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/2310.05597v3
Can language models learn analogical reasoning? Investigating training objectives and comparisons to human performance
Molly R. Petersen, Lonneke van der Plas
cs.CL
While analogies are a common way to evaluate word embeddings in NLP, it is also of interest to investigate whether or not analogical reasoning is a task in itself that can be learned. In this paper, we test several ways to learn basic analogical reasoning, specifically focusing on analogies that are more typical of wha...
http://arxiv.org/pdf/2310.05597v3
2023-10-22T09:17:30
2023-10-09T10:34:38
{ "title": [ -0.018096924, -0.008430481, -0.0030231476, -0.010192871, -0.037109375, -0.012565613, -0.030853271, -0.024093628, -0.00023674965, 0.03387451, -0.012336731, 0.0024585724, 0.031341553, -0.020507812, -0.034942627, 0.010375977, -0.021606445, ...
http://arxiv.org/abs/2310.05592v2
InterroLang: Exploring NLP Models and Datasets through Dialogue-based Explanations
Nils Feldhus, Qianli Wang, Tatiana Anikina, Sahil Chopra, Cennet Oguz, Sebastian Möller
cs.CL, cs.AI, cs.HC
While recently developed NLP explainability methods let us open the black box in various ways (Madsen et al., 2022), a missing ingredient in this endeavor is an interactive tool offering a conversational interface. Such a dialogue system can help users explore datasets and models with explanations in a contextualized m...
http://arxiv.org/pdf/2310.05592v2
2023-10-23T14:01:26
2023-10-09T10:27:26
{ "title": [ -0.032806396, -0.02748108, -0.04208374, -0.07519531, -0.013519287, 0.0031051636, -0.06274414, 0.010673523, 0.0010099411, -0.014595032, -0.02545166, 0.00831604, -0.043548584, -0.01763916, -0.030014038, -0.026855469, -0.022994995, 0.05...
http://arxiv.org/abs/2310.05589v1
DRIN: Dynamic Relation Interactive Network for Multimodal Entity Linking
Shangyu Xing, Fei Zhao, Zhen Wu, Chunhui Li, Jianbing Zhang, Xinyu Dai
cs.CL, cs.MM
Multimodal Entity Linking (MEL) is a task that aims to link ambiguous mentions within multimodal contexts to referential entities in a multimodal knowledge base. Recent methods for MEL adopt a common framework: they first interact and fuse the text and image to obtain representations of the mention and entity respectiv...
http://arxiv.org/pdf/2310.05589v1
2023-10-09T10:21:42
2023-10-09T10:21:42
{ "title": [ -0.0008802414, 0.0058288574, -0.060180664, -0.08557129, 0.02720642, -0.00605011, -0.037200928, 0.00063085556, -0.0058174133, 0.022277832, 0.015197754, -0.0035972595, 0.004261017, -0.032043457, -0.019729614, 0.007789612, -0.01663208, ...
http://arxiv.org/abs/2310.05553v1
Regulation and NLP (RegNLP): Taming Large Language Models
Catalina Goanta, Nikolaos Aletras, Ilias Chalkidis, Sofia Ranchordas, Gerasimos Spanakis
cs.CL
The scientific innovation in Natural Language Processing (NLP) and more broadly in artificial intelligence (AI) is at its fastest pace to date. As large language models (LLMs) unleash a new era of automation, important debates emerge regarding the benefits and risks of their development, deployment and use. Currently, ...
http://arxiv.org/pdf/2310.05553v1
2023-10-09T09:22:40
2023-10-09T09:22:40
{ "title": [ -0.029785156, -0.024276733, -0.048980713, -0.048583984, -0.005077362, -0.0064735413, -0.036346436, -0.0053138733, -0.014976501, 0.013168335, -0.023254395, 0.008865356, -0.0020313263, -0.021240234, -0.02355957, 0.008773804, -0.023010254, ...
http://arxiv.org/abs/2310.05513v1
Findings of the 2023 ML-SUPERB Challenge: Pre-Training and Evaluation over More Languages and Beyond
Jiatong Shi, William Chen, Dan Berrebbi, Hsiu-Hsuan Wang, Wei-Ping Huang, En-Pei Hu, Ho-Lam Chuang, Xuankai Chang, Yuxun Tang, Shang-Wen Li, Abdelrahman Mohamed, Hung-yi Lee, Shinji Watanabe
cs.SD, cs.CL, eess.AS
The 2023 Multilingual Speech Universal Performance Benchmark (ML-SUPERB) Challenge expands upon the acclaimed SUPERB framework, emphasizing self-supervised models in multilingual speech recognition and language identification. The challenge comprises a research track focused on applying ML-SUPERB to specific multilingu...
http://arxiv.org/pdf/2310.05513v1
2023-10-09T08:30:01
2023-10-09T08:30:01
{ "title": [ 0.006351471, -0.022888184, 0.006061554, -0.048034668, -0.025268555, 0.008224487, -0.048431396, -0.0030765533, 0.00079631805, 0.011306763, 0.005695343, 0.016540527, 0.010444641, -0.059936523, 0.038238525, 0.015151978, 0.004196167, 0.0...
