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/2308.01862v1
Wider and Deeper LLM Networks are Fairer LLM Evaluators
Xinghua Zhang, Bowen Yu, Haiyang Yu, Yangyu Lv, Tingwen Liu, Fei Huang, Hongbo Xu, Yongbin Li
cs.CL
Measuring the quality of responses generated by LLMs is a challenging task, particularly when it comes to evaluating whether the response is aligned with human preference. A novel approach involves using the LLM itself to make evaluation and stabilizing the results through multiple independent evaluations, similar to a...
http://arxiv.org/pdf/2308.01862v1
2023-08-03T16:38:34
2023-08-03T16:38:34
{ "title": [ -0.0071105957, -0.01449585, -0.04437256, -0.02684021, -0.051971436, -0.018539429, -0.050933838, 0.011428833, -0.038024902, 0.0037441254, -0.0496521, 0.025512695, -0.010604858, -0.039215088, 0.051605225, -0.02444458, -0.032287598, 0.0...
http://arxiv.org/abs/2308.01861v2
ClassEval: A Manually-Crafted Benchmark for Evaluating LLMs on Class-level Code Generation
Xueying Du, Mingwei Liu, Kaixin Wang, Hanlin Wang, Junwei Liu, Yixuan Chen, Jiayi Feng, Chaofeng Sha, Xin Peng, Yiling Lou
cs.CL, cs.AI
In this work, we make the first attempt to evaluate LLMs in a more challenging code generation scenario, i.e. class-level code generation. We first manually construct the first class-level code generation benchmark ClassEval of 100 class-level Python code generation tasks with approximately 500 person-hours. Based on i...
http://arxiv.org/pdf/2308.01861v2
2023-08-14T09:07:00
2023-08-03T16:31:02
{ "title": [ -0.043273926, -0.024841309, -0.03173828, -0.04788208, -0.026504517, -0.00504303, -0.06640625, -0.008117676, -0.03189087, 0.016555786, -0.032806396, 0.052337646, -0.013710022, -0.017074585, 0.027740479, -0.008117676, -0.017913818, 0.0...
http://arxiv.org/abs/2308.01849v1
Curricular Transfer Learning for Sentence Encoded Tasks
Jader Martins Camboim de Sá, Matheus Ferraroni Sanches, Rafael Roque de Souza, Júlio Cesar dos Reis, Leandro Aparecido Villas
cs.CL, cs.LG
Fine-tuning language models in a downstream task is the standard approach for many state-of-the-art methodologies in the field of NLP. However, when the distribution between the source task and target task drifts, \textit{e.g.}, conversational environments, these gains tend to be diminished. This article proposes a seq...
http://arxiv.org/pdf/2308.01849v1
2023-08-03T16:18:19
2023-08-03T16:18:19
{ "title": [ -0.027252197, -0.01713562, -0.028579712, -0.008399963, -0.0206604, -0.018951416, 0.001168251, -0.006011963, -0.028549194, 0.025100708, -0.004837036, 0.01335144, -0.011962891, -0.014518738, -0.019058228, -0.016357422, -0.028076172, 0....
http://arxiv.org/abs/2308.01846v1
XNLP: An Interactive Demonstration System for Universal Structured NLP
Hao Fei, Meishan Zhang, Min Zhang, Tat-Seng Chua
cs.CL
Structured Natural Language Processing (XNLP) is an important subset of NLP that entails understanding the underlying semantic or syntactic structure of texts, which serves as a foundational component for many downstream applications. Despite certain recent efforts to explore universal solutions for specific categories...
http://arxiv.org/pdf/2308.01846v1
2023-08-03T16:13:05
2023-08-03T16:13:05
{ "title": [ -0.022109985, -0.012145996, -0.028503418, -0.0803833, -0.00907135, 0.018127441, -0.06982422, -0.011070251, 0.014312744, 0.0069122314, -0.03668213, 0.00920105, -0.015258789, -0.006919861, -0.03503418, -0.0045928955, -0.014152527, 0.04...
http://arxiv.org/abs/2308.01834v1
The Capability of Large Language Models to Measure Psychiatric Functioning
Isaac R. Galatzer-Levy, Daniel McDuff, Vivek Natarajan, Alan Karthikesalingam, Matteo Malgaroli
cs.CL, cs.AI, cs.LG
The current work investigates the capability of Large language models (LLMs) that are explicitly trained on large corpuses of medical knowledge (Med-PaLM 2) to predict psychiatric functioning from patient interviews and clinical descriptions without being trained to do so. To assess this, n = 145 depression and n =115 ...
http://arxiv.org/pdf/2308.01834v1
2023-08-03T15:52:27
2023-08-03T15:52:27
{ "title": [ -0.01763916, 0.0028514862, -0.025634766, -0.055419922, 0.0025558472, -0.007381439, -0.046905518, -0.018112183, 0.010620117, 0.019134521, 0.00409317, -0.020767212, -0.007232666, -0.05303955, -0.03515625, 0.010635376, -0.00009149313, 0...
http://arxiv.org/abs/2308.01831v1
Many-to-Many Spoken Language Translation via Unified Speech and Text Representation Learning with Unit-to-Unit Translation
Minsu Kim, Jeongsoo Choi, Dahun Kim, Yong Man Ro
cs.CL, eess.AS, eess.SP
In this paper, we propose a method to learn unified representations of multilingual speech and text with a single model, especially focusing on the purpose of speech synthesis. We represent multilingual speech audio with speech units, the quantized representations of speech features encoded from a self-supervised speec...
http://arxiv.org/pdf/2308.01831v1
2023-08-03T15:47:04
2023-08-03T15:47:04
{ "title": [ -0.0064582825, 0.00466156, -0.032806396, -0.072143555, 0.001490593, 0.010429382, -0.060150146, 0.019744873, -0.030776978, 0.055419922, -0.020568848, 0.03604126, 0.03024292, -0.01486969, -0.016601562, 0.004673004, -0.0058670044, 0.028...
http://arxiv.org/abs/2308.01825v2
Scaling Relationship on Learning Mathematical Reasoning with Large Language Models
Zheng Yuan, Hongyi Yuan, Chengpeng Li, Guanting Dong, Keming Lu, Chuanqi Tan, Chang Zhou, Jingren Zhou
cs.CL
Mathematical reasoning is a challenging task for large language models (LLMs), while the scaling relationship of it with respect to LLM capacity is under-explored. In this paper, we investigate how the pre-training loss, supervised data amount, and augmented data amount influence the reasoning performances of a supervi...
http://arxiv.org/pdf/2308.01825v2
2023-09-13T03:57:29
2023-08-03T15:34:01
{ "title": [ -0.02192688, -0.008552551, -0.035583496, -0.045959473, -0.008171082, -0.015304565, -0.024551392, 0.015975952, -0.013717651, 0.01461792, -0.01864624, -0.012245178, -0.014602661, -0.035217285, -0.017456055, -0.015312195, -0.009284973, ...
http://arxiv.org/abs/2308.01785v1
Lexicon and Rule-based Word Lemmatization Approach for the Somali Language
Shafie Abdi Mohamed, Muhidin Abdullahi Mohamed
cs.CL
Lemmatization is a Natural Language Processing (NLP) technique used to normalize text by changing morphological derivations of words to their root forms. It is used as a core pre-processing step in many NLP tasks including text indexing, information retrieval, and machine learning for NLP, among others. This paper pion...
http://arxiv.org/pdf/2308.01785v1
2023-08-03T14:31:57
2023-08-03T14:31:57
{ "title": [ -0.017578125, 0.009147644, -0.037719727, -0.054229736, -0.0050239563, 0.0051841736, -0.052093506, 0.01966858, -0.027938843, 0.0390625, -0.059539795, 0.0030136108, 0.03552246, -0.009132385, -0.008468628, -0.008979797, -0.027618408, 0....
http://arxiv.org/abs/2308.01776v2
Does Correction Remain A Problem For Large Language Models?
Xiaowu Zhang, Xiaotian Zhang, Cheng Yang, Hang Yan, Xipeng Qiu
cs.CL
As large language models, such as GPT, continue to advance the capabilities of natural language processing (NLP), the question arises: does the problem of correction still persist? This paper investigates the role of correction in the context of large language models by conducting two experiments. The first experiment ...
http://arxiv.org/pdf/2308.01776v2
2023-08-14T09:15:42
2023-08-03T14:09:31
{ "title": [ -0.021759033, 0.00036096573, -0.012199402, -0.014923096, -0.028930664, -0.008384705, -0.03363037, -0.012527466, 0.017562866, 0.03237915, -0.044891357, -0.02696228, 0.013580322, -0.018920898, -0.025146484, -0.008453369, -0.009399414, ...
http://arxiv.org/abs/2308.01741v1
Supply chain emission estimation using large language models
Ayush Jain, Manikandan Padmanaban, Jagabondhu Hazra, Shantanu Godbole, Kommy Weldemariam
cs.CL
Large enterprises face a crucial imperative to achieve the Sustainable Development Goals (SDGs), especially goal 13, which focuses on combating climate change and its impacts. To mitigate the effects of climate change, reducing enterprise Scope 3 (supply chain emissions) is vital, as it accounts for more than 90\% of t...
http://arxiv.org/pdf/2308.01741v1
2023-08-03T13:06:37
2023-08-03T13:06:37
{ "title": [ -0.05581665, -0.006099701, -0.053771973, -0.06329346, -0.0088272095, 0.012519836, -0.062561035, 0.0033111572, -0.005622864, 0.021743774, -0.019378662, -0.006099701, -0.03869629, -0.023452759, -0.032806396, -0.02458191, -0.0012102127, ...
http://arxiv.org/abs/2308.01734v1
Ambient Adventures: Teaching ChatGPT on Developing Complex Stories
Zexin Chen, Eric Zhou, Kenneth Eaton, Xiangyu Peng, Mark Riedl
cs.CL, cs.HC
Imaginative play is an area of creativity that could allow robots to engage with the world around them in a much more personified way. Imaginary play can be seen as taking real objects and locations and using them as imaginary objects and locations in virtual scenarios. We adopted the story generation capability of lar...
http://arxiv.org/pdf/2308.01734v1
2023-08-03T12:52:49
2023-08-03T12:52:49
{ "title": [ -0.011650085, -0.024414062, -0.007068634, -0.001077652, -0.01625061, -0.00022506714, 0.01083374, -0.012901306, -0.026763916, 0.016662598, -0.014907837, 0.023239136, 0.0044021606, -0.03567505, -0.0005970001, 0.0021953583, -0.028961182, ...
http://arxiv.org/abs/2308.02575v1
Is GPT-4 a reliable rater? Evaluating Consistency in GPT-4 Text Ratings
Veronika Hackl, Alexandra Elena Müller, Michael Granitzer, Maximilian Sailer
cs.CL, cs.AI
This study investigates the consistency of feedback ratings generated by OpenAI's GPT-4, a state-of-the-art artificial intelligence language model, across multiple iterations, time spans and stylistic variations. The model rated responses to tasks within the Higher Education (HE) subject domain of macroeconomics in ter...
http://arxiv.org/pdf/2308.02575v1
2023-08-03T12:47:17
2023-08-03T12:47:17
{ "title": [ -0.03704834, 0.02960205, -0.036193848, -0.008003235, 0.0026397705, -0.008277893, -0.029251099, -0.055847168, -0.023391724, 0.012886047, -0.024520874, -0.0076446533, -0.018127441, -0.03857422, -0.023864746, -0.018722534, -0.021621704, ...
http://arxiv.org/abs/2308.01727v1
Local Large Language Models for Complex Structured Medical Tasks
V. K. Cody Bumgardner, Aaron Mullen, Sam Armstrong, Caylin Hickey, Jeff Talbert
cs.CL, cs.AI
This paper introduces an approach that combines the language reasoning capabilities of large language models (LLMs) with the benefits of local training to tackle complex, domain-specific tasks. Specifically, the authors demonstrate their approach by extracting structured condition codes from pathology reports. The prop...
