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/2309.08872v2
PDFTriage: Question Answering over Long, Structured Documents
Jon Saad-Falcon, Joe Barrow, Alexa Siu, Ani Nenkova, David Seunghyun Yoon, Ryan A. Rossi, Franck Dernoncourt
cs.CL, cs.AI, cs.LG
Large Language Models (LLMs) have issues with document question answering (QA) in situations where the document is unable to fit in the small context length of an LLM. To overcome this issue, most existing works focus on retrieving the relevant context from the document, representing them as plain text. However, docume...
http://arxiv.org/pdf/2309.08872v2
2023-11-08T05:09:28
2023-09-16T04:29:05
{ "title": [ -0.0020751953, -0.031982422, -0.0390625, -0.06713867, -0.02418518, -0.012779236, -0.04055786, 0.013015747, 0.006702423, 0.03781128, -0.020507812, -0.024230957, 0.049194336, -0.020629883, -0.03552246, -0.011703491, 0.0036849976, 0.016...
http://arxiv.org/abs/2309.08868v1
MHLAT: Multi-hop Label-wise Attention Model for Automatic ICD Coding
Junwen Duan, Han Jiang, Ying Yu
cs.CL, cs.AI
International Classification of Diseases (ICD) coding is the task of assigning ICD diagnosis codes to clinical notes. This can be challenging given the large quantity of labels (nearly 9,000) and lengthy texts (up to 8,000 tokens). However, unlike the single-pass reading process in previous works, humans tend to read t...
http://arxiv.org/pdf/2309.08868v1
2023-09-16T04:13:57
2023-09-16T04:13:57
{ "title": [ 0.005180359, -0.001080513, -0.025817871, -0.04586792, 0.00579834, -0.00818634, -0.0236969, -0.00983429, 0.004611969, 0.050048828, -0.008811951, 0.008102417, 0.039855957, -0.027908325, -0.017364502, 0.02720642, -0.003019333, 0.0062789...
http://arxiv.org/abs/2309.08845v1
Has Sentiment Returned to the Pre-pandemic Level? A Sentiment Analysis Using U.S. College Subreddit Data from 2019 to 2022
Tian Yan, Fang Liu
cs.CL, cs.CY
As impact of COVID-19 pandemic winds down, both individuals and society gradually return to pre-pandemic activities. This study aims to explore how people's emotions have changed from the pre-pandemic during the pandemic to post-emergency period and whether it has returned to pre-pandemic level. We collected Reddit dat...
http://arxiv.org/pdf/2309.08845v1
2023-09-16T02:57:30
2023-09-16T02:57:30
{ "title": [ 0.019515991, 0.010810852, -0.04006958, 0.02041626, 0.0029659271, -0.04031372, 0.049194336, -0.00075387955, 0.049987793, 0.019836426, -0.030288696, -0.0066375732, 0.012008667, -0.015396118, -0.018875122, -0.014305115, 0.011123657, -0....
http://arxiv.org/abs/2309.08836v1
Bias and Fairness in Chatbots: An Overview
Jintang Xue, Yun-Cheng Wang, Chengwei Wei, Xiaofeng Liu, Jonghye Woo, C. -C. Jay Kuo
cs.CL, cs.AI, cs.CY
Chatbots have been studied for more than half a century. With the rapid development of natural language processing (NLP) technologies in recent years, chatbots using large language models (LLMs) have received much attention nowadays. Compared with traditional ones, modern chatbots are more powerful and have been used i...
http://arxiv.org/pdf/2309.08836v1
2023-09-16T02:01:18
2023-09-16T02:01:18
{ "title": [ 0.0034599304, -0.014541626, -0.02809143, -0.03781128, 0.0021762848, -0.013023376, -0.0010175705, 0.0038089752, -0.00894928, 0.046417236, -0.04660034, 0.0049972534, 0.02734375, -0.003850937, -0.0418396, -0.0065307617, -0.02947998, 0.0...
http://arxiv.org/abs/2310.18316v1
Cognitive modeling and learning with sparse binary hypervectors
Zhonghao Yang
cs.AI, cs.CL
Following the general theoretical framework of VSA (Vector Symbolic Architecture), a cognitive model with the use of sparse binary hypervectors is proposed. In addition, learning algorithms are introduced to bootstrap the model from incoming data stream, with much improved transparency and efficiency. Mimicking human c...
http://arxiv.org/pdf/2310.18316v1
2023-09-16T01:58:51
2023-09-16T01:58:51
{ "title": [ -0.02053833, -0.00894928, -0.022506714, -0.031082153, -0.014335632, -0.0011072159, -0.034118652, -0.027297974, -0.003643036, 0.035095215, -0.011756897, -0.0051841736, -0.0036087036, -0.024520874, -0.04953003, -0.0013895035, 0.013191223, ...
http://arxiv.org/abs/2309.08832v1
SLIDE: Reference-free Evaluation for Machine Translation using a Sliding Document Window
Vikas Raunak, Tom Kocmi, Matt Post
cs.CL, cs.AI
Reference-based metrics that operate at the sentence level typically outperform quality estimation metrics, which have access only to the source and system output. This is unsurprising, since references resolve ambiguities that may be present in the source. We investigate whether additional source context can effective...
http://arxiv.org/pdf/2309.08832v1
2023-09-16T01:30:58
2023-09-16T01:30:58
{ "title": [ -0.031555176, -0.017333984, -0.017074585, -0.04840088, -0.01789856, 0.00762558, -0.031143188, 0.005302429, -0.002407074, 0.043273926, -0.04510498, 0.008735657, 0.027511597, -0.0082473755, -0.048614502, -0.0043144226, -0.015655518, 0....
http://arxiv.org/abs/2309.08827v1
S3-DST: Structured Open-Domain Dialogue Segmentation and State Tracking in the Era of LLMs
Sarkar Snigdha Sarathi Das, Chirag Shah, Mengting Wan, Jennifer Neville, Longqi Yang, Reid Andersen, Georg Buscher, Tara Safavi
cs.CL, cs.AI
The traditional Dialogue State Tracking (DST) problem aims to track user preferences and intents in user-agent conversations. While sufficient for task-oriented dialogue systems supporting narrow domain applications, the advent of Large Language Model (LLM)-based chat systems has introduced many real-world intricacies ...
http://arxiv.org/pdf/2309.08827v1
2023-09-16T00:59:23
2023-09-16T00:59:23
{ "title": [ -0.012557983, -0.005252838, -0.042877197, -0.088378906, -0.008399963, -0.0017299652, -0.06640625, -0.0118255615, -0.029510498, -0.007987976, -0.0088272095, 0.02470398, -0.040863037, -0.031799316, 0.0002462864, -0.015167236, -0.0061302185...
http://arxiv.org/abs/2309.10687v2
EchoPrompt: Instructing the Model to Rephrase Queries for Improved In-context Learning
Rajasekhar Reddy Mekala, Yasaman Razeghi, Sameer Singh
cs.CL
Language models are achieving impressive performance on various tasks by aggressively adopting inference-time prompting techniques, such as zero-shot and few-shot prompting. In this work, we introduce EchoPrompt, a simple yet effective approach that prompts the model to rephrase its queries before answering them. EchoP...
http://arxiv.org/pdf/2309.10687v2
2023-10-13T03:38:57
2023-09-16T00:55:08
{ "title": [ 0.027801514, -0.004119873, -0.038024902, -0.05126953, 0.023010254, -0.029327393, -0.008331299, -0.0256958, 0.008544922, 0.038879395, -0.05065918, 0.020614624, 0.006011963, -0.008773804, -0.027801514, 0.002462387, -0.011390686, 0.0062...
http://arxiv.org/abs/2309.08777v1
An Empirical Study on Instance Selection Strategies in Self-training for Sentiment Analysis
Haochen Liu, Sai Krishna Rallabandi, Yijing Wu, Parag Pravin Dakle, Preethi Raghavan
cs.CL
Sentiment analysis is a crucial task in natural language processing that involves identifying and extracting subjective sentiment from text. Self-training has recently emerged as an economical and efficient technique for developing sentiment analysis models by leveraging a small amount of labeled data and a larger amou...
http://arxiv.org/pdf/2309.08777v1
2023-09-15T21:42:46
2023-09-15T21:42:46
{ "title": [ 0.026031494, -0.0060539246, -0.016983032, -0.020080566, 0.0038757324, -0.030639648, -0.0017223358, -0.03955078, -0.008110046, 0.022537231, -0.0022010803, 0.012397766, -0.03753662, -0.022949219, -0.052490234, -0.020858765, -0.0001437664, ...
http://arxiv.org/abs/2309.08741v1
AlbNER: A Corpus for Named Entity Recognition in Albanian
Erion Çano
cs.CL, cs.AI, cs.LG
Scarcity of resources such as annotated text corpora for under-resourced languages like Albanian is a serious impediment in computational linguistics and natural language processing research. This paper presents AlbNER, a corpus of 900 sentences with labeled named entities, collected from Albanian Wikipedia articles. P...
http://arxiv.org/pdf/2309.08741v1
2023-09-15T20:03:19
2023-09-15T20:03:19
{ "title": [ -0.029541016, -0.010002136, -0.02960205, -0.05618286, 0.00308609, 0.010421753, -0.046966553, -0.0017366409, 0.0053367615, 0.036499023, -0.034484863, 0.010986328, 0.03845215, -0.0028324127, -0.030181885, 0.01763916, -0.0060653687, 0.0...
http://arxiv.org/abs/2309.09992v1
OpenAI Cribbed Our Tax Example, But Can GPT-4 Really Do Tax?
Andrew Blair-Stanek, Nils Holzenberger, Benjamin Van Durme
cs.AI, cs.CL, I.2.7; I.2.0
The authors explain where OpenAI got the tax law example in its livestream demonstration of GPT-4, why GPT-4 got the wrong answer, and how it fails to reliably calculate taxes.
http://arxiv.org/pdf/2309.09992v1
2023-09-15T20:00:27
2023-09-15T20:00:27
{ "title": [ 0.005897522, 0.009712219, -0.0068244934, 0.012176514, -0.007282257, -0.021514893, 0.03189087, -0.045166016, -0.01991272, 0.005760193, -0.015388489, -0.003206253, -0.0018033981, -0.046325684, 0.0060043335, -0.05343628, -0.00806427, -0...
http://arxiv.org/abs/2309.08730v2
MusiLingo: Bridging Music and Text with Pre-trained Language Models for Music Captioning and Query Response
Zihao Deng, Yinghao Ma, Yudong Liu, Rongchen Guo, Ge Zhang, Wenhu Chen, Wenhao Huang, Emmanouil Benetos
eess.AS, cs.AI, cs.CL, cs.MM, cs.SD
Large Language Models (LLMs) have shown immense potential in multimodal applications, yet the convergence of textual and musical domains remains relatively unexplored. To address this gap, we present MusiLingo, a novel system for music caption generation and music-related query responses. MusiLingo employs a single pro...
http://arxiv.org/pdf/2309.08730v2
2023-10-12T21:28:02
2023-09-15T19:31:40
{ "title": [ -0.005302429, -0.018310547, -0.04425049, -0.06542969, -0.003704071, -0.013267517, -0.042877197, 0.023406982, -0.00024843216, 0.05783081, -0.024414062, 0.01689148, 0.04006958, -0.012626648, -0.034729004, 0.0076026917, -0.011985779, 0....
http://arxiv.org/abs/2309.08714v1
Generating Semantic Graph Corpora with Graph Expansion Grammar
Eric Andersson, Johanna Björklund, Frank Drewes, Anna Jonsson
cs.FL, cs.CL, F.4.3; I.2.7
We introduce Lovelace, a tool for creating corpora of semantic graphs. The system uses graph expansion grammar as a representational language, thus allowing users to craft a grammar that describes a corpus with desired properties. When given such grammar as input, the system generates a set of output graphs that are we...
