578, -0.024017334, 0.019821167, -0.020736694, 0.006679535, 0.010986328, -0.034820557, 0.02784729, -0.0058059692, -0.013435364, 0.052...
http://arxiv.org/abs/2306.09631v1
AUGUST: an Automatic Generation Understudy for Synthesizing Conversational Recommendation Datasets
Yu Lu, Junwei Bao, Zichen Ma, Xiaoguang Han, Youzheng Wu, Shuguang Cui, Xiaodong He
cs.CL
High-quality data is essential for conversational recommendation systems and serves as the cornerstone of the network architecture development and training strategy design. Existing works contribute heavy human efforts to manually labeling or designing and extending recommender dialogue templates. However, they suffer ...
http://arxiv.org/pdf/2306.09631v1
2023-06-16T05:27:14
2023-06-16T05:27:14
{ "title": [ -0.007282257, 0.0025024414, -0.04107666, -0.057128906, -0.0031642914, -0.00793457, -0.05908203, -0.003660202, 0.01222229, 0.032348633, -0.028823853, 0.036865234, -0.010238647, -0.003145218, -0.043029785, 0.00970459, -0.016433716, 0.0...
http://arxiv.org/abs/2306.09607v1
Listener Model for the PhotoBook Referential Game with CLIPScores as Implicit Reference Chain
Shih-Lun Wu, Yi-Hui Chou, Liangze Li
cs.CL
PhotoBook is a collaborative dialogue game where two players receive private, partially-overlapping sets of images and resolve which images they have in common. It presents machines with a great challenge to learn how people build common ground around multimodal context to communicate effectively. Methods developed in ...
http://arxiv.org/pdf/2306.09607v1
2023-06-16T03:41:14
2023-06-16T03:41:14
{ "title": [ 0.03012085, -0.005493164, 0.0011310577, -0.018753052, -0.048736572, -0.009613037, -0.019714355, -0.024932861, 0.018295288, 0.023727417, -0.0107803345, -0.012878418, -0.01651001, -0.03152466, -0.036956787, -0.030899048, -0.006828308, ...
http://arxiv.org/abs/2306.09597v2
Clickbait Detection via Large Language Models
Han Wang, Yi Zhu, Ye Wang, Yun Li, Yunhao Yuan, Jipeng Qiang
cs.CL, cs.AI
Clickbait, which aims to induce users with some surprising and even thrilling headlines for increasing click-through rates, permeates almost all online content publishers, such as news portals and social media. Recently, Large Language Models (LLMs) have emerged as a powerful instrument and achieved tremendous success ...
http://arxiv.org/pdf/2306.09597v2
2023-10-15T09:32:15
2023-06-16T02:49:20
{ "title": [ 0.008033752, -0.008598328, -0.008308411, -0.039093018, 0.006023407, -0.01071167, -0.022277832, -0.017730713, -0.0014038086, 0.03768921, -0.021743774, 0.010231018, 0.012672424, -0.0073890686, -0.03881836, 0.0030097961, -0.026687622, 0...
http://arxiv.org/abs/2306.09594v1
CMLM-CSE: Based on Conditional MLM Contrastive Learning for Sentence Embeddings
Wei Zhang, Xu Chen
cs.CL, cs.AI
Traditional comparative learning sentence embedding directly uses the encoder to extract sentence features, and then passes in the comparative loss function for learning. However, this method pays too much attention to the sentence body and ignores the influence of some words in the sentence on the sentence semantics. ...
http://arxiv.org/pdf/2306.09594v1
2023-06-16T02:39:45
2023-06-16T02:39:45
{ "title": [ -0.0063819885, -0.0017976761, -0.02909851, -0.044952393, -0.007858276, -0.009033203, -0.03289795, -0.007068634, 0.0035095215, 0.020812988, -0.02017212, 0.043273926, -0.004386902, -0.018966675, -0.021499634, 0.0050811768, -0.027862549, ...
http://arxiv.org/abs/2306.14907v1
Clickbait Classification and Spoiling Using Natural Language Processing
Adhitya Thirumala, Elisa Ferracane
cs.CL, I.2.7
Clickbait is the practice of engineering titles to incentivize readers to click through to articles. Such titles with sensationalized language reveal as little information as possible. Occasionally, clickbait will be intentionally misleading, so natural language processing (NLP) can scan the article and answer the ques...
http://arxiv.org/pdf/2306.14907v1
2023-06-16T01:45:57
2023-06-16T01:45:57
{ "title": [ -0.00048661232, -0.0178833, -0.011550903, -0.042816162, -0.0036582947, -0.000042557716, -0.008781433, -0.010643005, -0.013687134, 0.042144775, -0.030197144, 0.016113281, 0.019500732, -0.0010290146, -0.029907227, -0.004711151, -0.02418518...
http://arxiv.org/abs/2306.09572v1
How do different tokenizers perform on downstream tasks in scriptio continua languages?: A case study in Japanese
Takuro Fujii, Koki Shibata, Atsuki Yamaguchi, Terufumi Morishita, Yasuhiro Sogawa
cs.CL, cs.AI
This paper investigates the effect of tokenizers on the downstream performance of pretrained language models (PLMs) in scriptio continua languages where no explicit spaces exist between words, using Japanese as a case study. The tokenizer for such languages often consists of a morphological analyzer and a subword token...
http://arxiv.org/pdf/2306.09572v1
2023-06-16T01:22:32
2023-06-16T01:22:32
{ "title": [ -0.038269043, -0.015571594, -0.04547119, -0.064819336, -0.016784668, 0.0039138794, -0.08935547, 0.018692017, -0.011276245, 0.041381836, -0.04397583, 0.020385742, 0.012435913, -0.004135132, -0.04699707, -0.033416748, -0.023345947, 0.0...
http://arxiv.org/abs/2306.09562v1
Reproducibility in NLP: What Have We Learned from the Checklist?
