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/2309.11564v1
Hierarchical reinforcement learning with natural language subgoals
Arun Ahuja, Kavya Kopparapu, Rob Fergus, Ishita Dasgupta
cs.LG, cs.CL
Hierarchical reinforcement learning has been a compelling approach for achieving goal directed behavior over long sequences of actions. However, it has been challenging to implement in realistic or open-ended environments. A main challenge has been to find the right space of sub-goals over which to instantiate a hierar...
http://arxiv.org/pdf/2309.11564v1
2023-09-20T18:03:04
2023-09-20T18:03:04
{ "title": [ -0.016708374, 0.0067596436, -0.030685425, -0.059661865, 0.021377563, -0.0069999695, -0.016204834, -0.013771057, -0.028671265, 0.021774292, -0.034423828, 0.022521973, -0.03704834, -0.0055656433, -0.02293396, -0.02822876, -0.019943237, ...
http://arxiv.org/abs/2309.11499v1
DreamLLM: Synergistic Multimodal Comprehension and Creation
Runpei Dong, Chunrui Han, Yuang Peng, Zekun Qi, Zheng Ge, Jinrong Yang, Liang Zhao, Jianjian Sun, Hongyu Zhou, Haoran Wei, Xiangwen Kong, Xiangyu Zhang, Kaisheng Ma, Li Yi
cs.CV, cs.CL, cs.LG
This paper presents DreamLLM, a learning framework that first achieves versatile Multimodal Large Language Models (MLLMs) empowered with frequently overlooked synergy between multimodal comprehension and creation. DreamLLM operates on two fundamental principles. The first focuses on the generative modeling of both lang...
http://arxiv.org/pdf/2309.11499v1
2023-09-20T17:58:05
2023-09-20T17:58:05
{ "title": [ -0.042877197, -0.0010032654, -0.038208008, -0.054473877, -0.014144897, 0.016921997, -0.0748291, -0.024658203, 0.001241684, 0.03253174, -0.015220642, 0.003578186, 0.0051651, -0.018676758, -0.03060913, 0.0043678284, -0.015098572, 0.034...
http://arxiv.org/abs/2309.11495v2
Chain-of-Verification Reduces Hallucination in Large Language Models
Shehzaad Dhuliawala, Mojtaba Komeili, Jing Xu, Roberta Raileanu, Xian Li, Asli Celikyilmaz, Jason Weston
cs.CL, cs.AI
Generation of plausible yet incorrect factual information, termed hallucination, is an unsolved issue in large language models. We study the ability of language models to deliberate on the responses they give in order to correct their mistakes. We develop the Chain-of-Verification (CoVe) method whereby the model first ...
http://arxiv.org/pdf/2309.11495v2
2023-09-25T15:25:49
2023-09-20T17:50:55
{ "title": [ -0.002948761, -0.0015268326, -0.041992188, -0.08392334, -0.0023880005, -0.0099487305, -0.05355835, 0.0061187744, -0.026672363, 0.038482666, -0.019302368, 0.019439697, 0.022888184, -0.010215759, -0.029373169, -0.0016336441, -0.021957397, ...
http://arxiv.org/abs/2309.11489v2
Text2Reward: Automated Dense Reward Function Generation for Reinforcement Learning
Tianbao Xie, Siheng Zhao, Chen Henry Wu, Yitao Liu, Qian Luo, Victor Zhong, Yanchao Yang, Tao Yu
cs.LG, cs.AI, cs.CL, cs.RO
Designing reward functions is a longstanding challenge in reinforcement learning (RL); it requires specialized knowledge or domain data, leading to high costs for development. To address this, we introduce Text2Reward, a data-free framework that automates the generation of dense reward functions based on large language...
http://arxiv.org/pdf/2309.11489v2
2023-09-21T15:17:09
2023-09-20T17:39:13
{ "title": [ -0.0020198822, 0.00182724, -0.012435913, -0.01939392, 0.011688232, -0.0075263977, 0.009529114, -0.040130615, -0.015792847, 0.017852783, -0.020462036, 0.03186035, -0.041381836, -0.011734009, -0.031173706, -0.015197754, -0.012519836, 0...
http://arxiv.org/abs/2309.11439v1
Controlled Generation with Prompt Insertion for Natural Language Explanations in Grammatical Error Correction
Masahiro Kaneko, Naoaki Okazaki
cs.CL
In Grammatical Error Correction (GEC), it is crucial to ensure the user's comprehension of a reason for correction. Existing studies present tokens, examples, and hints as to the basis for correction but do not directly explain the reasons for corrections. Although methods that use Large Language Models (LLMs) to provi...
http://arxiv.org/pdf/2309.11439v1
2023-09-20T16:14:10
2023-09-20T16:14:10
{ "title": [ -0.0009784698, -0.012397766, -0.012786865, -0.06542969, -0.01914978, 0.012763977, -0.043670654, -0.0016317368, -0.011138916, 0.010108948, -0.019638062, 0.009216309, 0.0068969727, -0.028259277, -0.014831543, -0.005470276, -0.039794922, ...
http://arxiv.org/abs/2309.11436v2
You Only Look at Screens: Multimodal Chain-of-Action Agents
Zhuosheng Zhang, Aston Zhang
cs.CL, cs.AI, cs.HC
Autonomous user interface (UI) agents aim to facilitate task automation by interacting with the user interface without manual intervention. Recent studies have investigated eliciting the capabilities of large language models (LLMs) for effective engagement in diverse environments. To align with the input-output require...
http://arxiv.org/pdf/2309.11436v2
2023-09-21T03:00:07
2023-09-20T16:12:32
{ "title": [ 0.0071411133, -0.025756836, -0.005012512, -0.00095558167, -0.03829956, -0.010810852, 0.024154663, 0.006374359, 0.0021705627, 0.0018587112, -0.056610107, 0.01687622, -0.0058631897, -0.030151367, 0.006187439, -0.04373169, -0.005115509, ...
http://arxiv.org/abs/2309.13075v1
SCREWS: A Modular Framework for Reasoning with Revisions
Kumar Shridhar, Harsh Jhamtani, Hao Fang, Benjamin Van Durme, Jason Eisner, Patrick Xia
cs.AI, cs.CL, cs.LG
Large language models (LLMs) can improve their accuracy on various tasks through iteratively refining and revising their output based on feedback. We observe that these revisions can introduce errors, in which case it is better to roll back to a previous result. Further, revisions are typically homogeneous: they use th...
http://arxiv.org/pdf/2309.13075v1
2023-09-20T15:59:54
2023-09-20T15:59:54
{ "title": [ -0.046142578, -0.0234375, -0.0657959, -0.11047363, 0.0132369995, 0.02960205, -0.06854248, 0.035827637, -0.0059814453, 0.011940002, -0.028167725, -0.0059051514, 0.0009703636, -0.014862061, -0.02357483, -0.006378174, -0.00028562546, 0....
http://arxiv.org/abs/2309.11419v1
Kosmos-2.5: A Multimodal Literate Model
Tengchao Lv, Yupan Huang, Jingye Chen, Lei Cui, Shuming Ma, Yaoyao Chang, Shaohan Huang, Wenhui Wang, Li Dong, Weiyao Luo, Shaoxiang Wu, Guoxin Wang, Cha Zhang, Furu Wei
cs.CL, cs.CV
We present Kosmos-2.5, a multimodal literate model for machine reading of text-intensive images. Pre-trained on large-scale text-intensive images, Kosmos-2.5 excels in two distinct yet cooperative transcription tasks: (1) generating spatially-aware text blocks, where each block of text is assigned its spatial coordinat...
http://arxiv.org/pdf/2309.11419v1
2023-09-20T15:50:08
2023-09-20T15:50:08
{ "title": [ -0.03555298, 0.003692627, -0.027648926, -0.0073051453, -0.036743164, -0.005306244, -0.051818848, -0.039276123, 0.0104599, 0.035003662, -0.0041236877, -0.020706177, 0.014213562, -0.032196045, -0.033203125, 0.010261536, -0.019348145, 0...
http://arxiv.org/abs/2309.11385v1
Safurai 001: New Qualitative Approach for Code LLM Evaluation
Davide Cifarelli, Leonardo Boiardi, Alessandro Puppo
cs.CL
This paper presents Safurai-001, a new Large Language Model (LLM) with significant potential in the domain of coding assistance. Driven by recent advancements in coding LLMs, Safurai-001 competes in performance with the latest models like WizardCoder [Xu et al., 2023], PanguCoder [Shen et al., 2023] and Phi-1 [Gunaseka...
http://arxiv.org/pdf/2309.11385v1
2023-09-20T15:11:32
2023-09-20T15:11:32
{ "title": [ -0.07348633, -0.015129089, -0.053833008, -0.033813477, -0.021438599, -0.0010080338, -0.060058594, 0.0022716522, 0.010375977, 0.009994507, -0.024795532, 0.012237549, -0.025604248, -0.03491211, -0.0149383545, -0.00082588196, -0.038208008, ...
http://arxiv.org/abs/2309.11384v1
Long-Form End-to-End Speech Translation via Latent Alignment Segmentation
Peter Polák, Ondřej Bojar
cs.CL, cs.AI, cs.SD, eess.AS
Current simultaneous speech translation models can process audio only up to a few seconds long. Contemporary datasets provide an oracle segmentation into sentences based on human-annotated transcripts and translations. However, the segmentation into sentences is not available in the real world. Current speech segmentat...
http://arxiv.org/pdf/2309.11384v1
2023-09-20T15:10:12
2023-09-20T15:10:12
{ "title": [ -0.030578613, 0.003791809, -0.031921387, -0.05886841, 0.0028514862, -0.0005502701, -0.03945923, 0.014862061, -0.023834229, 0.051208496, 0.015304565, 0.015960693, 0.017654419, -0.03540039, -0.025009155, 0.021255493, 0.008552551, 0.004...
http://arxiv.org/abs/2309.11382v1
Discuss Before Moving: Visual Language Navigation via Multi-expert Discussions
Yuxing Long, Xiaoqi Li, Wenzhe Cai, Hao Dong
cs.RO, cs.AI, cs.CL, cs.CV
Visual language navigation (VLN) is an embodied task demanding a wide range of skills encompassing understanding, perception, and planning. For such a multifaceted challenge, previous VLN methods totally rely on one model's own thinking to make predictions within one round. However, existing models, even the most advan...
http://arxiv.org/pdf/2309.11382v1
2023-09-20T15:04:49
2023-09-20T15:04:49
{ "title": [ -0.0044441223, -0.0044517517, -0.017486572, -0.04953003, -0.0006213188, -0.005584717, -0.0309906, -0.026382446, -0.01612854, 0.01991272, -0.008529663, 0.010643005, -0.018325806, -0.017196655, -0.023086548, -0.018249512, 0.008613586, ...
http://arxiv.org/abs/2309.11381v1
Studying Lobby Influence in the European Parliament
Aswin Suresh, Lazar Radojevic, Francesco Salvi, Antoine Magron, Victor Kristof, Matthias Grossglauser
cs.CL, cs.CE, cs.CY, cs.SI
We present a method based on natural language processing (NLP), for studying the influence of interest groups (lobbies) in the law-making process in the European Parliament (EP). We collect and analyze novel datasets of lobbies' position papers and speeches made by members of the EP (MEPs). By comparing these texts on ...
http://arxiv.org/pdf/2309.11381v1
2023-09-20T15:03:30
2023-09-20T15:03:30
{ "title": [ 0.040130615, 0.012786865, -0.04019165, -0.0068130493, -0.0050811768, -0.043640137, 0.045013428, -0.012588501, -0.056396484, 0.046173096, -0.055358887, 0.036895752, 0.020736694, -0.011985779, 0.013008118, -0.028671265, 0.008056641, 0....
http://arxiv.org/abs/2309.11379v1
Incremental Blockwise Beam Search for Simultaneous Speech Translation with Controllable Quality-Latency Tradeoff
Peter Polák, Brian Yan, Shinji Watanabe, Alex Waibel, Ondřej Bojar
cs.CL, cs.AI, cs.SD, eess.AS
Blockwise self-attentional encoder models have recently emerged as one promising end-to-end approach to simultaneous speech translation. These models employ a blockwise beam search with hypothesis reliability scoring to determine when to wait for more input speech before translating further. However, this method mainta...