http://arxiv.org/pdf/2309.08714v1
2023-09-15T19:10:19
2023-09-15T19:10:19
{ "title": [ -0.04006958, -0.012954712, -0.072143555, -0.0791626, 0.0017242432, 0.011772156, -0.070617676, 0.020935059, -0.0053100586, 0.017990112, -0.02268982, 0.018310547, -0.025146484, -0.018295288, -0.030166626, -0.0034255981, -0.016906738, 0...
http://arxiv.org/abs/2309.08708v1
Frustratingly Simple Memory Efficiency for Pre-trained Language Models via Dynamic Embedding Pruning
Miles Williams, Nikolaos Aletras
cs.CL
The extensive memory footprint of pre-trained language models (PLMs) can hinder deployment in memory-constrained settings, such as cloud environments or on-device. PLMs use embedding matrices to represent extensive vocabularies, forming a large proportion of the model parameters. While previous work towards parameter-e...
http://arxiv.org/pdf/2309.08708v1
2023-09-15T19:00:00
2023-09-15T19:00:00
{ "title": [ -0.00762558, -0.009796143, -0.025543213, -0.072021484, -0.009529114, -0.0052261353, -0.03933716, 0.011238098, 0.011787415, 0.02633667, -0.022476196, 0.008758545, 0.01902771, -0.017868042, -0.030288696, 0.013496399, 0.007293701, 0.050...
http://arxiv.org/abs/2309.08695v1
Resolving Legalese: A Multilingual Exploration of Negation Scope Resolution in Legal Documents
Ramona Christen, Anastassia Shaitarova, Matthias Stürmer, Joel Niklaus
cs.CL, cs.AI, cs.LG, 68T50, I.2
Resolving the scope of a negation within a sentence is a challenging NLP task. The complexity of legal texts and the lack of annotated in-domain negation corpora pose challenges for state-of-the-art (SotA) models when performing negation scope resolution on multilingual legal data. Our experiments demonstrate that mode...
http://arxiv.org/pdf/2309.08695v1
2023-09-15T18:38:06
2023-09-15T18:38:06
{ "title": [ -0.025497437, -0.03060913, -0.008087158, -0.0758667, -0.022262573, 0.017547607, -0.0680542, 0.012756348, 0.0033798218, 0.04852295, -0.046783447, 0.008659363, 0.03756714, -0.003730774, -0.0385437, -0.0011024475, -0.02154541, 0.0629882...
http://arxiv.org/abs/2309.08674v1
Fake News Detectors are Biased against Texts Generated by Large Language Models
Jinyan Su, Terry Yue Zhuo, Jonibek Mansurov, Di Wang, Preslav Nakov
cs.CL, cs.AI
The spread of fake news has emerged as a critical challenge, undermining trust and posing threats to society. In the era of Large Language Models (LLMs), the capability to generate believable fake content has intensified these concerns. In this study, we present a novel paradigm to evaluate fake news detectors in scena...
http://arxiv.org/pdf/2309.08674v1
2023-09-15T18:04:40
2023-09-15T18:04:40
{ "title": [ -0.0050239563, 0.0054626465, -0.025100708, -0.009185791, -0.018981934, -0.021560669, -0.0030918121, -0.03857422, -0.029281616, 0.037628174, -0.037353516, 0.010681152, 0.016021729, -0.015129089, -0.013374329, -0.011039734, -0.027496338, ...
http://arxiv.org/abs/2309.08600v3
Sparse Autoencoders Find Highly Interpretable Features in Language Models
Hoagy Cunningham, Aidan Ewart, Logan Riggs, Robert Huben, Lee Sharkey
cs.LG, cs.CL
One of the roadblocks to a better understanding of neural networks' internals is \textit{polysemanticity}, where neurons appear to activate in multiple, semantically distinct contexts. Polysemanticity prevents us from identifying concise, human-understandable explanations for what neural networks are doing internally. ...
http://arxiv.org/pdf/2309.08600v3
2023-10-04T13:17:38
2023-09-15T17:56:55
{ "title": [ -0.01979065, 0.003566742, -0.04901123, -0.06124878, -0.0063285828, -0.012931824, -0.072753906, 0.0037384033, 0.0048599243, 0.029708862, -0.0024356842, -0.013389587, 0.024917603, -0.025115967, -0.03225708, 0.023422241, 0.0049095154, 0...
http://arxiv.org/abs/2309.08594v1
"Merge Conflicts!" Exploring the Impacts of External Distractors to Parametric Knowledge Graphs
Cheng Qian, Xinran Zhao, Sherry Tongshuang Wu
cs.CL
Large language models (LLMs) acquire extensive knowledge during pre-training, known as their parametric knowledge. However, in order to remain up-to-date and align with human instructions, LLMs inevitably require external knowledge during their interactions with users. This raises a crucial question: How will LLMs resp...
http://arxiv.org/pdf/2309.08594v1
2023-09-15T17:47:59
2023-09-15T17:47:59
{ "title": [ 0.0022640228, -0.0069847107, -0.024154663, -0.07531738, -0.011169434, -8.34465e-7, -0.047088623, -0.0119018555, -0.008110046, 0.011398315, -0.056488037, -0.012290955, -0.0019741058, -0.004486084, -0.023635864, -0.047332764, 0.012466431, ...
http://arxiv.org/abs/2309.08591v1
Are Multilingual LLMs Culturally-Diverse Reasoners? An Investigation into Multicultural Proverbs and Sayings
Chen Cecilia Liu, Fajri Koto, Timothy Baldwin, Iryna Gurevych
cs.CL
Large language models (LLMs) are highly adept at question answering and reasoning tasks, but when reasoning in situational context, human expectations vary depending on the relevant cultural common ground. As human languages are associated with diverse cultures, LLMs should also be culturally-diverse reasoners. In this...
http://arxiv.org/pdf/2309.08591v1
2023-09-15T17:45:28
2023-09-15T17:45:28
{ "title": [ -0.038269043, -0.02558899, -0.019256592, -0.025177002, -0.028213501, 0.0063667297, -0.03756714, 0.0031089783, -0.019882202, 0.020446777, -0.022354126, 0.02909851, -0.0054473877, -0.021224976, -0.014076233, -0.0075645447, -0.020019531, ...
http://arxiv.org/abs/2309.08590v1
Neural Machine Translation Models Can Learn to be Few-shot Learners
Raphael Reinauer, Patrick Simianer, Kaden Uhlig, Johannes E. M. Mosig, Joern Wuebker
cs.CL
The emergent ability of Large Language Models to use a small number of examples to learn to perform in novel domains and tasks, also called in-context learning (ICL). In this work, we show that a much smaller model can be trained to perform ICL by fine-tuning towards a specialized training objective, exemplified on the...
http://arxiv.org/pdf/2309.08590v1
2023-09-15T17:44:21
2023-09-15T17:44:21
{ "title": [ -0.0028915405, -0.010726929, -0.00233078, -0.0041007996, -0.046966553, -0.01360321, -0.01524353, -0.010940552, -0.01727295, 0.034332275, -0.024017334, -0.0046195984, -0.0038814545, -0.028823853, -0.034057617, -0.012649536, 0.013221741, ...
http://arxiv.org/abs/2309.08589v2
Chain-of-Thought Reasoning is a Policy Improvement Operator
Hugh Zhang, David C. Parkes
cs.LG, cs.AI, cs.CL
Large language models have astounded the world with fascinating new capabilities. However, they currently lack the ability to teach themselves new skills, relying instead on large amounts of human-generated training data. We introduce SECToR (Self-Education via Chain-of-Thought Reasoning), a proof-of-concept demonstrat...
http://arxiv.org/pdf/2309.08589v2
2023-11-08T18:10:10
2023-09-15T17:44:17
{ "title": [ 0.08105469, -0.00044560432, 0.00026249886, -0.07293701, -0.010650635, -0.03656006, -0.011405945, -0.009010315, -0.0690918, 0.0035133362, -0.026275635, -0.005130768, -0.003944397, -0.05895996, 0.0625, -0.031951904, -0.0065727234, 0.02...
http://arxiv.org/abs/2309.08583v1
ICLEF: In-Context Learning with Expert Feedback for Explainable Style Transfer
Arkadiy Saakyan, Smaranda Muresan
cs.CL
While state-of-the-art language models excel at the style transfer task, current work does not address explainability of style transfer systems. Explanations could be generated using large language models such as GPT-3.5 and GPT-4, but the use of such complex systems is inefficient when smaller, widely distributed, and...
http://arxiv.org/pdf/2309.08583v1
2023-09-15T17:41:14
2023-09-15T17:41:14
{ "title": [ -0.01965332, -0.012924194, -0.037231445, -0.03967285, 0.00077724457, -0.00007992983, -0.028625488, -0.010643005, -0.007926941, 0.020645142, -0.0022068024, 0.023544312, 0.012161255, -0.034179688, 0.013000488, 0.014076233, 0.0074005127, ...
http://arxiv.org/abs/2309.08573v1
Casteist but Not Racist? Quantifying Disparities in Large Language Model Bias between India and the West
Khyati Khandelwal, Manuel Tonneau, Andrew M. Bean, Hannah Rose Kirk, Scott A. Hale
cs.CL, cs.CY
Large Language Models (LLMs), now used daily by millions of users, can encode societal biases, exposing their users to representational harms. A large body of scholarship on LLM bias exists but it predominantly adopts a Western-centric frame and attends comparatively less to bias levels and potential harms in the Globa...
http://arxiv.org/pdf/2309.08573v1
2023-09-15T17:38:41
2023-09-15T17:38:41
{ "title": [ -0.02281189, 0.002090454, -0.042755127, -0.021957397, -0.006149292, -0.028686523, -0.028457642, -0.02015686, -0.020706177, 0.010398865, -0.03250122, 0.0129470825, 0.005050659, -0.022537231, -0.02255249, -0.027938843, -0.027282715, 0....
http://arxiv.org/abs/2309.08565v1
How Transferable are Attribute Controllers on Pretrained Multilingual Translation Models?
Danni Liu, Jan Niehues
cs.CL, cs.AI
Customizing machine translation models to comply with fine-grained attributes such as formality has seen tremendous progress recently. However, current approaches mostly rely on at least some supervised data with attribute annotation. Data scarcity therefore remains a bottleneck to democratizing such customization poss...
http://arxiv.org/pdf/2309.08565v1
2023-09-15T17:33:24
2023-09-15T17:33:24
{ "title": [ -0.00096321106, -0.015434265, 0.011207581, -0.035064697, -0.044036865, 0.0064582825, -0.044525146, -0.01763916, -0.023727417, 0.05709839, 0.026245117, -0.0014247894, 0.052093506, -0.048431396, -0.0037899017, 0.019088745, 0.0073013306, ...
http://arxiv.org/abs/2309.08551v1
Augmenting conformers with structured state space models for online speech recognition
Haozhe Shan, Albert Gu, Zhong Meng, Weiran Wang, Krzysztof Choromanski, Tara Sainath
cs.CL, cs.SD, eess.AS
Online speech recognition, where the model only accesses context to the left, is an important and challenging use case for ASR systems. In this work, we investigate augmenting neural encoders for online ASR by incorporating structured state-space sequence models (S4), which are a family of models that provide a paramet...
http://arxiv.org/pdf/2309.08551v1
2023-09-15T17:14:17
2023-09-15T17:14:17
{ "title": [ -0.017791748, -0.0021686554, -0.0435791, -0.11065674, -0.00037527084, 0.019042969, -0.08154297, 0.027267456, 0.005340576, 0.013130188, -0.047698975, 0.016220093, -0.024719238, -0.0065193176, -0.030563354, -0.020233154, 0.008613586, 0...
http://arxiv.org/abs/2309.08541v1
When do Generative Query and Document Expansions Fail? A Comprehensive Study Across Methods, Retrievers, and Datasets
Orion Weller, Kyle Lo, David Wadden, Dawn Lawrie, Benjamin Van Durme, Arman Cohan, Luca Soldaini
cs.IR, cs.AI, cs.CL
Using large language models (LMs) for query or document expansion can improve generalization in information retrieval. However, it is unknown whether these techniques are universally beneficial or only effective in specific settings, such as for particular retrieval models, dataset domains, or query types. To answer th...