Ian Magnusson, Noah A. Smith, Jesse Dodge
cs.CL
Scientific progress in NLP rests on the reproducibility of researchers' claims. The *CL conferences created the NLP Reproducibility Checklist in 2020 to be completed by authors at submission to remind them of key information to include. We provide the first analysis of the Checklist by examining 10,405 anonymous respon...
http://arxiv.org/pdf/2306.09562v1
2023-06-16T00:39:25
2023-06-16T00:39:25
{ "title": [ -0.021087646, -0.010261536, -0.027374268, -0.0473938, 0.0034770966, 0.0038528442, -0.039642334, -0.022003174, -0.009513855, 0.011772156, -0.033996582, 0.018127441, -0.017791748, -0.01687622, -0.0357666, 0.0015115738, -0.033569336, 0....
http://arxiv.org/abs/2307.01201v1
Schema-learning and rebinding as mechanisms of in-context learning and emergence
Sivaramakrishnan Swaminathan, Antoine Dedieu, Rajkumar Vasudeva Raju, Murray Shanahan, Miguel Lazaro-Gredilla, Dileep George
cs.CL, cs.AI
In-context learning (ICL) is one of the most powerful and most unexpected capabilities to emerge in recent transformer-based large language models (LLMs). Yet the mechanisms that underlie it are poorly understood. In this paper, we demonstrate that comparable ICL capabilities can be acquired by an alternative sequence ...
http://arxiv.org/pdf/2307.01201v1
2023-06-16T00:29:19
2023-06-16T00:29:19
{ "title": [ -0.008407593, 0.0018663406, -0.027450562, -0.030456543, 0.0070648193, -0.02218628, -0.00089740753, -0.021514893, -0.004497528, 0.0423584, -0.0038585663, 0.0004067421, 0.009056091, -0.040496826, -0.030639648, 0.01902771, 0.014030457, ...
http://arxiv.org/abs/2306.09544v1
Building blocks for complex tasks: Robust generative event extraction for radiology reports under domain shifts
Sitong Zhou, Meliha Yetisgen, Mari Ostendorf
cs.CL
This paper explores methods for extracting information from radiology reports that generalize across exam modalities to reduce requirements for annotated data. We demonstrate that multi-pass T5-based text-to-text generative models exhibit better generalization across exam modalities compared to approaches that employ B...
http://arxiv.org/pdf/2306.09544v1
2023-06-15T23:16:58
2023-06-15T23:16:58
{ "title": [ 0.005607605, 0.019241333, -0.057922363, -0.07110596, 0.033081055, -0.015975952, -0.04623413, -0.016113281, 0.0049934387, 0.015686035, 0.01965332, 0.01235199, -0.027740479, -0.042633057, -0.03918457, 0.012298584, -0.010223389, 0.00315...
http://arxiv.org/abs/2306.09539v4
Block-State Transformers
Mahan Fathi, Jonathan Pilault, Orhan Firat, Christopher Pal, Pierre-Luc Bacon, Ross Goroshin
cs.CL, cs.LG
State space models (SSMs) have shown impressive results on tasks that require modeling long-range dependencies and efficiently scale to long sequences owing to their subquadratic runtime complexity. Originally designed for continuous signals, SSMs have shown superior performance on a plethora of tasks, in vision and au...
http://arxiv.org/pdf/2306.09539v4
2023-10-30T15:44:04
2023-06-15T22:48:08
{ "title": [ 0.034942627, 0.01878357, -0.024398804, -0.064697266, -0.014709473, -0.007873535, -0.05807495, -0.016220093, -0.006084442, 0.057037354, -0.030471802, -0.005695343, 0.03186035, -0.03390503, 0.013191223, 0.02696228, 0.027374268, 0.00531...
http://arxiv.org/abs/2306.10073v1
Thrilled by Your Progress! Large Language Models (GPT-4) No Longer Struggle to Pass Assessments in Higher Education Programming Courses
Jaromir Savelka, Arav Agarwal, Marshall An, Chris Bogart, Majd Sakr
cs.CY, cs.AI, cs.CL, cs.SE
This paper studies recent developments in large language models' (LLM) abilities to pass assessments in introductory and intermediate Python programming courses at the postsecondary level. The emergence of ChatGPT resulted in heated debates of its potential uses (e.g., exercise generation, code explanation) as well as ...
http://arxiv.org/pdf/2306.10073v1
2023-06-15T22:12:34
2023-06-15T22:12:34
{ "title": [ 0.013587952, -0.02305603, 0.01637268, 0.0061836243, -0.022155762, -0.0070724487, 0.0030174255, -0.021652222, 0.04647827, 0.020507812, 0.00031256676, -0.022994995, 0.001876831, -0.05255127, -0.01687622, -0.014274597, -0.017745972, -0....
http://arxiv.org/abs/2306.09525v2
Explaining Legal Concepts with Augmented Large Language Models (GPT-4)
Jaromir Savelka, Kevin D. Ashley, Morgan A. Gray, Hannes Westermann, Huihui Xu
cs.CL, cs.AI
Interpreting the meaning of legal open-textured terms is a key task of legal professionals. An important source for this interpretation is how the term was applied in previous court cases. In this paper, we evaluate the performance of GPT-4 in generating factually accurate, clear and relevant explanations of terms in l...
http://arxiv.org/pdf/2306.09525v2
2023-06-22T11:30:13
2023-06-15T21:58:18
{ "title": [ -0.019104004, -0.009628296, -0.010795593, -0.04598999, -0.0065574646, 0.007980347, -0.035186768, -0.016830444, 0.0069999695, 0.014251709, -0.028213501, 0.006454468, 0.0017251968, -0.01576233, -0.025787354, 0.0023536682, -0.027420044, ...
http://arxiv.org/abs/2306.09519v1
Relation-Aware Network with Attention-Based Loss for Few-Shot Knowledge Graph Completion
Qiao Qiao, Yuepei Li, Kang Zhou, Qi Li
cs.CL, cs.AI, cs.LG
Few-shot knowledge graph completion (FKGC) task aims to predict unseen facts of a relation with few-shot reference entity pairs. Current approaches randomly select one negative sample for each reference entity pair to minimize a margin-based ranking loss, which easily leads to a zero-loss problem if the negative sample...
http://arxiv.org/pdf/2306.09519v1
2023-06-15T21:41:43
2023-06-15T21:41:43
{ "title": [ -0.008293152, -0.025894165, -0.04864502, -0.06573486, -0.018371582, 0.008102417, -0.050323486, 0.035827637, 0.02758789, 0.03253174, -0.03262329, 0.012756348, 0.0031776428, 0.0018453598, -0.043884277, 0.0020008087, 0.0107803345, 0.005...