http://arxiv.org/abs/2310.05506v2
Query and Response Augmentation Cannot Help Out-of-domain Math Reasoning Generalization
Chengpeng Li, Zheng Yuan, Hongyi Yuan, Guanting Dong, Keming Lu, Jiancan Wu, Chuanqi Tan, Xiang Wang, Chang Zhou
cs.CL, cs.AI, cs.LG
In math reasoning with large language models (LLMs), fine-tuning data augmentation by query evolution and diverse reasoning paths is empirically verified effective, profoundly narrowing the gap between open-sourced LLMs and cutting-edge proprietary LLMs. In this paper, we conduct an investigation for such data augmenta...
http://arxiv.org/pdf/2310.05506v2
2023-11-01T08:41:02
2023-10-09T08:18:58
{ "title": [ 0.045440674, -0.0003581047, -0.0141067505, -0.036376953, -0.0022468567, -0.025115967, 0.053253174, 0.004787445, -0.009841919, 0.031433105, -0.046569824, 0.013328552, -0.0072402954, -0.0118637085, -0.013206482, -0.02583313, 0.010025024, ...
http://arxiv.org/abs/2310.05502v1
XAL: EXplainable Active Learning Makes Classifiers Better Low-resource Learners
Yun Luo, Zhen Yang, Fandong Meng, Yingjie Li, Fang Guo, Qinglin Qi, Jie Zhou, Yue Zhang
cs.CL
Active learning aims to construct an effective training set by iteratively curating the most informative unlabeled data for annotation, which is practical in low-resource tasks. Most active learning techniques in classification rely on the model's uncertainty or disagreement to choose unlabeled data. However, previous ...
http://arxiv.org/pdf/2310.05502v1
2023-10-09T08:07:04
2023-10-09T08:07:04
{ "title": [ -0.010017395, 0.011833191, -0.023208618, -0.052215576, 0.0069770813, -0.012611389, -0.03955078, -0.04046631, -0.014511108, 0.0012760162, -0.013771057, 0.02017212, -0.02154541, -0.042541504, 0.014060974, 0.00548172, -0.010604858, 0.04...
http://arxiv.org/abs/2310.05492v2
How Abilities in Large Language Models are Affected by Supervised Fine-tuning Data Composition
Guanting Dong, Hongyi Yuan, Keming Lu, Chengpeng Li, Mingfeng Xue, Dayiheng Liu, Wei Wang, Zheng Yuan, Chang Zhou, Jingren Zhou
cs.CL, cs.AI, cs.LG
Large language models (LLMs) with enormous pre-training tokens and parameter amounts emerge abilities, including math reasoning, code generation, and instruction following. These abilities are further enhanced by supervised fine-tuning (SFT). The open-source community has studied on ad-hoc SFT for each ability, while p...
http://arxiv.org/pdf/2310.05492v2
2023-11-01T07:11:37
2023-10-09T07:56:16
{ "title": [ -0.0088272095, -0.0041046143, -0.04324341, -0.076293945, -0.013946533, -0.009284973, -0.059387207, 0.022735596, -0.0055389404, 0.03564453, -0.02255249, -0.00091314316, -0.00042271614, -0.004787445, -0.054840088, -0.006160736, -0.02751159...
http://arxiv.org/abs/2310.05484v1
IDTraffickers: An Authorship Attribution Dataset to link and connect Potential Human-Trafficking Operations on Text Escort Advertisements
Vageesh Saxena, Benjamin Bashpole, Gijs Van Dijck, Gerasimos Spanakis
cs.CL, cs.CY, cs.LG
Human trafficking (HT) is a pervasive global issue affecting vulnerable individuals, violating their fundamental human rights. Investigations reveal that a significant number of HT cases are associated with online advertisements (ads), particularly in escort markets. Consequently, identifying and connecting HT vendors ...
http://arxiv.org/pdf/2310.05484v1
2023-10-09T07:43:57
2023-10-09T07:43:57
{ "title": [ 0.0211792, 0.014785767, -0.008712769, -0.006706238, 0.016983032, -0.012641907, 0.0042915344, -0.043182373, -0.015457153, 0.012290955, -0.024047852, 0.03048706, -0.018310547, -0.016448975, 0.0012054443, -0.022827148, -0.031311035, 0.0...
http://arxiv.org/abs/2310.05481v1
Cabbage Sweeter than Cake? Analysing the Potential of Large Language Models for Learning Conceptual Spaces
Usashi Chatterjee, Amit Gajbhiye, Steven Schockaert
cs.CL, cs.AI
The theory of Conceptual Spaces is an influential cognitive-linguistic framework for representing the meaning of concepts. Conceptual spaces are constructed from a set of quality dimensions, which essentially correspond to primitive perceptual features (e.g. hue or size). These quality dimensions are usually learned fr...