http://arxiv.org/pdf/2309.08541v1
2023-09-15T17:05:43
2023-09-15T17:05:43
{ "title": [ 0.007598877, -0.012016296, -0.030563354, -0.05633545, 0.0034618378, -0.020523071, -0.0340271, -0.017044067, 0.026260376, 0.0340271, -0.068481445, 0.02305603, 0.011985779, 0.001616478, -0.04296875, -0.0033512115, 0.002008438, 0.015869...
http://arxiv.org/abs/2309.08532v1
Connecting Large Language Models with Evolutionary Algorithms Yields Powerful Prompt Optimizers
Qingyan Guo, Rui Wang, Junliang Guo, Bei Li, Kaitao Song, Xu Tan, Guoqing Liu, Jiang Bian, Yujiu Yang
cs.CL, cs.AI
Large Language Models (LLMs) excel in various tasks, but they rely on carefully crafted prompts that often demand substantial human effort. To automate this process, in this paper, we propose a novel framework for discrete prompt optimization, called EvoPrompt, which borrows the idea of evolutionary algorithms (EAs) as...
http://arxiv.org/pdf/2309.08532v1
2023-09-15T16:50:09
2023-09-15T16:50:09
{ "title": [ -0.013069153, -0.012039185, -0.012023926, -0.042053223, -0.032287598, -0.012054443, -0.062683105, -0.01133728, -0.012237549, 0.03086853, 0.009300232, 0.0045051575, 0.003112793, -0.038360596, -0.007686615, 0.009315491, -0.0038318634, ...
http://arxiv.org/abs/2309.08531v1
Towards Practical and Efficient Image-to-Speech Captioning with Vision-Language Pre-training and Multi-modal Tokens
Minsu Kim, Jeongsoo Choi, Soumi Maiti, Jeong Hun Yeo, Shinji Watanabe, Yong Man Ro
cs.CV, cs.CL, eess.AS, eess.IV
In this paper, we propose methods to build a powerful and efficient Image-to-Speech captioning (Im2Sp) model. To this end, we start with importing the rich knowledge related to image comprehension and language modeling from a large-scale pre-trained vision-language model into Im2Sp. We set the output of the proposed Im...
http://arxiv.org/pdf/2309.08531v1
2023-09-15T16:48:34
2023-09-15T16:48:34
{ "title": [ 0.021972656, -0.007293701, -0.023956299, -0.07299805, 0.0046539307, -0.026016235, -0.024307251, -0.0040245056, 0.006713867, 0.04763794, -0.027328491, 0.026184082, 0.014343262, -0.008117676, -0.039611816, -0.0039253235, 0.0054359436, ...
http://arxiv.org/abs/2309.08503v1
HealthFC: A Dataset of Health Claims for Evidence-Based Medical Fact-Checking
Juraj Vladika, Phillip Schneider, Florian Matthes
cs.CL, cs.AI
Seeking health-related advice on the internet has become a common practice in the digital era. Determining the trustworthiness of medical claims found online and finding appropriate evidence for this information is increasingly challenging. Fact-checking has emerged as an approach to assess the veracity of factual clai...
http://arxiv.org/pdf/2309.08503v1
2023-09-15T16:05:48
2023-09-15T16:05:48
{ "title": [ -0.005809784, -0.01133728, -0.06060791, -0.081970215, 0.013923645, -0.0181427, -0.03982544, 0.014472961, -0.038726807, 0.03466797, -0.019882202, 0.048797607, 0.0006175041, -0.0184021, -0.0057754517, -0.003753662, -0.00793457, 0.03991...
http://arxiv.org/abs/2309.08491v1
Using Large Language Models for Knowledge Engineering (LLMKE): A Case Study on Wikidata
Bohui Zhang, Ioannis Reklos, Nitisha Jain, Albert Meroño Peñuela, Elena Simperl
cs.CL, cs.AI
In this work, we explore the use of Large Language Models (LLMs) for knowledge engineering tasks in the context of the ISWC 2023 LM-KBC Challenge. For this task, given subject and relation pairs sourced from Wikidata, we utilize pre-trained LLMs to produce the relevant objects in string format and link them to their re...
http://arxiv.org/pdf/2309.08491v1
2023-09-15T15:51:14
2023-09-15T15:51:14
{ "title": [ -0.055236816, -0.024765015, -0.07873535, -0.04663086, 0.0052986145, -0.0008454323, -0.051605225, 0.019897461, 0.0002541542, 0.032104492, 0.00088596344, -0.026168823, 0.0002696514, -0.020523071, -0.05935669, -0.0024471283, -0.0048942566, ...
http://arxiv.org/abs/2309.08469v1
SilverRetriever: Advancing Neural Passage Retrieval for Polish Question Answering
Piotr Rybak, Maciej Ogrodniczuk
cs.CL, cs.IR
Modern open-domain question answering systems often rely on accurate and efficient retrieval components to find passages containing the facts necessary to answer the question. Recently, neural retrievers have gained popularity over lexical alternatives due to their superior performance. However, most of the work concer...
http://arxiv.org/pdf/2309.08469v1
2023-09-15T15:19:53
2023-09-15T15:19:53
{ "title": [ -0.012199402, -0.03527832, -0.009338379, -0.026779175, -0.037963867, -0.0076026917, -0.018371582, -0.00096702576, 0.023651123, 0.057037354, -0.016799927, -0.0049591064, 0.019500732, -0.0023612976, -0.07733154, 0.018554688, 0.003900528, ...
http://arxiv.org/abs/2309.08650v1
Adversarial Attacks on Tables with Entity Swap
Aneta Koleva, Martin Ringsquandl, Volker Tresp
cs.CL, cs.AI, cs.CR
The capabilities of large language models (LLMs) have been successfully applied in the context of table representation learning. The recently proposed tabular language models have reported state-of-the-art results across various tasks for table interpretation. However, a closer look into the datasets commonly used for ...
http://arxiv.org/pdf/2309.08650v1
2023-09-15T15:03:33
2023-09-15T15:03:33
{ "title": [ -0.002456665, 0.022720337, -0.042175293, -0.032684326, 0.03466797, -0.024398804, 0.014259338, -0.018692017, -0.030883789, 0.054138184, 0.01776123, 0.012611389, -0.005104065, -0.024612427, -0.036224365, 0.010543823, 0.00053167343, 0.0...
http://arxiv.org/abs/2309.08454v1
Mixture Encoder Supporting Continuous Speech Separation for Meeting Recognition
Peter Vieting, Simon Berger, Thilo von Neumann, Christoph Boeddeker, Ralf Schlüter, Reinhold Haeb-Umbach
eess.AS, cs.CL, cs.LG, cs.SD
Many real-life applications of automatic speech recognition (ASR) require processing of overlapped speech. A commonmethod involves first separating the speech into overlap-free streams and then performing ASR on the resulting signals. Recently, the inclusion of a mixture encoder in the ASR model has been proposed. This...
http://arxiv.org/pdf/2309.08454v1
2023-09-15T14:57:28
2023-09-15T14:57:28
{ "title": [ -0.0077438354, 0.021835327, -0.10491943, -0.07489014, 0.015159607, -0.03842163, -0.04296875, 0.022628784, 0.002773285, 0.06488037, -0.014808655, 0.016540527, 0.05718994, -0.026763916, -0.019546509, 0.0262146, -0.0059127808, 0.0473327...
http://arxiv.org/abs/2309.08448v2
Advancing the Evaluation of Traditional Chinese Language Models: Towards a Comprehensive Benchmark Suite
Chan-Jan Hsu, Chang-Le Liu, Feng-Ting Liao, Po-Chun Hsu, Yi-Chang Chen, Da-shan Shiu
cs.CL
The evaluation of large language models is an essential task in the field of language understanding and generation. As language models continue to advance, the need for effective benchmarks to assess their performance has become imperative. In the context of Traditional Chinese, there is a scarcity of comprehensive and...
http://arxiv.org/pdf/2309.08448v2
2023-10-02T15:22:42
2023-09-15T14:52:23
{ "title": [ -0.0446167, 0.004722595, -0.023956299, -0.043670654, 0.0008678436, 0.009719849, -0.055114746, -0.005455017, 0.0042304993, 0.043670654, 0.009674072, -0.0031929016, 0.016540527, -0.03515625, -0.068603516, 0.021209717, -0.019882202, 0.0...
http://arxiv.org/abs/2309.08648v1
MAPLE: Mobile App Prediction Leveraging Large Language model Embeddings
Yonchanok Khaokaew, Hao Xue, Flora D. Salim
cs.CL, cs.AI
Despite the rapid advancement of mobile applications, predicting app usage remains a formidable challenge due to intricate user behaviours and ever-evolving contexts. To address these issues, this paper introduces the Mobile App Prediction Leveraging Large Language Model Embeddings (MAPLE) model. This innovative approa...
http://arxiv.org/pdf/2309.08648v1
2023-09-15T13:15:54
2023-09-15T13:15:54
{ "title": [ -0.023605347, -0.03189087, -0.01876831, -0.042785645, -0.013923645, 0.00036382675, -0.040222168, 0.005607605, -0.019317627, 0.021911621, -0.006866455, 0.020690918, -0.004169464, -0.021331787, -0.012336731, 0.0061798096, -0.026443481, ...
http://arxiv.org/abs/2309.08647v1
Intent Detection at Scale: Tuning a Generic Model using Relevant Intents
Nichal Narotamo, David Aparicio, Tiago Mesquita, Mariana Almeida
cs.CL, cs.AI
Accurately predicting the intent of customer support requests is vital for efficient support systems, enabling agents to quickly understand messages and prioritize responses accordingly. While different approaches exist for intent detection, maintaining separate client-specific or industry-specific models can be costly...
http://arxiv.org/pdf/2309.08647v1
2023-09-15T13:15:20
2023-09-15T13:15:20
{ "title": [ -0.005657196, 0.012817383, -0.05316162, -0.029159546, 0.019607544, -0.026611328, -0.017868042, -0.03994751, 0.015960693, 0.017181396, -0.019973755, -0.00623703, -0.0023880005, -0.028762817, -0.03225708, 0.004085541, -0.013000488, 0.0...
http://arxiv.org/abs/2309.08380v1
Unleashing Potential of Evidence in Knowledge-Intensive Dialogue Generation
Xianjie Wu, Jian Yang, Tongliang Li, Di Liang, Shiwei Zhang, Yiyang Du, Zhoujun Li
cs.CL
Incorporating external knowledge into dialogue generation (KIDG) is crucial for improving the correctness of response, where evidence fragments serve as knowledgeable snippets supporting the factual dialogue replies. However, introducing irrelevant content often adversely impacts reply quality and easily leads to hallu...
http://arxiv.org/pdf/2309.08380v1
2023-09-15T13:13:30
2023-09-15T13:13:30
{ "title": [ 0.005470276, -0.01386261, -0.050048828, -0.07611084, -0.012084961, -0.022079468, -0.045806885, 0.013366699, -0.020721436, 0.012794495, 0.015396118, 0.0070152283, -0.046966553, -0.0446167, -0.035003662, -0.0000049471855, 0.0042419434, ...
http://arxiv.org/abs/2309.08379v1
PatFig: Generating Short and Long Captions for Patent Figures
Dana Aubakirova, Kim Gerdes, Lufei Liu
cs.CV, cs.CL
This paper introduces Qatent PatFig, a novel large-scale patent figure dataset comprising 30,000+ patent figures from over 11,000 European patent applications. For each figure, this dataset provides short and long captions, reference numerals, their corresponding terms, and the minimal claim set that describes the inte...
http://arxiv.org/pdf/2309.08379v1
2023-09-15T13:10:36
2023-09-15T13:10:36
{ "title": [ -0.009147644, -0.01272583, -0.03753662, -0.020690918, -0.014968872, -0.021514893, -0.0050735474, -0.009613037, 0.008880615, 0.04937744, -0.026763916, 0.013130188, 0.017593384, -0.014785767, -0.043945312, -0.004463196, -0.01953125, 0....