http://arxiv.org/pdf/2309.11379v1
2023-09-20T14:59:06
2023-09-20T14:59:06
{ "title": [ -0.021621704, -0.017974854, -0.009376526, -0.059753418, -0.015716553, 0.010375977, -0.05517578, -0.0010738373, -0.022705078, 0.079956055, -0.008613586, 0.0077667236, 0.030410767, -0.011123657, -0.052734375, -0.0047721863, -0.012260437, ...
http://arxiv.org/abs/2309.11346v1
GECTurk: Grammatical Error Correction and Detection Dataset for Turkish
Atakan Kara, Farrin Marouf Sofian, Andrew Bond, Gözde Gül Şahin
cs.CL, cs.LG
Grammatical Error Detection and Correction (GEC) tools have proven useful for native speakers and second language learners. Developing such tools requires a large amount of parallel, annotated data, which is unavailable for most languages. Synthetic data generation is a common practice to overcome the scarcity of such ...
http://arxiv.org/pdf/2309.11346v1
2023-09-20T14:25:44
2023-09-20T14:25:44
{ "title": [ 0.0047416687, -0.0015897751, -0.03591919, -0.059295654, 0.020080566, -0.0051116943, -0.035583496, 0.00070667267, 0.012275696, 0.020370483, -0.033599854, 0.005973816, 0.01473999, -0.015220642, -0.038635254, 0.0005774498, -0.033111572, ...
http://arxiv.org/abs/2309.11341v1
Improving Article Classification with Edge-Heterogeneous Graph Neural Networks
Khang Ly, Yury Kashnitsky, Savvas Chamezopoulos, Valeria Krzhizhanovskaya
cs.LG, cs.CL
Classifying research output into context-specific label taxonomies is a challenging and relevant downstream task, given the volume of existing and newly published articles. We propose a method to enhance the performance of article classification by enriching simple Graph Neural Networks (GNN) pipelines with edge-hetero...
http://arxiv.org/pdf/2309.11341v1
2023-09-20T14:18:04
2023-09-20T14:18:04
{ "title": [ -0.030380249, 0.02508545, -0.052612305, -0.0079956055, 0.0093307495, -0.014259338, -0.030776978, -0.029556274, 0.008140564, 0.02772522, -0.02848816, 0.02607727, -0.030044556, 0.00774765, -0.034484863, 0.0020599365, -0.0048103333, 0.0...
http://arxiv.org/abs/2309.11338v1
TRAVID: An End-to-End Video Translation Framework
Prottay Kumar Adhikary, Bandaru Sugandhi, Subhojit Ghimire, Santanu Pal, Partha Pakray
cs.CL, cs.AI
In today's globalized world, effective communication with people from diverse linguistic backgrounds has become increasingly crucial. While traditional methods of language translation, such as written text or voice-only translations, can accomplish the task, they often fail to capture the complete context and nuanced i...
http://arxiv.org/pdf/2309.11338v1
2023-09-20T14:13:05
2023-09-20T14:13:05
{ "title": [ -0.03366089, -0.010185242, -0.030319214, -0.062072754, 0.00121212, 0.0025978088, -0.03829956, -0.008651733, -0.008422852, 0.05441284, -0.0010766983, -0.0020809174, 0.023529053, -0.018798828, -0.04522705, 0.011512756, 0.007472992, 0.0...
http://arxiv.org/abs/2309.11327v2
Leveraging Data Collection and Unsupervised Learning for Code-switched Tunisian Arabic Automatic Speech Recognition
Ahmed Amine Ben Abdallah, Ata Kabboudi, Amir Kanoun, Salah Zaiem
eess.AS, cs.CL, cs.LG, cs.SD
Crafting an effective Automatic Speech Recognition (ASR) solution for dialects demands innovative approaches that not only address the data scarcity issue but also navigate the intricacies of linguistic diversity. In this paper, we address the aforementioned ASR challenge, focusing on the Tunisian dialect. First, textu...
http://arxiv.org/pdf/2309.11327v2
2023-09-25T11:20:36
2023-09-20T13:56:27
{ "title": [ -0.024017334, -0.00932312, -0.031921387, -0.031097412, -0.0138168335, -0.0032596588, -0.017059326, 0.013763428, -0.014762878, 0.018875122, -0.03552246, -0.0073547363, -0.018966675, -0.035949707, -0.023223877, -0.027130127, 0.0045547485, ...
http://arxiv.org/abs/2309.11325v2
DISC-LawLLM: Fine-tuning Large Language Models for Intelligent Legal Services
Shengbin Yue, Wei Chen, Siyuan Wang, Bingxuan Li, Chenchen Shen, Shujun Liu, Yuxuan Zhou, Yao Xiao, Song Yun, Xuanjing Huang, Zhongyu Wei
cs.CL
We propose DISC-LawLLM, an intelligent legal system utilizing large language models (LLMs) to provide a wide range of legal services. We adopt legal syllogism prompting strategies to construct supervised fine-tuning datasets in the Chinese Judicial domain and fine-tune LLMs with legal reasoning capability. We augment L...
http://arxiv.org/pdf/2309.11325v2
2023-09-23T18:36:21
2023-09-20T13:50:26
{ "title": [ -0.055236816, -0.026870728, -0.046325684, -0.06903076, -0.00086927414, 0.018676758, -0.0692749, 0.019515991, 0.03878784, 0.038208008, -0.044189453, 0.0018415451, 0.035369873, 0.0040245056, -0.056732178, 0.02029419, -0.02330017, 0.044...
http://arxiv.org/abs/2309.11307v1
Rating Prediction in Conversational Task Assistants with Behavioral and Conversational-Flow Features
Rafael Ferreira, David Semedo, João Magalhães
cs.CL, cs.AI, I.2.7
Predicting the success of Conversational Task Assistants (CTA) can be critical to understand user behavior and act accordingly. In this paper, we propose TB-Rater, a Transformer model which combines conversational-flow features with user behavior features for predicting user ratings in a CTA scenario. In particular, we...
http://arxiv.org/pdf/2309.11307v1
2023-09-20T13:34:03
2023-09-20T13:34:03
{ "title": [ -0.017288208, -0.008041382, -0.028778076, -0.042510986, -0.013618469, 0.0047569275, -0.042755127, 0.0032310486, 0.02305603, 0.044433594, -0.018325806, -0.015670776, 0.017593384, -0.01689148, -0.060333252, 0.0019025803, -0.015655518, ...
http://arxiv.org/abs/2309.11295v1
CPLLM: Clinical Prediction with Large Language Models
Ofir Ben Shoham, Nadav Rappoport
cs.CL, cs.AI, cs.LG
We present Clinical Prediction with Large Language Models (CPLLM), a method that involves fine-tuning a pre-trained Large Language Model (LLM) for clinical disease prediction. We utilized quantization and fine-tuned the LLM using prompts, with the task of predicting whether patients will be diagnosed with a target dise...
http://arxiv.org/pdf/2309.11295v1
2023-09-20T13:24:12
2023-09-20T13:24:12
{ "title": [ -0.024597168, -0.016403198, -0.03955078, -0.089904785, 0.0020179749, 0.023422241, -0.08087158, -0.008888245, 0.0033569336, 0.025009155, 0.017807007, -0.0016450882, 0.0055770874, -0.04046631, -0.021011353, 0.026824951, -0.012084961, 0...
http://arxiv.org/abs/2309.11285v1
Overview of AuTexTification at IberLEF 2023: Detection and Attribution of Machine-Generated Text in Multiple Domains
Areg Mikael Sarvazyan, José Ángel González, Marc Franco-Salvador, Francisco Rangel, Berta Chulvi, Paolo Rosso
cs.CL, cs.AI, cs.LG
This paper presents the overview of the AuTexTification shared task as part of the IberLEF 2023 Workshop in Iberian Languages Evaluation Forum, within the framework of the SEPLN 2023 conference. AuTexTification consists of two subtasks: for Subtask 1, participants had to determine whether a text is human-authored or ha...
http://arxiv.org/pdf/2309.11285v1
2023-09-20T13:10:06
2023-09-20T13:10:06
{ "title": [ 0.013694763, 0.021484375, -0.02015686, -0.014404297, -0.003238678, -0.003835678, -0.047973633, -0.043060303, -0.05166626, 0.04824829, 0.013519287, 0.043121338, 0.03540039, -0.03262329, 0.034210205, 0.013046265, -0.028961182, 0.080200...
http://arxiv.org/abs/2309.11283v1
The Wizard of Curiosities: Enriching Dialogues with Fun Facts
Frederico Vicente, Rafael Ferreira, David Semedo, João Magalhães
cs.CL, I.2.7
Introducing curiosities in a conversation is a way to teach something new to the person in a pleasant and enjoyable way. Enriching dialogues with contextualized curiosities can improve the users' perception of a dialog system and their overall user experience. In this paper, we introduce a set of curated curiosities, t...
http://arxiv.org/pdf/2309.11283v1
2023-09-20T13:07:32
2023-09-20T13:07:32
{ "title": [ 0.011268616, -0.030227661, -0.022888184, -0.043395996, -0.030883789, -0.026550293, 0.0014181137, 0.007926941, -0.03793335, 0.03479004, 0.007843018, 0.004573822, 0.015068054, -0.046081543, -0.014411926, -0.0036144257, -0.0052871704, 0...
http://arxiv.org/abs/2309.11271v1
Grounded Complex Task Segmentation for Conversational Assistants
Rafael Ferreira, David Semedo, João Magalhães
cs.CL, cs.AI, I.2.7
Following complex instructions in conversational assistants can be quite daunting due to the shorter attention and memory spans when compared to reading the same instructions. Hence, when conversational assistants walk users through the steps of complex tasks, there is a need to structure the task into manageable piece...
http://arxiv.org/pdf/2309.11271v1
2023-09-20T12:55:46
2023-09-20T12:55:46
{ "title": [ -0.016738892, -0.0067100525, -0.046173096, -0.057556152, -0.00032114983, -0.0027332306, -0.049957275, 0.022827148, -0.013809204, 0.054840088, 0.0033092499, 0.020385742, 0.01876831, -0.015655518, -0.040039062, 0.01574707, -0.017578125, ...
http://arxiv.org/abs/2309.11259v1
Sequence-to-Sequence Spanish Pre-trained Language Models
Vladimir Araujo, Maria Mihaela Trusca, Rodrigo Tufiño, Marie-Francine Moens
cs.CL, cs.AI, cs.LG
In recent years, substantial advancements in pre-trained language models have paved the way for the development of numerous non-English language versions, with a particular focus on encoder-only and decoder-only architectures. While Spanish language models encompassing BERT, RoBERTa, and GPT have exhibited prowess in n...
http://arxiv.org/pdf/2309.11259v1
2023-09-20T12:35:19
2023-09-20T12:35:19
{ "title": [ 0.0020828247, -0.001709938, -0.016281128, -0.058013916, -0.025283813, 0.0061950684, -0.06085205, 0.017837524, -0.008476257, 0.043701172, -0.00856781, -0.0070228577, 0.039245605, -0.05328369, -0.0012807846, 0.015167236, 0.001206398, 0...
http://arxiv.org/abs/2309.11252v1
The Scenario Refiner: Grounding subjects in images at the morphological level
Claudia Tagliaferri, Sofia Axioti, Albert Gatt, Denis Paperno
cs.CL, cs.CV
Derivationally related words, such as "runner" and "running", exhibit semantic differences which also elicit different visual scenarios. In this paper, we ask whether Vision and Language (V\&L) models capture such distinctions at the morphological level, using a a new methodology and dataset. We compare the results fro...
http://arxiv.org/pdf/2309.11252v1
2023-09-20T12:23:06
2023-09-20T12:23:06
{ "title": [ 0.023452759, 0.0027942657, -0.017974854, -0.056274414, 0.009063721, -0.010795593, -0.016723633, -0.03768921, 0.012992859, 0.023483276, -0.019989014, 0.011909485, 0.0050811768, -0.02784729, -0.025756836, 0.0056495667, -0.0057029724, 0...