http://arxiv.org/pdf/2310.05481v1
2023-10-09T07:41:19
2023-10-09T07:41:19
{ "title": [ -0.019241333, -0.013313293, -0.028320312, -0.040374756, -0.017425537, -0.015914917, -0.041992188, -0.01133728, 0.013549805, 0.025604248, -0.03564453, 0.004337311, -0.002407074, -0.012802124, -0.044708252, -0.009246826, -0.01789856, 0...
http://arxiv.org/abs/2310.05470v1
Generative Judge for Evaluating Alignment
Junlong Li, Shichao Sun, Weizhe Yuan, Run-Ze Fan, Hai Zhao, Pengfei Liu
cs.CL, cs.AI
The rapid development of Large Language Models (LLMs) has substantially expanded the range of tasks they can address. In the field of Natural Language Processing (NLP), researchers have shifted their focus from conventional NLP tasks (e.g., sequence tagging and parsing) towards tasks that revolve around aligning with h...
http://arxiv.org/pdf/2310.05470v1
2023-10-09T07:27:15
2023-10-09T07:27:15
{ "title": [ -0.020843506, -0.006462097, -0.04309082, -0.052490234, -0.023284912, -0.0046958923, -0.0579834, -0.01586914, 0.037597656, 0.047027588, -0.013244629, -0.021743774, 0.0065956116, -0.032043457, -0.056610107, 0.013969421, 0.003982544, 0....
http://arxiv.org/abs/2310.05452v1
Explaining the Complex Task Reasoning of Large Language Models with Template-Content Structure
Haotong Yang, Fanxu Meng, Zhouchen Lin, Muhan Zhang
cs.AI, cs.CL
The continuous evolution of pre-trained large language models with ever-growing parameters and corpus sizes has augmented their capacity to solve complex tasks. This ability, which obviates the necessity for task-specific training or fine-tuning, relies on providing the model with a language description or some task ex...
http://arxiv.org/pdf/2310.05452v1
2023-10-09T06:57:45
2023-10-09T06:57:45
{ "title": [ -0.052703857, -0.019866943, -0.037597656, -0.07824707, -0.01828003, 0.02116394, -0.08093262, 0.02243042, -0.003435135, 0.02330017, -0.0061035156, -0.009887695, 0.003162384, -0.030212402, -0.019302368, -0.002254486, -0.009338379, 0.00...
http://arxiv.org/abs/2310.05450v1
Empower Nested Boolean Logic via Self-Supervised Curriculum Learning
Hongqiu Wu, Linfeng Liu, Hai Zhao, Min Zhang
cs.CL
Beyond the great cognitive powers showcased by language models, it is crucial to scrutinize whether their reasoning capabilities stem from strong generalization or merely exposure to relevant data. As opposed to constructing increasingly complex logic, this paper probes into the boolean logic, the root capability of a ...
http://arxiv.org/pdf/2310.05450v1
2023-10-09T06:54:02
2023-10-09T06:54:02
{ "title": [ -0.02079773, -0.005783081, -0.06402588, -0.0848999, 0.015029907, -0.021011353, -0.032318115, 0.036865234, -0.018112183, -0.0076904297, -0.010101318, 0.018325806, -0.053100586, -0.024505615, 0.01209259, -0.016815186, -0.016937256, 0.0...
http://arxiv.org/abs/2310.05442v2
Establishing Trustworthiness: Rethinking Tasks and Model Evaluation
Robert Litschko, Max Müller-Eberstein, Rob van der Goot, Leon Weber, Barbara Plank
cs.CL
Language understanding is a multi-faceted cognitive capability, which the Natural Language Processing (NLP) community has striven to model computationally for decades. Traditionally, facets of linguistic intelligence have been compartmentalized into tasks with specialized model architectures and corresponding evaluatio...
http://arxiv.org/pdf/2310.05442v2
2023-10-23T14:43:40
2023-10-09T06:32:10
{ "title": [ -0.026046753, 0.0021362305, -0.059692383, -0.03894043, 0.026977539, -0.008377075, -0.01348114, -0.021759033, -0.020599365, 0.011024475, 0.0105896, -0.0033817291, -0.037841797, -0.03744507, -0.038116455, -0.0038852692, -0.021102905, 0...
http://arxiv.org/abs/2310.05424v1
Fast and Robust Early-Exiting Framework for Autoregressive Language Models with Synchronized Parallel Decoding
Sangmin Bae, Jongwoo Ko, Hwanjun Song, Se-Young Yun
cs.CL
To tackle the high inference latency exhibited by autoregressive language models, previous studies have proposed an early-exiting framework that allocates adaptive computation paths for each token based on the complexity of generating the subsequent token. However, we observed several shortcomings, including performanc...