http://arxiv.org/abs/2306.09505v1
Wikibio: a Semantic Resource for the Intersectional Analysis of Biographical Events
Marco Antonio Stranisci, Rossana Damiano, Enrico Mensa, Viviana Patti, Daniele Radicioni, Tommaso Caselli
cs.CL
Biographical event detection is a relevant task for the exploration and comparison of the ways in which people's lives are told and represented. In this sense, it may support several applications in digital humanities and in works aimed at exploring bias about minoritized groups. Despite that, there are no corpora and ...
http://arxiv.org/pdf/2306.09505v1
2023-06-15T20:59:37
2023-06-15T20:59:37
{ "title": [ -0.018005371, -0.025680542, -0.03894043, -0.09246826, 0.004501343, -0.00022661686, -0.052886963, 0.008270264, 0.0016832352, 0.041931152, -0.01625061, 0.02911377, 0.019821167, -0.012512207, -0.038391113, 0.012664795, -0.016143799, 0.0...
http://arxiv.org/abs/2306.10070v2
Opportunities and Challenges for ChatGPT and Large Language Models in Biomedicine and Health
Shubo Tian, Qiao Jin, Lana Yeganova, Po-Ting Lai, Qingqing Zhu, Xiuying Chen, Yifan Yang, Qingyu Chen, Won Kim, Donald C. Comeau, Rezarta Islamaj, Aadit Kapoor, Xin Gao, Zhiyong Lu
cs.CY, cs.AI, cs.CL, q-bio.QM
ChatGPT has drawn considerable attention from both the general public and domain experts with its remarkable text generation capabilities. This has subsequently led to the emergence of diverse applications in the field of biomedicine and health. In this work, we examine the diverse applications of large language models...
http://arxiv.org/pdf/2306.10070v2
2023-10-17T03:29:04
2023-06-15T20:19:08
{ "title": [ -0.027069092, -0.022598267, -0.026138306, -0.036376953, 0.0006136894, 0.0074653625, -0.045196533, -0.007789612, -0.02041626, 0.024871826, 0.00687027, 0.0079574585, -0.0038604736, -0.036712646, -0.016494751, 0.019210815, -0.016189575, ...
http://arxiv.org/abs/2306.09479v1
Inverse Scaling: When Bigger Isn't Better
Ian R. McKenzie, Alexander Lyzhov, Michael Pieler, Alicia Parrish, Aaron Mueller, Ameya Prabhu, Euan McLean, Aaron Kirtland, Alexis Ross, Alisa Liu, Andrew Gritsevskiy, Daniel Wurgaft, Derik Kauffman, Gabriel Recchia, Jiacheng Liu, Joe Cavanagh, Max Weiss, Sicong Huang, The Floating Droid, Tom Tseng, Tomasz Korbak, Xud...
cs.CL, cs.AI, cs.CY
Work on scaling laws has found that large language models (LMs) show predictable improvements to overall loss with increased scale (model size, training data, and compute). Here, we present evidence for the claim that LMs may show inverse scaling, or worse task performance with increased scale, e.g., due to flaws in th...
http://arxiv.org/pdf/2306.09479v1
2023-06-15T20:11:23
2023-06-15T20:11:23
{ "title": [ -0.025878906, -0.0038928986, -0.0206604, 0.018447876, -0.037139893, -0.015197754, -0.01423645, -0.052703857, -0.030700684, 0.03250122, -0.044647217, -0.017623901, -0.017715454, -0.03286743, -0.030914307, -0.044921875, 0.023605347, 0....
http://arxiv.org/abs/2306.09442v3
Explore, Establish, Exploit: Red Teaming Language Models from Scratch
Stephen Casper, Jason Lin, Joe Kwon, Gatlen Culp, Dylan Hadfield-Menell
cs.CL, cs.AI, cs.LG
Deploying large language models (LMs) can pose hazards from harmful outputs such as toxic or false text. Prior work has introduced automated tools that elicit harmful outputs to identify these risks. While this is a valuable step toward securing models, these approaches rely on a pre-existing way to efficiently classif...
http://arxiv.org/pdf/2306.09442v3
2023-10-11T00:37:33
2023-06-15T18:49:50
{ "title": [ -0.031555176, -0.020706177, -0.028289795, -0.052368164, -0.010063171, 0.004058838, -0.031433105, -0.0075035095, -0.035247803, 0.016464233, -0.002046585, 0.008460999, -0.026367188, -0.034942627, -0.020217896, -0.017166138, -0.005393982, ...
http://arxiv.org/abs/2306.09343v1
SIGHT: A Large Annotated Dataset on Student Insights Gathered from Higher Education Transcripts
Rose E. Wang, Pawan Wirawarn, Noah Goodman, Dorottya Demszky
cs.CL, cs.AI
Lectures are a learning experience for both students and teachers. Students learn from teachers about the subject material, while teachers learn from students about how to refine their instruction. However, online student feedback is unstructured and abundant, making it challenging for teachers to learn and improve. We...
http://arxiv.org/pdf/2306.09343v1
2023-06-15T17:59:47
2023-06-15T17:59:47
{ "title": [ -0.0064888, -0.01222229, -0.059814453, -0.04019165, -0.008918762, -0.03515625, -0.011703491, 0.01524353, 0.025909424, 0.026885986, -0.040252686, -0.019683838, 0.010215759, -0.016845703, -0.028869629, -0.02809143, 0.0024356842, 0.0050...
http://arxiv.org/abs/2306.09342v1
PaReprop: Fast Parallelized Reversible Backpropagation
Tyler Zhu, Karttikeya Mangalam
cs.LG, cs.CL, cs.CV
The growing size of datasets and deep learning models has made faster and memory-efficient training crucial. Reversible transformers have recently been introduced as an exciting new method for extremely memory-efficient training, but they come with an additional computation overhead of activation re-computation in the ...
http://arxiv.org/pdf/2306.09342v1
2023-06-15T17:59:32
2023-06-15T17:59:32
{ "title": [ -0.026351929, 0.0010957718, -0.03213501, -0.06713867, 0.012863159, 0.0050735474, -0.027893066, -0.018569946, 0.015853882, 0.019638062, -0.0054359436, -0.0006222725, -0.0006799698, -0.02671814, -0.04849243, 0.0025558472, 0.0012521744, ...
http://arxiv.org/abs/2306.09340v1
Span-Selective Linear Attention Transformers for Effective and Robust Schema-Guided Dialogue State Tracking
Björn Bebensee, Haejun Lee
cs.CL, cs.LG
In schema-guided dialogue state tracking models estimate the current state of a conversation using natural language descriptions of the service schema for generalization to unseen services. Prior generative approaches which decode slot values sequentially do not generalize well to variations in schema, while discrimina...