http://arxiv.org/abs/2309.08377v1
DiaCorrect: Error Correction Back-end For Speaker Diarization
Jiangyu Han, Federico Landini, Johan Rohdin, Mireia Diez, Lukas Burget, Yuhang Cao, Heng Lu, Jan Cernocky
eess.AS, cs.CL, cs.SD
In this work, we propose an error correction framework, named DiaCorrect, to refine the output of a diarization system in a simple yet effective way. This method is inspired by error correction techniques in automatic speech recognition. Our model consists of two parallel convolutional encoders and a transform-based de...
http://arxiv.org/pdf/2309.08377v1
2023-09-15T13:08:12
2023-09-15T13:08:12
{ "title": [ -0.021636963, -0.006668091, -0.03744507, -0.034576416, -0.015045166, -0.005264282, -0.042938232, -0.014854431, -0.012229919, 0.061584473, -0.033111572, 0.009017944, 0.055908203, -0.022994995, -0.019241333, 0.0041236877, -0.005302429, ...
http://arxiv.org/abs/2309.08351v1
Headless Language Models: Learning without Predicting with Contrastive Weight Tying
Nathan Godey, Éric de la Clergerie, Benoît Sagot
cs.CL
Self-supervised pre-training of language models usually consists in predicting probability distributions over extensive token vocabularies. In this study, we propose an innovative method that shifts away from probability prediction and instead focuses on reconstructing input embeddings in a contrastive fashion via Cons...
http://arxiv.org/pdf/2309.08351v1
2023-09-15T12:20:00
2023-09-15T12:20:00
{ "title": [ -0.029846191, -0.015327454, -0.03479004, -0.064086914, -0.0025405884, 0.0033092499, -0.02545166, 0.018127441, 0.002149582, 0.029663086, -0.009681702, -0.002746582, -0.0098724365, -0.013725281, -0.047454834, 0.0034160614, -0.0039405823, ...
http://arxiv.org/abs/2309.08347v1
Reward Engineering for Generating Semi-structured Explanation
Jiuzhou Han, Wray Buntine, Ehsan Shareghi
cs.CL
Semi-structured explanation depicts the implicit process of a reasoner with an explicit representation. This explanation highlights how available information in a specific query is supplemented with information a reasoner produces from its internal weights towards generating an answer. Despite the recent improvements i...
http://arxiv.org/pdf/2309.08347v1
2023-09-15T12:10:03
2023-09-15T12:10:03
{ "title": [ -0.016098022, -0.015487671, -0.017944336, -0.04336548, -0.004737854, -0.0029563904, -0.014587402, -0.0040397644, -0.028701782, 0.009422302, 0.03074646, -0.0048103333, -0.034942627, -0.04675293, -0.018341064, -0.0045928955, -0.005836487, ...
http://arxiv.org/abs/2309.08345v1
Data Distribution Bottlenecks in Grounding Language Models to Knowledge Bases
Yiheng Shu, Zhiwei Yu
cs.CL, cs.AI
Language models (LMs) have already demonstrated remarkable abilities in understanding and generating both natural and formal language. Despite these advances, their integration with real-world environments such as large-scale knowledge bases (KBs) remains an underdeveloped area, affecting applications such as semantic ...
http://arxiv.org/pdf/2309.08345v1
2023-09-15T12:06:45
2023-09-15T12:06:45
{ "title": [ -0.04864502, 0.0046463013, -0.057647705, -0.07507324, 0.005619049, 0.011184692, -0.06451416, 0.0028533936, 0.019851685, 0.030456543, -0.060821533, 0.003124237, 0.00491333, -0.0015935898, -0.035186768, -0.0009160042, 0.00079631805, 0....
http://arxiv.org/abs/2309.08325v1
Distributional Inclusion Hypothesis and Quantifications: Probing Hypernymy in Functional Distributional Semantics
Chun Hei Lo, Guy Emerson
cs.CL
Functional Distributional Semantics (FDS) models the meaning of words by truth-conditional functions. This provides a natural representation for hypernymy, but no guarantee that it is learnt when FDS models are trained on a corpus. We demonstrate that FDS models learn hypernymy when a corpus strictly follows the Distri...
http://arxiv.org/pdf/2309.08325v1
2023-09-15T11:28:52
2023-09-15T11:28:52
{ "title": [ -0.012275696, 0.024276733, -0.047729492, -0.039978027, 0.029876709, -0.012924194, -0.0028858185, -0.01386261, -0.007232666, 0.03930664, -0.022262573, 0.031585693, -0.01108551, -0.00078487396, -0.052490234, 0.0071983337, -0.018844604, ...
http://arxiv.org/abs/2309.08316v1
Bridging Topic, Domain, and Language Shifts: An Evaluation of Comprehensive Out-of-Distribution Scenarios
Andreas Waldis, Iryna Gurevych
cs.CL
Language models (LMs) excel in in-distribution (ID) scenarios where train and test data are independent and identically distributed. However, their performance often degrades in real-world applications like argument mining. Such degradation happens when new topics emerge, or other text domains and languages become rele...
http://arxiv.org/pdf/2309.08316v1
2023-09-15T11:15:47
2023-09-15T11:15:47
{ "title": [ 0.014472961, -0.021377563, -0.007144928, -0.055145264, 0.0025939941, -0.009819031, -0.018997192, -0.02330017, 0.015686035, 0.019104004, 0.025177002, 0.015930176, -0.0013952255, -0.050354004, -0.00083875656, 0.019042969, -0.012489319, ...
http://arxiv.org/abs/2309.08303v1
Self-Consistent Narrative Prompts on Abductive Natural Language Inference
Chunkit Chan, Xin Liu, Tsz Ho Chan, Jiayang Cheng, Yangqiu Song, Ginny Wong, Simon See
cs.CL
Abduction has long been seen as crucial for narrative comprehension and reasoning about everyday situations. The abductive natural language inference ($\alpha$NLI) task has been proposed, and this narrative text-based task aims to infer the most plausible hypothesis from the candidates given two observations. However, ...
http://arxiv.org/pdf/2309.08303v1
2023-09-15T10:48:10
2023-09-15T10:48:10
{ "title": [ -0.018173218, -0.005355835, -0.021850586, -0.043762207, -0.012512207, 0.0038471222, -0.0137786865, -0.012786865, -0.040222168, -0.0040512085, -0.009109497, 0.0076942444, -0.028381348, -0.0289917, 0.00041794777, -0.0132751465, -0.01038360...
http://arxiv.org/abs/2309.08646v2
CoCA: Fusing position embedding with Collinear Constrained Attention for fine-tuning free context window extending
Shiyi Zhu, Jing Ye, Wei Jiang, Qi Zhang, Yifan Wu, Jianguo Li
cs.LG, cs.AI, cs.CL
Self-attention and position embedding are two key modules in Transformer based LLMs. The potential relationship among them are far from well studied, especially for context window extending. In this paper, we introduce collinear constrained relationship to fuse RoPE and self-attention, and name it as Collinear Constrai...
http://arxiv.org/pdf/2309.08646v2
2023-11-08T08:45:45
2023-09-15T09:36:51
{ "title": [ 0.012496948, -0.015403748, -0.03781128, -0.06536865, -0.0033473969, -0.010643005, -0.00945282, -0.008628845, 0.032165527, 0.04901123, -0.035705566, 0.005393982, 0.016845703, 0.004299164, -0.046783447, -0.000095784664, -0.009109497, 0...
http://arxiv.org/abs/2309.08272v1
Structural Self-Supervised Objectives for Transformers
Luca Di Liello
cs.CL, cs.IR
This thesis focuses on improving the pre-training of natural language models using unsupervised raw data to make them more efficient and aligned with downstream applications. In the first part, we introduce three alternative pre-training objectives to BERT's Masked Language Modeling (MLM), namely Random Token Substit...
http://arxiv.org/pdf/2309.08272v1
2023-09-15T09:30:45
2023-09-15T09:30:45
{ "title": [ 0.0149002075, 0.03527832, -0.027862549, -0.04550171, 0.0090789795, -0.011444092, -0.023971558, -0.028411865, -0.037750244, 0.023544312, -0.008216858, 0.0049972534, -0.032592773, -0.034454346, 0.0012226105, 0.0057678223, 0.0068244934, ...
http://arxiv.org/abs/2309.08255v1
Cross-lingual Knowledge Distillation via Flow-based Voice Conversion for Robust Polyglot Text-To-Speech
Dariusz Piotrowski, Renard Korzeniowski, Alessio Falai, Sebastian Cygert, Kamil Pokora, Georgi Tinchev, Ziyao Zhang, Kayoko Yanagisawa
eess.AS, cs.CL, cs.LG, cs.SD
In this work, we introduce a framework for cross-lingual speech synthesis, which involves an upstream Voice Conversion (VC) model and a downstream Text-To-Speech (TTS) model. The proposed framework consists of 4 stages. In the first two stages, we use a VC model to convert utterances in the target locale to the voice o...
http://arxiv.org/pdf/2309.08255v1
2023-09-15T09:03:14
2023-09-15T09:03:14
{ "title": [ -0.008224487, -0.012718201, -0.049316406, -0.11212158, 0.005241394, 0.0020256042, -0.0904541, 0.02758789, 0.019714355, 0.05734253, -0.029373169, 0.0052719116, 0.04019165, -0.008361816, -0.05114746, 0.008102417, 0.00011283159, 0.01977...
http://arxiv.org/abs/2309.08210v1
Investigating Answerability of LLMs for Long-Form Question Answering
Meghana Moorthy Bhat, Rui Meng, Ye Liu, Yingbo Zhou, Semih Yavuz
cs.CL
As we embark on a new era of LLMs, it becomes increasingly crucial to understand their capabilities, limitations, and differences. Toward making further progress in this direction, we strive to build a deeper understanding of the gaps between massive LLMs (e.g., ChatGPT) and smaller yet effective open-source LLMs and t...
http://arxiv.org/pdf/2309.08210v1
2023-09-15T07:22:56
2023-09-15T07:22:56
{ "title": [ -0.035858154, 0.00003618002, -0.031402588, -0.03463745, -0.011413574, 0.0022354126, -0.032562256, -0.0038795471, 0.015808105, 0.007423401, -0.030960083, 0.002878189, -0.025985718, -0.013420105, -0.049346924, -0.017654419, -0.008926392, ...
http://arxiv.org/abs/2309.08187v1
Encoded Summarization: Summarizing Documents into Continuous Vector Space for Legal Case Retrieval
Vu Tran, Minh Le Nguyen, Satoshi Tojo, Ken Satoh
cs.CL
We present our method for tackling a legal case retrieval task by introducing our method of encoding documents by summarizing them into continuous vector space via our phrase scoring framework utilizing deep neural networks. On the other hand, we explore the benefits from combining lexical features and latent features ...
http://arxiv.org/pdf/2309.08187v1
2023-09-15T06:33:49
2023-09-15T06:33:49
{ "title": [ 0.006477356, -0.008453369, -0.026641846, -0.047180176, 0.00022232533, -0.012763977, -0.012420654, -0.0056533813, 0.01335144, 0.04647827, -0.04321289, 0.008071899, 0.04067993, -0.010345459, -0.0579834, 0.00006812811, -0.001209259, 0.0...
http://arxiv.org/abs/2309.08185v1
Multilingual Sentence-Level Semantic Search using Meta-Distillation Learning
Meryem M'hamdi, Jonathan May, Franck Dernoncourt, Trung Bui, Seunghyun Yoon
cs.CL
Multilingual semantic search is the task of retrieving relevant contents to a query expressed in different language combinations. This requires a better semantic understanding of the user's intent and its contextual meaning. Multilingual semantic search is less explored and more challenging than its monolingual or bili...
http://arxiv.org/pdf/2309.08185v1
2023-09-15T06:22:37
2023-09-15T06:22:37
{ "title": [ -0.006828308, -0.0074539185, -0.013046265, -0.051483154, -0.009269714, 0.015106201, -0.060791016, -0.01676941, 0.0440979, 0.05166626, -0.01159668, -0.008865356, 0.0027580261, -0.0022068024, -0.070617676, 0.01007843, -0.015541077, 0.0...