http://arxiv.org/pdf/2310.05424v1
2023-10-09T05:53:05
2023-10-09T05:53:05
{ "title": [ -0.043304443, 0.0015268326, -0.03552246, -0.06335449, -0.012756348, 0.006916046, -0.0758667, 0.0007610321, -0.0070991516, 0.07220459, -0.008880615, 0.0059928894, 0.026489258, -0.009567261, -0.06829834, 0.022216797, 0.010612488, 0.029...
http://arxiv.org/abs/2310.05421v1
Automating Customer Service using LangChain: Building custom open-source GPT Chatbot for organizations
Keivalya Pandya, Mehfuza Holia
cs.CL, cs.CY, cs.LG
In the digital age, the dynamics of customer service are evolving, driven by technological advancements and the integration of Large Language Models (LLMs). This research paper introduces a groundbreaking approach to automating customer service using LangChain, a custom LLM tailored for organizations. The paper explore...
http://arxiv.org/pdf/2310.05421v1
2023-10-09T05:35:10
2023-10-09T05:35:10
{ "title": [ 0.0007452965, -0.0046157837, -0.034942627, -0.058013916, -0.0132369995, -0.00039243698, -0.03164673, 0.016647339, -0.037963867, 0.04550171, -0.03692627, 0.004524231, 0.013175964, -0.026641846, -0.0066604614, -0.04525757, 0.013572693, ...
http://arxiv.org/abs/2310.05418v1
Humanoid Agents: Platform for Simulating Human-like Generative Agents
Zhilin Wang, Yu Ying Chiu, Yu Cheung Chiu
cs.CL, cs.AI, cs.HC
Just as computational simulations of atoms, molecules and cells have shaped the way we study the sciences, true-to-life simulations of human-like agents can be valuable tools for studying human behavior. We propose Humanoid Agents, a system that guides Generative Agents to behave more like humans by introducing three e...
http://arxiv.org/pdf/2310.05418v1
2023-10-09T05:30:42
2023-10-09T05:30:42
{ "title": [ -0.023406982, -0.005443573, -0.022354126, -0.03475952, 0.0006375313, -0.000078201294, -0.0060539246, -0.02607727, 0.0014066696, 0.043640137, -0.027297974, 0.016815186, 0.020202637, -0.0077781677, -0.04269409, 0.011375427, -0.017211914, ...
http://arxiv.org/abs/2310.05404v1
mBBC: Exploring the Multilingual Maze
Sina Bagheri Nezhad, Ameeta Agrawal
cs.CL, I.2.7
Multilingual language models have gained significant attention in recent years, enabling the development of applications that cater to diverse linguistic contexts. In this paper, we present a comprehensive evaluation of three prominent multilingual language models: mBERT, XLM-R, and GPT-3. Using the self-supervised tas...
http://arxiv.org/pdf/2310.05404v1
2023-10-09T04:48:14
2023-10-09T04:48:14
{ "title": [ -0.0064697266, -0.03768921, 0.004310608, 0.009979248, -0.07684326, 0.0033817291, -0.043151855, -0.03390503, -0.019439697, 0.05734253, 0.013916016, -0.019119263, 0.042266846, -0.05621338, -0.010734558, -0.0066184998, 0.0011911392, 0.0...
http://arxiv.org/abs/2310.05388v2
GROVE: A Retrieval-augmented Complex Story Generation Framework with A Forest of Evidence
Zhihua Wen, Zhiliang Tian, Wei Wu, Yuxin Yang, Yanqi Shi, Zhen Huang, Dongsheng Li
cs.CL
Conditional story generation is significant in human-machine interaction, particularly in producing stories with complex plots. While Large language models (LLMs) perform well on multiple NLP tasks, including story generation, it is challenging to generate stories with both complex and creative plots. Existing methods ...
http://arxiv.org/pdf/2310.05388v2
2023-10-24T01:37:46
2023-10-09T03:55:55
{ "title": [ -0.021453857, -0.02178955, -0.049102783, -0.071777344, -0.018157959, 0.014083862, -0.04058838, 0.014701843, -0.023468018, 0.028793335, -0.005115509, 0.027069092, -0.015281677, -0.020263672, -0.030807495, 0.001033783, -0.026275635, 0....
http://arxiv.org/abs/2310.05381v1
CCAE: A Corpus of Chinese-based Asian Englishes
Yang Liu, Melissa Xiaohui Qin, Long Wang, Chao Huang
cs.CL, cs.AI, cs.LG, 68T50, I.2.7
Language models have been foundations in various scenarios of NLP applications, but it has not been well applied in language variety studies, even for the most popular language like English. This paper represents one of the few initial efforts to utilize the NLP technology in the paradigm of World Englishes, specifical...
http://arxiv.org/pdf/2310.05381v1
2023-10-09T03:34:15
2023-10-09T03:34:15
{ "title": [ -0.047424316, -0.015579224, -0.064208984, -0.0041389465, -0.009765625, -0.02583313, -0.01285553, 0.006511688, -0.016296387, 0.04458618, -0.0067367554, 0.013633728, 0.050689697, -0.05126953, -0.019836426, 0.029541016, -0.011116028, 0....