http://arxiv.org/pdf/2306.09340v1
2023-06-15T17:59:31
2023-06-15T17:59:31
{ "title": [ 0.01878357, -0.002193451, -0.02822876, -0.07055664, -0.0079422, -0.008682251, -0.041900635, 0.0001718998, -0.009262085, 0.008323669, -0.017303467, 0.0076904297, 0.0046310425, -0.024841309, -0.0056610107, -0.012893677, -0.009307861, 0...
http://arxiv.org/abs/2306.09328v1
WizMap: Scalable Interactive Visualization for Exploring Large Machine Learning Embeddings
Zijie J. Wang, Fred Hohman, Duen Horng Chau
cs.LG, cs.CL, cs.CV, cs.HC
Machine learning models often learn latent embedding representations that capture the domain semantics of their training data. These embedding representations are valuable for interpreting trained models, building new models, and analyzing new datasets. However, interpreting and using embeddings can be challenging due ...
http://arxiv.org/pdf/2306.09328v1
2023-06-15T17:58:04
2023-06-15T17:58:04
{ "title": [ -0.006385803, -0.01335907, -0.035339355, -0.028381348, 0.0003569126, -0.024291992, 0.017532349, -0.016281128, 0.013137817, 0.0030612946, -0.02507019, -0.0040740967, -0.03387451, -0.022277832, -0.036224365, -0.015289307, 0.0011377335, ...
http://arxiv.org/abs/2306.09313v1
Lexical Speaker Error Correction: Leveraging Language Models for Speaker Diarization Error Correction
Rohit Paturi, Sundararajan Srinivasan, Xiang Li
eess.AS, cs.AI, cs.CL, cs.LG
Speaker diarization (SD) is typically used with an automatic speech recognition (ASR) system to ascribe speaker labels to recognized words. The conventional approach reconciles outputs from independently optimized ASR and SD systems, where the SD system typically uses only acoustic information to identify the speakers ...
http://arxiv.org/pdf/2306.09313v1
2023-06-15T17:47:41
2023-06-15T17:47:41
{ "title": [ -0.012649536, 0.010314941, -0.028244019, -0.03955078, -0.011749268, -0.0017776489, -0.05340576, -0.023132324, 0.004272461, 0.055358887, -0.042541504, -0.0010738373, 0.044189453, -0.0023212433, -0.041870117, 0.013679504, -0.015556335, ...
http://arxiv.org/abs/2306.09312v2
Semantic HELM: A Human-Readable Memory for Reinforcement Learning
Fabian Paischer, Thomas Adler, Markus Hofmarcher, Sepp Hochreiter
cs.LG, cs.AI, cs.CL, stat.ML
Reinforcement learning agents deployed in the real world often have to cope with partially observable environments. Therefore, most agents employ memory mechanisms to approximate the state of the environment. Recently, there have been impressive success stories in mastering partially observable environments, mostly in ...
http://arxiv.org/pdf/2306.09312v2
2023-10-27T10:34:20
2023-06-15T17:47:31
{ "title": [ -0.018661499, -0.0030975342, -0.039123535, -0.027694702, 0.007926941, -0.016738892, 0.00029158592, -0.019683838, -0.023773193, 0.018661499, -0.030929565, 0.014427185, -0.03451538, -0.0066490173, -0.030883789, -0.014663696, 0.00044727325,...
http://arxiv.org/abs/2306.09308v1
Matching Pairs: Attributing Fine-Tuned Models to their Pre-Trained Large Language Models
Myles Foley, Ambrish Rawat, Taesung Lee, Yufang Hou, Gabriele Picco, Giulio Zizzo
cs.CL, cs.AI, cs.CR
The wide applicability and adaptability of generative large language models (LLMs) has enabled their rapid adoption. While the pre-trained models can perform many tasks, such models are often fine-tuned to improve their performance on various downstream applications. However, this leads to issues over violation of mode...
http://arxiv.org/pdf/2306.09308v1
2023-06-15T17:42:48
2023-06-15T17:42:48
{ "title": [ -0.016098022, -0.012397766, -0.028411865, -0.07928467, 0.0037899017, 0.008911133, -0.06982422, -0.0069732666, -0.0070152283, 0.021591187, -0.0042037964, -0.00032281876, -0.009635925, -0.025817871, -0.031463623, -0.0013628006, -0.01029968...
http://arxiv.org/abs/2306.09307v1
Quality and Efficiency of Manual Annotation: Pre-annotation Bias
Marie Mikulová, Milan Straka, Jan Štěpánek, Barbora Štěpánková, Jan Hajič
cs.CL
This paper presents an analysis of annotation using an automatic pre-annotation for a mid-level annotation complexity task -- dependency syntax annotation. It compares the annotation efforts made by annotators using a pre-annotated version (with a high-accuracy parser) and those made by fully manual annotation. The aim...
http://arxiv.org/pdf/2306.09307v1
2023-06-15T17:41:14
2023-06-15T17:41:14
{ "title": [ -0.015235901, 0.009338379, -0.021484375, -0.04824829, -0.016738892, -0.011474609, -0.04928589, -0.048431396, 0.008369446, 0.03036499, -0.023330688, 0.000091314316, 0.011756897, -0.03488159, -0.01838684, -0.0015182495, 0.0043640137, 0...
http://arxiv.org/abs/2306.09306v2
Propagating Knowledge Updates to LMs Through Distillation
Shankar Padmanabhan, Yasumasa Onoe, Michael J. Q. Zhang, Greg Durrett, Eunsol Choi
cs.CL
Modern language models have the capacity to store and use immense amounts of knowledge about real-world entities, but it remains unclear how to update such knowledge stored in model parameters. While prior methods for updating knowledge in LMs successfully inject atomic facts, updated LMs fail to make inferences based ...
http://arxiv.org/pdf/2306.09306v2
2023-10-31T00:29:12
2023-06-15T17:39:50
{ "title": [ -0.004005432, 0.016296387, -0.07019043, -0.08532715, 0.012794495, -0.011238098, -0.057281494, -0.0023670197, 0.016433716, 0.032836914, -0.05731201, -0.03955078, 0.011726379, -0.020950317, -0.041748047, -0.04055786, 0.029815674, -0.00...