http://arxiv.org/abs/2309.11235v1
OpenChat: Advancing Open-source Language Models with Mixed-Quality Data
Guan Wang, Sijie Cheng, Xianyuan Zhan, Xiangang Li, Sen Song, Yang Liu
cs.CL
Nowadays, open-source large language models like LLaMA have emerged. Recent developments have incorporated supervised fine-tuning (SFT) and reinforcement learning fine-tuning (RLFT) to align these models with human goals. However, SFT methods treat all training data with mixed quality equally, while RLFT methods requir...
http://arxiv.org/pdf/2309.11235v1
2023-09-20T11:54:40
2023-09-20T11:54:40
{ "title": [ -0.024169922, -0.008010864, -0.02784729, -0.07116699, -0.00016450882, 0.015144348, -0.07305908, -0.005554199, -0.011894226, 0.042053223, -0.004398346, 0.011978149, 0.020614624, -0.015602112, -0.042144775, 0.020599365, -0.008003235, 0...
http://arxiv.org/abs/2309.11210v1
Speak While You Think: Streaming Speech Synthesis During Text Generation
Avihu Dekel, Slava Shechtman, Raul Fernandez, David Haws, Zvi Kons, Ron Hoory
eess.AS, cs.CL, cs.SD
Large Language Models (LLMs) demonstrate impressive capabilities, yet interaction with these models is mostly facilitated through text. Using Text-To-Speech to synthesize LLM outputs typically results in notable latency, which is impractical for fluent voice conversations. We propose LLM2Speech, an architecture to synt...
http://arxiv.org/pdf/2309.11210v1
2023-09-20T11:00:15
2023-09-20T11:00:15
{ "title": [ -0.04083252, -0.016662598, -0.03189087, -0.06335449, -0.011909485, 0.014640808, -0.055419922, 0.027420044, -0.008796692, 0.046081543, -0.025344849, 0.023529053, 0.016662598, -0.010231018, -0.039733887, 0.005821228, -0.017959595, 0.04...
http://arxiv.org/abs/2309.11206v2
Retrieve-Rewrite-Answer: A KG-to-Text Enhanced LLMs Framework for Knowledge Graph Question Answering
Yike Wu, Nan Hu, Sheng Bi, Guilin Qi, Jie Ren, Anhuan Xie, Wei Song
cs.CL, cs.AI
Despite their competitive performance on knowledge-intensive tasks, large language models (LLMs) still have limitations in memorizing all world knowledge especially long tail knowledge. In this paper, we study the KG-augmented language model approach for solving the knowledge graph question answering (KGQA) task that r...
http://arxiv.org/pdf/2309.11206v2
2023-09-21T04:43:46
2023-09-20T10:42:08
{ "title": [ -0.0713501, -0.036224365, -0.04574585, -0.052520752, -0.057647705, 0.024993896, -0.087646484, 0.020019531, 0.0030841827, 0.038024902, -0.029785156, 0.027160645, 0.011398315, -0.002445221, -0.046142578, -0.0009493828, 0.011894226, -0....
http://arxiv.org/abs/2309.11197v1
The Languini Kitchen: Enabling Language Modelling Research at Different Scales of Compute
Aleksandar Stanić, Dylan Ashley, Oleg Serikov, Louis Kirsch, Francesco Faccio, Jürgen Schmidhuber, Thomas Hofmann, Imanol Schlag
cs.LG, cs.CL
The Languini Kitchen serves as both a research collective and codebase designed to empower researchers with limited computational resources to contribute meaningfully to the field of language modelling. We introduce an experimental protocol that enables model comparisons based on equivalent compute, measured in acceler...
http://arxiv.org/pdf/2309.11197v1
2023-09-20T10:31:17
2023-09-20T10:31:17
{ "title": [ -0.057739258, -0.014854431, -0.0647583, -0.04776001, -0.016616821, 0.0018005371, -0.06011963, 0.014556885, 0.00017929077, 0.047302246, -0.012397766, -0.007320404, 0.03439331, -0.021347046, -0.05029297, 0.023345947, -0.009986877, 0.02...
http://arxiv.org/abs/2309.11166v2
Are Large Language Models Really Robust to Word-Level Perturbations?
Haoyu Wang, Guozheng Ma, Cong Yu, Ning Gui, Linrui Zhang, Zhiqi Huang, Suwei Ma, Yongzhe Chang, Sen Zhang, Li Shen, Xueqian Wang, Peilin Zhao, Dacheng Tao
cs.CL, cs.AI
The swift advancement in the scales and capabilities of Large Language Models (LLMs) positions them as promising tools for a variety of downstream tasks. In addition to the pursuit of better performance and the avoidance of violent feedback on a certain prompt, to ensure the responsibility of the LLM, much attention is...
http://arxiv.org/pdf/2309.11166v2
2023-09-27T09:53:16
2023-09-20T09:23:46
{ "title": [ -0.015045166, 0.0046195984, -0.033843994, -0.056488037, -0.003194809, -0.006538391, -0.053100586, -0.0017642975, 0.0064926147, 0.025039673, -0.039764404, 0.004699707, 0.008216858, -0.012680054, -0.044647217, 0.0011749268, -0.003643036, ...
http://arxiv.org/abs/2309.11165v1
Assessment of Pre-Trained Models Across Languages and Grammars
Alberto Muñoz-Ortiz, David Vilares, Carlos Gómez-Rodríguez
cs.CL
We present an approach for assessing how multilingual large language models (LLMs) learn syntax in terms of multi-formalism syntactic structures. We aim to recover constituent and dependency structures by casting parsing as sequence labeling. To do so, we select a few LLMs and study them on 13 diverse UD treebanks for ...
http://arxiv.org/pdf/2309.11165v1
2023-09-20T09:23:36
2023-09-20T09:23:36
{ "title": [ 0.0027809143, 0.002670288, -0.012283325, -0.06097412, -0.013336182, -0.0016031265, -0.076171875, -0.0048561096, -0.008300781, 0.023284912, -0.022338867, -0.0010786057, 0.016311646, -0.047821045, -0.005886078, 0.009315491, 0.002603531, ...
http://arxiv.org/abs/2309.11143v1
CoT-BERT: Enhancing Unsupervised Sentence Representation through Chain-of-Thought
Bowen Zhang, Kehua Chang, Chunping Li
cs.CL, cs.AI
Unsupervised sentence representation learning aims to transform input sentences into fixed-length vectors enriched with intricate semantic information while obviating the reliance on labeled data. Recent progress within this field, propelled by contrastive learning and prompt engineering, has significantly bridged the ...
http://arxiv.org/pdf/2309.11143v1
2023-09-20T08:42:06
2023-09-20T08:42:06
{ "title": [ 0.013572693, -0.0048446655, -0.0037879944, -0.039398193, -0.0059394836, -0.0005598068, -0.029022217, -0.019317627, 0.011268616, 0.036071777, -0.027633667, 0.011169434, 0.028244019, -0.0011329651, -0.044708252, 0.013847351, -0.022827148, ...
http://arxiv.org/abs/2309.11142v1
Prototype of a robotic system to assist the learning process of English language with text-generation through DNN
Carlos Morales-Torres, Mario Campos-Soberanis, Diego Campos-Sobrino
cs.CL
In the last ongoing years, there has been a significant ascending on the field of Natural Language Processing (NLP) for performing multiple tasks including English Language Teaching (ELT). An effective strategy to favor the learning process uses interactive devices to engage learners in their self-learning process. In ...
http://arxiv.org/pdf/2309.11142v1
2023-09-20T08:39:51
2023-09-20T08:39:51
{ "title": [ 0.007194519, -0.010925293, -0.021865845, 0.0050201416, -0.04510498, -0.03274536, -0.010154724, -0.0065231323, -0.048339844, 0.0418396, -0.03894043, 0.00095129013, 0.0317688, -0.029708862, 0.016952515, -0.04586792, 0.012908936, 0.0342...
http://arxiv.org/abs/2309.11127v1
Language-Oriented Communication with Semantic Coding and Knowledge Distillation for Text-to-Image Generation
Hyelin Nam, Jihong Park, Jinho Choi, Mehdi Bennis, Seong-Lyun Kim
eess.SP, cs.AI, cs.CL
By integrating recent advances in large language models (LLMs) and generative models into the emerging semantic communication (SC) paradigm, in this article we put forward to a novel framework of language-oriented semantic communication (LSC). In LSC, machines communicate using human language messages that can be inter...
http://arxiv.org/pdf/2309.11127v1
2023-09-20T08:19:05
2023-09-20T08:19:05
{ "title": [ -0.019332886, -0.019805908, -0.03274536, -0.05178833, -0.01360321, 0.0006275177, -0.04888916, -0.005329132, 0.03540039, 0.03881836, -0.010726929, -0.0038032532, 0.028396606, -0.025527954, -0.030288696, 0.002571106, 0.014663696, 0.015...
http://arxiv.org/abs/2309.11104v1
AttentionMix: Data augmentation method that relies on BERT attention mechanism
Dominik Lewy, Jacek Mańdziuk
cs.CL, cs.AI
The Mixup method has proven to be a powerful data augmentation technique in Computer Vision, with many successors that perform image mixing in a guided manner. One of the interesting research directions is transferring the underlying Mixup idea to other domains, e.g. Natural Language Processing (NLP). Even though there...
http://arxiv.org/pdf/2309.11104v1
2023-09-20T07:18:53
2023-09-20T07:18:53
{ "title": [ 0.034698486, -0.0040664673, -0.049865723, -0.05734253, -0.007896423, -0.028640747, -0.025024414, 0.007534027, 0.009307861, 0.04650879, -0.045043945, -0.011512756, 0.07312012, -0.017196655, -0.009132385, 0.0003862381, -0.007873535, 0....
http://arxiv.org/abs/2309.11093v1
K-pop Lyric Translation: Dataset, Analysis, and Neural-Modelling
Haven Kim, Jongmin Jung, Dasaem Jeong, Juhan Nam
cs.CL, cs.LG, cs.MM
Lyric translation, a field studied for over a century, is now attracting computational linguistics researchers. We identified two limitations in previous studies. Firstly, lyric translation studies have predominantly focused on Western genres and languages, with no previous study centering on K-pop despite its populari...
http://arxiv.org/pdf/2309.11093v1
2023-09-20T06:54:55
2023-09-20T06:54:55
{ "title": [ -0.041168213, -0.015022278, -0.026245117, -0.024246216, -0.012512207, 0.008964539, -0.038116455, 0.010253906, -0.019226074, 0.058044434, -0.018234253, 0.00007426739, 0.029785156, -0.0062026978, -0.047668457, -0.0011768341, -0.0027694702,...
http://arxiv.org/abs/2309.11082v1
Dual-Modal Attention-Enhanced Text-Video Retrieval with Triplet Partial Margin Contrastive Learning
Chen Jiang, Hong Liu, Xuzheng Yu, Qing Wang, Yuan Cheng, Jia Xu, Zhongyi Liu, Qingpei Guo, Wei Chu, Ming Yang, Yuan Qi
cs.CV, cs.CL, cs.MM
In recent years, the explosion of web videos makes text-video retrieval increasingly essential and popular for video filtering, recommendation, and search. Text-video retrieval aims to rank relevant text/video higher than irrelevant ones. The core of this task is to precisely measure the cross-modal similarity between ...
http://arxiv.org/pdf/2309.11082v1
2023-09-20T06:08:11
2023-09-20T06:08:11
{ "title": [ 0.05267334, 0.031341553, -0.0126953125, -0.05331421, 0.033050537, -0.02999878, 0.032836914, -0.033203125, 0.039398193, 0.051483154, -0.053100586, 0.020019531, 0.0126571655, 0.015258789, -0.066833496, -0.00299263, -0.008300781, 0.0186...
http://arxiv.org/abs/2309.11065v1
UniPCM: Universal Pre-trained Conversation Model with Task-aware Automatic Prompt
Yucheng Cai, Wentao Ma, Yuchuan Wu, Shuzheng Si, Yuan Shao, Zhijian Ou, Yongbin Li
cs.CL
Recent research has shown that multi-task pre-training greatly improves the model's robustness and transfer ability, which is crucial for building a high-quality dialog system. However, most previous works on multi-task pre-training rely heavily on human-defined input format or prompt, which is not optimal in quality a...