http://arxiv.org/pdf/2308.01727v1
2023-08-03T12:36:13
2023-08-03T12:36:13
{ "title": [ -0.035247803, -0.005382538, -0.03692627, -0.07849121, 0.006462097, 0.01586914, -0.08404541, -0.0061836243, 0.011436462, 0.028884888, 0.012748718, -0.017684937, -0.0003528595, -0.03805542, -0.03274536, 0.008758545, 0.015014648, 0.0094...
http://arxiv.org/abs/2308.01684v2
Baby's CoThought: Leveraging Large Language Models for Enhanced Reasoning in Compact Models
Zheyu Zhang, Han Yang, Bolei Ma, David Rügamer, Ercong Nie
cs.CL
Large Language Models (LLMs) demonstrate remarkable performance on a variety of natural language understanding (NLU) tasks, primarily due to their in-context learning ability. This ability could be applied to building babylike models, i.e. models at small scales, improving training efficiency. In this paper, we propose...
http://arxiv.org/pdf/2308.01684v2
2023-10-23T12:05:52
2023-08-03T10:52:52
{ "title": [ -0.03375244, -0.03656006, -0.047851562, -0.08013916, -0.00021874905, 0.0155181885, -0.05770874, 0.016082764, 0.023712158, 0.022109985, -0.020904541, -0.011749268, 0.0012216568, -0.019348145, -0.045166016, 0.0060806274, -0.0062675476, ...
http://arxiv.org/abs/2308.01681v3
NBIAS: A Natural Language Processing Framework for Bias Identification in Text
Shaina Raza, Muskan Garg, Deepak John Reji, Syed Raza Bashir, Chen Ding
cs.CL, cs.AI
Bias in textual data can lead to skewed interpretations and outcomes when the data is used. These biases could perpetuate stereotypes, discrimination, or other forms of unfair treatment. An algorithm trained on biased data may end up making decisions that disproportionately impact a certain group of people. Therefore, ...
http://arxiv.org/pdf/2308.01681v3
2023-08-29T12:20:15
2023-08-03T10:48:30
{ "title": [ -0.0008325577, -0.00092077255, -0.032104492, -0.042510986, 0.01612854, -0.00034689903, -0.02822876, -0.039215088, 0.00033569336, 0.021560669, -0.03555298, -0.0057411194, 0.0038757324, -0.0103302, -0.048828125, -0.017303467, -0.036071777,...
http://arxiv.org/abs/2308.02570v1
Learning Implicit Entity-object Relations by Bidirectional Generative Alignment for Multimodal NER
Feng Chen, Jiajia Liu, Kaixiang Ji, Wang Ren, Jian Wang, Jingdong Wang
cs.LG, cs.AI, cs.CL, cs.CV
The challenge posed by multimodal named entity recognition (MNER) is mainly two-fold: (1) bridging the semantic gap between text and image and (2) matching the entity with its associated object in image. Existing methods fail to capture the implicit entity-object relations, due to the lack of corresponding annotation. ...
http://arxiv.org/pdf/2308.02570v1
2023-08-03T10:37:20
2023-08-03T10:37:20
{ "title": [ 0.007980347, 0.0076675415, -0.038909912, -0.06878662, -0.0035209656, 0.0013370514, -0.04876709, -0.0038089752, 0.029342651, 0.015945435, -0.03262329, 0.01727295, -0.00036931038, -0.021148682, -0.013748169, 0.002904892, -0.025756836, ...
http://arxiv.org/abs/2308.01666v1
Evaluating ChatGPT text-mining of clinical records for obesity monitoring
Ivo S. Fins, Heather Davies, Sean Farrell, Jose R. Torres, Gina Pinchbeck, Alan D. Radford, Peter-John Noble
cs.IR, cs.CL
Background: Veterinary clinical narratives remain a largely untapped resource for addressing complex diseases. Here we compare the ability of a large language model (ChatGPT) and a previously developed regular expression (RegexT) to identify overweight body condition scores (BCS) in veterinary narratives. Methods: BCS ...
http://arxiv.org/pdf/2308.01666v1
2023-08-03T10:11:42
2023-08-03T10:11:42
{ "title": [ -0.0050315857, -0.01234436, 0.0068130493, -0.023788452, -0.014129639, 0.018936157, -0.022613525, -0.027954102, -0.0011634827, 0.02519226, -0.039123535, 0.027374268, -0.027816772, -0.017578125, -0.036590576, -0.0056152344, -0.030181885, ...
http://arxiv.org/abs/2308.02569v1
BioBERT Based SNP-traits Associations Extraction from Biomedical Literature
Mohammad Dehghani, Behrouz Bokharaeian, Zahra Yazdanparast
cs.CL
Scientific literature contains a considerable amount of information that provides an excellent opportunity for developing text mining methods to extract biomedical relationships. An important type of information is the relationship between singular nucleotide polymorphisms (SNP) and traits. In this paper, we present a ...
http://arxiv.org/pdf/2308.02569v1
2023-08-03T09:40:27
2023-08-03T09:40:27
{ "title": [ -0.036224365, 0.022491455, -0.035827637, -0.038208008, -0.0031833649, 0.0056915283, -0.04168701, -0.02746582, -0.007534027, 0.030319214, 0.027130127, 0.0016450882, -0.01789856, -0.03314209, -0.037322998, 0.009429932, -0.023757935, 0....
http://arxiv.org/abs/2308.01589v1
Holy Grail 2.0: From Natural Language to Constraint Models
Dimos Tsouros, Hélène Verhaeghe, Serdar Kadıoğlu, Tias Guns
cs.AI, cs.CL, cs.HC
Twenty-seven years ago, E. Freuder highlighted that "Constraint programming represents one of the closest approaches computer science has yet made to the Holy Grail of programming: the user states the problem, the computer solves it". Nowadays, CP users have great modeling tools available (like Minizinc and CPMpy), all...
http://arxiv.org/pdf/2308.01589v1
2023-08-03T07:48:02
2023-08-03T07:48:02
{ "title": [ -0.055114746, -0.018310547, -0.036987305, -0.06335449, 0.005470276, 0.033203125, -0.0670166, -0.0107040405, 0.0115356445, 0.012611389, -0.015129089, -0.0104522705, -0.010437012, -0.018493652, -0.03338623, -0.003036499, -0.024765015, ...
http://arxiv.org/abs/2308.02565v1
SimTeG: A Frustratingly Simple Approach Improves Textual Graph Learning
Keyu Duan, Qian Liu, Tat-Seng Chua, Shuicheng Yan, Wei Tsang Ooi, Qizhe Xie, Junxian He
cs.CL, cs.AI
Textual graphs (TGs) are graphs whose nodes correspond to text (sentences or documents), which are widely prevalent. The representation learning of TGs involves two stages: (i) unsupervised feature extraction and (ii) supervised graph representation learning. In recent years, extensive efforts have been devoted to the ...
http://arxiv.org/pdf/2308.02565v1
2023-08-03T07:00:04
2023-08-03T07:00:04
{ "title": [ -0.013641357, -0.018707275, -0.040161133, -0.055145264, -0.02003479, -0.0011873245, -0.051605225, -0.00844574, -0.0026340485, 0.0284729, -0.010551453, 0.010444641, -0.0146484375, -0.021865845, -0.040283203, 0.0005569458, 0.00006502867, ...
http://arxiv.org/abs/2308.01552v1
InterAct: Exploring the Potentials of ChatGPT as a Cooperative Agent
Po-Lin Chen, Cheng-Shang Chang
cs.AI, cs.CL, cs.LG
This research paper delves into the integration of OpenAI's ChatGPT into embodied agent systems, evaluating its influence on interactive decision-making benchmark. Drawing a parallel to the concept of people assuming roles according to their unique strengths, we introduce InterAct. In this approach, we feed ChatGPT wit...
http://arxiv.org/pdf/2308.01552v1
2023-08-03T06:19:58
2023-08-03T06:19:58
{ "title": [ -0.005241394, -0.010498047, -0.01676941, -0.023803711, 0.008628845, -0.007106781, 0.0016231537, -0.025177002, -0.014907837, 0.031341553, -0.021591187, 0.0317688, -0.022140503, -0.007396698, -0.044799805, 0.00083112717, -0.03857422, 0...
http://arxiv.org/abs/2308.01544v2
Multimodal Neurons in Pretrained Text-Only Transformers
Sarah Schwettmann, Neil Chowdhury, Samuel Klein, David Bau, Antonio Torralba
cs.CV, cs.CL
Language models demonstrate remarkable capacity to generalize representations learned in one modality to downstream tasks in other modalities. Can we trace this ability to individual neurons? We study the case where a frozen text transformer is augmented with vision using a self-supervised visual encoder and a single l...
http://arxiv.org/pdf/2308.01544v2
2023-10-01T23:24:13
2023-08-03T05:27:12
{ "title": [ 0.0040512085, 0.013015747, -0.015731812, -0.037078857, -0.005039215, -0.015129089, -0.011634827, -0.022506714, -0.0041046143, 0.038726807, -0.0009422302, -0.0006790161, 0.0042037964, -0.041015625, -0.019302368, 0.014289856, -0.004852295,...
http://arxiv.org/abs/2308.01535v1
Comparing scalable strategies for generating numerical perspectives
Hancheng Cao, Sofia Eleni Spatharioti, Daniel G. Goldstein, Jake M. Hofman
cs.HC, cs.CL, cs.CY
Numerical perspectives help people understand extreme and unfamiliar numbers (e.g., \$330 billion is about \$1,000 per person in the United States). While research shows perspectives to be helpful, generating them at scale is challenging both because it is difficult to identify what makes some analogies more helpful th...
http://arxiv.org/pdf/2308.01535v1
2023-08-03T04:35:46
2023-08-03T04:35:46
{ "title": [ 0.005619049, 0.0022277832, -0.017196655, -0.070129395, -0.0088272095, -0.00466156, -0.043640137, -0.01184082, 0.0022506714, 0.021865845, -0.0031223297, 0.0017166138, -0.037384033, -0.037628174, -0.020477295, 0.00077962875, 0.021270752, ...
http://arxiv.org/abs/2308.01497v2
Large Language Model Displays Emergent Ability to Interpret Novel Literary Metaphors
Nicholas Ichien, Dušan Stamenković, Keith J. Holyoak
cs.CL
Recent advances in the performance of large language models (LLMs) have sparked debate over whether, given sufficient training, high-level human abilities emerge in such generic forms of artificial intelligence (AI). Despite the exceptional performance of LLMs on a wide range of tasks involving natural language process...
http://arxiv.org/pdf/2308.01497v2
2023-10-13T15:51:46
2023-08-03T01:46:27
{ "title": [ 0.00014090538, 0.0013866425, -0.028686523, 0.0063667297, -0.04916382, -0.023330688, -0.010597229, -0.009918213, -0.010040283, 0.039093018, -0.005428314, -0.031677246, 0.07116699, -0.05038452, -0.0046081543, 0.007293701, 0.0062332153, ...
http://arxiv.org/abs/2308.01479v1
Investigating Reinforcement Learning for Communication Strategies in a Task-Initiative Setting
Baber Khalid, Matthew Stone
cs.CL
Many conversational domains require the system to present nuanced information to users. Such systems must follow up what they say to address clarification questions and repair misunderstandings. In this work, we explore this interactive strategy in a referential communication task. Using simulation, we analyze the comm...
http://arxiv.org/pdf/2308.01479v1
2023-08-03T00:10:23
2023-08-03T00:10:23
{ "title": [ -0.011459351, 0.0061302185, -0.044769287, -0.008644104, 0.010848999, -0.034820557, 0.007820129, -0.013999939, -0.05545044, 0.01776123, 0.004798889, 0.010314941, -0.042114258, -0.029190063, -0.015464783, -0.03756714, -0.019927979, 0.0...
http://arxiv.org/abs/2308.01472v1
Reverse Stable Diffusion: What prompt was used to generate this image?