http://arxiv.org/abs/2309.08182v2
Using Large Language Model to Solve and Explain Physics Word Problems Approaching Human Level
Jingzhe Ding, Yan Cen, Xinyuan Wei
cs.CL, cs.AI, I.2.7
Our work demonstrates that large language model (LLM) pre-trained on texts can not only solve pure math word problems, but also physics word problems, whose solution requires calculation and inference based on prior physical knowledge. We collect and annotate the first physics word problem dataset-PhysQA, which contain...
http://arxiv.org/pdf/2309.08182v2
2023-09-20T07:08:53
2023-09-15T06:13:06
{ "title": [ -0.015434265, -0.004611969, -0.021209717, -0.016998291, -0.021530151, -0.022766113, -0.02229309, -0.01979065, -0.02104187, 0.0345459, 0.004486084, -0.005672455, 0.020446777, -0.03137207, -0.0065345764, 0.010437012, -0.004310608, 0.02...
http://arxiv.org/abs/2309.08181v1
Large Language Models for Failure Mode Classification: An Investigation
Michael Stewart, Melinda Hodkiewicz, Sirui Li
cs.CL
In this paper we present the first investigation into the effectiveness of Large Language Models (LLMs) for Failure Mode Classification (FMC). FMC, the task of automatically labelling an observation with a corresponding failure mode code, is a critical task in the maintenance domain as it reduces the need for reliabili...
http://arxiv.org/pdf/2309.08181v1
2023-09-15T06:13:01
2023-09-15T06:13:01
{ "title": [ -0.016937256, -0.013473511, -0.025665283, -0.08947754, 0.00017380714, 0.013580322, -0.08746338, 0.020553589, -0.004798889, 0.027252197, -0.00081300735, -0.0064315796, -0.008094788, -0.035491943, -0.021072388, -0.0047912598, -0.014671326,...
http://arxiv.org/abs/2309.08173v1
FedJudge: Federated Legal Large Language Model
Linan Yue, Qi Liu, Yichao Du, Weibo Gao, Ye Liu, Fangzhou Yao
cs.CL
Large Language Models (LLMs) have gained prominence in the field of Legal Intelligence, offering potential applications in assisting legal professionals and laymen. However, the centralized training of these Legal LLMs raises data privacy concerns, as legal data is distributed among various institutions containing sens...
http://arxiv.org/pdf/2309.08173v1
2023-09-15T05:45:44
2023-09-15T05:45:44
{ "title": [ -0.04638672, -0.021057129, -0.07574463, -0.09844971, 0.017242432, 0.019866943, -0.060455322, 0.04776001, -0.010597229, 0.049072266, -0.041931152, 0.008888245, 0.04611206, -0.0014505386, -0.021118164, 0.013259888, -0.024734497, 0.0438...
http://arxiv.org/abs/2309.08172v1
LASER: LLM Agent with State-Space Exploration for Web Navigation
Kaixin Ma, Hongming Zhang, Hongwei Wang, Xiaoman Pan, Dong Yu
cs.CL
Large language models (LLMs) have been successfully adapted for interactive decision-making tasks like web navigation. While achieving decent performance, previous methods implicitly assume a forward-only execution mode for the model, where they only provide oracle trajectories as in-context examples to teach the model...
http://arxiv.org/pdf/2309.08172v1
2023-09-15T05:44:08
2023-09-15T05:44:08
{ "title": [ -0.028640747, -0.024749756, -0.030303955, -0.037597656, -0.01058197, -0.008087158, -0.02104187, 0.0027770996, -0.010185242, 0.024414062, -0.0413208, 0.02810669, -0.015731812, -0.011062622, -0.016616821, -0.010002136, -0.0140686035, 0...
http://arxiv.org/abs/2309.08168v1
Draft & Verify: Lossless Large Language Model Acceleration via Self-Speculative Decoding
Jun Zhang, Jue Wang, Huan Li, Lidan Shou, Ke Chen, Gang Chen, Sharad Mehrotra
cs.CL
We present a novel inference scheme, self-speculative decoding, for accelerating Large Language Models (LLMs) without the need for an auxiliary model. This approach is characterized by a two-stage process: drafting and verification. The drafting stage generates draft tokens at a slightly lower quality but more quickly,...
http://arxiv.org/pdf/2309.08168v1
2023-09-15T05:34:32
2023-09-15T05:34:32
{ "title": [ -0.0597229, -0.025482178, -0.054870605, -0.07470703, -0.013702393, 0.008544922, -0.0713501, 0.041534424, -0.023132324, 0.058166504, 0.010665894, -0.010360718, 0.02772522, -0.02658081, -0.059295654, 0.017440796, 0.011413574, 0.0310668...
http://arxiv.org/abs/2309.08163v1
Investigating the Applicability of Self-Assessment Tests for Personality Measurement of Large Language Models
Akshat Gupta, Xiaoyang Song, Gopala Anumanchipalli
cs.CL, cs.AI
As large language models (LLM) evolve in their capabilities, various recent studies have tried to quantify their behavior using psychological tools created to study human behavior. One such example is the measurement of "personality" of LLMs using personality self-assessment tests. In this paper, we take three such stu...
http://arxiv.org/pdf/2309.08163v1
2023-09-15T05:19:39
2023-09-15T05:19:39
{ "title": [ -0.046966553, -0.0061912537, -0.0395813, -0.02758789, -0.008262634, -0.008178711, -0.025115967, -0.0015506744, -0.00724411, 0.036254883, 0.012313843, -0.0012407303, -0.009719849, -0.03326416, -0.053894043, 0.013343811, -0.011398315, ...
http://arxiv.org/abs/2310.05964v2
Exploring Embeddings for Measuring Text Relatedness: Unveiling Sentiments and Relationships in Online Comments
Anthony Olakangil, Cindy Wang, Justin Nguyen, Qunbo Zhou, Kaavya Jethwa, Jason Li, Aryan Narendra, Nishk Patel, Arjun Rajaram
cs.CL
After the COVID-19 pandemic caused internet usage to grow by 70%, there has been an increased number of people all across the world using social media. Applications like Twitter, Meta Threads, YouTube, and Reddit have become increasingly pervasive, leaving almost no digital space where public opinion is not expressed. ...
http://arxiv.org/pdf/2310.05964v2
2023-10-30T18:38:39
2023-09-15T04:57:23
{ "title": [ 0.0057525635, 0.0025348663, -0.021499634, -0.055267334, 0.008674622, -0.0067825317, -0.03414917, -0.015655518, 0.019454956, 0.028930664, -0.041900635, 0.02015686, 0.017227173, 0.012954712, -0.04699707, -0.0043258667, -0.011688232, 0....
http://arxiv.org/abs/2309.08156v2
RADE: Reference-Assisted Dialogue Evaluation for Open-Domain Dialogue
Zhengliang Shi, Weiwei Sun, Shuo Zhang, Zhen Zhang, Pengjie Ren, Zhaochun Ren
cs.CL
Evaluating open-domain dialogue systems is challenging for reasons such as the one-to-many problem, i.e., many appropriate responses other than just the golden response. As of now, automatic evaluation methods need better consistency with humans, while reliable human evaluation can be time- and cost-intensive. To this ...
http://arxiv.org/pdf/2309.08156v2
2023-09-18T00:43:47
2023-09-15T04:47:19
{ "title": [ -0.0129776, -0.0022220612, -0.033721924, -0.06124878, 0.010383606, -0.0028839111, -0.041748047, -0.008384705, -0.004486084, 0.02772522, -0.006401062, 0.0079422, 0.004425049, -0.020614624, -0.043182373, -0.000787735, -0.030029297, 0.0...
http://arxiv.org/abs/2309.08150v1
Unimodal Aggregation for CTC-based Speech Recognition
Ying Fang, Xiaofei Li
cs.CL, cs.SD, eess.AS
This paper works on non-autoregressive automatic speech recognition. A unimodal aggregation (UMA) is proposed to segment and integrate the feature frames that belong to the same text token, and thus to learn better feature representations for text tokens. The frame-wise features and weights are both derived from an enc...
http://arxiv.org/pdf/2309.08150v1
2023-09-15T04:34:40
2023-09-15T04:34:40
{ "title": [ -0.011566162, -0.00045084953, -0.057403564, -0.083496094, -0.005771637, 0.0087509155, -0.081726074, 0.024215698, 0.002986908, 0.06100464, -0.044921875, 0.013244629, 0.028793335, 0.006313324, -0.062042236, 0.0048294067, -0.01878357, 0...
http://arxiv.org/abs/2309.08140v1
PromptTTS++: Controlling Speaker Identity in Prompt-Based Text-to-Speech Using Natural Language Descriptions
Reo Shimizu, Ryuichi Yamamoto, Masaya Kawamura, Yuma Shirahata, Hironori Doi, Tatsuya Komatsu, Kentaro Tachibana
eess.AS, cs.CL, cs.LG, cs.SD
We propose PromptTTS++, a prompt-based text-to-speech (TTS) synthesis system that allows control over speaker identity using natural language descriptions. To control speaker identity within the prompt-based TTS framework, we introduce the concept of speaker prompt, which describes voice characteristics (e.g., gender-n...
http://arxiv.org/pdf/2309.08140v1
2023-09-15T04:11:37
2023-09-15T04:11:37
{ "title": [ 0.007396698, 0.000049471855, -0.040527344, -0.06530762, 0.015853882, -0.004348755, -0.032470703, 0.00016915798, -0.01486969, 0.03189087, -0.024337769, 0.014984131, 0.030593872, -0.020492554, -0.021255493, -0.003982544, -0.012878418, ...
http://arxiv.org/abs/2309.08141v1
Audio Difference Learning for Audio Captioning
Tatsuya Komatsu, Yusuke Fujita, Kazuya Takeda, Tomoki Toda
eess.AS, cs.CL, cs.LG, cs.SD, eess.SP
This study introduces a novel training paradigm, audio difference learning, for improving audio captioning. The fundamental concept of the proposed learning method is to create a feature representation space that preserves the relationship between audio, enabling the generation of captions that detail intricate audio i...
http://arxiv.org/pdf/2309.08141v1
2023-09-15T04:11:37
2023-09-15T04:11:37
{ "title": [ -0.010169983, 0.0017690659, -0.04083252, -0.040100098, -0.002922058, -0.02319336, -0.0013837814, 0.008583069, -0.027496338, 0.04660034, -0.045166016, 0.03729248, -0.00038337708, 0.0024147034, -0.019439697, -0.0024356842, -0.022659302, ...
http://arxiv.org/abs/2311.04905v1
Detecting Relevant Information in High-Volume Chat Logs: Keyphrase Extraction for Grooming and Drug Dealing Forensic Analysis
Jeovane Honório Alves, Horácio A. C. G. Pedroso, Rafael Honorio Venetikides, Joel E. M. Köster, Luiz Rodrigo Grochocki, Cinthia O. A. Freitas, Jean Paul Barddal
cs.CL, cs.AI, cs.LG
The growing use of digital communication platforms has given rise to various criminal activities, such as grooming and drug dealing, which pose significant challenges to law enforcement and forensic experts. This paper presents a supervised keyphrase extraction approach to detect relevant information in high-volume cha...
http://arxiv.org/pdf/2311.04905v1
2023-09-15T03:18:31
2023-09-15T03:18:31
{ "title": [ -0.022598267, -0.016677856, -0.039215088, -0.05456543, -0.0051651, 0.0020141602, -0.025314331, 0.029296875, -0.007987976, 0.06591797, 0.017730713, 0.014480591, 0.040039062, -0.024246216, -0.04522705, 0.017303467, -0.028060913, 0.0565...
http://arxiv.org/abs/2309.13065v1
Personality Profiling: How informative are social media profiles in predicting personal information?