http://arxiv.org/pdf/2309.11065v1
2023-09-20T05:05:40
2023-09-20T05:05:40
{ "title": [ -0.013008118, -0.00016868114, -0.046051025, -0.07287598, -0.003206253, -0.0010757446, -0.07318115, 0.000415802, 0.01335144, 0.022750854, -0.010383606, 0.003124237, 0.011474609, -0.028030396, -0.03451538, 0.0178833, -0.009376526, 0.01...
http://arxiv.org/abs/2309.11063v1
XATU: A Fine-grained Instruction-based Benchmark for Explainable Text Updates
Haopeng Zhang, Hayate Iso, Sairam Gurajada, Nikita Bhutani
cs.CL
Text editing is a crucial task that involves modifying text to better align with user intents. However, existing text editing benchmark datasets have limitations in providing only coarse-grained instructions. Consequently, although the edited output may seem reasonable, it often deviates from the intended changes outli...
http://arxiv.org/pdf/2309.11063v1
2023-09-20T04:58:59
2023-09-20T04:58:59
{ "title": [ -0.04147339, -0.01486969, -0.018157959, -0.0670166, -0.012817383, 0.019744873, -0.07409668, 0.0021305084, 0.0013141632, 0.049865723, -0.025039673, 0.03289795, -0.0023555756, 0.011421204, -0.048675537, 0.0062217712, 0.00434494, 0.0253...
http://arxiv.org/abs/2309.11054v2
Design of Chain-of-Thought in Math Problem Solving
Zhanming Jie, Trung Quoc Luong, Xinbo Zhang, Xiaoran Jin, Hang Li
cs.CL, cs.AI, cs.LG, cs.PL
Chain-of-Thought (CoT) plays a crucial role in reasoning for math problem solving. We conduct a comprehensive examination of methods for designing CoT, comparing conventional natural language CoT with various program CoTs, including the self-describing program, the comment-describing program, and the non-describing pro...
http://arxiv.org/pdf/2309.11054v2
2023-09-30T15:51:38
2023-09-20T04:17:28
{ "title": [ 0.024963379, -0.01020813, -0.017456055, 0.012611389, -0.028381348, -0.053588867, 0.05114746, -0.015274048, -0.028671265, 0.017654419, -0.036315918, -0.010147095, -0.004081726, -0.03012085, 0.000620842, -0.031433105, 0.013343811, 0.00...
http://arxiv.org/abs/2309.11052v2
fakenewsbr: A Fake News Detection Platform for Brazilian Portuguese
Luiz Giordani, Gilsiley Darú, Rhenan Queiroz, Vitor Buzinaro, Davi Keglevich Neiva, Daniel Camilo Fuentes Guzmán, Marcos Jardel Henriques, Oilson Alberto Gonzatto Junior, Francisco Louzada
cs.CL, cs.LG, stat.ML
The proliferation of fake news has become a significant concern in recent times due to its potential to spread misinformation and manipulate public opinion. This paper presents a comprehensive study on detecting fake news in Brazilian Portuguese, focusing on journalistic-type news. We propose a machine learning-based a...
http://arxiv.org/pdf/2309.11052v2
2023-09-21T00:35:12
2023-09-20T04:10:03
{ "title": [ -0.029464722, 0.00048279762, -0.030853271, -0.03616333, 0.009735107, 0.015930176, -0.024139404, -0.018753052, -0.032440186, 0.043029785, -0.0154418945, 0.0060806274, 0.010467529, -0.015945435, -0.034362793, -0.0077590942, -0.04232788, ...
http://arxiv.org/abs/2309.11049v2
Localize, Retrieve and Fuse: A Generalized Framework for Free-Form Question Answering over Tables
Wenting Zhao, Ye Liu, Yao Wan, Yibo Wang, Zhongfen Deng, Philip S. Yu
cs.CL
Question answering on tabular data (a.k.a TableQA), which aims at generating answers to questions grounded on a provided table, has gained significant attention recently. Prior work primarily produces concise factual responses through information extraction from individual or limited table cells, lacking the ability to...
http://arxiv.org/pdf/2309.11049v2
2023-09-21T10:57:08
2023-09-20T03:52:34
{ "title": [ -0.0028591156, -0.011390686, -0.03390503, -0.072021484, 0.016921997, 0.0051727295, -0.034454346, -0.0013971329, 0.025131226, 0.034606934, -0.026794434, 0.0039749146, 0.0039100647, 0.008682251, -0.0692749, -0.0012578964, 0.0064811707, ...
http://arxiv.org/abs/2309.11046v1
Heterogeneous Entity Matching with Complex Attribute Associations using BERT and Neural Networks
Shitao Wang, Jiamin Lu
cs.CL
Across various domains, data from different sources such as Baidu Baike and Wikipedia often manifest in distinct forms. Current entity matching methodologies predominantly focus on homogeneous data, characterized by attributes that share the same structure and concise attribute values. However, this orientation poses c...
http://arxiv.org/pdf/2309.11046v1
2023-09-20T03:49:57
2023-09-20T03:49:57
{ "title": [ 0.0007915497, 0.016159058, -0.045532227, -0.061920166, 0.03363037, -0.005809784, -0.016693115, -0.012664795, 0.027496338, 0.044006348, -0.00006788969, -0.009849548, 0.051879883, -0.017669678, -0.033966064, 0.026931763, -0.02684021, 0...
http://arxiv.org/abs/2309.11042v1
Making Small Language Models Better Multi-task Learners with Mixture-of-Task-Adapters
Yukang Xie, Chengyu Wang, Junbing Yan, Jiyong Zhou, Feiqi Deng, Jun Huang
cs.CL, cs.AI
Recently, Large Language Models (LLMs) have achieved amazing zero-shot learning performance over a variety of Natural Language Processing (NLP) tasks, especially for text generative tasks. Yet, the large size of LLMs often leads to the high computational cost of model training and online deployment. In our work, we pre...
http://arxiv.org/pdf/2309.11042v1
2023-09-20T03:39:56
2023-09-20T03:39:56
{ "title": [ -0.04058838, 0.0013504028, -0.046875, -0.03363037, -0.01084137, -0.01058197, -0.047546387, -0.00045728683, -0.023895264, 0.04525757, 0.0021343231, 0.0014820099, -0.00080394745, -0.0141067505, -0.04510498, -0.0011463165, -0.016464233, ...
http://arxiv.org/abs/2309.11027v1
Named Entity Recognition via Machine Reading Comprehension: A Multi-Task Learning Approach
Yibo Wang, Wenting Zhao, Yao Wan, Zhongfen Deng, Philip S. Yu
cs.CL
Named Entity Recognition (NER) aims to extract and classify entity mentions in the text into pre-defined types (e.g., organization or person name). Recently, many works have been proposed to shape the NER as a machine reading comprehension problem (also termed MRC-based NER), in which entity recognition is achieved by ...
http://arxiv.org/pdf/2309.11027v1
2023-09-20T03:15:05
2023-09-20T03:15:05
{ "title": [ -0.036224365, 0.00034952164, -0.04675293, -0.067993164, 0.008094788, 0.0023536682, -0.05899048, -0.0005426407, 0.025558472, 0.02192688, -0.03579712, 0.006389618, 0.0005259514, 0.0026340485, -0.048919678, 0.003540039, -0.012321472, 0....
http://arxiv.org/abs/2309.11000v1
Towards Joint Modeling of Dialogue Response and Speech Synthesis based on Large Language Model
Xinyu Zhou, Delong Chen, Yudong Chen
cs.CL, cs.SD, eess.AS
This paper explores the potential of constructing an AI spoken dialogue system that "thinks how to respond" and "thinks how to speak" simultaneously, which more closely aligns with the human speech production process compared to the current cascade pipeline of independent chatbot and Text-to-Speech (TTS) modules. We hy...
http://arxiv.org/pdf/2309.11000v1
2023-09-20T01:48:27
2023-09-20T01:48:27
{ "title": [ -0.030075073, -0.009086609, -0.039276123, -0.07196045, -0.0057373047, 0.012802124, -0.085998535, 0.024749756, -0.012069702, 0.03289795, -0.00070667267, -0.012519836, 0.0043907166, -0.03164673, -0.044067383, -0.00004029274, -0.021118164, ...
http://arxiv.org/abs/2309.10966v5
MBR and QE Finetuning: Training-time Distillation of the Best and Most Expensive Decoding Methods
Mara Finkelstein, Subhajit Naskar, Mehdi Mirzazadeh, Apurva Shah, Markus Freitag
cs.CL
Recent research in decoding methods for Natural Language Generation (NLG) tasks has shown that MAP decoding is not optimal, because model probabilities do not always align with human preferences. Stronger decoding methods, including Quality Estimation (QE) reranking and Minimum Bayes' Risk (MBR) decoding, have since be...
http://arxiv.org/pdf/2309.10966v5
2023-10-11T23:59:52
2023-09-19T23:39:07
{ "title": [ -0.008674622, -0.023117065, 0.0025177002, -0.043640137, -0.033416748, -0.011680603, -0.007648468, -0.016952515, -0.0446167, 0.058166504, 0.010025024, 0.03237915, 0.018539429, -0.033721924, -0.012229919, 0.0055160522, 0.0049705505, 0....
http://arxiv.org/abs/2309.10954v1
In-Context Learning for Text Classification with Many Labels
Aristides Milios, Siva Reddy, Dzmitry Bahdanau
cs.CL, cs.LG
In-context learning (ICL) using large language models for tasks with many labels is challenging due to the limited context window, which makes it difficult to fit a sufficient number of examples in the prompt. In this paper, we use a pre-trained dense retrieval model to bypass this limitation, giving the model only a p...
http://arxiv.org/pdf/2309.10954v1
2023-09-19T22:41:44
2023-09-19T22:41:44
{ "title": [ -0.0031776428, 0.0023269653, -0.04522705, -0.051849365, 0.022476196, -0.018096924, 0.0022697449, 0.0011510849, 0.012702942, 0.021102905, -0.03677368, 0.023208618, -0.016311646, -0.0011405945, -0.03237915, -0.0071907043, -0.012237549, ...
http://arxiv.org/abs/2309.10952v1
LMDX: Language Model-based Document Information Extraction and Localization
Vincent Perot, Kai Kang, Florian Luisier, Guolong Su, Xiaoyu Sun, Ramya Sree Boppana, Zilong Wang, Jiaqi Mu, Hao Zhang, Nan Hua
cs.CL, cs.AI, cs.LG
Large Language Models (LLM) have revolutionized Natural Language Processing (NLP), improving state-of-the-art on many existing tasks and exhibiting emergent capabilities. However, LLMs have not yet been successfully applied on semi-structured document information extraction, which is at the core of many document proces...
http://arxiv.org/pdf/2309.10952v1
2023-09-19T22:32:56
2023-09-19T22:32:56
{ "title": [ -0.052581787, -0.009666443, -0.028457642, -0.036621094, -0.004650116, 0.012001038, -0.053955078, -0.024032593, 0.0146102905, 0.032440186, -0.029907227, 0.0053710938, 0.020553589, -0.0005173683, -0.039215088, 0.013130188, -0.00806427, ...
http://arxiv.org/abs/2309.10945v1
Benchmarks for Pirá 2.0, a Reading Comprehension Dataset about the Ocean, the Brazilian Coast, and Climate Change
Paulo Pirozelli, Marcos M. José, Igor Silveira, Flávio Nakasato, Sarajane M. Peres, Anarosa A. F. Brandão, Anna H. R. Costa, Fabio G. Cozman
cs.CL, cs.AI
Pir\'a is a reading comprehension dataset focused on the ocean, the Brazilian coast, and climate change, built from a collection of scientific abstracts and reports on these topics. This dataset represents a versatile language resource, particularly useful for testing the ability of current machine learning models to a...
http://arxiv.org/pdf/2309.10945v1
2023-09-19T21:56:45
2023-09-19T21:56:45
{ "title": [ -0.06964111, -0.026519775, -0.012207031, 0.012298584, -0.07409668, 0.013969421, -0.051696777, 0.0028362274, 0.024795532, 0.027038574, -0.055236816, 0.01260376, -0.030044556, 0.0041923523, -0.030776978, -0.043182373, -0.010765076, 0.0...