http://arxiv.org/abs/2310.05378v2
Transcending the Attention Paradigm: Representation Learning from Geospatial Social Media Data
Nick DiSanto, Anthony Corso, Benjamin Sanders, Gavin Harding
cs.CL, cs.SI
While transformers have pioneered attention-driven architectures as a cornerstone of research, their dependence on explicitly contextual information underscores limitations in their abilities to tacitly learn overarching textual themes. This study investigates social media data as a source of distributed patterns, chal...
http://arxiv.org/pdf/2310.05378v2
2023-10-30T07:56:21
2023-10-09T03:27:05
{ "title": [ 0.037139893, 0.012336731, -0.021591187, -0.03579712, 0.024932861, -0.02709961, 0.008132935, -0.021636963, 0.0093307495, 0.034576416, -0.05014038, -0.0035514832, -0.018920898, 0.0020828247, -0.055664062, -0.0262146, 0.008071899, 0.049...
http://arxiv.org/abs/2310.05374v3
Improving End-to-End Speech Processing by Efficient Text Data Utilization with Latent Synthesis
Jianqiao Lu, Wenyong Huang, Nianzu Zheng, Xingshan Zeng, Yu Ting Yeung, Xiao Chen
cs.CL, cs.LG, cs.SD, eess.AS
Training a high performance end-to-end speech (E2E) processing model requires an enormous amount of labeled speech data, especially in the era of data-centric artificial intelligence. However, labeled speech data are usually scarcer and more expensive for collection, compared to textual data. We propose Latent Synthesi...
http://arxiv.org/pdf/2310.05374v3
2023-10-24T06:48:55
2023-10-09T03:10:49
{ "title": [ -0.010620117, 0.009246826, -0.04949951, -0.087402344, 0.02557373, -0.0025863647, -0.044647217, 0.02468872, -0.003112793, 0.047973633, -0.029525757, 0.032196045, 0.013832092, -0.011756897, -0.0345459, 0.019317627, -0.012466431, 0.0289...
http://arxiv.org/abs/2310.05364v3
Universal Multi-modal Entity Alignment via Iteratively Fusing Modality Similarity Paths
Bolin Zhu, Xiaoze Liu, Xin Mao, Zhuo Chen, Lingbing Guo, Tao Gui, Qi Zhang
cs.CL, cs.AI
The objective of Entity Alignment (EA) is to identify equivalent entity pairs from multiple Knowledge Graphs (KGs) and create a more comprehensive and unified KG. The majority of EA methods have primarily focused on the structural modality of KGs, lacking exploration of multi-modal information. A few multi-modal EA met...
http://arxiv.org/pdf/2310.05364v3
2023-10-13T09:47:08
2023-10-09T02:50:54
{ "title": [ -0.022323608, 0.0135269165, -0.059570312, -0.08465576, 0.026275635, -0.0033435822, -0.068237305, -0.0104904175, 0.03414917, 0.05731201, 0.021957397, 0.012062073, 0.06738281, -0.029251099, -0.024032593, 0.050750732, -0.011947632, 0.02...
http://arxiv.org/abs/2310.05352v1
A Glance is Enough: Extract Target Sentence By Looking at A keyword
Ying Shi, Dong Wang, Lantian Li, Jiqing Han
cs.CL, cs.SD, eess.AS
This paper investigates the possibility of extracting a target sentence from multi-talker speech using only a keyword as input. For example, in social security applications, the keyword might be "help", and the goal is to identify what the person who called for help is articulating while ignoring other speakers. To add...
http://arxiv.org/pdf/2310.05352v1
2023-10-09T02:28:19
2023-10-09T02:28:19
{ "title": [ -0.0037765503, -0.028366089, -0.018341064, 0.002193451, -0.044952393, -0.034606934, -0.009056091, -0.020187378, 0.0066070557, 0.02166748, -0.03302002, 0.017318726, -0.0001078248, -0.0135650635, -0.0154418945, -0.011024475, 0.014419556, ...
http://arxiv.org/abs/2310.05344v1
SteerLM: Attribute Conditioned SFT as an (User-Steerable) Alternative to RLHF
Yi Dong, Zhilin Wang, Makesh Narsimhan Sreedhar, Xianchao Wu, Oleksii Kuchaiev
cs.CL, cs.AI, cs.LG
Model alignment with human preferences is an essential step in making Large Language Models (LLMs) helpful and consistent with human values. It typically consists of supervised fine-tuning (SFT) and reinforcement learning from human feedback (RLHF) stages. However, RLHF faces inherent limitations stemming from a comple...
http://arxiv.org/pdf/2310.05344v1
2023-10-09T02:11:21
2023-10-09T02:11:21
{ "title": [ 0.007587433, -0.00522995, -0.03878784, -0.056854248, -0.027923584, -0.00242424, -0.057006836, 0.0011377335, -0.032470703, 0.03387451, 0.018356323, -0.0234375, 0.039001465, -0.056030273, 0.011734009, -0.0035495758, -0.019927979, 0.022...