Joshua Watt, Jonathan Tuke, Lewis Mitchell
cs.CL, cs.SI
Personality profiling has been utilised by companies for targeted advertising, political campaigns and vaccine campaigns. However, the accuracy and versatility of such models still remains relatively unknown. Consequently, we aim to explore the extent to which peoples' online digital footprints can be used to profile t...
http://arxiv.org/pdf/2309.13065v1
2023-09-15T03:09:43
2023-09-15T03:09:43
{ "title": [ -0.0021896362, 0.028930664, -0.038024902, -0.043304443, 0.030471802, -0.014511108, -0.0019140244, -0.024749756, -0.05911255, 0.029281616, -0.012290955, 0.0418396, 0.0001988411, -0.015266418, 0.010696411, -0.008354187, 0.006996155, 0....
http://arxiv.org/abs/2309.13064v1
InvestLM: A Large Language Model for Investment using Financial Domain Instruction Tuning
Yi Yang, Yixuan Tang, Kar Yan Tam
q-fin.GN, cs.AI, cs.CL, cs.LG
We present a new financial domain large language model, InvestLM, tuned on LLaMA-65B (Touvron et al., 2023), using a carefully curated instruction dataset related to financial investment. Inspired by less-is-more-for-alignment (Zhou et al., 2023), we manually curate a small yet diverse instruction dataset, covering a w...
http://arxiv.org/pdf/2309.13064v1
2023-09-15T02:59:31
2023-09-15T02:59:31
{ "title": [ -0.059509277, -0.01802063, -0.049743652, -0.03869629, -0.01676941, -0.017715454, -0.043762207, 0.020248413, -0.01449585, 0.02407837, -0.01374054, -0.004425049, -0.009208679, -0.021057129, -0.03274536, -0.020492554, -0.032440186, 0.05...
http://arxiv.org/abs/2309.08100v1
Research on Joint Representation Learning Methods for Entity Neighborhood Information and Description Information
Le Xiao, Xin Shan, Yuhua Wang, Miaolei Deng
cs.CL, cs.AI
To address the issue of poor embedding performance in the knowledge graph of a programming design course, a joint represen-tation learning model that combines entity neighborhood infor-mation and description information is proposed. Firstly, a graph at-tention network is employed to obtain the features of entity neigh-...
http://arxiv.org/pdf/2309.08100v1
2023-09-15T01:38:07
2023-09-15T01:38:07
{ "title": [ 0.010414124, 0.011016846, -0.060516357, -0.06677246, 0.032440186, -0.02458191, -0.009849548, 0.0024223328, -0.022644043, 0.04324341, -0.032196045, 0.021240234, 0.026000977, -0.015716553, 0.0038757324, 0.006072998, -0.008117676, 0.053...
http://arxiv.org/abs/2309.08099v1
Characterizing the temporal dynamics of universal speech representations for generalizable deepfake detection
Yi Zhu, Saurabh Powar, Tiago H. Falk
cs.SD, cs.CL, eess.AS
Existing deepfake speech detection systems lack generalizability to unseen attacks (i.e., samples generated by generative algorithms not seen during training). Recent studies have explored the use of universal speech representations to tackle this issue and have obtained inspiring results. These works, however, have fo...
http://arxiv.org/pdf/2309.08099v1
2023-09-15T01:37:45
2023-09-15T01:37:45
{ "title": [ 0.02935791, 0.030578613, -0.05407715, -0.08258057, 0.039031982, -0.021621704, -0.020401001, 0.011077881, -0.022262573, 0.060150146, 0.0070648193, 0.032592773, 0.025024414, -0.014839172, -0.029632568, 0.025619507, -0.02255249, 0.05770...
http://arxiv.org/abs/2309.08069v1
Connecting the Dots in News Analysis: A Cross-Disciplinary Survey of Media Bias and Framing
Gisela Vallejo, Timothy Baldwin, Lea Frermann
cs.CL
The manifestation and effect of bias in news reporting have been central topics in the social sciences for decades, and have received increasing attention in the NLP community recently. While NLP can help to scale up analyses or contribute automatic procedures to investigate the impact of biased news in society, we arg...
http://arxiv.org/pdf/2309.08069v1
2023-09-14T23:57:55
2023-09-14T23:57:55
{ "title": [ -0.011146545, -0.0013256073, -0.026428223, -0.031707764, 0.023880005, -0.015182495, 0.019515991, -0.03741455, -0.043060303, 0.035705566, -0.014335632, 0.029434204, -0.016586304, -0.0060195923, -0.037384033, -0.019226074, -0.02658081, ...
http://arxiv.org/abs/2309.08047v1
Investigating Gender Bias in News Summarization
Julius Steen, Katja Markert
cs.CL
Summarization is an important application of large language models (LLMs). Most previous evaluation of summarization models has focused on their performance in content selection, grammaticality and coherence. However, it is well known that LLMs reproduce and reinforce harmful social biases. This raises the question: Do...
http://arxiv.org/pdf/2309.08047v1
2023-09-14T22:20:27
2023-09-14T22:20:27
{ "title": [ -0.027191162, -0.0035953522, -0.049316406, -0.032836914, -0.00982666, -0.019927979, -0.021514893, -0.008331299, -0.004047394, 0.062805176, -0.026046753, 0.001124382, 0.009552002, 0.017684937, -0.07537842, -0.009292603, -0.027542114, ...
http://arxiv.org/abs/2309.08030v2
AV2Wav: Diffusion-Based Re-synthesis from Continuous Self-supervised Features for Audio-Visual Speech Enhancement
Ju-Chieh Chou, Chung-Ming Chien, Karen Livescu
eess.AS, cs.CL, cs.SD
Speech enhancement systems are typically trained using pairs of clean and noisy speech. In audio-visual speech enhancement (AVSE), there is not as much ground-truth clean data available; most audio-visual datasets are collected in real-world environments with background noise and reverberation, hampering the developmen...
http://arxiv.org/pdf/2309.08030v2
2023-10-24T15:43:11
2023-09-14T21:07:53
{ "title": [ -0.015594482, 0.013641357, -0.044036865, -0.055023193, 0.009346008, 0.0066452026, -0.024627686, 0.0054626465, -0.017868042, 0.04598999, -0.02268982, 0.017471313, 0.028762817, -0.026397705, -0.038208008, 0.0013847351, -0.011421204, 0....
http://arxiv.org/abs/2309.13063v1
Using Large Language Models to Generate, Validate, and Apply User Intent Taxonomies
Chirag Shah, Ryen W. White, Reid Andersen, Georg Buscher, Scott Counts, Sarkar Snigdha Sarathi Das, Ali Montazer, Sathish Manivannan, Jennifer Neville, Xiaochuan Ni, Nagu Rangan, Tara Safavi, Siddharth Suri, Mengting Wan, Leijie Wang, Longqi Yang
cs.IR, cs.AI, cs.CL
Log data can reveal valuable information about how users interact with web search services, what they want, and how satisfied they are. However, analyzing user intents in log data is not easy, especially for new forms of web search such as AI-driven chat. To understand user intents from log data, we need a way to label...
http://arxiv.org/pdf/2309.13063v1
2023-09-14T20:46:48
2023-09-14T20:46:48
{ "title": [ 0.009803772, 0.0015182495, -0.048461914, -0.05886841, 0.0061149597, -0.02154541, -0.052886963, -0.01838684, 0.00617218, 0.03616333, -0.0071372986, -0.009857178, 0.022521973, -0.03213501, -0.037963867, 0.014854431, -0.022399902, 0.038...
http://arxiv.org/abs/2309.08008v1
An Empirical Evaluation of Prompting Strategies for Large Language Models in Zero-Shot Clinical Natural Language Processing
Sonish Sivarajkumar, Mark Kelley, Alyssa Samolyk-Mazzanti, Shyam Visweswaran, Yanshan Wang
cs.CL, cs.AI
Large language models (LLMs) have shown remarkable capabilities in Natural Language Processing (NLP), especially in domains where labeled data is scarce or expensive, such as clinical domain. However, to unlock the clinical knowledge hidden in these LLMs, we need to design effective prompts that can guide them to perfo...
http://arxiv.org/pdf/2309.08008v1
2023-09-14T19:35:00
2023-09-14T19:35:00
{ "title": [ -0.0032997131, -0.028244019, -0.022354126, -0.06555176, -0.025772095, -0.0016345978, -0.07165527, -0.0012073517, -0.0062217712, 0.0149383545, -0.01777649, 0.008659363, 0.0069770813, -0.03466797, -0.019042969, 0.0019454956, 0.0035419464, ...
http://arxiv.org/abs/2309.08007v1
DiariST: Streaming Speech Translation with Speaker Diarization
Mu Yang, Naoyuki Kanda, Xiaofei Wang, Junkun Chen, Peidong Wang, Jian Xue, Jinyu Li, Takuya Yoshioka
eess.AS, cs.CL, cs.SD
End-to-end speech translation (ST) for conversation recordings involves several under-explored challenges such as speaker diarization (SD) without accurate word time stamps and handling of overlapping speech in a streaming fashion. In this work, we propose DiariST, the first streaming ST and SD solution. It is built up...
http://arxiv.org/pdf/2309.08007v1
2023-09-14T19:33:27
2023-09-14T19:33:27
{ "title": [ -0.021087646, 0.0047683716, -0.024246216, -0.05895996, -0.003326416, 0.008834839, -0.0418396, 0.007980347, -0.020904541, 0.066833496, -0.017929077, 0.010444641, 0.038208008, -0.009223938, -0.036621094, 0.00907135, 0.0060577393, 0.005...
http://arxiv.org/abs/2309.07998v1
Exploring the Impact of Human Evaluator Group on Chat-Oriented Dialogue Evaluation
Sarah E. Finch, James D. Finch, Jinho D. Choi
cs.CL
Human evaluation has been widely accepted as the standard for evaluating chat-oriented dialogue systems. However, there is a significant variation in previous work regarding who gets recruited as evaluators. Evaluator groups such as domain experts, university students, and professional annotators have been used to asse...
http://arxiv.org/pdf/2309.07998v1
2023-09-14T19:19:50
2023-09-14T19:19:50
{ "title": [ 0.009880066, -0.0009069443, -0.027435303, -0.050872803, -0.0037899017, -0.018325806, -0.0065307617, 0.0057907104, -0.010513306, 0.04623413, -0.0044555664, -0.0035915375, -0.011169434, -0.014602661, -0.076416016, -0.015975952, -0.03244018...
http://arxiv.org/abs/2309.07990v1
Leveraging Contextual Information for Effective Entity Salience Detection
Rajarshi Bhowmik, Marco Ponza, Atharva Tendle, Anant Gupta, Rebecca Jiang, Xingyu Lu, Qian Zhao, Daniel Preotiuc-Pietro
cs.CL
In text documents such as news articles, the content and key events usually revolve around a subset of all the entities mentioned in a document. These entities, often deemed as salient entities, provide useful cues of the aboutness of a document to a reader. Identifying the salience of entities was found helpful in sev...
http://arxiv.org/pdf/2309.07990v1
2023-09-14T19:04:40
2023-09-14T19:04:40
{ "title": [ 0.035247803, 0.008331299, -0.026535034, -0.04901123, 0.035888672, -0.03378296, 0.015670776, -0.03564453, 0.04144287, 0.039886475, -0.018966675, -0.0030975342, 0.03225708, -0.0047073364, -0.06317139, 0.020553589, -0.010475159, 0.03674...
http://arxiv.org/abs/2309.07915v2
MMICL: Empowering Vision-language Model with Multi-Modal In-Context Learning
Haozhe Zhao, Zefan Cai, Shuzheng Si, Xiaojian Ma, Kaikai An, Liang Chen, Zixuan Liu, Sheng Wang, Wenjuan Han, Baobao Chang
cs.CL, cs.AI, cs.CV
Since the resurgence of deep learning, vision-language models (VLMs) enhanced by large language models (LLMs) have grown exponentially in popularity. However, while LLMs can utilize extensive background knowledge and task information with in-context learning, most VLMs still struggle with understanding complex multi-mo...