http://arxiv.org/abs/2309.10931v1
A Family of Pretrained Transformer Language Models for Russian
Dmitry Zmitrovich, Alexander Abramov, Andrey Kalmykov, Maria Tikhonova, Ekaterina Taktasheva, Danil Astafurov, Mark Baushenko, Artem Snegirev, Tatiana Shavrina, Sergey Markov, Vladislav Mikhailov, Alena Fenogenova
cs.CL
Nowadays, Transformer language models (LMs) represent a fundamental component of the NLP research methodologies and applications. However, the development of such models specifically for the Russian language has received little attention. This paper presents a collection of 13 Russian Transformer LMs based on the encod...
http://arxiv.org/pdf/2309.10931v1
2023-09-19T21:07:52
2023-09-19T21:07:52
{ "title": [ 0.007858276, 0.00856781, -0.020187378, -0.06365967, 0.019546509, 0.005168915, -0.03289795, -0.002122879, -0.026367188, 0.03704834, -0.011405945, 0.0062217712, 0.036315918, -0.037200928, -0.013793945, 0.019622803, 0.0023460388, 0.0438...
http://arxiv.org/abs/2309.10929v1
Specializing Small Language Models towards Complex Style Transfer via Latent Attribute Pre-Training
Ruiqi Xu, Yongfeng Huang, Xin Chen, Lin Zhang
cs.CL
In this work, we introduce the concept of complex text style transfer tasks, and constructed complex text datasets based on two widely applicable scenarios. Our dataset is the first large-scale data set of its kind, with 700 rephrased sentences and 1,000 sentences from the game Genshin Impact. While large language mode...
http://arxiv.org/pdf/2309.10929v1
2023-09-19T21:01:40
2023-09-19T21:01:40
{ "title": [ -0.022613525, -0.016281128, -0.035247803, -0.04385376, -0.025375366, -0.0027599335, -0.04724121, 0.0017557144, -0.024932861, 0.0501709, 0.016677856, 0.027832031, 0.038726807, -0.03817749, 0.0018205643, 0.032592773, 0.010246277, 0.023...
http://arxiv.org/abs/2309.10926v1
Semi-Autoregressive Streaming ASR With Label Context
Siddhant Arora, George Saon, Shinji Watanabe, Brian Kingsbury
cs.CL, cs.SD, eess.AS
Non-autoregressive (NAR) modeling has gained significant interest in speech processing since these models achieve dramatically lower inference time than autoregressive (AR) models while also achieving good transcription accuracy. Since NAR automatic speech recognition (ASR) models must wait for the completion of the en...
http://arxiv.org/pdf/2309.10926v1
2023-09-19T20:55:58
2023-09-19T20:55:58
{ "title": [ -0.0131073, 0.013305664, -0.08294678, -0.09564209, 0.02432251, -0.01889038, -0.04232788, 0.046783447, -0.01979065, 0.048309326, -0.019744873, -0.0013246536, 0.01725769, -0.017990112, -0.020904541, -0.00016450882, 0.022537231, 0.01617...
http://arxiv.org/abs/2309.10923v1
Semi-automatic staging area for high-quality structured data extraction from scientific literature
Luca Foppiano, Tomoya Mato, Kensei Terashima, Pedro Ortiz Suarez, Taku Tou, Chikako Sakai, Wei-Sheng Wang, Toshiyuki Amagasa, Yoshihiko Takano, Masashi Ishii
cs.CL, cond-mat.supr-con, cs.DB, cs.LG
In this study, we propose a staging area for ingesting new superconductors' experimental data in SuperCon that is machine-collected from scientific articles. Our objective is to enhance the efficiency of updating SuperCon while maintaining or enhancing the data quality. We present a semi-automatic staging area driven b...
http://arxiv.org/pdf/2309.10923v1
2023-09-19T20:53:13
2023-09-19T20:53:13
{ "title": [ -0.045959473, 0.013298035, -0.06762695, -0.08654785, 0.013465881, -0.0055618286, -0.05126953, 0.0049705505, 0.0041160583, 0.03149414, 0.005504608, 0.0026340485, -0.023803711, -0.03189087, -0.050598145, 0.007598877, -0.0035820007, -0....
http://arxiv.org/abs/2309.10917v1
End-to-End Speech Recognition Contextualization with Large Language Models
Egor Lakomkin, Chunyang Wu, Yassir Fathullah, Ozlem Kalinli, Michael L. Seltzer, Christian Fuegen
eess.AS, cs.AI, cs.CL, cs.LG, cs.SD
In recent years, Large Language Models (LLMs) have garnered significant attention from the research community due to their exceptional performance and generalization capabilities. In this paper, we introduce a novel method for contextualizing speech recognition models incorporating LLMs. Our approach casts speech recog...
http://arxiv.org/pdf/2309.10917v1
2023-09-19T20:28:57
2023-09-19T20:28:57
{ "title": [ -0.0068855286, 0.019241333, -0.054595947, -0.09472656, 0.02684021, 0.0019035339, -0.07434082, 0.013877869, -0.020889282, 0.0513916, -0.021636963, 0.000096559525, 0.0129776, -0.012771606, -0.03164673, 0.00061893463, -0.0059814453, 0.0...
http://arxiv.org/abs/2309.10916v3
What Learned Representations and Influence Functions Can Tell Us About Adversarial Examples
Shakila Mahjabin Tonni, Mark Dras
cs.LG, cs.CL
Adversarial examples, deliberately crafted using small perturbations to fool deep neural networks, were first studied in image processing and more recently in NLP. While approaches to detecting adversarial examples in NLP have largely relied on search over input perturbations, image processing has seen a range of techn...
http://arxiv.org/pdf/2309.10916v3
2023-10-10T08:45:00
2023-09-19T20:28:24
{ "title": [ 0.030227661, 0.004722595, -0.0067329407, -0.03387451, 0.009048462, -0.02067566, 0.015235901, -0.03439331, -0.03970337, 0.0075263977, -0.012184143, 0.019317627, -0.043884277, -0.02670288, -0.015434265, -0.025131226, 0.010406494, 0.043...
http://arxiv.org/abs/2310.12989v1
Enhancing Health Data Interoperability with Large Language Models: A FHIR Study
Yikuan Li, Hanyin Wang, Halid Yerebakan, Yoshihisa Shinagawa, Yuan Luo
cs.CL, cs.AI
In this study, we investigated the ability of the large language model (LLM) to enhance healthcare data interoperability. We leveraged the LLM to convert clinical texts into their corresponding FHIR resources. Our experiments, conducted on 3,671 snippets of clinical text, demonstrated that the LLM not only streamlines ...
http://arxiv.org/pdf/2310.12989v1
2023-09-19T20:09:35
2023-09-19T20:09:35
{ "title": [ -0.020339966, -0.0019388199, -0.07922363, -0.10662842, 0.037200928, -0.0018606186, -0.070251465, 0.022735596, -0.008140564, 0.029769897, -0.016784668, -0.012290955, -0.0031547546, -0.023132324, -0.03338623, -0.011436462, -0.0037117004, ...
http://arxiv.org/abs/2309.10898v1
RedPenNet for Grammatical Error Correction: Outputs to Tokens, Attentions to Spans
Bohdan Didenko, Andrii Sameliuk
cs.CL
The text editing tasks, including sentence fusion, sentence splitting and rephrasing, text simplification, and Grammatical Error Correction (GEC), share a common trait of dealing with highly similar input and output sequences. This area of research lies at the intersection of two well-established fields: (i) fully auto...
http://arxiv.org/pdf/2309.10898v1
2023-09-19T19:48:30
2023-09-19T19:48:30
{ "title": [ 0.004459381, -0.009361267, -0.030548096, -0.028579712, -0.013580322, -0.0019044876, -0.021896362, -0.015670776, 0.0023498535, 0.007858276, -0.044525146, -0.019332886, 0.024246216, -0.037384033, -0.010551453, -0.027755737, -0.025360107, ...
http://arxiv.org/abs/2309.10891v1
Self-Augmentation Improves Zero-Shot Cross-Lingual Transfer
Fei Wang, Kuan-Hao Huang, Kai-Wei Chang, Muhao Chen
cs.CL, cs.AI, cs.LG
Zero-shot cross-lingual transfer is a central task in multilingual NLP, allowing models trained in languages with more sufficient training resources to generalize to other low-resource languages. Earlier efforts on this task use parallel corpora, bilingual dictionaries, or other annotated alignment data to improve cros...
http://arxiv.org/pdf/2309.10891v1
2023-09-19T19:30:56
2023-09-19T19:30:56
{ "title": [ 0.046661377, -0.00086164474, -0.016738892, -0.023849487, -0.024871826, -0.024002075, 0.0066719055, -0.0017061234, -0.02192688, 0.056488037, 0.009002686, 0.017974854, 0.0158844, -0.047058105, -0.0021018982, 0.022323608, 0.0043182373, ...
http://arxiv.org/abs/2309.10880v1
Classifying Organizations for Food System Ontologies using Natural Language Processing
Tianyu Jiang, Sonia Vinogradova, Nathan Stringham, E. Louise Earl, Allan D. Hollander, Patrick R. Huber, Ellen Riloff, R. Sandra Schillo, Giorgio A. Ubbiali, Matthew Lange
cs.CL, cs.AI, cs.CY, cs.IR, H.3.1; I.2.7; J.3; J.4; K.4.3
Our research explores the use of natural language processing (NLP) methods to automatically classify entities for the purpose of knowledge graph population and integration with food system ontologies. We have created NLP models that can automatically classify organizations with respect to categories associated with env...
http://arxiv.org/pdf/2309.10880v1
2023-09-19T19:07:48
2023-09-19T19:07:48
{ "title": [ -0.0072631836, 0.008613586, -0.035491943, -0.058563232, 0.013793945, 0.00894928, -0.04220581, -0.015899658, -0.012016296, 0.009185791, -0.043182373, 0.00029802322, -0.0058174133, -0.015991211, -0.013305664, -0.02909851, -0.030548096, ...
http://arxiv.org/abs/2309.10818v2
SlimPajama-DC: Understanding Data Combinations for LLM Training
Zhiqiang Shen, Tianhua Tao, Liqun Ma, Willie Neiswanger, Zhengzhong Liu, Hongyi Wang, Bowen Tan, Joel Hestness, Natalia Vassilieva, Daria Soboleva, Eric Xing
cs.CL, cs.AI
This paper aims to understand the impacts of various data combinations (e.g., web text, wikipedia, github, books) on the training of large language models using SlimPajama. SlimPajama is a rigorously deduplicated, multi-source dataset, which has been refined and further deduplicated to 627B tokens from the extensive 1....
http://arxiv.org/pdf/2309.10818v2
2023-10-09T18:30:48
2023-09-19T17:59:54
{ "title": [ -0.074279785, -0.012756348, -0.08874512, -0.06878662, -0.019882202, -0.0000461936, -0.08843994, 0.02897644, 0.0013904572, 0.013824463, -0.020492554, 0.022491455, -0.031234741, -0.03918457, -0.0042915344, -0.0025348663, -0.015945435, ...
http://arxiv.org/abs/2309.10814v1
Natural Language Embedded Programs for Hybrid Language Symbolic Reasoning
Tianhua Zhang, Jiaxin Ge, Hongyin Luo, Yung-Sung Chuang, Mingye Gao, Yuan Gong, Xixin Wu, Yoon Kim, Helen Meng, James Glass
cs.CL
How can we perform computations over natural language representations to solve tasks that require symbolic and numeric reasoning? We propose natural language embedded programs (NLEP) as a unifying framework for addressing math/symbolic reasoning, natural language understanding, and instruction following tasks. Our appr...
http://arxiv.org/pdf/2309.10814v1
2023-09-19T17:54:21
2023-09-19T17:54:21
{ "title": [ -0.032318115, -0.02432251, -0.025512695, -0.04586792, -0.021835327, 0.008529663, -0.032196045, 0.0095825195, -0.015823364, 0.024032593, -0.048187256, -0.010063171, 0.009056091, -0.022705078, -0.012908936, -0.022232056, -0.00907135, 0...
http://arxiv.org/abs/2309.10811v1
Modeling interdisciplinary interactions among Physics, Mathematics & Computer Science
Rima Hazra, Mayank Singh, Pawan Goyal, Bibhas Adhikari, Animesh Mukherjee
cs.DL, cs.CL
Interdisciplinarity has over the recent years have gained tremendous importance and has become one of the key ways of doing cutting edge research. In this paper we attempt to model the citation flow across three different fields -- Physics (PHY), Mathematics (MA) and Computer Science (CS). For instance, is there a spec...