Florinel-Alin Croitoru, Vlad Hondru, Radu Tudor Ionescu, Mubarak Shah
cs.CV, cs.CL, cs.LG
Text-to-image diffusion models such as Stable Diffusion have recently attracted the interest of many researchers, and inverting the diffusion process can play an important role in better understanding the generative process and how to engineer prompts in order to obtain the desired images. To this end, we introduce the...
http://arxiv.org/pdf/2308.01472v1
2023-08-02T23:39:29
2023-08-02T23:39:29
{ "title": [ -0.0093688965, -0.005844116, -0.028121948, 0.02861023, -0.037994385, -0.024368286, 0.011222839, -0.035888672, 0.015090942, 0.010726929, 0.004962921, -0.005432129, -0.005493164, -0.033599854, -0.001704216, 0.0031490326, 0.0158844, 0.0...
http://arxiv.org/abs/2308.01936v2
Why Do We Need Neuro-symbolic AI to Model Pragmatic Analogies?
Thilini Wijesiriwardene, Amit Sheth, Valerie L. Shalin, Amitava Das
cs.AI, cs.CL
A hallmark of intelligence is the ability to use a familiar domain to make inferences about a less familiar domain, known as analogical reasoning. In this article, we delve into the performance of Large Language Models (LLMs) in dealing with progressively complex analogies expressed in unstructured text. We discuss ana...
http://arxiv.org/pdf/2308.01936v2
2023-09-12T16:33:15
2023-08-02T21:13:38
{ "title": [ 0.020324707, -0.022659302, 0.054016113, 0.029449463, -0.048706055, -0.003944397, 0.03050232, -0.039367676, 0.0023345947, 0.049041748, -0.04550171, -0.021118164, 0.047332764, -0.011306763, -0.019866943, -0.0096588135, 0.005180359, 0.0...
http://arxiv.org/abs/2308.01408v1
UPB at IberLEF-2023 AuTexTification: Detection of Machine-Generated Text using Transformer Ensembles
Andrei-Alexandru Preda, Dumitru-Clementin Cercel, Traian Rebedea, Costin-Gabriel Chiru
cs.CL, I.2.7
This paper describes the solutions submitted by the UPB team to the AuTexTification shared task, featured as part of IberLEF-2023. Our team participated in the first subtask, identifying text documents produced by large language models instead of humans. The organizers provided a bilingual dataset for this subtask, com...
http://arxiv.org/pdf/2308.01408v1
2023-08-02T20:08:59
2023-08-02T20:08:59
{ "title": [ -0.0069503784, 0.020767212, -0.028884888, -0.0031337738, 0.004940033, -0.015281677, -0.025650024, -0.046203613, -0.010513306, 0.057006836, -0.0146102905, 0.022201538, 0.04522705, -0.028778076, -0.0075912476, 0.022201538, -0.029724121, ...
http://arxiv.org/abs/2308.03782v1
Bio+Clinical BERT, BERT Base, and CNN Performance Comparison for Predicting Drug-Review Satisfaction
Yue Ling
cs.CL, cs.AI, cs.LG
The objective of this study is to develop natural language processing (NLP) models that can analyze patients' drug reviews and accurately classify their satisfaction levels as positive, neutral, or negative. Such models would reduce the workload of healthcare professionals and provide greater insight into patients' qua...
http://arxiv.org/pdf/2308.03782v1
2023-08-02T20:01:38
2023-08-02T20:01:38
{ "title": [ -0.00806427, -0.018722534, -0.0121536255, -0.030395508, -0.041137695, -0.007610321, -0.03955078, 0.00844574, 0.019317627, 0.021087646, -0.06896973, 0.0055122375, 0.007659912, -0.010276794, -0.036010742, -0.020263672, -0.012123108, 0....
http://arxiv.org/abs/2308.01391v1
Optimizing Machine Translation through Prompt Engineering: An Investigation into ChatGPT's Customizability
Masaru Yamada
cs.CL
This paper explores the influence of integrating the purpose of the translation and the target audience into prompts on the quality of translations produced by ChatGPT. Drawing on previous translation studies, industry practices, and ISO standards, the research underscores the significance of the pre-production phase i...
http://arxiv.org/pdf/2308.01391v1
2023-08-02T19:11:04
2023-08-02T19:11:04
{ "title": [ -0.04345703, -0.02204895, 0.005760193, -0.005592346, -0.037506104, 0.021347046, -0.032470703, -0.010467529, -0.031677246, 0.03414917, 0.011558533, 0.023269653, -0.0051231384, -0.025360107, -0.025360107, 0.01574707, -0.013374329, 0.03...
http://arxiv.org/abs/2308.01320v1
DeepSpeed-Chat: Easy, Fast and Affordable RLHF Training of ChatGPT-like Models at All Scales
Zhewei Yao, Reza Yazdani Aminabadi, Olatunji Ruwase, Samyam Rajbhandari, Xiaoxia Wu, Ammar Ahmad Awan, Jeff Rasley, Minjia Zhang, Conglong Li, Connor Holmes, Zhongzhu Zhou, Michael Wyatt, Molly Smith, Lev Kurilenko, Heyang Qin, Masahiro Tanaka, Shuai Che, Shuaiwen Leon Song, Yuxiong He
cs.LG, cs.AI, cs.CL
ChatGPT-like models have revolutionized various applications in artificial intelligence, from summarization and coding to translation, matching or even surpassing human performance. However, the current landscape lacks an accessible, efficient, and cost-effective end-to-end RLHF (Reinforcement Learning with Human Feedb...
http://arxiv.org/pdf/2308.01320v1
2023-08-02T18:49:57
2023-08-02T18:49:57
{ "title": [ -0.011177063, -0.021240234, -0.025665283, -0.047210693, -0.0287323, -0.014350891, -0.04248047, -0.012794495, -0.0037155151, 0.019454956, -0.01574707, 0.015342712, 0.0011644363, -0.037994385, -0.009666443, 0.0033512115, -0.009750366, ...
http://arxiv.org/abs/2308.01368v1
Empirical Translation Process Research: Past and Possible Future Perspectives
Michael Carl
cs.CL, cs.AI, cs.IT, math.IT
Over the past four decades, efforts have been made to develop and evaluate models for Empirical Translation Process Research (TPR), yet a comprehensive framework remains elusive. This article traces the evolution of empirical TPR within the CRITT TPR-DB tradition and proposes the Free Energy Principle (FEP) and Active ...
http://arxiv.org/pdf/2308.01368v1
2023-08-02T18:22:49
2023-08-02T18:22:49
{ "title": [ -0.053100586, -0.01777649, -0.020828247, -0.016998291, -0.0018606186, 0.005012512, -0.017456055, -0.015838623, -0.038757324, 0.031585693, 0.011367798, 0.0146102905, 0.0143966675, -0.03982544, -0.028457642, 0.025299072, -0.0061035156, ...
http://arxiv.org/abs/2308.01313v2
More Context, Less Distraction: Zero-shot Visual Classification by Inferring and Conditioning on Contextual Attributes
Bang An, Sicheng Zhu, Michael-Andrei Panaitescu-Liess, Chaithanya Kumar Mummadi, Furong Huang
cs.CV, cs.AI, cs.CL, cs.LG
Vision-language models like CLIP are widely used in zero-shot image classification due to their ability to understand various visual concepts and natural language descriptions. However, how to fully leverage CLIP's unprecedented human-like understanding capabilities to achieve better performance is still an open questi...
http://arxiv.org/pdf/2308.01313v2
2023-10-08T21:56:53
2023-08-02T17:57:25
{ "title": [ 0.028869629, 0.015357971, -0.05343628, -0.055755615, 0.033111572, -0.037078857, 0.005340576, -0.009429932, 0.0146865845, 0.029876709, -0.034210205, 0.013122559, -0.016052246, -0.011192322, -0.049591064, -0.011833191, 0.015914917, 0.0...
http://arxiv.org/abs/2308.01284v2
Fighting Fire with Fire: Can ChatGPT Detect AI-generated Text?
Amrita Bhattacharjee, Huan Liu
cs.CL, cs.AI
Large language models (LLMs) such as ChatGPT are increasingly being used for various use cases, including text content generation at scale. Although detection methods for such AI-generated text exist already, we investigate ChatGPT's performance as a detector on such AI-generated text, inspired by works that use ChatGP...
http://arxiv.org/pdf/2308.01284v2
2023-08-17T22:34:38
2023-08-02T17:11:37
{ "title": [ -0.016723633, 0.014907837, -0.031555176, -0.004840851, -0.0017871857, -0.015899658, -0.0025787354, -0.031173706, -0.032287598, 0.027420044, -0.023406982, 0.04336548, -0.0072746277, -0.024368286, 0.009315491, -0.017822266, -0.020904541, ...
http://arxiv.org/abs/2308.01264v1
Exploring the psychology of GPT-4's Moral and Legal Reasoning
Guilherme F. C. F. Almeida, José Luiz Nunes, Neele Engelmann, Alex Wiegmann, Marcelo de Araújo
cs.AI, cs.CL
Large language models have been used as the foundation of highly sophisticated artificial intelligences, capable of delivering human-like responses to probes about legal and moral issues. However, these models are unreliable guides to their own inner workings, and even the engineering teams behind their creation are un...
http://arxiv.org/pdf/2308.01264v1
2023-08-02T16:36:58
2023-08-02T16:36:58
{ "title": [ 0.021255493, 0.0056114197, -0.0043907166, 0.011802673, -0.022720337, -0.039978027, 0.010940552, -0.029724121, -0.0552063, 0.021133423, -0.0053482056, -0.013656616, 0.005935669, -0.04949951, 0.023620605, -0.021057129, -0.017715454, 0....
http://arxiv.org/abs/2308.01263v2
XSTest: A Test Suite for Identifying Exaggerated Safety Behaviours in Large Language Models
Paul Röttger, Hannah Rose Kirk, Bertie Vidgen, Giuseppe Attanasio, Federico Bianchi, Dirk Hovy
cs.CL, cs.AI
Without proper safeguards, large language models will readily follow malicious instructions and generate toxic content. This risk motivates safety efforts such as red-teaming and large-scale feedback learning, which aim to make models both helpful and harmless. However, there is a tension between these two objectives, ...
http://arxiv.org/pdf/2308.01263v2
2023-10-17T16:21:55
2023-08-02T16:30:40
{ "title": [ 0.00042152405, 0.0010299683, -0.039398193, -0.09667969, 0.024719238, -0.0055389404, -0.053894043, 0.005962372, -0.013587952, 0.022613525, -0.015136719, 0.021743774, 0.013946533, -0.025634766, -0.016738892, 0.01234436, -0.024383545, 0...
http://arxiv.org/abs/2308.01240v1
Evaluating Instruction-Tuned Large Language Models on Code Comprehension and Generation
Zhiqiang Yuan, Junwei Liu, Qiancheng Zi, Mingwei Liu, Xin Peng, Yiling Lou
cs.CL, cs.AI
In this work, we evaluate 10 open-source instructed LLMs on four representative code comprehension and generation tasks. We have the following main findings. First, for the zero-shot setting, instructed LLMs are very competitive on code comprehension and generation tasks and sometimes even better than small SOTA models...
http://arxiv.org/pdf/2308.01240v1
2023-08-02T15:54:22
2023-08-02T15:54:22
{ "title": [ -0.047790527, -0.03842163, -0.018859863, -0.051940918, -0.03451538, 0.016403198, -0.06890869, 0.017028809, 0.0017242432, 0.023162842, -0.032836914, -0.0023975372, -0.019058228, -0.009429932, -0.04736328, -0.011772156, -0.024246216, 0...
http://arxiv.org/abs/2308.01327v1
Careful Whisper -- leveraging advances in automatic speech recognition for robust and interpretable aphasia subtype classification
Laurin Wagner, Mario Zusag, Theresa Bloder
cs.SD, cs.CL, cs.LG, eess.AS
This paper presents a fully automated approach for identifying speech anomalies from voice recordings to aid in the assessment of speech impairments. By combining Connectionist Temporal Classification (CTC) and encoder-decoder-based automatic speech recognition models, we generate rich acoustic and clean transcripts. W...
http://arxiv.org/pdf/2308.01327v1
2023-08-02T15:53:59
2023-08-02T15:53:59
{ "title": [ -0.012741089, 0.010643005, -0.04434204, -0.049865723, 0.0021858215, -0.010070801, -0.037231445, -0.011795044, -0.03616333, 0.034118652, -0.020370483, 0.004753113, -0.012680054, -0.03881836, -0.027038574, -0.016052246, -0.0060920715, ...
http://arxiv.org/abs/2308.01236v2
Grounded Image Text Matching with Mismatched Relation Reasoning
Yu Wu, Yana Wei, Haozhe Wang, Yongfei Liu, Sibei Yang, Xuming He
cs.CV, cs.CL
This paper introduces Grounded Image Text Matching with Mismatched Relation (GITM-MR), a novel visual-linguistic joint task that evaluates the relation understanding capabilities of transformer-based pre-trained models. GITM-MR requires a model to first determine if an expression describes an image, then localize refer...
http://arxiv.org/pdf/2308.01236v2
2023-08-04T17:51:57
2023-08-02T15:44:36
{ "title": [ -0.010444641, 0.013214111, -0.02861023, -0.028961182, 0.0010585785, -0.005584717, -0.010787964, -0.01826477, 0.028015137, 0.032043457, -0.03149414, -0.0011835098, 0.00038599968, -0.008338928, -0.03250122, -0.009162903, -0.0032653809, ...
http://arxiv.org/abs/2308.01223v1
Do Multilingual Language Models Think Better in English?