http://arxiv.org/pdf/2309.07915v2
2023-10-02T14:46:01
2023-09-14T17:59:17
{ "title": [ -0.014328003, -0.012870789, -0.05606079, -0.057891846, -0.0007791519, -0.016433716, -0.05029297, -0.006439209, 0.008155823, 0.03475952, -0.031082153, -0.010597229, 0.02809143, -0.023284912, -0.03262329, -0.014930725, 0.0015125275, 0....
http://arxiv.org/abs/2309.07900v1
Ambiguity-Aware In-Context Learning with Large Language Models
Lingyu Gao, Aditi Chaudhary, Krishna Srinivasan, Kazuma Hashimoto, Karthik Raman, Michael Bendersky
cs.CL, cs.IR
In-context learning (ICL) i.e. showing LLMs only a few task-specific demonstrations has led to downstream gains with no task-specific fine-tuning required. However, LLMs are sensitive to the choice of prompts, and therefore a crucial research question is how to select good demonstrations for ICL. One effective strategy...
http://arxiv.org/pdf/2309.07900v1
2023-09-14T17:48:34
2023-09-14T17:48:34
{ "title": [ -0.017990112, 0.0078048706, -0.060638428, -0.05682373, 0.009529114, -0.012001038, -0.044311523, 0.00504303, 0.016220093, 0.025146484, -0.03866577, -0.011276245, 0.011505127, -0.0138549805, -0.03918457, -0.0017795563, -0.0075416565, 0...
http://arxiv.org/abs/2309.08638v1
Anchor Points: Benchmarking Models with Much Fewer Examples
Rajan Vivek, Kawin Ethayarajh, Diyi Yang, Douwe Kiela
cs.CL
Modern language models often exhibit powerful but brittle behavior, leading to the development of larger and more diverse benchmarks to reliably assess their behavior. Here, we suggest that model performance can be benchmarked and elucidated with much smaller evaluation sets. We first show that in six popular language ...
http://arxiv.org/pdf/2309.08638v1
2023-09-14T17:45:51
2023-09-14T17:45:51
{ "title": [ -0.011199951, 0.0045051575, -0.025848389, -0.001200676, 0.00028777122, -0.015556335, 0.0021305084, -0.05706787, -0.007926941, 0.025512695, -0.0340271, 0.012771606, -0.032318115, -0.025009155, -0.02947998, -0.0022335052, 0.017486572, ...
http://arxiv.org/abs/2309.07875v2
Safety-Tuned LLaMAs: Lessons From Improving the Safety of Large Language Models that Follow Instructions
Federico Bianchi, Mirac Suzgun, Giuseppe Attanasio, Paul Röttger, Dan Jurafsky, Tatsunori Hashimoto, James Zou
cs.CL
Training large language models to follow instructions makes them perform better on a wide range of tasks, generally becoming more helpful. However, a perfectly helpful model will follow even the most malicious instructions and readily generate harmful content. In this paper, we raise concerns over the safety of models ...
http://arxiv.org/pdf/2309.07875v2
2023-09-25T15:45:13
2023-09-14T17:23:37
{ "title": [ -0.024932861, -0.0038108826, -0.03945923, -0.095336914, 0.015556335, 0.014144897, -0.07446289, 0.0061531067, -0.019638062, 0.025405884, -0.018508911, 0.026657104, 0.0035247803, -0.023162842, -0.0069084167, 0.0096206665, -0.009399414, ...
http://arxiv.org/abs/2309.07870v2
Agents: An Open-source Framework for Autonomous Language Agents
Wangchunshu Zhou, Yuchen Eleanor Jiang, Long Li, Jialong Wu, Tiannan Wang, Shi Qiu, Jintian Zhang, Jing Chen, Ruipu Wu, Shuai Wang, Shiding Zhu, Jiyu Chen, Wentao Zhang, Ningyu Zhang, Huajun Chen, Peng Cui, Mrinmaya Sachan
cs.CL
Recent advances on large language models (LLMs) enable researchers and developers to build autonomous language agents that can automatically solve various tasks and interact with environments, humans, and other agents using natural language interfaces. We consider language agents as a promising direction towards artifi...
http://arxiv.org/pdf/2309.07870v2
2023-10-08T14:43:38
2023-09-14T17:18:25
{ "title": [ -0.027145386, -0.009689331, -0.025543213, -0.05618286, 0.009719849, 0.008804321, -0.022537231, -0.006046295, -0.0061187744, 0.03933716, -0.021362305, 0.010231018, 0.011039734, -0.0110321045, -0.042022705, 0.0059013367, -0.017562866, ...
http://arxiv.org/abs/2309.07864v3
The Rise and Potential of Large Language Model Based Agents: A Survey
Zhiheng Xi, Wenxiang Chen, Xin Guo, Wei He, Yiwen Ding, Boyang Hong, Ming Zhang, Junzhe Wang, Senjie Jin, Enyu Zhou, Rui Zheng, Xiaoran Fan, Xiao Wang, Limao Xiong, Yuhao Zhou, Weiran Wang, Changhao Jiang, Yicheng Zou, Xiangyang Liu, Zhangyue Yin, Shihan Dou, Rongxiang Weng, Wensen Cheng, Qi Zhang, Wenjuan Qin, Yongyan...
cs.AI, cs.CL
For a long time, humanity has pursued artificial intelligence (AI) equivalent to or surpassing the human level, with AI agents considered a promising vehicle for this pursuit. AI agents are artificial entities that sense their environment, make decisions, and take actions. Many efforts have been made to develop intelli...
http://arxiv.org/pdf/2309.07864v3
2023-09-19T08:29:18
2023-09-14T17:12:03
{ "title": [ -0.034973145, -0.021484375, -0.035308838, -0.064575195, 0.0013971329, 0.0068359375, -0.048431396, 0.013687134, 0.00605011, 0.04147339, -0.010345459, -0.004714966, 0.018630981, -0.013046265, -0.047180176, 0.01399231, -0.020263672, 0.0...
http://arxiv.org/abs/2309.07861v1
CiwaGAN: Articulatory information exchange
Gašper Beguš, Thomas Lu, Alan Zhou, Peter Wu, Gopala K. Anumanchipalli
cs.SD, cs.AI, cs.CL, eess.AS
Humans encode information into sounds by controlling articulators and decode information from sounds using the auditory apparatus. This paper introduces CiwaGAN, a model of human spoken language acquisition that combines unsupervised articulatory modeling with an unsupervised model of information exchange through the a...
http://arxiv.org/pdf/2309.07861v1
2023-09-14T17:10:39
2023-09-14T17:10:39
{ "title": [ -0.03857422, -0.028274536, -0.043518066, -0.025848389, -0.044281006, -0.0052948, -0.04159546, 0.020141602, -0.009384155, 0.037506104, -0.016235352, -0.01033783, 0.016342163, -0.052520752, -0.022994995, -0.007144928, -0.0014371872, 0....
http://arxiv.org/abs/2309.07852v1
ExpertQA: Expert-Curated Questions and Attributed Answers
Chaitanya Malaviya, Subin Lee, Sihao Chen, Elizabeth Sieber, Mark Yatskar, Dan Roth
cs.CL, cs.AI
As language models are adapted by a more sophisticated and diverse set of users, the importance of guaranteeing that they provide factually correct information supported by verifiable sources is critical across fields of study & professions. This is especially the case for high-stakes fields, such as medicine and law, ...
http://arxiv.org/pdf/2309.07852v1
2023-09-14T16:54:34
2023-09-14T16:54:34
{ "title": [ -0.013595581, -0.0069084167, -0.0637207, -0.05392456, 0.0023937225, -0.016937256, -0.012962341, -0.005355835, -0.011734009, 0.023895264, 0.0022621155, 0.005695343, -0.0015354156, -0.048797607, -0.017242432, 0.017944336, 0.031433105, ...
http://arxiv.org/abs/2309.07822v2
CATfOOD: Counterfactual Augmented Training for Improving Out-of-Domain Performance and Calibration
Rachneet Sachdeva, Martin Tutek, Iryna Gurevych
cs.CL
In recent years, large language models (LLMs) have shown remarkable capabilities at scale, particularly at generating text conditioned on a prompt. In our work, we investigate the use of LLMs to augment training data of small language models~(SLMs) with automatically generated counterfactual~(CF) instances -- i.e. mini...
http://arxiv.org/pdf/2309.07822v2
2023-09-15T07:57:55
2023-09-14T16:16:40
{ "title": [ 0.010612488, -0.00025463104, -0.016464233, -0.010757446, -0.001083374, -0.02293396, 0.011131287, -0.033691406, -0.023117065, 0.020614624, -0.016403198, 0.011741638, -0.008415222, -0.036743164, -0.019866943, -0.024429321, -0.024124146, ...
http://arxiv.org/abs/2309.07812v2
Text Classification of Cancer Clinical Trial Eligibility Criteria
Yumeng Yang, Soumya Jayaraj, Ethan B Ludmir, Kirk Roberts
cs.CL, cs.LG
Automatic identification of clinical trials for which a patient is eligible is complicated by the fact that trial eligibility is stated in natural language. A potential solution to this problem is to employ text classification methods for common types of eligibility criteria. In this study, we focus on seven common exc...
http://arxiv.org/pdf/2309.07812v2
2023-09-15T21:59:56
2023-09-14T15:59:16
{ "title": [ -0.010757446, 0.029556274, -0.046722412, -0.0121536255, -0.006095886, -0.026550293, -0.023727417, -0.027755737, -0.060424805, 0.04248047, -0.04510498, 0.016525269, -0.005092621, -0.013031006, -0.0023765564, -0.03201294, -0.007537842, ...
http://arxiv.org/abs/2309.07804v1
Pop Quiz! Do Pre-trained Code Models Possess Knowledge of Correct API Names?
Terry Yue Zhuo, Xiaoning Du, Zhenchang Xing, Jiamou Sun, Haowei Quan, Li Li, Liming Zhu
cs.SE, cs.CL
Recent breakthroughs in pre-trained code models, such as CodeBERT and Codex, have shown their superior performance in various downstream tasks. The correctness and unambiguity of API usage among these code models are crucial for achieving desirable program functionalities, requiring them to learn various API fully qual...
http://arxiv.org/pdf/2309.07804v1
2023-09-14T15:46:41
2023-09-14T15:46:41
{ "title": [ -0.029449463, -0.025634766, -0.04815674, -0.013389587, -0.048309326, -0.021972656, -0.028121948, 0.008125305, -0.004535675, 0.01586914, -0.018066406, 0.005836487, -0.0012369156, -0.040527344, 0.012290955, -0.036193848, 0.022827148, 0...
http://arxiv.org/abs/2309.07797v2
The Dynamical Principles of Storytelling
Isidoros Doxas, James Meiss, Steven Bottone, Tom Strelich, Andrew Plummer, Adrienne Breland, Simon Dennis, Kathy Garvin-Doxas, Michael Klymkowsky
cs.CL
When considering the opening part of 1800 short stories, we find that the first dozen paragraphs of the average narrative follow an action principle as defined in arXiv:2309.06600. When the order of the paragraphs is shuffled, the average no longer exhibits this property. The findings show that there is a preferential ...
http://arxiv.org/pdf/2309.07797v2
2023-09-29T15:42:54
2023-09-14T15:36:10
{ "title": [ 0.026046753, -0.021942139, -0.019546509, -0.013008118, -0.048980713, -0.029342651, -0.017837524, 0.007160187, -0.029754639, 0.03555298, 0.0026130676, -0.007041931, 0.00667572, -0.037322998, 0.008361816, -0.011154175, 0.019317627, 0.0...
http://arxiv.org/abs/2309.08637v3
TextBind: Multi-turn Interleaved Multimodal Instruction-following in the Wild
Huayang Li, Siheng Li, Deng Cai, Longyue Wang, Lemao Liu, Taro Watanabe, Yujiu Yang, Shuming Shi
cs.CL, cs.AI
Large language models with instruction-following abilities have revolutionized the field of artificial intelligence. These models show exceptional generalizability to tackle various real-world tasks through their natural language interfaces. However, their performance heavily relies on high-quality exemplar data, which...