http://arxiv.org/pdf/2309.10811v1
2023-09-19T17:52:50
2023-09-19T17:52:50
{ "title": [ 0.014831543, 0.020690918, -0.03982544, -0.018157959, -0.01537323, -0.045898438, 0.01979065, -0.017471313, -0.037475586, 0.03527832, 0.0062675476, -5.364418e-7, 0.021057129, -0.04397583, 0.009963989, -0.017242432, 0.010658264, 0.02200...
http://arxiv.org/abs/2309.10809v1
Semantic Text Compression for Classification
Emrecan Kutay, Aylin Yener
cs.IT, cs.CL, math.IT
We study semantic compression for text where meanings contained in the text are conveyed to a source decoder, e.g., for classification. The main motivator to move to such an approach of recovering the meaning without requiring exact reconstruction is the potential resource savings, both in storage and in conveying the ...
http://arxiv.org/pdf/2309.10809v1
2023-09-19T17:50:57
2023-09-19T17:50:57
{ "title": [ -0.04928589, -0.011024475, -0.04147339, -0.043640137, -0.00289917, 0.009162903, -0.03253174, 0.006198883, 0.012283325, 0.044952393, -0.005264282, 0.0008382797, -0.0065078735, -0.014839172, -0.057250977, 0.007194519, 0.013473511, 0.02...
http://arxiv.org/abs/2309.10783v1
Language as the Medium: Multimodal Video Classification through text only
Laura Hanu, Anita L. Verő, James Thewlis
cs.CV, cs.AI, cs.CL
Despite an exciting new wave of multimodal machine learning models, current approaches still struggle to interpret the complex contextual relationships between the different modalities present in videos. Going beyond existing methods that emphasize simple activities or objects, we propose a new model-agnostic approach ...
http://arxiv.org/pdf/2309.10783v1
2023-09-19T17:32:21
2023-09-19T17:32:21
{ "title": [ -0.025299072, 0.010276794, -0.04937744, -0.031921387, 0.014793396, -0.020370483, -0.012779236, -0.013069153, 0.006755829, 0.02558899, -0.03262329, 0.014579773, -0.01586914, -0.0060539246, -0.045166016, -0.015068054, -0.019454956, 0.0...
http://arxiv.org/abs/2309.10772v1
Interactive Distillation of Large Single-Topic Corpora of Scientific Papers
Nicholas Solovyev, Ryan Barron, Manish Bhattarai, Maksim E. Eren, Kim O. Rasmussen, Boian S. Alexandrov
cs.IR, cs.CL, cs.DL, cs.LG
Highly specific datasets of scientific literature are important for both research and education. However, it is difficult to build such datasets at scale. A common approach is to build these datasets reductively by applying topic modeling on an established corpus and selecting specific topics. A more robust but time-co...
http://arxiv.org/pdf/2309.10772v1
2023-09-19T17:18:36
2023-09-19T17:18:36
{ "title": [ -0.0031166077, 0.014503479, -0.043304443, -0.06945801, 0.01386261, -0.002855301, -0.051330566, -0.015945435, 0.03933716, 0.033569336, -0.019927979, -0.00434494, -0.01007843, -0.01121521, -0.047943115, 0.006996155, -0.0063095093, 0.02...
http://arxiv.org/abs/2309.10770v1
FRASIMED: a Clinical French Annotated Resource Produced through Crosslingual BERT-Based Annotation Projection
Jamil Zaghir, Mina Bjelogrlic, Jean-Philippe Goldman, Soukaïna Aananou, Christophe Gaudet-Blavignac, Christian Lovis
cs.CL, cs.AI
Natural language processing (NLP) applications such as named entity recognition (NER) for low-resource corpora do not benefit from recent advances in the development of large language models (LLMs) where there is still a need for larger annotated datasets. This research article introduces a methodology for generating t...
http://arxiv.org/pdf/2309.10770v1
2023-09-19T17:17:28
2023-09-19T17:17:28
{ "title": [ -0.015258789, -0.033416748, -0.008384705, -0.0859375, -0.037597656, 0.03439331, -0.088378906, 0.009529114, -0.0065612793, 0.05630493, -0.0004825592, -0.002910614, 0.020339966, -0.041290283, -0.01979065, 0.02027893, -0.00057554245, 0....
http://arxiv.org/abs/2309.10744v1
Evaluating large language models' ability to understand metaphor and sarcasm using a screening test for Asperger syndrome
Hiromu Yakura
cs.CL, cs.AI, cs.HC
Metaphors and sarcasm are precious fruits of our highly-evolved social communication skills. However, children with Asperger syndrome are known to have difficulties in comprehending sarcasm, even if they possess a certain level of verbal IQ sufficient for understanding metaphors. Given that, a screening test that score...
http://arxiv.org/pdf/2309.10744v1
2023-09-19T16:41:19
2023-09-19T16:41:19
{ "title": [ -0.029769897, 0.0005707741, -0.0058174133, -0.0032215118, -0.03111267, -0.002708435, -0.02357483, -0.029373169, 0.03060913, 0.025024414, -0.030593872, -0.043518066, 0.045928955, -0.026779175, -0.053741455, -0.012550354, -0.024230957, ...
http://arxiv.org/abs/2309.10738v2
MelodyGLM: Multi-task Pre-training for Symbolic Melody Generation
Xinda Wu, Zhijie Huang, Kejun Zhang, Jiaxing Yu, Xu Tan, Tieyao Zhang, Zihao Wang, Lingyun Sun
cs.SD, cs.AI, cs.CL, cs.IR, cs.MM, eess.AS
Pre-trained language models have achieved impressive results in various music understanding and generation tasks. However, existing pre-training methods for symbolic melody generation struggle to capture multi-scale, multi-dimensional structural information in note sequences, due to the domain knowledge discrepancy bet...
http://arxiv.org/pdf/2309.10738v2
2023-09-20T10:56:07
2023-09-19T16:34:24
{ "title": [ -0.0071372986, -0.028762817, -0.035003662, -0.09075928, -0.0116119385, -0.0004761219, -0.05517578, 0.014228821, -0.001004219, 0.046966553, 0.004081726, 0.008598328, 0.015129089, -0.020248413, -0.038238525, 0.010284424, -0.0077323914, ...
http://arxiv.org/abs/2309.10706v2
OpenBA: An Open-sourced 15B Bilingual Asymmetric seq2seq Model Pre-trained from Scratch
Juntao Li, Zecheng Tang, Yuyang Ding, Pinzheng Wang, Pei Guo, Wangjie You, Dan Qiao, Wenliang Chen, Guohong Fu, Qiaoming Zhu, Guodong Zhou, Min Zhang
cs.CL
Large language models (LLMs) with billions of parameters have demonstrated outstanding performance on various natural language processing tasks. This report presents OpenBA, an open-sourced 15B bilingual asymmetric seq2seq model, to contribute an LLM variant to the Chinese-oriented open-source model community. We enhan...
http://arxiv.org/pdf/2309.10706v2
2023-10-01T18:21:58
2023-09-19T15:46:40
{ "title": [ -0.03805542, 0.0076560974, -0.024612427, -0.043792725, -0.025863647, 0.00869751, -0.07611084, -0.036224365, -0.019348145, 0.04284668, 0.027297974, -0.021911621, 0.020492554, -0.06463623, -0.006401062, 0.008735657, 0.011329651, 0.0430...
http://arxiv.org/abs/2309.10691v2
MINT: Evaluating LLMs in Multi-turn Interaction with Tools and Language Feedback
Xingyao Wang, Zihan Wang, Jiateng Liu, Yangyi Chen, Lifan Yuan, Hao Peng, Heng Ji
cs.CL, cs.AI, cs.LG
To solve complex tasks, large language models (LLMs) often require multiple rounds of interactions with the user, sometimes assisted by external tools. However, current evaluation protocols often emphasize benchmark performance with single-turn exchanges, neglecting the nuanced interactions among the user, LLMs, and ex...
http://arxiv.org/pdf/2309.10691v2
2023-10-12T04:07:56
2023-09-19T15:25:42
{ "title": [ -0.038726807, -0.016601562, -0.027236938, -0.021835327, -0.029815674, -0.00027942657, -0.058441162, -0.013519287, -0.015609741, 0.023788452, -0.030044556, 0.022277832, -0.030075073, -0.017349243, -0.032958984, -0.024291992, -0.024749756,...
http://arxiv.org/abs/2309.10677v2
Estimating Contamination via Perplexity: Quantifying Memorisation in Language Model Evaluation
Yucheng Li
cs.CL, cs.AI
Data contamination in model evaluation is getting increasingly prevalent as the massive training corpora of large language models often unintentionally include benchmark samples. Therefore, contamination analysis has became an inevitable part of reliable model evaluation. However, existing method of contamination analy...
http://arxiv.org/pdf/2309.10677v2
2023-09-27T01:15:49
2023-09-19T15:02:58
{ "title": [ -0.0040397644, -0.0053482056, -0.0027580261, -0.06402588, -0.014923096, -0.002494812, -0.05178833, -0.027618408, -0.015670776, 0.03933716, -0.025680542, 0.007293701, 0.0011224747, -0.010749817, -0.06933594, -0.016357422, -0.023757935, ...
http://arxiv.org/abs/2309.10668v1
Language Modeling Is Compression
Grégoire Delétang, Anian Ruoss, Paul-Ambroise Duquenne, Elliot Catt, Tim Genewein, Christopher Mattern, Jordi Grau-Moya, Li Kevin Wenliang, Matthew Aitchison, Laurent Orseau, Marcus Hutter, Joel Veness
cs.LG, cs.AI, cs.CL, cs.IT, math.IT
It has long been established that predictive models can be transformed into lossless compressors and vice versa. Incidentally, in recent years, the machine learning community has focused on training increasingly large and powerful self-supervised (language) models. Since these large language models exhibit impressive p...
http://arxiv.org/pdf/2309.10668v1
2023-09-19T14:50:38
2023-09-19T14:50:38
{ "title": [ -0.04537964, -0.022155762, -0.019470215, -0.03491211, -0.064331055, 0.021560669, -0.07196045, 0.027160645, -0.011703491, 0.04711914, 0.009132385, -0.045074463, 0.008285522, -0.033477783, -0.02583313, 0.0039596558, 0.032318115, -0.008...
http://arxiv.org/abs/2309.10661v2
NusaWrites: Constructing High-Quality Corpora for Underrepresented and Extremely Low-Resource Languages
Samuel Cahyawijaya, Holy Lovenia, Fajri Koto, Dea Adhista, Emmanuel Dave, Sarah Oktavianti, Salsabil Maulana Akbar, Jhonson Lee, Nuur Shadieq, Tjeng Wawan Cenggoro, Hanung Wahyuning Linuwih, Bryan Wilie, Galih Pradipta Muridan, Genta Indra Winata, David Moeljadi, Alham Fikri Aji, Ayu Purwarianti, Pascale Fung
cs.CL, cs.AI
Democratizing access to natural language processing (NLP) technology is crucial, especially for underrepresented and extremely low-resource languages. Previous research has focused on developing labeled and unlabeled corpora for these languages through online scraping and document translation. While these methods have ...
http://arxiv.org/pdf/2309.10661v2
2023-09-20T02:15:50
2023-09-19T14:42:33
{ "title": [ -0.025756836, -0.007259369, -0.055908203, -0.07507324, 0.00730896, 0.0064353943, -0.06311035, 0.002632141, -0.0016326904, 0.023208618, -0.03427124, 0.0071907043, -0.0013380051, -0.007709503, -0.04232788, -0.0063209534, -0.020492554, ...
http://arxiv.org/abs/2309.10654v2
CFGPT: Chinese Financial Assistant with Large Language Model
Jiangtong Li, Yuxuan Bian, Guoxuan Wang, Yang Lei, Dawei Cheng, Zhijun Ding, Changjun Jiang
cs.CL, cs.AI, cs.CE
Large language models (LLMs) have demonstrated great potential in natural language processing tasks within the financial domain. In this work, we present a Chinese Financial Generative Pre-trained Transformer framework, named CFGPT, which includes a dataset~(CFData) for pre-training and supervised fine-tuning, a financ...