http://arxiv.org/abs/2310.05338v1
Negative Object Presence Evaluation (NOPE) to Measure Object Hallucination in Vision-Language Models
Holy Lovenia, Wenliang Dai, Samuel Cahyawijaya, Ziwei Ji, Pascale Fung
cs.CV, cs.CL
Object hallucination poses a significant challenge in vision-language (VL) models, often leading to the generation of nonsensical or unfaithful responses with non-existent objects. However, the absence of a general measurement for evaluating object hallucination in VL models has hindered our understanding and ability t...
http://arxiv.org/pdf/2310.05338v1
2023-10-09T01:52:27
2023-10-09T01:52:27
{ "title": [ 0.016586304, 0.009933472, -0.022735596, -0.029556274, -0.007446289, -0.03591919, -0.0044403076, -0.032470703, 0.030899048, 0.03717041, -0.024551392, -0.0026245117, 0.023208618, -0.030349731, -0.06414795, 0.006969452, -0.003786087, 0....
http://arxiv.org/abs/2310.05318v2
Resolving the Imbalance Issue in Hierarchical Disciplinary Topic Inference via LLM-based Data Augmentation
Xunxin Cai, Meng Xiao, Zhiyuan Ning, Yuanchun Zhou
cs.CL
In addressing the imbalanced issue of data within the realm of Natural Language Processing, text data augmentation methods have emerged as pivotal solutions. This data imbalance is prevalent in the research proposals submitted during the funding application process. Such imbalances, resulting from the varying popularit...
http://arxiv.org/pdf/2310.05318v2
2023-10-15T03:34:40
2023-10-09T00:45:20
{ "title": [ -0.033477783, 0.01725769, -0.08880615, -0.042022705, -0.0042648315, -0.013175964, -0.045410156, -0.01739502, 0.008049011, -0.0035533905, -0.019470215, 0.04699707, -0.020492554, -0.027282715, 0.0042152405, 0.00472641, -0.035003662, 0....
http://arxiv.org/abs/2310.05317v5
Task-Adaptive Tokenization: Enhancing Long-Form Text Generation Efficacy in Mental Health and Beyond
Siyang Liu, Naihao Deng, Sahand Sabour, Yilin Jia, Minlie Huang, Rada Mihalcea
cs.CL, cs.AI, 68, I.2.7
We propose task-adaptive tokenization as a way to adapt the generation pipeline to the specifics of a downstream task and enhance long-form generation in mental health. Inspired by insights from cognitive science, our task-adaptive tokenizer samples variable segmentations from multiple outcomes, with sampling probabili...
http://arxiv.org/pdf/2310.05317v5
2023-11-13T14:35:22
2023-10-09T00:20:59
{ "title": [ -0.024429321, -0.0044174194, -0.052978516, -0.051513672, 0.00349617, -0.0029659271, -0.04937744, -0.0015420914, -0.01083374, 0.033416748, -0.0047721863, 0.015777588, 0.0057258606, -0.026748657, -0.032714844, -0.0024719238, -0.010391235, ...
http://arxiv.org/abs/2310.05295v2
Visual Storytelling with Question-Answer Plans
Danyang Liu, Mirella Lapata, Frank Keller
cs.CL
Visual storytelling aims to generate compelling narratives from image sequences. Existing models often focus on enhancing the representation of the image sequence, e.g., with external knowledge sources or advanced graph structures. Despite recent progress, the stories are often repetitive, illogical, and lacking in det...
http://arxiv.org/pdf/2310.05295v2
2023-10-17T22:43:08
2023-10-08T21:45:34
{ "title": [ 0.022003174, -0.01474762, -0.019500732, -0.04244995, -0.059753418, -0.017593384, -0.0053482056, 0.016326904, -0.020599365, 0.007904053, -0.031585693, 0.008361816, -0.032592773, -0.04147339, 0.008094788, -0.02897644, 0.043273926, -0.0...
http://arxiv.org/abs/2310.05294v1
Hi Guys or Hi Folks? Benchmarking Gender-Neutral Machine Translation with the GeNTE Corpus
Andrea Piergentili, Beatrice Savoldi, Dennis Fucci, Matteo Negri, Luisa Bentivogli
cs.CL
Gender inequality is embedded in our communication practices and perpetuated in translation technologies. This becomes particularly apparent when translating into grammatical gender languages, where machine translation (MT) often defaults to masculine and stereotypical representations by making undue binary gender assu...
http://arxiv.org/pdf/2310.05294v1
2023-10-08T21:44:00
2023-10-08T21:44:00
{ "title": [ -0.057373047, -0.0035991669, -0.034851074, -0.03262329, -0.0077667236, 0.002248764, -0.04156494, 0.008308411, 0.012741089, 0.074279785, -0.011001587, 0.011802673, 0.06542969, -0.0018072128, -0.055725098, 0.033050537, -0.028045654, 0....