http://arxiv.org/abs/2306.09299v2
Can Language Models Teach Weaker Agents? Teacher Explanations Improve Students via Personalization
Swarnadeep Saha, Peter Hase, Mohit Bansal
cs.CL, cs.AI, cs.LG
A hallmark property of explainable AI models is the ability to teach other agents, communicating knowledge of how to perform a task. While Large Language Models perform complex reasoning by generating explanations for their predictions, it is unclear whether they also make good teachers for weaker agents. To address th...
http://arxiv.org/pdf/2306.09299v2
2023-11-14T05:24:06
2023-06-15T17:27:20
{ "title": [ 0.0069007874, -0.00096797943, -0.004383087, -0.020019531, -0.034210205, -0.0073890686, -0.02331543, -0.0143585205, -0.062408447, 0.027282715, -0.0066337585, 0.015594482, -0.00013566017, -0.022949219, 0.008293152, -0.029953003, -0.0137405...
http://arxiv.org/abs/2306.09296v2
KoLA: Carefully Benchmarking World Knowledge of Large Language Models
Jifan Yu, Xiaozhi Wang, Shangqing Tu, Shulin Cao, Daniel Zhang-Li, Xin Lv, Hao Peng, Zijun Yao, Xiaohan Zhang, Hanming Li, Chunyang Li, Zheyuan Zhang, Yushi Bai, Yantao Liu, Amy Xin, Nianyi Lin, Kaifeng Yun, Linlu Gong, Jianhui Chen, Zhili Wu, Yunjia Qi, Weikai Li, Yong Guan, Kaisheng Zeng, Ji Qi, Hailong Jin, Jinxin L...
cs.CL
The unprecedented performance of large language models (LLMs) necessitates improvements in evaluations. Rather than merely exploring the breadth of LLM abilities, we believe meticulous and thoughtful designs are essential to thorough, unbiased, and applicable evaluations. Given the importance of world knowledge to LLMs...
http://arxiv.org/pdf/2306.09296v2
2023-07-06T17:25:10
2023-06-15T17:20:46
{ "title": [ -0.046569824, -0.023651123, -0.055847168, -0.070129395, -0.011650085, -0.0029067993, -0.064453125, 0.019821167, 0.00041031837, 0.023880005, -0.024505615, 0.02935791, 0.0075263977, -0.010986328, -0.026809692, 0.017425537, -0.000562191, ...
http://arxiv.org/abs/2306.09288v1
DaMuEL: A Large Multilingual Dataset for Entity Linking
David Kubeša, Milan Straka
cs.CL
We present DaMuEL, a large Multilingual Dataset for Entity Linking containing data in 53 languages. DaMuEL consists of two components: a knowledge base that contains language-agnostic information about entities, including their claims from Wikidata and named entity types (PER, ORG, LOC, EVENT, BRAND, WORK_OF_ART, MANUF...
http://arxiv.org/pdf/2306.09288v1
2023-06-15T17:15:52
2023-06-15T17:15:52
{ "title": [ -0.005329132, -0.012069702, -0.052764893, -0.090148926, 0.037475586, 0.0008149147, -0.04333496, 0.0079193115, 0.0028514862, 0.03112793, 0.002418518, 0.0154418945, 0.03765869, -0.015731812, -0.009498596, 0.028060913, -0.015174866, 0.0...
http://arxiv.org/abs/2306.09237v2
SCALE: Scaling up the Complexity for Advanced Language Model Evaluation
Vishvaksenan Rasiah, Ronja Stern, Veton Matoshi, Matthias Stürmer, Ilias Chalkidis, Daniel E. Ho, Joel Niklaus
cs.CL, cs.AI, cs.LG, 68T50, I.2
Recent strides in Large Language Models (LLMs) have saturated many NLP benchmarks (even professional domain-specific ones), emphasizing the need for novel, more challenging novel ones to properly assess LLM capabilities. In this paper, we introduce a novel NLP benchmark that poses challenges to current LLMs across four...
http://arxiv.org/pdf/2306.09237v2
2023-09-01T18:00:57
2023-06-15T16:19:15
{ "title": [ -0.03427124, -0.011207581, -0.04208374, -0.067871094, 0.0034656525, 0.008415222, -0.064208984, -0.0008883476, -0.0021400452, 0.026306152, -0.015579224, 0.0069122314, -0.0004518032, -0.03286743, -0.042938232, 0.013793945, -0.020111084, ...
http://arxiv.org/abs/2306.09390v1
ChatGPT for Suicide Risk Assessment on Social Media: Quantitative Evaluation of Model Performance, Potentials and Limitations
Hamideh Ghanadian, Isar Nejadgholi, Hussein Al Osman
cs.CL, cs.AI, cs.LG
This paper presents a novel framework for quantitatively evaluating the interactive ChatGPT model in the context of suicidality assessment from social media posts, utilizing the University of Maryland Reddit suicidality dataset. We conduct a technical evaluation of ChatGPT's performance on this task using Zero-Shot and...
http://arxiv.org/pdf/2306.09390v1
2023-06-15T16:01:30
2023-06-15T16:01:30
{ "title": [ 0.025253296, 0.0016651154, 0.0034656525, -0.026519775, 0.012435913, -0.007347107, 0.007255554, -0.027679443, -0.022888184, 0.04864502, -0.01777649, 0.043762207, 0.011276245, -0.009666443, -0.03161621, 0.010261536, -0.028335571, 0.030...
http://arxiv.org/abs/2306.09212v1
CMMLU: Measuring massive multitask language understanding in Chinese
Haonan Li, Yixuan Zhang, Fajri Koto, Yifei Yang, Hai Zhao, Yeyun Gong, Nan Duan, Timothy Baldwin
cs.CL
As the capabilities of large language models (LLMs) continue to advance, evaluating their performance becomes increasingly crucial and challenging. This paper aims to bridge this gap by introducing CMMLU, a comprehensive Chinese benchmark that covers various subjects, including natural science, social sciences, enginee...
http://arxiv.org/pdf/2306.09212v1
2023-06-15T15:49:51
2023-06-15T15:49:51
{ "title": [ -0.033233643, -0.008514404, -0.034851074, -0.04284668, -0.019195557, -0.0056152344, -0.046173096, 0.003479004, -0.010818481, 0.023117065, -0.006465912, 0.010498047, -0.003894806, -0.02835083, -0.03729248, 0.00083065033, -0.024230957, ...
http://arxiv.org/abs/2306.10067v2
Domain-specific ChatBots for Science using Embeddings
Kevin G. Yager
cs.CL, cs.AI, I.2.7
Large language models (LLMs) have emerged as powerful machine-learning systems capable of handling a myriad of tasks. Tuned versions of these systems have been turned into chatbots that can respond to user queries on a vast diversity of topics, providing informative and creative replies. However, their application to p...
http://arxiv.org/pdf/2306.10067v2
2023-08-24T20:24:13
2023-06-15T15:26:20
{ "title": [ -0.014305115, -0.018188477, -0.017456055, -0.04071045, -0.007926941, 0.004470825, -0.013801575, 0.0041503906, 0.0037498474, 0.041625977, -0.0041618347, -0.0011262894, 0.005088806, -0.014778137, -0.041625977, 0.0039405823, -0.011360168, ...
http://arxiv.org/abs/2306.09170v3
Can ChatGPT pass the Vietnamese National High School Graduation Examination?