http://arxiv.org/pdf/2309.10654v2
2023-09-22T09:52:07
2023-09-19T14:34:01
{ "title": [ -0.035583496, -0.013366699, -0.05203247, -0.029220581, -0.024398804, -0.007686615, -0.045715332, 0.0001295805, 0.007385254, 0.047943115, -0.027618408, -0.02468872, 0.066345215, -0.033691406, -0.026489258, 0.001991272, -0.011787415, 0...
http://arxiv.org/abs/2309.10621v1
Large language models can accurately predict searcher preferences
Paul Thomas, Seth Spielman, Nick Craswell, Bhaskar Mitra
cs.IR, cs.AI, cs.CL, cs.LG
Relevance labels, which indicate whether a search result is valuable to a searcher, are key to evaluating and optimising search systems. The best way to capture the true preferences of users is to ask them for their careful feedback on which results would be useful, but this approach does not scale to produce a large n...
http://arxiv.org/pdf/2309.10621v1
2023-09-19T13:55:39
2023-09-19T13:55:39
{ "title": [ 0.056732178, 0.004928589, 0.0033874512, -0.03729248, -0.031951904, -0.033843994, -0.011787415, -0.01399231, 0.00831604, 0.033721924, -0.022659302, -0.03704834, 0.032592773, -0.05581665, 0.0057525635, -0.007858276, -0.0070877075, 0.04...
http://arxiv.org/abs/2309.10608v1
Improving Medical Dialogue Generation with Abstract Meaning Representations
Bohao Yang, Chen Tang, Chenghua Lin
cs.CL
Medical Dialogue Generation serves a critical role in telemedicine by facilitating the dissemination of medical expertise to patients. Existing studies focus on incorporating textual representations, which have limited their ability to represent the semantics of text, such as ignoring important medical entities. To enh...
http://arxiv.org/pdf/2309.10608v1
2023-09-19T13:31:49
2023-09-19T13:31:49
{ "title": [ 0.000033438206, -0.00299263, -0.037750244, -0.06921387, 0.0034255981, 0.0035190582, -0.064208984, 0.0016593933, 0.0032424927, 0.041046143, 0.018844604, 0.0099487305, -0.008720398, -0.029251099, -0.049346924, 0.018753052, -0.012428284, ...
http://arxiv.org/abs/2309.10604v1
FRACAS: A FRench Annotated Corpus of Attribution relations in newS
Ange Richard, Laura Alonzo-Canul, François Portet
cs.CL
Quotation extraction is a widely useful task both from a sociological and from a Natural Language Processing perspective. However, very little data is available to study this task in languages other than English. In this paper, we present a manually annotated corpus of 1676 newswire texts in French for quotation extrac...
http://arxiv.org/pdf/2309.10604v1
2023-09-19T13:19:54
2023-09-19T13:19:54
{ "title": [ -0.021499634, -0.021453857, -0.047058105, -0.051483154, -0.015686035, 0.0029201508, -0.030303955, -0.0010223389, -0.014587402, 0.032196045, -0.031143188, 0.03012085, 0.0072746277, -0.010932922, -0.022338867, -0.01335144, -0.026885986, ...
http://arxiv.org/abs/2309.10598v1
Unsupervised Deep Cross-Language Entity Alignment
Chuanyu Jiang, Yiming Qian, Lijun Chen, Yang Gu, Xia Xie
cs.CL, cs.LG
Cross-lingual entity alignment is the task of finding the same semantic entities from different language knowledge graphs. In this paper, we propose a simple and novel unsupervised method for cross-language entity alignment. We utilize the deep learning multi-language encoder combined with a machine translator to encod...
http://arxiv.org/pdf/2309.10598v1
2023-09-19T13:12:48
2023-09-19T13:12:48
{ "title": [ -0.022415161, 0.0015144348, -0.044830322, -0.080078125, 0.02835083, 0.005317688, -0.060455322, -0.0037994385, 0.038024902, 0.036865234, -0.0020942688, 0.0000041127205, 0.044555664, -0.015357971, -0.026351929, 0.03829956, -0.0076789856, ...
http://arxiv.org/abs/2309.10567v1
Multimodal Modeling For Spoken Language Identification
Shikhar Bharadwaj, Min Ma, Shikhar Vashishth, Ankur Bapna, Sriram Ganapathy, Vera Axelrod, Siddharth Dalmia, Wei Han, Yu Zhang, Daan van Esch, Sandy Ritchie, Partha Talukdar, Jason Riesa
cs.CL, cs.LG, cs.SD, eess.AS
Spoken language identification refers to the task of automatically predicting the spoken language in a given utterance. Conventionally, it is modeled as a speech-based language identification task. Prior techniques have been constrained to a single modality; however in the case of video data there is a wealth of other ...
http://arxiv.org/pdf/2309.10567v1
2023-09-19T12:21:39
2023-09-19T12:21:39
{ "title": [ -0.023422241, 0.02279663, -0.05770874, -0.047821045, 0.008331299, -0.0129470825, -0.06549072, -0.015060425, -0.016235352, 0.03289795, 0.0021762848, -0.013023376, 0.00059747696, -0.0340271, -0.044403076, -0.012748718, -0.015792847, 0....
http://arxiv.org/abs/2309.10551v1
A Neighbourhood-Aware Differential Privacy Mechanism for Static Word Embeddings
Danushka Bollegala, Shuichi Otake, Tomoya Machide, Ken-ichi Kawarabayashi
cs.LG, cs.AI, cs.CL, cs.CR
We propose a Neighbourhood-Aware Differential Privacy (NADP) mechanism considering the neighbourhood of a word in a pretrained static word embedding space to determine the minimal amount of noise required to guarantee a specified privacy level. We first construct a nearest neighbour graph over the words using their emb...
http://arxiv.org/pdf/2309.10551v1
2023-09-19T11:58:08
2023-09-19T11:58:08
{ "title": [ -0.00037765503, -0.0143585205, -0.028762817, -0.05496216, 0.011688232, -0.0037269592, -0.0063667297, 0.014907837, 0.0012378693, 0.013153076, -0.02571106, 0.025283813, 0.008773804, -0.012680054, -0.00030875206, 0.010169983, 0.0027008057, ...
http://arxiv.org/abs/2309.10544v1
Model Leeching: An Extraction Attack Targeting LLMs
Lewis Birch, William Hackett, Stefan Trawicki, Neeraj Suri, Peter Garraghan
cs.LG, cs.AI, cs.CL, cs.CR
Model Leeching is a novel extraction attack targeting Large Language Models (LLMs), capable of distilling task-specific knowledge from a target LLM into a reduced parameter model. We demonstrate the effectiveness of our attack by extracting task capability from ChatGPT-3.5-Turbo, achieving 73% Exact Match (EM) similari...
http://arxiv.org/pdf/2309.10544v1
2023-09-19T11:45:29
2023-09-19T11:45:29
{ "title": [ -0.035461426, 0.007873535, -0.041381836, -0.021408081, -0.020965576, -0.009407043, -0.0357666, -0.015487671, -0.0435791, 0.035583496, -0.0034942627, 0.033203125, -0.031311035, -0.032928467, -0.017959595, -0.009841919, -0.0027503967, ...
http://arxiv.org/abs/2309.10539v1
OpenMSD: Towards Multilingual Scientific Documents Similarity Measurement
Yang Gao, Ji Ma, Ivan Korotkov, Keith Hall, Dana Alon, Don Metzler
cs.CL, cs.AI
We develop and evaluate multilingual scientific documents similarity measurement models in this work. Such models can be used to find related works in different languages, which can help multilingual researchers find and explore papers more efficiently. We propose the first multilingual scientific documents dataset, Op...
http://arxiv.org/pdf/2309.10539v1
2023-09-19T11:38:39
2023-09-19T11:38:39
{ "title": [ -0.057556152, -0.010002136, -0.03439331, -0.049560547, -0.00044965744, 0.025848389, -0.05368042, -0.020767212, 0.015350342, 0.057678223, 0.018493652, -0.0005197525, 0.019836426, -0.017532349, -0.045288086, 0.024658203, -0.0036773682, ...
http://arxiv.org/abs/2309.10526v1
NSOAMT -- New Search Only Approach to Machine Translation
João Luís, Diogo Cardoso, José Marques, Luís Campos
cs.CL
Translation automation mechanisms and tools have been developed for several years to bring people who speak different languages together. A "new search only approach to machine translation" was adopted to tackle some of the slowness and inaccuracy of the other technologies. The idea is to develop a solution that, by in...
http://arxiv.org/pdf/2309.10526v1
2023-09-19T11:12:21
2023-09-19T11:12:21
{ "title": [ -0.021911621, -0.029373169, -0.0005707741, -0.039398193, -0.036071777, 0.0115737915, -0.024353027, 0.0037727356, -0.0028934479, 0.052337646, -0.0076828003, 0.0049972534, 0.043884277, -0.037994385, -0.02458191, 0.025802612, -0.002702713, ...
http://arxiv.org/abs/2309.10524v1
Harnessing the Zero-Shot Power of Instruction-Tuned Large Language Model in End-to-End Speech Recognition
Yosuke Higuchi, Tetsuji Ogawa, Tetsunori Kobayashi
eess.AS, cs.CL, cs.SD
We present a novel integration of an instruction-tuned large language model (LLM) and end-to-end automatic speech recognition (ASR). Modern LLMs can perform a wide range of linguistic tasks within zero-shot learning when provided with a precise instruction or a prompt to guide the text generation process towards the de...
http://arxiv.org/pdf/2309.10524v1
2023-09-19T11:10:50
2023-09-19T11:10:50
{ "title": [ -0.013938904, -0.017593384, -0.037384033, -0.08557129, -0.010002136, -0.0007367134, -0.062194824, 0.033294678, -0.011421204, 0.027923584, -0.016601562, -0.0010871887, -0.0057411194, -0.016799927, -0.036468506, -0.004611969, -0.0024414062...
http://arxiv.org/abs/2309.10506v1
Enhancing Open-Domain Table Question Answering via Syntax- and Structure-aware Dense Retrieval
Nengzheng Jin, Dongfang Li, Junying Chen, Joanna Siebert, Qingcai Chen
cs.CL
Open-domain table question answering aims to provide answers to a question by retrieving and extracting information from a large collection of tables. Existing studies of open-domain table QA either directly adopt text retrieval methods or consider the table structure only in the encoding layer for table retrieval, whi...
http://arxiv.org/pdf/2309.10506v1
2023-09-19T10:40:09
2023-09-19T10:40:09
{ "title": [ 0.018585205, -0.004135132, -0.01777649, -0.07727051, 0.0015859604, 0.0077171326, -0.041503906, 0.0037517548, 0.027175903, 0.039855957, -0.029205322, 0.0046081543, 0.033325195, 0.005630493, -0.057891846, 0.0038490295, 0.010192871, -0....
http://arxiv.org/abs/2309.10492v1
An Evaluation of GPT-4 on the ETHICS Dataset
Sergey Rodionov, Zarathustra Amadeus Goertzel, Ben Goertzel
cs.CL, cs.AI
This report summarizes a short study of the performance of GPT-4 on the ETHICS dataset. The ETHICS dataset consists of five sub-datasets covering different fields of ethics: Justice, Deontology, Virtue Ethics, Utilitarianism, and Commonsense Ethics. The moral judgments were curated so as to have a high degree of agreem...
http://arxiv.org/pdf/2309.10492v1
2023-09-19T10:01:50
2023-09-19T10:01:50
{ "title": [ 0.016326904, 0.019683838, -0.03338623, -0.026000977, 0.031204224, -0.03894043, -0.00083589554, -0.030792236, -0.018173218, 0.019104004, -0.02330017, 0.0030975342, -0.02859497, -0.026168823, -0.028045654, 0.0024414062, -0.03036499, 0....
http://arxiv.org/abs/2309.10456v1
Improving Speaker Diarization using Semantic Information: Joint Pairwise Constraints Propagation
Luyao Cheng, Siqi Zheng, Qinglin Zhang, Hui Wang, Yafeng Chen, Qian Chen, Shiliang Zhang
cs.SD, cs.CL, eess.AS
Speaker diarization has gained considerable attention within speech processing research community. Mainstream speaker diarization rely primarily on speakers' voice characteristics extracted from acoustic signals and often overlook the potential of semantic information. Considering the fact that speech signals can effic...