Julen Etxaniz, Gorka Azkune, Aitor Soroa, Oier Lopez de Lacalle, Mikel Artetxe
cs.CL, cs.AI, cs.LG
Translate-test is a popular technique to improve the performance of multilingual language models. This approach works by translating the input into English using an external machine translation system, and running inference over the translated input. However, these improvements can be attributed to the use of a separat...
http://arxiv.org/pdf/2308.01223v1
2023-08-02T15:29:22
2023-08-02T15:29:22
{ "title": [ -0.02998352, -0.01007843, -0.008102417, 0.007762909, -0.05709839, -0.0024700165, -0.04522705, -0.033966064, -0.01826477, 0.034301758, 0.000010311604, -0.013092041, 0.012809753, -0.04473877, -0.008010864, 0.0074882507, -0.022003174, 0...
http://arxiv.org/abs/2308.01210v1
Global Hierarchical Neural Networks using Hierarchical Softmax
Jetze Schuurmans, Flavius Frasincar
stat.ML, cs.CL, cs.LG
This paper presents a framework in which hierarchical softmax is used to create a global hierarchical classifier. The approach is applicable for any classification task where there is a natural hierarchy among classes. We show empirical results on four text classification datasets. In all datasets the hierarchical soft...
http://arxiv.org/pdf/2308.01210v1
2023-08-02T15:12:56
2023-08-02T15:12:56
{ "title": [ -0.023132324, 0.0013818741, -0.046722412, -0.064819336, 0.022247314, -0.008659363, -0.03616333, -0.002773285, -0.0047950745, 0.034240723, -0.008476257, 0.005065918, 0.0041618347, -0.028167725, -0.027832031, 0.019729614, -0.0016517639, ...
http://arxiv.org/abs/2308.01154v1
Arithmetic with Language Models: from Memorization to Computation
Davide Maltoni, Matteo Ferrara
cs.AI, cs.CL
A better understanding of the emergent computation and problem-solving capabilities of recent large language models is of paramount importance to further improve them and broaden their applicability. This work investigates how a language model, trained to predict the next token, can perform arithmetic computations gene...
http://arxiv.org/pdf/2308.01154v1
2023-08-02T13:58:37
2023-08-02T13:58:37
{ "title": [ -0.044158936, -0.007896423, -0.026794434, -0.05319214, -0.0040779114, 0.0051345825, -0.03439331, 0.013908386, 0.009422302, 0.03741455, -0.007843018, -0.0036964417, -0.008842468, -0.013626099, -0.0625, 0.008811951, -0.004463196, 0.018...
http://arxiv.org/abs/2308.01143v1
ADS-Cap: A Framework for Accurate and Diverse Stylized Captioning with Unpaired Stylistic Corpora
Kanzhi Cheng, Zheng Ma, Shi Zong, Jianbing Zhang, Xinyu Dai, Jiajun Chen
cs.CV, cs.CL
Generating visually grounded image captions with specific linguistic styles using unpaired stylistic corpora is a challenging task, especially since we expect stylized captions with a wide variety of stylistic patterns. In this paper, we propose a novel framework to generate Accurate and Diverse Stylized Captions (ADS-...
http://arxiv.org/pdf/2308.01143v1
2023-08-02T13:33:20
2023-08-02T13:33:20
{ "title": [ -0.018539429, 0.0016679764, -0.039215088, -0.04473877, -0.0018882751, -0.014587402, -0.037750244, -0.014961243, 0.0020885468, 0.049316406, -0.016448975, 0.026977539, 0.018585205, 0.0041122437, -0.035247803, 0.009490967, -0.00349617, ...
http://arxiv.org/abs/2308.01126v1
Beyond Generic: Enhancing Image Captioning with Real-World Knowledge using Vision-Language Pre-Training Model
Kanzhi Cheng, Wenpo Song, Zheng Ma, Wenhao Zhu, Zixuan Zhu, Jianbing Zhang
cs.CV, cs.CL, cs.MM
Current captioning approaches tend to generate correct but "generic" descriptions that lack real-world knowledge, e.g., named entities and contextual information. Considering that Vision-Language Pre-Training (VLP) models master massive such knowledge from large-scale web-harvested data, it is promising to utilize the ...
http://arxiv.org/pdf/2308.01126v1
2023-08-02T13:09:57
2023-08-02T13:09:57
{ "title": [ 0.017990112, -0.010520935, -0.030532837, -0.057861328, -0.0063171387, -0.038726807, -0.014053345, -0.008468628, 0.011917114, 0.044921875, -0.016525269, 0.0178833, 0.0036754608, -0.0067329407, -0.0463562, -0.00554657, 0.016906738, 0.0...
http://arxiv.org/abs/2308.01927v1
MultiEM: Efficient and Effective Unsupervised Multi-Table Entity Matching
Xiaocan Zeng, Pengfei Wang, Yuren Mao, Lu Chen, Xiaoze Liu, Yunjun Gao
cs.DB, cs.CL, cs.IR
Entity Matching (EM), which aims to identify all entity pairs referring to the same real-world entity from relational tables, is one of the most important tasks in real-world data management systems. Due to the labeling process of EM being extremely labor-intensive, unsupervised EM is more applicable than supervised EM...
http://arxiv.org/pdf/2308.01927v1
2023-08-02T11:39:19
2023-08-02T11:39:19
{ "title": [ -0.0124053955, 0.008544922, -0.066223145, -0.07159424, 0.048065186, -0.004459381, -0.019882202, -0.0038261414, 0.025787354, 0.035858154, -0.018234253, 0.017913818, 0.034057617, -0.0039100647, -0.034576416, 0.022781372, -0.0135269165, ...
http://arxiv.org/abs/2308.01080v1
Leveraging Few-Shot Data Augmentation and Waterfall Prompting for Response Generation
Lea Krause, Selene Báez Santamaría, Michiel van der Meer, Urja Khurana
cs.CL
This paper discusses our approaches for task-oriented conversational modelling using subjective knowledge, with a particular emphasis on response generation. Our methodology was shaped by an extensive data analysis that evaluated key factors such as response length, sentiment, and dialogue acts present in the provided ...
http://arxiv.org/pdf/2308.01080v1
2023-08-02T11:04:27
2023-08-02T11:04:27
{ "title": [ 0.03933716, -0.025360107, -0.04638672, -0.073791504, -0.012229919, -0.016937256, -0.027908325, 0.020111084, 0.0032215118, 0.015426636, -0.03289795, 0.030654907, -0.04824829, -0.011459351, -0.044433594, -0.022705078, 0.03744507, -0.01...
http://arxiv.org/abs/2308.02556v1
Industrial Memories: Exploring the Findings of Government Inquiries with Neural Word Embedding and Machine Learning
Susan Leavy, Emilie Pine, Mark T Keane
cs.CL, cs.CY
We present a text mining system to support the exploration of large volumes of text detailing the findings of government inquiries. Despite their historical significance and potential societal impact, key findings of inquiries are often hidden within lengthy documents and remain inaccessible to the general public. We t...
http://arxiv.org/pdf/2308.02556v1
2023-08-02T10:39:11
2023-08-02T10:39:11
{ "title": [ 0.0026950836, -0.010620117, -0.05218506, -0.02017212, -0.0005187988, -0.044891357, -0.00818634, 0.0070610046, -0.016540527, 0.008140564, -0.030319214, 0.02355957, -0.041259766, -0.021774292, -0.019058228, -0.020431519, -0.03152466, 0...
http://arxiv.org/abs/2308.01044v1
Chat Translation Error Detection for Assisting Cross-lingual Communications
Yunmeng Li, Jun Suzuki, Makoto Morishita, Kaori Abe, Ryoko Tokuhisa, Ana Brassard, Kentaro Inui
cs.CL, cs.AI
In this paper, we describe the development of a communication support system that detects erroneous translations to facilitate crosslingual communications due to the limitations of current machine chat translation methods. We trained an error detector as the baseline of the system and constructed a new Japanese-English...
http://arxiv.org/pdf/2308.01044v1
2023-08-02T09:38:29
2023-08-02T09:38:29
{ "title": [ -0.013259888, -0.0029392242, -0.009384155, -0.047729492, -0.0077323914, 0.009292603, -0.043640137, -0.017410278, 0.027069092, 0.055664062, -0.022903442, -0.0043258667, 0.05657959, -0.012535095, -0.047943115, 0.018798828, -0.0058784485, ...
http://arxiv.org/abs/2308.01018v1
SALTTS: Leveraging Self-Supervised Speech Representations for improved Text-to-Speech Synthesis
Ramanan Sivaguru, Vasista Sai Lodagala, S Umesh
cs.CL, cs.SD, eess.AS
While FastSpeech2 aims to integrate aspects of speech such as pitch, energy, and duration as conditional inputs, it still leaves scope for richer representations. As a part of this work, we leverage representations from various Self-Supervised Learning (SSL) models to enhance the quality of the synthesized speech. In p...
http://arxiv.org/pdf/2308.01018v1
2023-08-02T08:59:52
2023-08-02T08:59:52
{ "title": [ -0.006427765, -0.0028839111, -0.03564453, -0.084228516, 0.008331299, 0.013671875, -0.062805176, 0.030151367, -0.022262573, 0.052246094, -0.022399902, 0.025253296, 0.01524353, -0.0119018555, -0.039764404, 0.0028820038, -0.013786316, 0...
http://arxiv.org/abs/2308.00946v2
Teaching Smaller Language Models To Generalise To Unseen Compositional Questions
Tim Hartill, Neset Tan, Michael Witbrock, Patricia J. Riddle
cs.CL, cs.AI
We equip a smaller Language Model to generalise to answering challenging compositional questions that have not been seen in training. To do so we propose a combination of multitask supervised pretraining on up to 93 tasks designed to instill diverse reasoning abilities, and a dense retrieval system that aims to retriev...
http://arxiv.org/pdf/2308.00946v2
2023-08-21T00:28:44
2023-08-02T05:00:12
{ "title": [ -0.033294678, -0.015014648, -0.027679443, -0.044158936, -0.035339355, 0.0077667236, -0.06573486, -0.004802704, 0.0021018982, 0.019714355, -0.01979065, -0.005706787, -0.027526855, -0.01902771, -0.058898926, -0.024276733, -0.016647339, ...