Xuan-Quy Dao, Ngoc-Bich Le, Xuan-Dung Phan, Bac-Bien Ngo
cs.CL, cs.HC
This research article highlights the potential of AI-powered chatbots in education and presents the results of using ChatGPT, a large language model, to complete the Vietnamese National High School Graduation Examination (VNHSGE). The study dataset included 30 essays in the literature test case and 1,700 multiple-choic...
http://arxiv.org/pdf/2306.09170v3
2023-07-10T11:22:20
2023-06-15T14:47:03
{ "title": [ 0.019927979, -0.017211914, 0.003112793, 0.0031375885, -0.0418396, -0.00074338913, -0.0057029724, -0.03656006, -0.040802002, 0.038848877, 0.0053520203, 0.0030899048, -0.01146698, -0.04031372, 0.010131836, -0.016067505, -0.033477783, 0...
http://arxiv.org/abs/2306.09169v1
Opportunities for Large Language Models and Discourse in Engineering Design
Jan Göpfert, Jann M. Weinand, Patrick Kuckertz, Detlef Stolten
cs.CL, cs.AI, cs.CE
In recent years, large language models have achieved breakthroughs on a wide range of benchmarks in natural language processing and continue to increase in performance. Recently, the advances of large language models have raised interest outside the natural language processing community and could have a large impact on...
http://arxiv.org/pdf/2306.09169v1
2023-06-15T14:46:44
2023-06-15T14:46:44
{ "title": [ -0.04550171, -0.017791748, -0.047424316, -0.026107788, -0.0014715195, -0.017578125, -0.03652954, -0.00070619583, -0.0051879883, 0.011878967, 0.007423401, -0.014816284, -0.022628784, -0.0413208, -0.018966675, 0.005004883, -0.0034599304, ...
http://arxiv.org/abs/2306.09144v1
On the $k$-Hamming and $k$-Edit Distances
Chiara Epifanio, Luca Forlizzi, Francesca Marzi, Filippo Mignosi, Giuseppe Placidi, Matteo Spezialetti
cs.CC, cs.CL, cs.DS
In this paper we consider the weighted $k$-Hamming and $k$-Edit distances, that are natural generalizations of the classical Hamming and Edit distances. As main results of this paper we prove that for any $k\geq 2$ the DECIS-$k$-Hamming problem is $\mathbb{P}$-SPACE-complete and the DECIS-$k$-Edit problem is NEXPTIME-c...
http://arxiv.org/pdf/2306.09144v1
2023-06-15T14:03:28
2023-06-15T14:03:28
{ "title": [ -0.06561279, -0.0030441284, -0.043945312, -0.029281616, -0.00082445145, 0.022949219, -0.0345459, -0.026931763, -0.003786087, 0.017486572, -0.00059604645, -0.009513855, 0.0010662079, -0.02960205, -0.031173706, 0.011955261, -0.003835678, ...
http://arxiv.org/abs/2306.09114v1
Relational Temporal Graph Reasoning for Dual-task Dialogue Language Understanding
Bowen Xing, Ivor W. Tsang
cs.CL, cs.AI
Dual-task dialog language understanding aims to tackle two correlative dialog language understanding tasks simultaneously via leveraging their inherent correlations. In this paper, we put forward a new framework, whose core is relational temporal graph reasoning.We propose a speaker-aware temporal graph (SATG) and a du...
http://arxiv.org/pdf/2306.09114v1
2023-06-15T13:19:08
2023-06-15T13:19:08
{ "title": [ -0.025390625, -0.029006958, -0.05041504, -0.10455322, -0.021743774, 0.0033950806, -0.08935547, 0.050476074, -0.012832642, 0.020126343, 0.0070991516, -0.00579834, -0.018966675, -0.037750244, -0.03488159, -0.018600464, -0.0090408325, 0...
http://arxiv.org/abs/2306.09108v1
KUCST at CheckThat 2023: How good can we be with a generic model?
Manex Agirrezabal
cs.CL
In this paper we present our method for tasks 2 and 3A at the CheckThat2023 shared task. We make use of a generic approach that has been used to tackle a diverse set of tasks, inspired by authorship attribution and profiling. We train a number of Machine Learning models and our results show that Gradient Boosting perfo...
http://arxiv.org/pdf/2306.09108v1
2023-06-15T13:09:40
2023-06-15T13:09:40
{ "title": [ 0.013214111, 0.0006971359, 0.009147644, -0.017929077, -0.010437012, 0.01512146, -0.027908325, -0.035339355, 0.0008983612, 0.053649902, -0.024627686, -0.0009860992, 0.0625, -0.025360107, 0.005634308, -0.0023212433, 0.025146484, 0.0101...
http://arxiv.org/abs/2306.09093v1
Macaw-LLM: Multi-Modal Language Modeling with Image, Audio, Video, and Text Integration
Chenyang Lyu, Minghao Wu, Longyue Wang, Xinting Huang, Bingshuai Liu, Zefeng Du, Shuming Shi, Zhaopeng Tu
cs.CL, cs.AI, cs.CV
Although instruction-tuned large language models (LLMs) have exhibited remarkable capabilities across various NLP tasks, their effectiveness on other data modalities beyond text has not been fully studied. In this work, we propose Macaw-LLM, a novel multi-modal LLM that seamlessly integrates visual, audio, and textual ...