http://arxiv.org/pdf/2309.10456v1
2023-09-19T09:13:30
2023-09-19T09:13:30
{ "title": [ -0.010108948, 0.005760193, -0.027175903, -0.038757324, -0.020309448, 0.001294136, -0.054107666, -0.019683838, -0.024536133, 0.04562378, -0.022613525, 0.0015211105, 0.02746582, -0.018997192, -0.0016040802, -0.0055389404, -0.0073432922, ...
http://arxiv.org/abs/2309.10447v2
Toward Unified Controllable Text Generation via Regular Expression Instruction
Xin Zheng, Hongyu Lin, Xianpei Han, Le Sun
cs.CL, cs.AI
Controllable text generation is a fundamental aspect of natural language generation, with numerous methods proposed for different constraint types. However, these approaches often require significant architectural or decoding modifications, making them challenging to apply to additional constraints or resolve different...
http://arxiv.org/pdf/2309.10447v2
2023-09-20T02:18:06
2023-09-19T09:05:14
{ "title": [ -0.037475586, -0.023284912, -0.04537964, -0.06951904, -0.011665344, 0.01146698, -0.06161499, 0.021942139, -0.023971558, 0.038726807, -0.006626129, 0.025115967, -0.0026416779, -0.010238647, -0.03265381, 0.0046577454, -0.014633179, 0.0...
http://arxiv.org/abs/2309.10444v2
Exploring Self-Reinforcement for Improving Learnersourced Multiple-Choice Question Explanations with Large Language Models
Qiming Bao, Juho Leinonen, Alex Yuxuan Peng, Wanjun Zhong, Tim Pistotti, Alice Huang, Paul Denny, Michael Witbrock, Jiamou Liu
cs.AI, cs.CL
Learnersourcing involves students generating and sharing learning resources with their peers. When learnersourcing multiple-choice questions, creating explanations for the generated questions is a crucial step as it facilitates a deeper understanding of the related concepts. However, it is often difficult for students ...
http://arxiv.org/pdf/2309.10444v2
2023-10-19T11:01:19
2023-09-19T09:04:15
{ "title": [ 0.0015592575, -0.005455017, -0.0031051636, -0.028625488, -0.018463135, -0.01146698, -0.006454468, -0.008911133, -0.01776123, 0.009429932, 0.0054893494, 0.0090408325, -0.047210693, -0.016906738, -0.031280518, -0.02079773, -0.016296387, ...
http://arxiv.org/abs/2309.10435v1
Reformulating Sequential Recommendation: Learning Dynamic User Interest with Content-enriched Language Modeling
Junzhe Jiang, Shang Qu, Mingyue Cheng, Qi Liu
cs.IR, cs.CL
Recommender systems are essential for online applications, and sequential recommendation has enjoyed significant prevalence due to its expressive ability to capture dynamic user interests. However, previous sequential modeling methods still have limitations in capturing contextual information. The primary reason for th...
http://arxiv.org/pdf/2309.10435v1
2023-09-19T08:54:47
2023-09-19T08:54:47
{ "title": [ 0.01977539, -0.0023899078, -0.029373169, -0.05569458, 0.007904053, -0.012123108, -0.006412506, -0.011528015, -0.0087509155, 0.033325195, -0.011413574, 0.019958496, -0.019332886, -0.019088745, -0.033813477, 0.007335663, -0.012039185, ...
http://arxiv.org/abs/2309.10433v1
Writer-Defined AI Personas for On-Demand Feedback Generation
Karim Benharrak, Tim Zindulka, Florian Lehmann, Hendrik Heuer, Daniel Buschek
cs.HC, cs.CL, H.5.2; I.2.7
Compelling writing is tailored to its audience. This is challenging, as writers may struggle to empathize with readers, get feedback in time, or gain access to the target group. We propose a concept that generates on-demand feedback, based on writer-defined AI personas of any target audience. We explore this concept wi...
http://arxiv.org/pdf/2309.10433v1
2023-09-19T08:49:35
2023-09-19T08:49:35
{ "title": [ -0.028640747, -0.01637268, -0.023864746, -0.03591919, -0.033325195, 0.0040283203, -0.017913818, -0.0022068024, -0.0037879944, 0.032836914, -0.039398193, 0.014060974, 0.009254456, -0.007633209, -0.02696228, 0.0013074875, -0.00422287, ...
http://arxiv.org/abs/2309.10413v1
PICK: Polished & Informed Candidate Scoring for Knowledge-Grounded Dialogue Systems
Bryan Wilie, Yan Xu, Willy Chung, Samuel Cahyawijaya, Holy Lovenia, Pascale Fung
cs.CL
Grounding dialogue response generation on external knowledge is proposed to produce informative and engaging responses. However, current knowledge-grounded dialogue (KGD) systems often fail to align the generated responses with human-preferred qualities due to several issues like hallucination and the lack of coherence...
http://arxiv.org/pdf/2309.10413v1
2023-09-19T08:27:09
2023-09-19T08:27:09
{ "title": [ -0.017852783, -0.00006276369, -0.047058105, -0.04949951, 0.011619568, -0.014045715, -0.022750854, -0.0014543533, 0.010063171, -0.00041675568, -0.021072388, -0.007045746, -0.010101318, -0.035064697, -0.03994751, -0.011306763, -0.018600464...
http://arxiv.org/abs/2309.10400v2
PoSE: Efficient Context Window Extension of LLMs via Positional Skip-wise Training
Dawei Zhu, Nan Yang, Liang Wang, Yifan Song, Wenhao Wu, Furu Wei, Sujian Li
cs.CL, cs.LG
Large Language Models (LLMs) are trained with a pre-defined context length, restricting their use in scenarios requiring long inputs. Previous efforts for adapting LLMs to a longer length usually requires fine-tuning with this target length (Full-length fine-tuning), suffering intensive training cost. To decouple train...
http://arxiv.org/pdf/2309.10400v2
2023-10-10T07:51:31
2023-09-19T08:03:38
{ "title": [ -0.028518677, -0.018569946, -0.038238525, -0.053894043, -0.0073547363, -0.002532959, -0.041229248, -0.014556885, 0.00015830994, 0.03375244, -0.050354004, 0.01727295, -0.0019512177, -0.014213562, -0.020492554, -0.019683838, -0.008270264, ...
http://arxiv.org/abs/2309.10359v1
Prompt, Condition, and Generate: Classification of Unsupported Claims with In-Context Learning
Peter Ebert Christensen, Srishti Yadav, Serge Belongie
cs.CL
Unsupported and unfalsifiable claims we encounter in our daily lives can influence our view of the world. Characterizing, summarizing, and -- more generally -- making sense of such claims, however, can be challenging. In this work, we focus on fine-grained debate topics and formulate a new task of distilling, from such...
http://arxiv.org/pdf/2309.10359v1
2023-09-19T06:42:37
2023-09-19T06:42:37
{ "title": [ 0.0423584, 0.016021729, -0.031799316, -0.06085205, 0.02772522, -0.031982422, 0.000114262104, -0.024993896, -0.016952515, -0.0020751953, -0.071777344, 0.004131317, -0.014099121, -0.029891968, 0.004524231, -0.04849243, -0.0121536255, 0...
http://arxiv.org/abs/2309.10346v1
Explaining Agent Behavior with Large Language Models
Xijia Zhang, Yue Guo, Simon Stepputtis, Katia Sycara, Joseph Campbell
cs.LG, cs.AI, cs.CL
Intelligent agents such as robots are increasingly deployed in real-world, safety-critical settings. It is vital that these agents are able to explain the reasoning behind their decisions to human counterparts, however, their behavior is often produced by uninterpretable models such as deep neural networks. We propose ...
http://arxiv.org/pdf/2309.10346v1
2023-09-19T06:13:24
2023-09-19T06:13:24
{ "title": [ -0.026916504, -0.012756348, -0.021636963, -0.055877686, -0.00014090538, 0.0055732727, -0.044433594, -0.009429932, -0.0076675415, 0.010314941, -0.016799927, -0.008926392, -0.018463135, -0.014602661, -0.02432251, -0.007987976, -0.016448975...
http://arxiv.org/abs/2309.13072v1
Weakly Supervised Reasoning by Neuro-Symbolic Approaches
Xianggen Liu, Zhengdong Lu, Lili Mou
cs.CL, cs.AI, cs.LG
Deep learning has largely improved the performance of various natural language processing (NLP) tasks. However, most deep learning models are black-box machinery, and lack explicit interpretation. In this chapter, we will introduce our recent progress on neuro-symbolic approaches to NLP, which combines different school...
http://arxiv.org/pdf/2309.13072v1
2023-09-19T06:10:51
2023-09-19T06:10:51
{ "title": [ -0.015052795, -0.004562378, 0.0041046143, -0.039520264, 0.0036621094, 0.017669678, -0.013320923, -0.017211914, -0.010650635, 0.022262573, -0.020950317, -0.013679504, 0.0015220642, -0.02746582, -0.0345459, -0.012939453, 0.0047416687, ...
http://arxiv.org/abs/2309.10339v1
KoBigBird-large: Transformation of Transformer for Korean Language Understanding
Kisu Yang, Yoonna Jang, Taewoo Lee, Jinwoo Seong, Hyungjin Lee, Hwanseok Jang, Heuiseok Lim
cs.CL
This work presents KoBigBird-large, a large size of Korean BigBird that achieves state-of-the-art performance and allows long sequence processing for Korean language understanding. Without further pretraining, we only transform the architecture and extend the positional encoding with our proposed Tapered Absolute Posit...
http://arxiv.org/pdf/2309.10339v1
2023-09-19T05:48:57
2023-09-19T05:48:57
{ "title": [ -0.023345947, -0.025039673, -0.021026611, -0.049560547, -0.035003662, 0.010658264, -0.07269287, -0.0135650635, -0.014862061, 0.04208374, -0.029312134, -0.00712204, 0.06347656, -0.031707764, -0.016036987, 0.002117157, -0.0052871704, 0...
http://arxiv.org/abs/2309.10326v2
QASnowball: An Iterative Bootstrapping Framework for High-Quality Question-Answering Data Generation
Kunlun Zhu, Shihao Liang, Xu Han, Zhi Zheng, Guoyang Zeng, Zhiyuan Liu, Maosong Sun
cs.CL, cs.AI
Recent years have witnessed the success of question answering (QA), especially its potential to be a foundation paradigm for tackling diverse NLP tasks. However, obtaining sufficient data to build an effective and stable QA system still remains an open problem. For this problem, we introduce an iterative bootstrapping ...
http://arxiv.org/pdf/2309.10326v2
2023-09-20T01:57:10
2023-09-19T05:20:36
{ "title": [ -0.016098022, 0.00051546097, -0.0501709, -0.05404663, 0.0066566467, -0.005332947, -0.039520264, -0.0055770874, 0.020187378, 0.012084961, -0.02330017, 0.010185242, -0.024276733, -0.008895874, -0.06100464, 0.00036025047, -0.00340271, 0...
http://arxiv.org/abs/2309.10313v3
Investigating the Catastrophic Forgetting in Multimodal Large Language Models
Yuexiang Zhai, Shengbang Tong, Xiao Li, Mu Cai, Qing Qu, Yong Jae Lee, Yi Ma
cs.CL, cs.AI, cs.LG
Following the success of GPT4, there has been a surge in interest in multimodal large language model (MLLM) research. This line of research focuses on developing general-purpose LLMs through fine-tuning pre-trained LLMs and vision models. However, catastrophic forgetting, a notorious phenomenon where the fine-tuned mod...
http://arxiv.org/pdf/2309.10313v3
2023-10-03T20:18:39
2023-09-19T04:51:13
{ "title": [ -0.03857422, 0.009162903, -0.05206299, -0.06524658, -0.0030517578, -0.003396988, -0.06036377, 0.01676941, 0.015655518, 0.051635742, -0.018936157, -0.0022773743, 0.032409668, -0.0058937073, -0.05038452, 0.010726929, -0.012664795, 0.03...