http://arxiv.org/abs/2308.00939v1
Feature-aware conditional GAN for category text generation
Xinze Li, Kezhi Mao, Fanfan Lin, Zijian Feng
cs.CL, cs.AI
Category text generation receives considerable attentions since it is beneficial for various natural language processing tasks. Recently, the generative adversarial network (GAN) has attained promising performance in text generation, attributed to its adversarial training process. However, there are several issues in t...
http://arxiv.org/pdf/2308.00939v1
2023-08-02T04:43:54
2023-08-02T04:43:54
{ "title": [ -0.028289795, 0.012718201, -0.05331421, -0.05722046, 0.01574707, 0.0012359619, -0.049713135, 0.0065078735, 0.014144897, 0.01576233, -0.017974854, 0.0031375885, 0.0058937073, -0.025039673, -0.03579712, -0.0061035156, -0.018066406, 0.0...
http://arxiv.org/abs/2308.00878v1
DiactTOD: Learning Generalizable Latent Dialogue Acts for Controllable Task-Oriented Dialogue Systems
Qingyang Wu, James Gung, Raphael Shu, Yi Zhang
cs.CL
Dialogue act annotations are important to improve response generation quality in task-oriented dialogue systems. However, it can be challenging to use dialogue acts to control response generation in a generalizable way because different datasets and tasks may have incompatible annotations. While alternative methods tha...
http://arxiv.org/pdf/2308.00878v1
2023-08-01T23:29:16
2023-08-01T23:29:16
{ "title": [ -0.011672974, -0.009521484, -0.051818848, -0.051361084, 0.00831604, -0.008720398, -0.03161621, 0.012260437, 0.010185242, 0.028182983, -0.027633667, 0.015182495, -0.006504059, -0.0066108704, -0.04269409, -0.00016701221, -0.011772156, ...
http://arxiv.org/abs/2308.03679v1
Proceedings Modalities in substructural logics: Applications at the interfaces of logic, language and computation
Michael Moortgat, Mehrnoosh Sadrzadeh
cs.LO, cs.CL
By calling into question the implicit structural rules that are taken for granted in classical logic, substructural logics have brought to the fore new forms of reasoning with applications in many interdisciplinary areas of interest. Modalities, in the substructural setting, provide the tools to control and finetune th...
http://arxiv.org/pdf/2308.03679v1
2023-08-01T22:40:19
2023-08-01T22:40:19
{ "title": [ -0.022827148, 0.008148193, -0.060668945, -0.08312988, 0.021102905, -0.008651733, -0.06878662, 0.0004234314, 0.0043678284, 0.020050049, -0.00048422813, -0.021224976, 0.01424408, -0.041534424, -0.021850586, 0.0017757416, -0.018966675, ...
http://arxiv.org/abs/2308.02548v1
Aspect based sentimental analysis for travellers' reviews
Mohammed Saad M Alaydaa, Jun Li, Karl Jinkins
cs.CL
Airport service quality evaluation is commonly found on social media, including Google Maps. This valuable for airport management in order to enhance the quality of services provided. However; prior studies either provide general review for topics discussed by travellers or provide sentimental value to tag the entire r...
http://arxiv.org/pdf/2308.02548v1
2023-08-01T21:23:02
2023-08-01T21:23:02
{ "title": [ 0.009651184, -0.008544922, 0.0032672882, -0.022109985, -0.008430481, 0.00041151047, -0.0015687943, -0.025985718, 0.009475708, 0.03741455, -0.026931763, -0.0158844, -0.0053634644, -0.01586914, -0.060821533, -0.011672974, 0.007255554, ...
http://arxiv.org/abs/2308.00802v2
GRDD: A Dataset for Greek Dialectal NLP
Stergios Chatzikyriakidis, Chatrine Qwaider, Ilias Kolokousis, Christina Koula, Dimitris Papadakis, Efthymia Sakellariou
cs.CL
In this paper, we present a dataset for the computational study of a number of Modern Greek dialects. It consists of raw text data from four dialects of Modern Greek, Cretan, Pontic, Northern Greek and Cypriot Greek. The dataset is of considerable size, albeit imbalanced, and presents the first attempt to create large ...
http://arxiv.org/pdf/2308.00802v2
2023-09-13T14:43:45
2023-08-01T19:34:18
{ "title": [ -0.0026817322, -0.011604309, -0.020523071, -0.059692383, 0.0059661865, 0.007068634, -0.02330017, 0.003435135, 0.03439331, 0.035247803, -0.04574585, 0.0048599243, 0.026367188, 0.0035934448, -0.06530762, 0.016021729, 0.003019333, 0.017...
http://arxiv.org/abs/2308.00762v1
Self-Supervised Contrastive BERT Fine-tuning for Fusion-based Reviewed-Item Retrieval
Mohammad Mahdi Abdollah Pour, Parsa Farinneya, Armin Toroghi, Anton Korikov, Ali Pesaranghader, Touqir Sajed, Manasa Bharadwaj, Borislav Mavrin, Scott Sanner
cs.IR, cs.CL, cs.LG
As natural language interfaces enable users to express increasingly complex natural language queries, there is a parallel explosion of user review content that can allow users to better find items such as restaurants, books, or movies that match these expressive queries. While Neural Information Retrieval (IR) methods ...
http://arxiv.org/pdf/2308.00762v1
2023-08-01T18:01:21
2023-08-01T18:01:21
{ "title": [ 0.019989014, 0.0039367676, 0.029037476, -0.040283203, -0.016998291, 0.004283905, -0.0064811707, -0.031311035, 0.016082764, 0.052459717, -0.04888916, 0.022583008, 0.0068740845, 0.0134887695, -0.07495117, -0.002878189, -0.012123108, 0....
http://arxiv.org/abs/2308.00755v2
The Bias Amplification Paradox in Text-to-Image Generation
Preethi Seshadri, Sameer Singh, Yanai Elazar
cs.LG, cs.CL, cs.CV, cs.CY
Bias amplification is a phenomenon in which models exacerbate biases or stereotypes present in the training data. In this paper, we study bias amplification in the text-to-image domain using Stable Diffusion by comparing gender ratios in training vs. generated images. We find that the model appears to amplify gender-oc...
http://arxiv.org/pdf/2308.00755v2
2023-11-15T08:51:11
2023-08-01T18:00:08
{ "title": [ -0.0041999817, 0.012771606, -0.022079468, -0.022644043, -0.00044035912, -0.0066070557, -0.017471313, -0.040283203, 0.012336731, 0.043701172, -0.036956787, 0.008995056, 0.006538391, -0.010978699, -0.034057617, -0.0020065308, -0.0060806274...
http://arxiv.org/abs/2308.00683v1
CodeBPE: Investigating Subtokenization Options for Large Language Model Pretraining on Source Code
Nadezhda Chirkova, Sergey Troshin
cs.LG, cs.CL, cs.SE
Recent works have widely adopted large language model pretraining for source code, suggested source code-specific pretraining objectives and investigated the applicability of various Transformer-based language model architectures for source code. This work investigates another important aspect of such models, namely th...
http://arxiv.org/pdf/2308.00683v1
2023-08-01T17:40:48
2023-08-01T17:40:48
{ "title": [ -0.025024414, -0.0287323, -0.028671265, -0.06414795, -0.011260986, 0.003042221, -0.05065918, 0.015144348, -0.015106201, 0.025039673, -0.016662598, 0.008728027, 0.021453857, -0.03463745, -0.00447464, 0.0027103424, -0.016723633, 0.0263...
http://arxiv.org/abs/2308.00675v1
Tool Documentation Enables Zero-Shot Tool-Usage with Large Language Models
Cheng-Yu Hsieh, Si-An Chen, Chun-Liang Li, Yasuhisa Fujii, Alexander Ratner, Chen-Yu Lee, Ranjay Krishna, Tomas Pfister
cs.CL, cs.AI, cs.CV, cs.LG
Today, large language models (LLMs) are taught to use new tools by providing a few demonstrations of the tool's usage. Unfortunately, demonstrations are hard to acquire, and can result in undesirable biased usage if the wrong demonstration is chosen. Even in the rare scenario that demonstrations are readily available, ...
http://arxiv.org/pdf/2308.00675v1
2023-08-01T17:21:38
2023-08-01T17:21:38
{ "title": [ 0.023910522, -0.037200928, -0.05227661, -0.097473145, -0.04522705, -0.018203735, -0.07965088, 0.04864502, 0.008331299, 0.043640137, -0.018081665, -0.025222778, 0.04473877, -0.04345703, 0.0057754517, 0.00095796585, 0.023513794, 0.0260...
http://arxiv.org/abs/2308.00624v1
JIANG: Chinese Open Foundation Language Model
Qinhua Duan, Wenchao Gu, Yujia Chen, Wenxin Mao, Zewen Tian, Hui Cao
cs.CL, cs.AI
With the advancements in large language model technology, it has showcased capabilities that come close to those of human beings across various tasks. This achievement has garnered significant interest from companies and scientific research institutions, leading to substantial investments in the research and developmen...
http://arxiv.org/pdf/2308.00624v1
2023-08-01T15:51:41
2023-08-01T15:51:41
{ "title": [ -0.056243896, -0.018203735, -0.04159546, -0.012878418, -0.056884766, 0.013420105, -0.0770874, -0.020202637, -0.046875, 0.038848877, 0.0010929108, -0.009643555, 0.052368164, -0.056121826, 0.023406982, -0.0045547485, -0.030410767, 0.04...
http://arxiv.org/abs/2308.02541v1
Towards More Human-like AI Communication: A Review of Emergent Communication Research
Nicolo' Brandizzi
cs.CL, cs.AI, cs.HC, cs.MA, I.2.6; I.2.7; I.2.11
In the recent shift towards human-centric AI, the need for machines to accurately use natural language has become increasingly important. While a common approach to achieve this is to train large language models, this method presents a form of learning misalignment where the model may not capture the underlying structu...
http://arxiv.org/pdf/2308.02541v1
2023-08-01T14:43:10
2023-08-01T14:43:10
{ "title": [ -0.042266846, -0.0036849976, -0.026062012, -0.0034751892, 0.004173279, -0.008163452, 0.015266418, -0.022598267, -0.008995056, 0.03527832, -0.00365448, 0.014541626, 0.010368347, -0.017471313, -0.042266846, 0.008384705, -0.02192688, 0....
http://arxiv.org/abs/2308.02539v1
CoSMo: A constructor specification language for Abstract Wikipedia's content selection process
Kutz Arrieta, Pablo R. Fillottrani, C. Maria Keet
cs.CL, I.2.4; H.2.3
Representing snippets of information abstractly is a task that needs to be performed for various purposes, such as database view specification and the first stage in the natural language generation pipeline for generative AI from structured input, i.e., the content selection stage to determine what needs to be verbalis...
http://arxiv.org/pdf/2308.02539v1
2023-08-01T13:57:23
2023-08-01T13:57:23
{ "title": [ -0.0079422, 0.009605408, -0.06124878, -0.051849365, 0.016494751, -0.019882202, -0.031982422, -0.019866943, -0.016357422, 0.020828247, -0.03744507, 0.029937744, 0.014350891, -0.014389038, 0.021728516, -0.013206482, -0.018600464, 0.047...
http://arxiv.org/abs/2308.00528v1
Unimodal Intermediate Training for Multimodal Meme Sentiment Classification
Muzhaffar Hazman, Susan McKeever, Josephine Griffith
cs.CL
Internet Memes remain a challenging form of user-generated content for automated sentiment classification. The availability of labelled memes is a barrier to developing sentiment classifiers of multimodal memes. To address the shortage of labelled memes, we propose to supplement the training of a multimodal meme classi...