http://arxiv.org/pdf/2306.09093v1
2023-06-15T12:45:25
2023-06-15T12:45:25
{ "title": [ -0.03326416, 0.0067749023, -0.04385376, -0.04574585, -0.0060653687, -0.0031967163, -0.06518555, -0.017608643, -0.00041651726, 0.036254883, -0.014450073, 0.025344849, 0.012374878, -0.019073486, -0.03366089, 0.004184723, -0.010787964, ...
http://arxiv.org/abs/2306.09085v1
COSA: Concatenated Sample Pretrained Vision-Language Foundation Model
Sihan Chen, Xingjian He, Handong Li, Xiaojie Jin, Jiashi Feng, Jing Liu
cs.CV, cs.AI, cs.CL, cs.LG, cs.MM
Due to the limited scale and quality of video-text training corpus, most vision-language foundation models employ image-text datasets for pretraining and primarily focus on modeling visually semantic representations while disregarding temporal semantic representations and correlations. To address this issue, we propose...
http://arxiv.org/pdf/2306.09085v1
2023-06-15T12:29:42
2023-06-15T12:29:42
{ "title": [ -0.017837524, -0.035888672, -0.053131104, -0.06713867, -0.03768921, -0.014778137, -0.067993164, 0.0052833557, -0.0005507469, 0.025985718, -0.018112183, -0.0018939972, 0.027832031, -0.042236328, 0.010322571, -0.0012655258, 0.008552551, ...
http://arxiv.org/abs/2306.09079v1
Web of Things and Trends in Agriculture: A Systematic Literature Review
Muhammad Shoaib Farooq, Shamyla Riaz, Atif Alvi
cs.IR, cs.CL
In the past few years, the Web of Things (WOT) became a beneficial game-changing technology within the Agriculture domain as it introduces innovative and promising solutions to the Internet of Things (IoT) agricultural applications problems by providing its services. WOT provides the support for integration, interopera...
http://arxiv.org/pdf/2306.09079v1
2023-06-15T12:20:30
2023-06-15T12:20:30
{ "title": [ -0.0016775131, -0.003189087, -0.023544312, -0.043395996, 0.02128601, -0.01209259, 0.0038032532, 0.007183075, -0.04446411, 0.039794922, -0.02293396, 0.012672424, 0.020339966, -0.020370483, -0.027938843, -0.0017633438, -0.011383057, 0....
http://arxiv.org/abs/2306.09066v1
A Bayesian approach to uncertainty in word embedding bias estimation
Alicja Dobrzeniecka, Rafal Urbaniak
cs.CL, cs.HC, cs.LG, stat.AP, stat.ME
Multiple measures, such as WEAT or MAC, attempt to quantify the magnitude of bias present in word embeddings in terms of a single-number metric. However, such metrics and the related statistical significance calculations rely on treating pre-averaged data as individual data points and employing bootstrapping techniques...
http://arxiv.org/pdf/2306.09066v1
2023-06-15T11:48:50
2023-06-15T11:48:50
{ "title": [ -0.021240234, 0.00092220306, -0.033050537, -0.011329651, -0.00003361702, -0.016647339, -0.020385742, -0.031585693, 0.022888184, 0.001490593, -0.050842285, 0.011199951, -0.006565094, -0.023742676, -0.031082153, -0.016952515, -0.020858765,...
http://arxiv.org/abs/2306.09064v1
Learning by Analogy: Diverse Questions Generation in Math Word Problem
Zihao Zhou, Maizhen Ning, Qiufeng Wang, Jie Yao, Wei Wang, Xiaowei Huang, Kaizhu Huang
cs.CL
Solving math word problem (MWP) with AI techniques has recently made great progress with the success of deep neural networks (DNN), but it is far from being solved. We argue that the ability of learning by analogy is essential for an MWP solver to better understand same problems which may typically be formulated in div...
http://arxiv.org/pdf/2306.09064v1
2023-06-15T11:47:07
2023-06-15T11:47:07
{ "title": [ -0.021530151, -0.006668091, -0.024337769, -0.004501343, -0.0058517456, -0.021942139, 0.025924683, -0.013008118, -0.009971619, 0.009719849, -0.019424438, 0.018707275, -0.036895752, -0.014961243, -0.022277832, -0.018569946, -0.0036239624, ...
http://arxiv.org/abs/2306.09049v1
Mapping Researcher Activity based on Publication Data by means of Transformers
Zineddine Bettouche, Andreas Fischer
cs.CL, cs.DL, cs.IR, cs.LG
Modern performance on several natural language processing (NLP) tasks has been enhanced thanks to the Transformer-based pre-trained language model BERT. We employ this concept to investigate a local publication database. Research papers are encoded and clustered to form a landscape view of the scientific topics, in whi...
http://arxiv.org/pdf/2306.09049v1
2023-06-15T11:13:54
2023-06-15T11:13:54
{ "title": [ 0.015556335, 0.0335083, -0.039123535, -0.004951477, 0.0065727234, -0.04260254, 0.018325806, -0.052246094, -0.028244019, 0.016098022, -0.014732361, 0.0011110306, -0.020248413, -0.047851562, -0.00044679642, -0.027359009, 0.0063591003, ...
http://arxiv.org/abs/2306.09030v2
DiPlomat: A Dialogue Dataset for Situated Pragmatic Reasoning
Hengli Li, Song-Chun Zhu, Zilong Zheng
cs.CL
Pragmatic reasoning plays a pivotal role in deciphering implicit meanings that frequently arise in real-life conversations and is essential for the development of communicative social agents. In this paper, we introduce a novel challenge, DiPlomat, aiming at benchmarking machines' capabilities on pragmatic reasoning an...
http://arxiv.org/pdf/2306.09030v2
2023-06-19T07:31:55
2023-06-15T10:41:23
{ "title": [ 0.003566742, -0.0178833, -0.04095459, -0.082458496, 0.0062942505, -0.0073165894, -0.04196167, 0.025772095, 0.015083313, 0.004917145, -0.02998352, -0.012039185, -0.02557373, -0.018005371, -0.040985107, -0.020858765, -0.024047852, 0.03...
http://arxiv.org/abs/2306.08999v1
Voting Booklet Bias: Stance Detection in Swiss Federal Communication
Eric Egli, Noah Mamié, Eyal Liron Dolev, Mathias Müller
cs.CL, cs.AI
In this study, we use recent stance detection methods to study the stance (for, against or neutral) of statements in official information booklets for voters. Our main goal is to answer the fundamental question: are topics to be voted on presented in a neutral way? To this end, we first train and compare several mode...