http://arxiv.org/abs/2310.05280v5
Are Personalized Stochastic Parrots More Dangerous? Evaluating Persona Biases in Dialogue Systems
Yixin Wan, Jieyu Zhao, Aman Chadha, Nanyun Peng, Kai-Wei Chang
cs.CL, cs.AI
Recent advancements in Large Language Models empower them to follow freeform instructions, including imitating generic or specific demographic personas in conversations. We define generic personas to represent demographic groups, such as "an Asian person", whereas specific personas may take the form of specific popular...
http://arxiv.org/pdf/2310.05280v5
2023-11-02T23:06:36
2023-10-08T21:03:18
{ "title": [ 0.019104004, -0.0018186569, -0.016967773, -0.03970337, 0.0027885437, -0.02017212, -0.004562378, -0.0076904297, -0.019302368, 0.02758789, -0.011756897, 0.030960083, -0.016555786, -0.018569946, -0.024902344, -0.014678955, -0.03338623, ...
http://arxiv.org/abs/2310.05276v1
Enhancing Pre-Trained Language Models with Sentence Position Embeddings for Rhetorical Roles Recognition in Legal Opinions
Anas Belfathi, Nicolas Hernandez, Laura Monceaux
cs.CL
The legal domain is a vast and complex field that involves a considerable amount of text analysis, including laws, legal arguments, and legal opinions. Legal practitioners must analyze these texts to understand legal cases, research legal precedents, and prepare legal documents. The size of legal opinions continues to ...
http://arxiv.org/pdf/2310.05276v1
2023-10-08T20:33:55
2023-10-08T20:33:55
{ "title": [ 0.0028514862, -0.0077400208, -0.030792236, -0.04888916, -0.02619934, -0.0016069412, -0.064941406, 0.010009766, 0.021560669, 0.025894165, -0.02911377, -0.005809784, 0.030517578, -0.0058021545, -0.050445557, -0.0025291443, -0.03378296, ...
http://arxiv.org/abs/2310.05253v2
Explainable Claim Verification via Knowledge-Grounded Reasoning with Large Language Models
Haoran Wang, Kai Shu
cs.CL, cs.AI, cs.LG
Claim verification plays a crucial role in combating misinformation. While existing works on claim verification have shown promising results, a crucial piece of the puzzle that remains unsolved is to understand how to verify claims without relying on human-annotated data, which is expensive to create at a large scale. ...
http://arxiv.org/pdf/2310.05253v2
2023-10-20T02:31:21
2023-10-08T18:04:05
{ "title": [ -0.020462036, 0.009925842, -0.047180176, -0.1060791, 0.023071289, 0.020309448, -0.07543945, 0.009185791, -0.01991272, 0.013626099, -0.009857178, -0.0019779205, 0.006881714, -0.026290894, -0.016418457, -0.013175964, -0.027328491, 0.05...
http://arxiv.org/abs/2310.05242v2
ChatRadio-Valuer: A Chat Large Language Model for Generalizable Radiology Report Generation Based on Multi-institution and Multi-system Data
Tianyang Zhong, Wei Zhao, Yutong Zhang, Yi Pan, Peixin Dong, Zuowei Jiang, Xiaoyan Kui, Youlan Shang, Li Yang, Yaonai Wei, Longtao Yang, Hao Chen, Huan Zhao, Yuxiao Liu, Ning Zhu, Yiwei Li, Yisong Wang, Jiaqi Yao, Jiaqi Wang, Ying Zeng, Lei He, Chao Zheng, Zhixue Zhang, Ming Li, Zhengliang Liu, Haixing Dai, Zihao Wu, L...
cs.CL, cs.AI
Radiology report generation, as a key step in medical image analysis, is critical to the quantitative analysis of clinically informed decision-making levels. However, complex and diverse radiology reports with cross-source heterogeneity pose a huge generalizability challenge to the current methods under massive data vo...
http://arxiv.org/pdf/2310.05242v2
2023-10-10T01:22:16
2023-10-08T17:23:17
{ "title": [ -0.01651001, -0.01864624, -0.03842163, -0.07788086, -0.03048706, 0.02810669, -0.078125, 0.02482605, 0.0124053955, 0.035247803, -0.012054443, -0.006931305, -0.029510498, -0.03189087, -0.045837402, -0.010795593, -0.0039863586, 0.024337...
http://arxiv.org/abs/2310.05235v1
XLS-R fine-tuning on noisy word boundaries for unsupervised speech segmentation into words
Robin Algayres, Pablo Diego-Simon, Benoit Sagot, Emmanuel Dupoux
cs.CL, cs.SD, eess.AS
Due to the absence of explicit word boundaries in the speech stream, the task of segmenting spoken sentences into word units without text supervision is particularly challenging. In this work, we leverage the most recent self-supervised speech models that have proved to quickly adapt to new tasks through fine-tuning, e...