http://arxiv.org/pdf/2308.00528v1
2023-08-01T13:14:10
2023-08-01T13:14:10
{ "title": [ 0.02180481, 0.018432617, -0.016143799, -0.02268982, -0.0018939972, -0.023223877, -0.0038547516, -0.022262573, 0.013183594, 0.04989624, -0.01777649, 0.013336182, 0.024993896, -0.01966858, -0.030944824, 0.017913818, -0.028564453, 0.065...
http://arxiv.org/abs/2308.00479v1
Retrieval Augmented Generation and Representative Vector Summarization for large unstructured textual data in Medical Education
S. S. Manathunga, Y. A. Illangasekara
cs.CL, cs.AI, H.3.1; J.3
Large Language Models are increasingly being used for various tasks including content generation and as chatbots. Despite their impressive performances in general tasks, LLMs need to be aligned when applying for domain specific tasks to mitigate the problems of hallucination and producing harmful answers. Retrieval Aug...
http://arxiv.org/pdf/2308.00479v1
2023-08-01T12:04:50
2023-08-01T12:04:50
{ "title": [ 0.015258789, 0.012321472, -0.014984131, -0.06530762, 0.007041931, 0.007827759, -0.03677368, -0.02456665, -0.009712219, 0.051818848, 0.0023384094, 0.011367798, 0.01701355, -0.024902344, -0.043945312, -0.006706238, 0.0029029846, 0.0047...
http://arxiv.org/abs/2308.00447v1
Structural Embeddings of Tools for Large Language Models
Eren Unlu
cs.AI, cs.CL
It is evident that the current state of Large Language Models (LLMs) necessitates the incorporation of external tools. The lack of straightforward algebraic and logical reasoning is well documented and prompted researchers to develop frameworks which allow LLMs to operate via external tools. The ontological nature of t...
http://arxiv.org/pdf/2308.00447v1
2023-08-01T10:46:09
2023-08-01T10:46:09
{ "title": [ -0.03652954, -0.01689148, -0.033966064, -0.06933594, -0.0075950623, 0.019607544, -0.06817627, 0.020645142, 0.018463135, 0.032684326, -0.019256592, -0.010498047, 0.018569946, -0.013298035, -0.04434204, 0.013198853, -0.008712769, 0.034...
http://arxiv.org/abs/2308.02537v1
ALE: A Simulation-Based Active Learning Evaluation Framework for the Parameter-Driven Comparison of Query Strategies for NLP
Philipp Kohl, Nils Freyer, Yoka Krämer, Henri Werth, Steffen Wolf, Bodo Kraft, Matthias Meinecke, Albert Zündorf
cs.CL, cs.AI
Supervised machine learning and deep learning require a large amount of labeled data, which data scientists obtain in a manual, and time-consuming annotation process. To mitigate this challenge, Active Learning (AL) proposes promising data points to annotators they annotate next instead of a subsequent or random sample...
http://arxiv.org/pdf/2308.02537v1
2023-08-01T10:42:11
2023-08-01T10:42:11
{ "title": [ -0.0013418198, -0.023757935, 0.0030269623, -0.045440674, -0.0010709763, 0.004550934, -0.0023441315, -0.0073013306, -0.01322937, 0.03604126, -0.0072021484, 0.043914795, -0.0062561035, -0.021362305, -0.014083862, 0.0069999695, -0.011558533...
http://arxiv.org/abs/2308.00436v3
SelfCheck: Using LLMs to Zero-Shot Check Their Own Step-by-Step Reasoning
Ning Miao, Yee Whye Teh, Tom Rainforth
cs.AI, cs.CL, cs.LG
The recent progress in large language models (LLMs), especially the invention of chain-of-thought prompting, has made it possible to automatically answer questions by stepwise reasoning. However, when faced with more complicated problems that require non-linear thinking, even the strongest LLMs make mistakes. To addres...
http://arxiv.org/pdf/2308.00436v3
2023-10-05T12:59:59
2023-08-01T10:31:36
{ "title": [ -0.019607544, -0.01058197, -0.051818848, -0.042755127, -0.025466919, -0.011917114, -0.050354004, 0.0082092285, -0.03353882, 0.019714355, -0.027770996, 0.023071289, -0.026321411, -0.04940796, 0.011138916, -0.02305603, 0.010368347, 0.0...
http://arxiv.org/abs/2308.00425v1
Discourse-Aware Text Simplification: From Complex Sentences to Linked Propositions
Christina Niklaus, Matthias Cetto, André Freitas, Siegfried Handschuh
cs.CL, cs.AI
Sentences that present a complex syntax act as a major stumbling block for downstream Natural Language Processing applications whose predictive quality deteriorates with sentence length and complexity. The task of Text Simplification (TS) may remedy this situation. It aims to modify sentences in order to make them easi...
http://arxiv.org/pdf/2308.00425v1
2023-08-01T10:10:59
2023-08-01T10:10:59
{ "title": [ -0.012832642, -0.0014247894, -0.02696228, -0.08532715, 0.0035953522, 0.018798828, -0.06109619, 0.0018396378, -0.0035133362, 0.047698975, 0.018722534, 0.025665283, -0.0043678284, -0.013763428, -0.05795288, 0.0074272156, -0.0058021545, ...
http://arxiv.org/abs/2308.06241v1
Covid-19 Public Sentiment Analysis for Indian Tweets Classification
Mohammad Maksood Akhter, Devpriya Kanojia
cs.CL, cs.SI
When any extraordinary event takes place in the world wide area, it is the social media that acts as the fastest carrier of the news along with the consequences dealt with that event. One can gather much information through social networks regarding the sentiments, behavior, and opinions of the people. In this paper, w...
http://arxiv.org/pdf/2308.06241v1
2023-08-01T09:29:55
2023-08-01T09:29:55
{ "title": [ 0.012023926, 0.011116028, -0.052001953, -0.05908203, 0.003189087, -0.01058197, -0.030654907, 0.002904892, -0.026184082, 0.055847168, -0.04321289, 0.02420044, 0.019241333, -0.012214661, -0.02319336, -0.020858765, -0.0039520264, 0.0328...
http://arxiv.org/abs/2308.00400v2
ZRIGF: An Innovative Multimodal Framework for Zero-Resource Image-Grounded Dialogue Generation
Bo Zhang, Jian Wang, Hui Ma, Bo Xu, Hongfei Lin
cs.CL, cs.MM
Image-grounded dialogue systems benefit greatly from integrating visual information, resulting in high-quality response generation. However, current models struggle to effectively utilize such information in zero-resource scenarios, mainly due to the disparity between image and text modalities. To overcome this challen...
http://arxiv.org/pdf/2308.00400v2
2023-08-02T13:53:30
2023-08-01T09:28:36
{ "title": [ 0.011444092, -0.0095825195, -0.056854248, -0.07897949, 0.003080368, -0.01576233, -0.048583984, 0.015167236, 0.01977539, 0.024230957, 0.0051841736, 0.0053520203, -0.027908325, -0.03338623, -0.05819702, -0.0013809204, -0.0073928833, 0....
http://arxiv.org/abs/2308.00399v1
Tackling Hallucinations in Neural Chart Summarization
Saad Obaid ul Islam, Iza Škrjanec, Ondřej Dušek, Vera Demberg
cs.CL, cs.LG
Hallucinations in text generation occur when the system produces text that is not grounded in the input. In this work, we tackle the problem of hallucinations in neural chart summarization. Our analysis shows that the target side of chart summarization training datasets often contains additional information, leading to...
http://arxiv.org/pdf/2308.00399v1
2023-08-01T09:26:40
2023-08-01T09:26:40
{ "title": [ 0.018600464, 0.0033073425, -0.020767212, -0.061340332, -0.0058784485, -0.008407593, -0.029724121, -0.014122009, 0.03692627, 0.049957275, -0.017623901, -0.0087890625, 0.0017471313, -0.002998352, -0.081970215, 0.01776123, 0.0158844, 0....
http://arxiv.org/abs/2308.00364v1
Fountain -- an intelligent contextual assistant combining knowledge representation and language models for manufacturing risk identification
Saurabh Kumar, Daniel Fuchs, Klaus Spindler
cs.CL, cs.AI
Deviations from the approved design or processes during mass production can lead to unforeseen risks. However, these changes are sometimes necessary due to changes in the product design characteristics or an adaptation in the manufacturing process. A major challenge is to identify these risks early in the workflow so t...
http://arxiv.org/pdf/2308.00364v1
2023-08-01T08:12:43
2023-08-01T08:12:43
{ "title": [ -0.0051345825, 0.018127441, -0.08227539, -0.04916382, 0.0020923615, -0.024963379, -0.03253174, 0.018081665, -0.0051727295, 0.04284668, -0.055755615, 0.010749817, 0.009979248, -0.0107040405, -0.022491455, -0.03579712, -0.0057678223, 0...
http://arxiv.org/abs/2308.00319v1
LimeAttack: Local Explainable Method for Textual Hard-Label Adversarial Attack
Hai Zhu, Zhaoqing Yang, Weiwei Shang, Yuren Wu
cs.CL
Natural language processing models are vulnerable to adversarial examples. Previous textual adversarial attacks adopt gradients or confidence scores to calculate word importance ranking and generate adversarial examples. However, this information is unavailable in the real world. Therefore, we focus on a more realistic...
http://arxiv.org/pdf/2308.00319v1
2023-08-01T06:30:37
2023-08-01T06:30:37
{ "title": [ -0.025817871, -0.0023841858, -0.022445679, -0.035980225, 0.0077934265, 0.001912117, -0.00945282, -0.00818634, -0.053771973, 0.018463135, 0.021728516, 0.042999268, -0.03805542, -0.032073975, -0.012512207, -0.010917664, -0.012504578, 0...
http://arxiv.org/abs/2308.00304v2
Skills-in-Context Prompting: Unlocking Compositionality in Large Language Models
Jiaao Chen, Xiaoman Pan, Dian Yu, Kaiqiang Song, Xiaoyang Wang, Dong Yu, Jianshu Chen
cs.CL
We consider the problem of eliciting compositional generalization capabilities in large language models (LLMs) with a novel type of prompting strategy. Compositional generalization empowers the LLMs to solve problems that are harder than the ones they have seen (i.e., easy-to-hard generalization), which is a critical r...
http://arxiv.org/pdf/2308.00304v2
2023-08-14T08:11:15
2023-08-01T05:54:12
{ "title": [ 0.006450653, -0.008255005, -0.021514893, -0.08404541, -0.004142761, -0.0006289482, -0.072021484, 0.0004105568, 0.00869751, 0.030090332, -0.026168823, 0.008308411, 0.024414062, -0.01651001, -0.030517578, 0.007896423, -0.0072402954, 0....
http://arxiv.org/abs/2308.02582v3
Adapt and Decompose: Efficient Generalization of Text-to-SQL via Domain Adapted Least-To-Most Prompting
Aseem Arora, Shabbirhussain Bhaisaheb, Harshit Nigam, Manasi Patwardhan, Lovekesh Vig, Gautam Shroff
cs.CL, cs.AI, cs.LG
Cross-domain and cross-compositional generalization of Text-to-SQL semantic parsing is a challenging task. Existing Large Language Model (LLM) based solutions rely on inference-time retrieval of few-shot exemplars from the training set to synthesize a run-time prompt for each Natural Language (NL) test query. In contra...
http://arxiv.org/pdf/2308.02582v3
2023-08-09T09:03:12
2023-08-01T05:31:36
{ "title": [ -0.012168884, -0.00705719, -0.03439331, -0.087646484, 0.018478394, 0.016601562, -0.0340271, 0.015037537, 0.028366089, 0.03643799, -0.03692627, 0.028457642, -0.0037136078, 0.005142212, -0.06036377, -0.015991211, 0.008964539, 0.0074844...
http://arxiv.org/abs/2308.00295v1
Making the V in Text-VQA Matter
Shamanthak Hegde, Soumya Jahagirdar, Shankar Gangisetty
cs.CV, cs.AI, cs.CL
Text-based VQA aims at answering questions by reading the text present in the images. It requires a large amount of scene-text relationship understanding compared to the VQA task. Recent studies have shown that the question-answer pairs in the dataset are more focused on the text present in the image but less importanc...