http://arxiv.org/pdf/2306.08999v1
2023-06-15T09:49:12
2023-06-15T09:49:12
{ "title": [ 0.011810303, 0.006000519, -0.055267334, -0.021118164, 0.013458252, -0.03945923, -0.014579773, -0.011894226, 0.020736694, 0.021347046, -0.052825928, -0.0037403107, 0.0019283295, -0.009719849, -0.03427124, -0.023086548, -0.023376465, 0...
http://arxiv.org/abs/2306.08997v2
Exploring the MIT Mathematics and EECS Curriculum Using Large Language Models
Sarah J. Zhang, Samuel Florin, Ariel N. Lee, Eamon Niknafs, Andrei Marginean, Annie Wang, Keith Tyser, Zad Chin, Yann Hicke, Nikhil Singh, Madeleine Udell, Yoon Kim, Tonio Buonassisi, Armando Solar-Lezama, Iddo Drori
cs.CL, cs.AI, cs.LG
We curate a comprehensive dataset of 4,550 questions and solutions from problem sets, midterm exams, and final exams across all MIT Mathematics and Electrical Engineering and Computer Science (EECS) courses required for obtaining a degree. We evaluate the ability of large language models to fulfill the graduation requi...
http://arxiv.org/pdf/2306.08997v2
2023-06-24T12:39:06
2023-06-15T09:48:14
{ "title": [ -0.011474609, -0.014556885, -0.012809753, 0.016220093, -0.020507812, -0.032073975, 0.017578125, -0.019607544, -0.017166138, 0.024765015, -0.030761719, -0.005619049, 0.004837036, -0.036834717, 0.0035305023, -0.013702393, -0.0113220215, ...
http://arxiv.org/abs/2306.08953v1
Rethinking Document-Level Relation Extraction: A Reality Check
Jing Li, Yequan Wang, Shuai Zhang, Min Zhang
cs.CL
Recently, numerous efforts have continued to push up performance boundaries of document-level relation extraction (DocRE) and have claimed significant progress in DocRE. In this paper, we do not aim at proposing a novel model for DocRE. Instead, we take a closer look at the field to see if these performance gains are a...
http://arxiv.org/pdf/2306.08953v1
2023-06-15T08:47:42
2023-06-15T08:47:42
{ "title": [ -0.010375977, -0.011886597, -0.04425049, -0.074035645, 0.01247406, -0.0018453598, -0.03994751, 0.0040130615, 0.020233154, 0.009346008, -0.025756836, 0.0023708344, -0.011428833, -0.0043296814, -0.049560547, -0.010932922, -0.018112183, ...
http://arxiv.org/abs/2306.08952v2
Towards Benchmarking and Improving the Temporal Reasoning Capability of Large Language Models
Qingyu Tan, Hwee Tou Ng, Lidong Bing
cs.CL, cs.AI
Reasoning about time is of fundamental importance. Many facts are time-dependent. For example, athletes change teams from time to time, and different government officials are elected periodically. Previous time-dependent question answering (QA) datasets tend to be biased in either their coverage of time spans or questi...
http://arxiv.org/pdf/2306.08952v2
2023-06-27T05:39:25
2023-06-15T08:44:41
{ "title": [ -0.015792847, -0.022354126, -0.04647827, -0.12335205, 0.011619568, 0.006210327, -0.07745361, 0.039520264, 0.012252808, 0.033050537, -0.012062073, -0.0041275024, 0.0062789917, -0.015838623, -0.0574646, 0.006362915, -0.013442993, 0.026...
http://arxiv.org/abs/2306.08937v3
DocumentNet: Bridging the Data Gap in Document Pre-Training
Lijun Yu, Jin Miao, Xiaoyu Sun, Jiayi Chen, Alexander G. Hauptmann, Hanjun Dai, Wei Wei
cs.CL, cs.IR
Document understanding tasks, in particular, Visually-rich Document Entity Retrieval (VDER), have gained significant attention in recent years thanks to their broad applications in enterprise AI. However, publicly available data have been scarce for these tasks due to strict privacy constraints and high annotation cost...
http://arxiv.org/pdf/2306.08937v3
2023-10-26T16:23:15
2023-06-15T08:21:15
{ "title": [ -0.018112183, -0.012283325, -0.052886963, -0.07458496, 0.0039901733, -0.0052604675, -0.05557251, -0.017364502, -0.0020923615, 0.017974854, -0.03781128, 0.0013608932, 0.015525818, -0.023086548, -0.021362305, -0.013412476, 0.0033740997, ...
http://arxiv.org/abs/2306.08925v1
Opinion Tree Parsing for Aspect-based Sentiment Analysis
Xiaoyi Bao, Xiaotong Jiang, Zhongqing Wang, Yue Zhang, Guodong Zhou
cs.CL, cs.AI
Extracting sentiment elements using pre-trained generative models has recently led to large improvements in aspect-based sentiment analysis benchmarks. However, these models always need large-scale computing resources, and they also ignore explicit modeling of structure between sentiment elements. To address these chal...
http://arxiv.org/pdf/2306.08925v1
2023-06-15T07:53:14
2023-06-15T07:53:14
{ "title": [ -0.009101868, -0.012741089, -0.018417358, -0.06750488, 0.00365448, 0.017440796, -0.034729004, 0.009048462, 0.025665283, 0.041748047, -0.043029785, -0.015823364, -0.0053634644, -0.0020923615, -0.0847168, -0.011276245, -0.016540527, 0....
http://arxiv.org/abs/2306.08920v1
Pushing the Limits of Unsupervised Unit Discovery for SSL Speech Representation
Ziyang Ma, Zhisheng Zheng, Guanrou Yang, Yu Wang, Chao Zhang, Xie Chen
cs.CL, cs.SD, eess.AS
The excellent generalization ability of self-supervised learning (SSL) for speech foundation models has garnered significant attention. HuBERT is a successful example that utilizes offline clustering to convert speech features into discrete units for a masked language modeling pretext task. However, simply clustering f...
http://arxiv.org/pdf/2306.08920v1
2023-06-15T07:45:12
2023-06-15T07:45:12
{ "title": [ -0.01171875, 0.015419006, -0.080322266, -0.076049805, 0.022720337, -0.020431519, -0.0579834, 0.03289795, -0.019927979, 0.057556152, -0.02508545, 0.009429932, 0.040893555, -0.01713562, -0.014808655, 0.014320374, -0.017181396, 0.047119...