http://arxiv.org/pdf/2309.08637v3
2023-11-08T09:33:31
2023-09-14T15:34:01
{ "title": [ -0.039154053, -0.016021729, -0.018661499, -0.061920166, -0.02772522, 0.01424408, -0.06933594, -0.0026397705, -0.016418457, 0.038116455, -0.03390503, 0.040161133, -0.0038661957, -0.0026226044, -0.018737793, -0.0008893013, -0.015182495, ...
http://arxiv.org/abs/2309.07794v1
Improving Multimodal Classification of Social Media Posts by Leveraging Image-Text Auxiliary tasks
Danae Sánchez Villegas, Daniel Preoţiuc-Pietro, Nikolaos Aletras
cs.CL, cs.LG, cs.SI
Effectively leveraging multimodal information from social media posts is essential to various downstream tasks such as sentiment analysis, sarcasm detection and hate speech classification. However, combining text and image information is challenging because of the idiosyncratic cross-modal semantics with hidden or comp...
http://arxiv.org/pdf/2309.07794v1
2023-09-14T15:30:59
2023-09-14T15:30:59
{ "title": [ 0.009613037, 0.024108887, -0.030548096, -0.024734497, 0.019607544, -0.027770996, -0.00079631805, -0.030700684, -0.0077323914, 0.042419434, -0.03479004, 0.041809082, -0.011123657, 0.012931824, -0.044006348, 0.0015239716, -0.0073127747, ...
http://arxiv.org/abs/2309.07773v2
Spoken Humanoid Embodied Conversational Agents in Mobile Serious Games: A Usability Assessment
Danai Korre, Judy Robertson
cs.HC, cs.CL, cs.MM, H.5.2; H.5.1
This paper presents an empirical investigation of the extent to which spoken Humanoid Embodied Conversational Agents (HECAs) can foster usability in mobile serious game (MSG) applications. The aim of the research is to assess the impact of multiple agents and illusion of humanness on the quality of the interaction. The...
http://arxiv.org/pdf/2309.07773v2
2023-09-15T15:42:02
2023-09-14T15:02:05
{ "title": [ -0.025726318, -0.0126953125, -0.008338928, 0.0149002075, -0.019210815, -0.004436493, 0.008399963, -0.016845703, -0.01159668, 0.047027588, -0.0040130615, 0.002658844, -0.0029945374, -0.013961792, -0.0496521, -0.0031337738, -0.038604736, ...
http://arxiv.org/abs/2309.07765v1
Echotune: A Modular Extractor Leveraging the Variable-Length Nature of Speech in ASR Tasks
Sizhou Chen, Songyang Gao, Sen Fang
cs.SD, cs.CL, eess.AS
The Transformer architecture has proven to be highly effective for Automatic Speech Recognition (ASR) tasks, becoming a foundational component for a plethora of research in the domain. Historically, many approaches have leaned on fixed-length attention windows, which becomes problematic for varied speech samples in dur...
http://arxiv.org/pdf/2309.07765v1
2023-09-14T14:51:51
2023-09-14T14:51:51
{ "title": [ -0.024230957, -0.010292053, -0.08087158, -0.059936523, -0.02230835, -0.024749756, -0.050750732, 0.030776978, -0.04586792, 0.031982422, 0.025543213, -0.0121154785, 0.010719299, -0.045715332, 0.0011224747, -0.0050315857, 0.0019483566, ...
http://arxiv.org/abs/2309.07759v1
PROGrasp: Pragmatic Human-Robot Communication for Object Grasping
Gi-Cheon Kang, Junghyun Kim, Jaein Kim, Byoung-Tak Zhang
cs.CL, cs.RO
Interactive Object Grasping (IOG) is the task of identifying and grasping the desired object via human-robot natural language interaction. Current IOG systems assume that a human user initially specifies the target object's category (e.g., bottle). Inspired by pragmatics, where humans often convey their intentions by r...
http://arxiv.org/pdf/2309.07759v1
2023-09-14T14:45:47
2023-09-14T14:45:47
{ "title": [ -0.018569946, 0.008117676, -0.03793335, -0.036468506, 0.017089844, -0.0056533813, -0.01939392, -0.023483276, -0.008148193, 0.028259277, -0.013885498, 0.004261017, 0.018829346, -0.024353027, -0.0129852295, -0.016815186, -0.00084781647, ...
http://arxiv.org/abs/2309.07755v1
Generative AI Text Classification using Ensemble LLM Approaches
Harika Abburi, Michael Suesserman, Nirmala Pudota, Balaji Veeramani, Edward Bowen, Sanmitra Bhattacharya
cs.CL, cs.AI
Large Language Models (LLMs) have shown impressive performance across a variety of Artificial Intelligence (AI) and natural language processing tasks, such as content creation, report generation, etc. However, unregulated malign application of these models can create undesirable consequences such as generation of fake ...
http://arxiv.org/pdf/2309.07755v1
2023-09-14T14:41:46
2023-09-14T14:41:46
{ "title": [ -0.044952393, 0.0024490356, -0.053222656, -0.03543091, -0.008674622, -0.0059776306, -0.03439331, -0.006801605, 0.012084961, 0.02619934, -0.035217285, 0.030075073, -0.026397705, -0.011360168, -0.054138184, -0.0051078796, -0.02885437, ...
http://arxiv.org/abs/2309.07739v1
The complementary roles of non-verbal cues for Robust Pronunciation Assessment
Yassine El Kheir, Shammur Absar Chowdhury, Ahmed Ali
cs.CL, cs.SD, eess.AS
Research on pronunciation assessment systems focuses on utilizing phonetic and phonological aspects of non-native (L2) speech, often neglecting the rich layer of information hidden within the non-verbal cues. In this study, we proposed a novel pronunciation assessment framework, IntraVerbalPA. % The framework innovativ...
http://arxiv.org/pdf/2309.07739v1
2023-09-14T14:18:07
2023-09-14T14:18:07
{ "title": [ 0.012214661, 0.018737793, -0.031143188, -0.04425049, -0.009536743, -0.015686035, -0.034179688, -0.0045204163, -0.01084137, 0.034210205, -0.03387451, 0.007499695, 0.012184143, -0.020263672, -0.044158936, -0.015045166, -0.0309906, 0.04...
http://arxiv.org/abs/2309.07733v1
Explaining Speech Classification Models via Word-Level Audio Segments and Paralinguistic Features
Eliana Pastor, Alkis Koudounas, Giuseppe Attanasio, Dirk Hovy, Elena Baralis
cs.CL, cs.SD, eess.AS
Recent advances in eXplainable AI (XAI) have provided new insights into how models for vision, language, and tabular data operate. However, few approaches exist for understanding speech models. Existing work focuses on a few spoken language understanding (SLU) tasks, and explanations are difficult to interpret for most...
http://arxiv.org/pdf/2309.07733v1
2023-09-14T14:12:34
2023-09-14T14:12:34
{ "title": [ -0.010505676, -0.0067977905, -0.02583313, -0.05935669, -0.015205383, -0.008728027, -0.03677368, 0.007911682, -0.024673462, 0.01914978, -0.009162903, -0.006629944, -0.018447876, -0.028244019, -0.020446777, -0.017196655, -0.024490356, ...
http://arxiv.org/abs/2309.08636v3
ChatGPT v Bard v Bing v Claude 2 v Aria v human-expert. How good are AI chatbots at scientific writing?
Edisa Lozić, Benjamin Štular
cs.CL, cs.AI, cs.CY, cs.ET, cs.HC, 68T01, I.2.0
Historical emphasis on writing mastery has shifted with advances in generative AI, especially in scientific writing. This study analysed six AI chatbots for scholarly writing in humanities and archaeology. Using methods that assessed factual correctness and scientific contribution, ChatGPT-4 showed the highest quantita...
http://arxiv.org/pdf/2309.08636v3
2023-10-16T14:24:02
2023-09-14T14:04:03
{ "title": [ -0.05722046, -0.023742676, -0.00687027, 0.037750244, -0.060516357, 0.011909485, -0.019943237, -0.030639648, -0.019256592, 0.05581665, 0.004333496, 0.020401001, 0.0060272217, -0.014465332, -0.026107788, -0.004135132, -0.01638794, 0.01...
http://arxiv.org/abs/2309.07727v1
PerPLM: Personalized Fine-tuning of Pretrained Language Models via Writer-specific Intermediate Learning and Prompts
Daisuke Oba, Naoki Yoshinaga, Masashi Toyoda
cs.CL
The meanings of words and phrases depend not only on where they are used (contexts) but also on who use them (writers). Pretrained language models (PLMs) are powerful tools for capturing context, but they are typically pretrained and fine-tuned for universal use across different writers. This study aims to improve the ...
http://arxiv.org/pdf/2309.07727v1
2023-09-14T14:03:48
2023-09-14T14:03:48
{ "title": [ -0.020080566, -0.011627197, -0.027755737, -0.051696777, -0.051513672, 0.0011978149, -0.08508301, -0.0018692017, -0.03427124, 0.016098022, -0.040802002, 0.04638672, -0.007888794, -0.023452759, 0.0098724365, -0.01776123, -0.018096924, ...
http://arxiv.org/abs/2309.07719v2
L1-aware Multilingual Mispronunciation Detection Framework
Yassine El Kheir, Shammur Absar Chowdhury, Ahmed Ali
cs.CL, cs.SD, eess.AS
The phonological discrepancies between a speaker's native (L1) and the non-native language (L2) serves as a major factor for mispronunciation. This paper introduces a novel multilingual MDD architecture, L1-MultiMDD, enriched with L1-aware speech representation. An end-to-end speech encoder is trained on the input sign...
http://arxiv.org/pdf/2309.07719v2
2023-09-21T15:26:33
2023-09-14T13:53:17
{ "title": [ -0.020996094, -0.00011819601, -0.05819702, -0.040130615, -0.003566742, -0.0022659302, -0.035003662, 0.014137268, 0.0035648346, 0.04525757, -0.024917603, 0.0067825317, 0.03253174, -0.0262146, -0.020950317, 0.015945435, -0.027359009, 0...
http://arxiv.org/abs/2309.07707v1
CoLLD: Contrastive Layer-to-layer Distillation for Compressing Multilingual Pre-trained Speech Encoders
Heng-Jui Chang, Ning Dong, Ruslan Mavlyutov, Sravya Popuri, Yu-An Chung
cs.CL, cs.SD, eess.AS
Large-scale self-supervised pre-trained speech encoders outperform conventional approaches in speech recognition and translation tasks. Due to the high cost of developing these large models, building new encoders for new tasks and deploying them to on-device applications are infeasible. Prior studies propose model comp...
http://arxiv.org/pdf/2309.07707v1
2023-09-14T13:38:02
2023-09-14T13:38:02
{ "title": [ -0.020080566, -0.008163452, -0.049957275, -0.045074463, -0.021453857, 0.0020179749, -0.04812622, 0.031143188, 0.0065841675, 0.056884766, 0.042053223, -0.015609741, 0.037963867, -0.05593872, -0.01335907, 0.021697998, -0.0045433044, 0....
http://arxiv.org/abs/2309.07694v1
Tree of Uncertain Thoughts Reasoning for Large Language Models
Shentong Mo, Miao Xin
cs.CL, cs.AI, cs.LG
While the recently introduced Tree of Thoughts (ToT) has heralded advancements in allowing Large Language Models (LLMs) to reason through foresight and backtracking for global decision-making, it has overlooked the inherent local uncertainties in intermediate decision points or "thoughts". These local uncertainties, in...
http://arxiv.org/pdf/2309.07694v1
2023-09-14T13:14:51
2023-09-14T13:14:51
{ "title": [ -0.043823242, -0.03161621, -0.036193848, -0.056793213, -0.036712646, 0.020980835, -0.06774902, 0.025756836, 0.020507812, 0.012588501, -0.04031372, -0.021118164, 0.010795593, -0.026412964, -0.035308838, -0.006313324, 0.00068950653, 0....