http://arxiv.org/pdf/2308.00295v1
2023-08-01T05:28:13
2023-08-01T05:28:13
{ "title": [ -0.01852417, -0.020111084, -0.057861328, -0.024780273, -0.06506348, -0.021850586, -0.040496826, 0.008422852, -0.020767212, 0.06512451, 0.003862381, -0.018585205, 0.027420044, -0.04147339, -0.02986145, -0.007820129, -0.0016508102, 0.0...
http://arxiv.org/abs/2308.00264v2
Multi-Modality Multi-Loss Fusion Network
Zehui Wu, Ziwei Gong, Jaywon Koo, Julia Hirschberg
cs.CL, cs.AI, cs.LG, cs.MM
In this work we investigate the optimal selection and fusion of features across multiple modalities and combine these in a neural network to improve emotion detection. We compare different fusion methods and examine the impact of multi-loss training within the multi-modality fusion network, identifying useful findings ...
http://arxiv.org/pdf/2308.00264v2
2023-09-12T02:40:08
2023-08-01T03:54:27
{ "title": [ -0.016143799, 0.008132935, -0.051361084, -0.014228821, 0.008590698, -0.030426025, -0.030670166, -0.027801514, -0.0069084167, 0.039764404, 0.0052948, -0.0025615692, 0.020568848, -0.05682373, 0.0026378632, 0.019958496, -0.0032520294, 0...
http://arxiv.org/abs/2308.00240v1
Towards Effective Ancient Chinese Translation: Dataset, Model, and Evaluation
Geyang Guo, Jiarong Yang, Fengyuan Lu, Jiaxin Qin, Tianyi Tang, Wayne Xin Zhao
cs.CL
Interpreting ancient Chinese has been the key to comprehending vast Chinese literature, tradition, and civilization. In this paper, we propose Erya for ancient Chinese translation. From a dataset perspective, we collect, clean, and classify ancient Chinese materials from various sources, forming the most extensive anci...
http://arxiv.org/pdf/2308.00240v1
2023-08-01T02:43:27
2023-08-01T02:43:27
{ "title": [ -0.04260254, 0.016174316, -0.02368164, -0.010505676, 0.011802673, -0.0059127808, -0.013969421, -0.022003174, 0.016525269, 0.039642334, -0.0050849915, -0.0021762848, 0.03503418, -0.030441284, -0.062805176, 0.026290894, 0.00055503845, ...
http://arxiv.org/abs/2308.01423v2
ChatMOF: An Autonomous AI System for Predicting and Generating Metal-Organic Frameworks
Yeonghun Kang, Jihan Kim
cs.CL, cs.AI, cs.LG, physics.chem-ph
ChatMOF is an autonomous Artificial Intelligence (AI) system that is built to predict and generate metal-organic frameworks (MOFs). By leveraging a large-scale language model (GPT-4 and GPT-3.5-turbo), ChatMOF extracts key details from textual inputs and delivers appropriate responses, thus eliminating the necessity fo...
http://arxiv.org/pdf/2308.01423v2
2023-08-25T15:13:46
2023-08-01T02:08:13
{ "title": [ -0.043670654, -0.022003174, -0.041015625, -0.051940918, -0.017623901, -0.0025024414, -0.05883789, 0.014724731, -0.018035889, 0.025802612, 0.00466156, 0.031463623, -0.016052246, -0.028503418, -0.015701294, 0.020370483, -0.024383545, 0...
http://arxiv.org/abs/2308.00221v2
Advancing Beyond Identification: Multi-bit Watermark for Large Language Models
KiYoon Yoo, Wonhyuk Ahn, Nojun Kwak
cs.CL, cs.AI, cs.CR
We propose a method to tackle misuses of large language models beyond the identification of machine-generated text. While existing methods focus on detection, some malicious misuses demand tracing the adversary user for counteracting them. To address this, we propose Multi-bit Watermark via Position Allocation, embeddi...
http://arxiv.org/pdf/2308.00221v2
2023-09-27T04:08:04
2023-08-01T01:27:40
{ "title": [ -0.0231781, 0.004295349, -0.0413208, -0.0657959, 0.014022827, -0.0010681152, -0.05645752, -0.0032787323, -0.001616478, 0.03857422, -0.005382538, -0.0036258698, 0.011856079, -0.023452759, -0.040161133, 0.015686035, -0.009925842, 0.031...
http://arxiv.org/abs/2308.00189v1
Generative Models as a Complex Systems Science: How can we make sense of large language model behavior?
Ari Holtzman, Peter West, Luke Zettlemoyer
cs.LG, cs.AI, cs.CL
Coaxing out desired behavior from pretrained models, while avoiding undesirable ones, has redefined NLP and is reshaping how we interact with computers. What was once a scientific engineering discipline-in which building blocks are stacked one on top of the other-is arguably already a complex systems science, in which ...
http://arxiv.org/pdf/2308.00189v1
2023-07-31T22:58:41
2023-07-31T22:58:41
{ "title": [ -0.025680542, -0.0046081543, -0.025482178, -0.07678223, -0.013145447, 0.004802704, -0.06347656, 0.0017499924, 0.001045227, 0.0158844, -0.017303467, 0.017486572, -0.0037822723, -0.032226562, -0.027908325, 0.0037002563, -0.008285522, 0...
http://arxiv.org/abs/2308.00165v1
Adversarially Robust Neural Legal Judgement Systems
Rohit Raj, V Susheela Devi
cs.CL, cs.AI
Legal judgment prediction is the task of predicting the outcome of court cases on a given text description of facts of cases. These tasks apply Natural Language Processing (NLP) techniques to predict legal judgment results based on facts. Recently, large-scale public datasets and NLP models have increased research in a...
http://arxiv.org/pdf/2308.00165v1
2023-07-31T21:44:48
2023-07-31T21:44:48
{ "title": [ -0.016937256, 0.0046195984, -0.026519775, -0.032165527, 0.006298065, -0.0012235641, -0.016021729, -0.020736694, 0.008956909, 0.030303955, -0.039154053, 0.0026168823, -0.009132385, -0.004890442, -0.05142212, -0.0046539307, -0.020706177, ...
http://arxiv.org/abs/2308.00158v5
Predictive Data Analytics with AI: assessing the need for post-editing of MT output by fine-tuning OpenAI LLMs
Serge Gladkoff, Gleb Erofeev, Irina Sorokina, Lifeng Han, Goran Nenadic
cs.CL, cs.AI
Translation Quality Evaluation (TQE) is an essential step of the modern translation production process. TQE is critical in assessing both machine translation (MT) and human translation (HT) quality without reference translations. The ability to evaluate or even simply estimate the quality of translation automatically m...
http://arxiv.org/pdf/2308.00158v5
2023-11-08T17:25:34
2023-07-31T21:13:30
{ "title": [ -0.024551392, 0.013908386, -0.0028305054, -0.004470825, -0.030349731, -0.0035743713, -0.00390625, -0.021697998, -0.018112183, 0.018569946, -0.040130615, 0.0034885406, -0.010604858, -0.039489746, 0.008659363, -0.0463562, 0.008758545, ...
http://arxiv.org/abs/2308.00157v1
Boosting Adverse Drug Event Normalization on Social Media: General-Purpose Model Initialization and Biomedical Semantic Text Similarity Benefit Zero-Shot Linking in Informal Contexts
François Remy, Simone Scaboro, Beatrice Portelli
cs.CL
Biomedical entity linking, also known as biomedical concept normalization, has recently witnessed the rise to prominence of zero-shot contrastive models. However, the pre-training material used for these models has, until now, largely consisted of specialist biomedical content such as MIMIC-III clinical notes (Johnson ...
http://arxiv.org/pdf/2308.00157v1
2023-07-31T21:12:06
2023-07-31T21:12:06
{ "title": [ 0.040008545, 0.0033168793, -0.017105103, -0.081970215, -0.01574707, -0.008178711, -0.061523438, 0.010887146, -0.044799805, 0.04257202, -0.006652832, 0.067993164, 0.03189087, -0.025543213, 0.037109375, 0.017623901, 0.0037612915, 0.023...
http://arxiv.org/abs/2308.00116v1
A Modular Ontology for MODS -- Metadata Object Description Schema
Rushrukh Rayan, Cogan Shimizu, Heidi Sieverding, Pascal Hitzler
cs.CL, cs.AI
The Metadata Object Description Schema (MODS) was developed to describe bibliographic concepts and metadata and is maintained by the Library of Congress. Its authoritative version is given as an XML schema based on an XML mindset which means that it has significant limitations for use in a knowledge graphs context. We ...
http://arxiv.org/pdf/2308.00116v1
2023-07-31T19:36:07
2023-07-31T19:36:07
{ "title": [ -0.011314392, -0.0015697479, -0.08013916, -0.11175537, 0.03173828, -0.0066947937, -0.049743652, 0.019577026, -0.018798828, 0.044403076, 0.008117676, -0.01033783, 0.03024292, -0.031219482, -0.02468872, 0.02218628, -0.009384155, 0.0362...
http://arxiv.org/abs/2308.00081v2
Towards Semantically Enriched Embeddings for Knowledge Graph Completion
Mehwish Alam, Frank van Harmelen, Maribel Acosta
cs.AI, cs.CL
Embedding based Knowledge Graph (KG) Completion has gained much attention over the past few years. Most of the current algorithms consider a KG as a multidirectional labeled graph and lack the ability to capture the semantics underlying the schematic information. In a separate development, a vast amount of information ...
http://arxiv.org/pdf/2308.00081v2
2023-08-02T07:34:24
2023-07-31T18:53:47
{ "title": [ -0.007965088, -0.012718201, -0.06573486, -0.103149414, 0.012649536, 0.008560181, -0.062316895, 0.030532837, 0.015571594, 0.029205322, -0.03717041, 0.025344849, 0.004333496, 0.007160187, -0.037261963, 0.0051727295, -0.011161804, 0.048...
http://arxiv.org/abs/2308.01399v1
Learning to Model the World with Language
Jessy Lin, Yuqing Du, Olivia Watkins, Danijar Hafner, Pieter Abbeel, Dan Klein, Anca Dragan
cs.CL, cs.AI, cs.LG
To interact with humans in the world, agents need to understand the diverse types of language that people use, relate them to the visual world, and act based on them. While current agents learn to execute simple language instructions from task rewards, we aim to build agents that leverage diverse language that conveys ...
http://arxiv.org/pdf/2308.01399v1
2023-07-31T17:57:49
2023-07-31T17:57:49
{ "title": [ -0.022857666, -0.024841309, -0.014595032, -0.037078857, -0.06829834, 0.001159668, -0.057281494, 0.011154175, -0.02053833, 0.0031337738, -0.03552246, -0.018661499, -0.018630981, -0.040527344, 0.006275177, -0.031951904, 0.025482178, 0....
http://arxiv.org/abs/2307.16888v2
Backdooring Instruction-Tuned Large Language Models with Virtual Prompt Injection
Jun Yan, Vikas Yadav, Shiyang Li, Lichang Chen, Zheng Tang, Hai Wang, Vijay Srinivasan, Xiang Ren, Hongxia Jin
cs.CL, cs.CR, cs.LG
Instruction-tuned Large Language Models (LLMs) have demonstrated remarkable abilities to modulate their responses based on human instructions. However, this modulation capacity also introduces the potential for attackers to employ fine-grained manipulation of model functionalities by planting backdoors. In this paper, ...
http://arxiv.org/pdf/2307.16888v2
2023-10-06T11:12:18
2023-07-31T17:56:00
{ "title": [ -0.034851074, -0.015075684, -0.025604248, -0.05078125, -0.024475098, 0.00036168098, -0.05392456, -0.0071907043, -0.042175293, 0.017974854, -0.012130737, 0.011100769, -0.022216797, -0.025680542, -0.013977051, -0.009437561, -0.0069503784, ...
http://arxiv.org/abs/2307.16883v1
HAGRID: A Human-LLM Collaborative Dataset for Generative Information-Seeking with Attribution
Ehsan Kamalloo, Aref Jafari, Xinyu Zhang, Nandan Thakur, Jimmy Lin
cs.CL, cs.IR
The rise of large language models (LLMs) had a transformative impact on search, ushering in a new era of search engines that are capable of generating search results in natural language text, imbued with citations for supporting sources. Building generative information-seeking models demands openly accessible datasets,...
http://arxiv.org/pdf/2307.16883v1
2023-07-31T17:49:18
2023-07-31T17:49:18
{ "title": [ -0.016967773, -0.01423645, -0.027755737, -0.021942139, -0.034057617, -0.009300232, -0.03164673, -0.012710571, -0.0118255615, 0.027801514, -0.017654419, 0.036956787, -0.032165527, -0.012023926, -0.012290955, -0.005329132, -0.02168274, ...
http://arxiv.org/abs/2307.16878v2
Contrastive Learning for API Aspect Analysis
G. M. Shahariar, Tahmid Hasan, Anindya Iqbal, Gias Uddin
cs.SE, cs.AI, cs.CL
We present a novel approach - CLAA - for API aspect detection in API reviews that utilizes transformer models trained with a supervised contrastive loss objective function. We evaluate CLAA using performance and impact analysis. For performance analysis, we utilized a benchmark dataset on developer discussions collecte...
http://arxiv.org/pdf/2307.16878v2
2023-08-14T16:40:31
2023-07-31T17:41:10
{ "title": [ 0.002073288, -0.01687622, -0.034088135, -0.07287598, 0.0064048767, -0.009262085, -0.031707764, 0.011001587, 0.028823853, 0.038970947, -0.02911377, 0.00856781, -0.012588501, -0.007713318, -0.060791016, -0.004825592, -0.01638794, 0.042...
http://arxiv.org/abs/2307.16877v1
Evaluating Correctness and Faithfulness of Instruction-Following Models for Question Answering
Vaibhav Adlakha, Parishad BehnamGhader, Xing Han Lu, Nicholas Meade, Siva Reddy
cs.CL, cs.AI
Retriever-augmented instruction-following models are attractive alternatives to fine-tuned approaches for information-seeking tasks such as question answering (QA). By simply prepending retrieved documents in its input along with an instruction, these models can be adapted to various information domains and tasks witho...
http://arxiv.org/pdf/2307.16877v1
2023-07-31T17:41:00
2023-07-31T17:41:00
{ "title": [ -0.007537842, 0.0054626465, -0.018371582, -0.056274414, 0.008430481, 0.0004682541, -0.034210205, -0.02017212, 0.012237549, 0.03253174, -0.018554688, 0.0019664764, -0.0043907166, 0.002866745, -0.0690918, -0.0022792816, -0.008583069, 0...
http://arxiv.org/abs/2308.00016v1
Alpha-GPT: Human-AI Interactive Alpha Mining for Quantitative Investment
Saizhuo Wang, Hang Yuan, Leon Zhou, Lionel M. Ni, Heung-Yeung Shum, Jian Guo
q-fin.CP, cs.AI, cs.CL
One of the most important tasks in quantitative investment research is mining new alphas (effective trading signals or factors). Traditional alpha mining methods, either hand-crafted factor synthesizing or algorithmic factor mining (e.g., search with genetic programming), have inherent limitations, especially in implem...
http://arxiv.org/pdf/2308.00016v1
2023-07-31T16:40:06
2023-07-31T16:40:06
{ "title": [ -0.010040283, -0.0020999908, -0.027450562, -0.0024261475, -0.0010738373, -0.022109985, 0.0084991455, -0.038024902, -0.017623901, 0.03466797, -0.025024414, 0.021774292, -0.039520264, -0.004627228, -0.06707764, -0.019195557, -0.012519836, ...
http://arxiv.org/abs/2307.16816v1
Defense of Adversarial Ranking Attack in Text Retrieval: Benchmark and Baseline via Detection
Xuanang Chen, Ben He, Le Sun, Yingfei Sun
cs.IR, cs.CL
Neural ranking models (NRMs) have undergone significant development and have become integral components of information retrieval (IR) systems. Unfortunately, recent research has unveiled the vulnerability of NRMs to adversarial document manipulations, potentially exploited by malicious search engine optimization practi...
http://arxiv.org/pdf/2307.16816v1
2023-07-31T16:31:24
2023-07-31T16:31:24
{ "title": [ 0.019546509, 0.014587402, -0.01473999, -0.027023315, 0.035064697, -0.013504028, 0.034606934, -0.030136108, -0.025009155, 0.038970947, 0.0062828064, 0.034118652, -0.022994995, -0.0013561249, -0.055541992, -0.011627197, -0.019439697, 0...
http://arxiv.org/abs/2307.16811v2
DoDo Learning: DOmain-DemOgraphic Transfer in Language Models for Detecting Abuse Targeted at Public Figures
Hannah Rose Kirk, Angus R. Williams, Liam Burke, Yi-Ling Chung, Ivan Debono, Pica Johansson, Francesca Stevens, Jonathan Bright, Scott A. Hale
cs.CL, cs.CY
Public figures receive a disproportionate amount of abuse on social media, impacting their active participation in public life. Automated systems can identify abuse at scale but labelling training data is expensive, complex and potentially harmful. So, it is desirable that systems are efficient and generalisable, handl...
http://arxiv.org/pdf/2307.16811v2
2023-08-21T10:20:02
2023-07-31T16:29:08
{ "title": [ 0.013381958, 0.009925842, -0.04135132, -0.060791016, 0.015541077, -0.00982666, -0.025283813, -0.000500679, -0.0060806274, 0.021209717, -0.024597168, 0.02456665, 0.028182983, -0.021240234, -0.010276794, -0.0042915344, -0.027938843, 0....
http://arxiv.org/abs/2308.04444v1
Changes in Policy Preferences in German Tweets during the COVID Pandemic
Felix Biessmann
cs.CY, cs.CL, cs.IR, cs.LG, cs.SI
Online social media have become an important forum for exchanging political opinions. In response to COVID measures citizens expressed their policy preferences directly on these platforms. Quantifying political preferences in online social media remains challenging: The vast amount of content requires scalable automate...
http://arxiv.org/pdf/2308.04444v1
2023-07-31T16:07:28
2023-07-31T16:07:28
{ "title": [ 0.0054473877, -0.007865906, -0.049804688, -0.028701782, -0.01953125, -0.023406982, 0.004425049, 0.01436615, -0.015853882, 0.03265381, -0.029953003, 0.030090332, 0.015617371, -0.029541016, 0.0042381287, 0.0023880005, 0.0003299713, 0.0...
http://arxiv.org/abs/2307.16795v1
Structural Transfer Learning in NL-to-Bash Semantic Parsers
Kyle Duffy, Satwik Bhattamishra, Phil Blunsom
cs.CL, cs.AI, cs.LG
Large-scale pre-training has made progress in many fields of natural language processing, though little is understood about the design of pre-training datasets. We propose a methodology for obtaining a quantitative understanding of structural overlap between machine translation tasks. We apply our methodology to the na...
http://arxiv.org/pdf/2307.16795v1
2023-07-31T16:02:09
2023-07-31T16:02:09
{ "title": [ -0.0137786865, -0.0028209686, -0.024734497, -0.06951904, -0.0011262894, 0.018981934, -0.06072998, 0.00724411, -0.016662598, 0.029724121, -0.01576233, 0.013114929, 0.00699234, -0.014755249, -0.026168823, -0.0016260147, -0.0017642975, ...
http://arxiv.org/abs/2307.16789v2
ToolLLM: Facilitating Large Language Models to Master 16000+ Real-world APIs
Yujia Qin, Shihao Liang, Yining Ye, Kunlun Zhu, Lan Yan, Yaxi Lu, Yankai Lin, Xin Cong, Xiangru Tang, Bill Qian, Sihan Zhao, Lauren Hong, Runchu Tian, Ruobing Xie, Jie Zhou, Mark Gerstein, Dahai Li, Zhiyuan Liu, Maosong Sun
cs.AI, cs.CL, cs.LG
Despite the advancements of open-source large language models (LLMs), e.g., LLaMA, they remain significantly limited in tool-use capabilities, i.e., using external tools (APIs) to fulfill human instructions. The reason is that current instruction tuning largely focuses on basic language tasks but ignores the tool-use d...
http://arxiv.org/pdf/2307.16789v2
2023-10-03T14:45:48
2023-07-31T15:56:53
{ "title": [ -0.028762817, -0.036743164, -0.05834961, -0.068115234, -0.01473999, -0.005004883, -0.07373047, 0.021362305, -0.007637024, 0.024429321, -0.0082092285, -0.004459381, 0.014434814, -0.024719238, -0.0178833, 0.0132369995, -0.0031032562, 0...
http://arxiv.org/abs/2307.16779v1
Lexically-Accelerated Dense Retrieval
Hrishikesh Kulkarni, Sean MacAvaney, Nazli Goharian, Ophir Frieder
cs.IR, cs.CL
Retrieval approaches that score documents based on learned dense vectors (i.e., dense retrieval) rather than lexical signals (i.e., conventional retrieval) are increasingly popular. Their ability to identify related documents that do not necessarily contain the same terms as those appearing in the user's query (thereby...
http://arxiv.org/pdf/2307.16779v1
2023-07-31T15:44:26
2023-07-31T15:44:26
{ "title": [ -0.0084991455, -0.006996155, -0.0357666, -0.07098389, 0.020309448, 0.0062942505, -0.027923584, -0.0042037964, 0.042999268, 0.045532227, -0.03567505, 0.013442993, 0.023666382, 0.0014381409, -0.039398193, 0.039855957, 0.0051612854, 0.0...
http://arxiv.org/abs/2307.16778v1
KoBBQ: Korean Bias Benchmark for Question Answering
Jiho Jin, Jiseon Kim, Nayeon Lee, Haneul Yoo, Alice Oh, Hwaran Lee
cs.CL, cs.AI
The BBQ (Bias Benchmark for Question Answering) dataset enables the evaluation of the social biases that language models (LMs) exhibit in downstream tasks. However, it is challenging to adapt BBQ to languages other than English as social biases are culturally dependent. In this paper, we devise a process to construct a...
http://arxiv.org/pdf/2307.16778v1
2023-07-31T15:44:15
2023-07-31T15:44:15
{ "title": [ -0.02746582, -0.01739502, -0.02633667, -0.047027588, -0.014945984, 0.0034217834, -0.04537964, -0.013977051, 0.019515991, 0.039031982, -0.029190063, 0.011199951, 0.021697998, 0.0030593872, -0.050720215, 0.009643555, -0.013221741, 0.03...
http://arxiv.org/abs/2307.16773v2
AsdKB: A Chinese Knowledge Base for the Early Screening and Diagnosis of Autism Spectrum Disorder
Tianxing Wu, Xudong Cao, Yipeng Zhu, Feiyue Wu, Tianling Gong, Yuxiang Wang, Shenqi Jing
cs.AI, cs.CL, cs.IR
To easily obtain the knowledge about autism spectrum disorder and help its early screening and diagnosis, we create AsdKB, a Chinese knowledge base on autism spectrum disorder. The knowledge base is built on top of various sources, including 1) the disease knowledge from SNOMED CT and ICD-10 clinical descriptions on me...
http://arxiv.org/pdf/2307.16773v2
2023-08-02T08:04:29
2023-07-31T15:40:45
{ "title": [ -0.023284912, -0.01210022, -0.0067977905, -0.04244995, -0.005718231, 0.017227173, -0.03237915, -0.024032593, 0.008338928, 0.021392822, -0.018676758, 0.0023422241, 0.006313324, -0.03994751, -0.023895264, 0.0099487305, 0.023361206, 0.0...