http://arxiv.org/pdf/2310.05235v1
2023-10-08T17:05:00
2023-10-08T17:05:00
{ "title": [ -0.026168823, -0.009498596, -0.030090332, -0.046325684, -0.011543274, 0.0050621033, -0.042755127, 0.010406494, -0.041412354, 0.024719238, -0.04864502, 0.04598999, -0.012458801, -0.007472992, -0.0033283234, -0.010803223, -0.0035209656, ...
http://arxiv.org/abs/2310.05231v1
MindfulDiary: Harnessing Large Language Model to Support Psychiatric Patients' Journaling
Taewan Kim, Seolyeong Bae, Hyun Ah Kim, Su-woo Lee, Hwajung Hong, Chanmo Yang, Young-Ho Kim
cs.HC, cs.AI, cs.CL, H.5.2; I.2.7
In the mental health domain, Large Language Models (LLMs) offer promising new opportunities, though their inherent complexity and low controllability have raised questions about their suitability in clinical settings. We present MindfulDiary, a mobile journaling app incorporating an LLM to help psychiatric patients doc...
http://arxiv.org/pdf/2310.05231v1
2023-10-08T17:00:04
2023-10-08T17:00:04
{ "title": [ -0.04272461, -0.025466919, -0.05508423, -0.058380127, -0.0071105957, -0.002588272, -0.05731201, 0.012016296, 0.026885986, 0.02947998, 0.0014533997, -0.016296387, 0.011711121, -0.03250122, -0.047027588, -0.0029792786, 0.012275696, 0.0...
http://arxiv.org/abs/2310.05224v1
Generative Spoken Language Model based on continuous word-sized audio tokens
Robin Algayres, Yossi Adi, Tu Anh Nguyen, Jade Copet, Gabriel Synnaeve, Benoit Sagot, Emmanuel Dupoux
cs.CL, cs.LG
In NLP, text language models based on words or subwords are known to outperform their character-based counterparts. Yet, in the speech community, the standard input of spoken LMs are 20ms or 40ms-long discrete units (shorter than a phoneme). Taking inspiration from word-based LM, we introduce a Generative Spoken Langua...
http://arxiv.org/pdf/2310.05224v1
2023-10-08T16:46:14
2023-10-08T16:46:14
{ "title": [ -0.018508911, -0.011672974, -0.059661865, -0.06561279, -0.013053894, -0.0043029785, -0.06359863, 0.03050232, -0.0064315796, 0.040222168, -0.023406982, -0.00025129318, 0.0032043457, -0.013015747, -0.051757812, -0.008491516, -0.012809753, ...
http://arxiv.org/abs/2310.05216v1
Probing Language Models from A Human Behavioral Perspective
Xintong Wang, Xiaoyu Li, Xingshan Li, Chris Biemann
cs.CL
Large Language Models (LLMs) have emerged as dominant foundational models in modern NLP. However, the understanding of their prediction process and internal mechanisms, such as feed-forward networks and multi-head self-attention, remains largely unexplored. In this study, we probe LLMs from a human behavioral perspecti...
http://arxiv.org/pdf/2310.05216v1
2023-10-08T16:16:21
2023-10-08T16:16:21
{ "title": [ -0.021987915, -0.010696411, -0.02432251, -0.048431396, -0.011581421, -0.014968872, -0.039093018, 0.004863739, -0.020828247, 0.027069092, -0.011779785, -0.00617218, -0.0018196106, -0.020004272, -0.043029785, 0.0059127808, -0.006416321, ...
http://arxiv.org/abs/2310.05210v1
TILFA: A Unified Framework for Text, Image, and Layout Fusion in Argument Mining
Qing Zong, Zhaowei Wang, Baixuan Xu, Tianshi Zheng, Haochen Shi, Weiqi Wang, Yangqiu Song, Ginny Y. Wong, Simon See
cs.AI, cs.CL
A main goal of Argument Mining (AM) is to analyze an author's stance. Unlike previous AM datasets focusing only on text, the shared task at the 10th Workshop on Argument Mining introduces a dataset including both text and images. Importantly, these images contain both visual elements and optical characters. Our new fra...
http://arxiv.org/pdf/2310.05210v1
2023-10-08T15:54:37
2023-10-08T15:54:37
{ "title": [ 0.02633667, -0.016647339, -0.006084442, -0.061187744, 0.007068634, 0.014213562, -0.017349243, -0.019973755, -0.012611389, 0.02078247, -0.022720337, 0.023391724, -0.02671814, -0.008239746, -0.037750244, -0.031097412, -0.021316528, 0.0...
http://arxiv.org/abs/2310.05209v1
Scaling Laws of RoPE-based Extrapolation
Xiaoran Liu, Hang Yan, Shuo Zhang, Chenxin An, Xipeng Qiu, Dahua Lin
cs.CL, cs.AI
The extrapolation capability of Large Language Models (LLMs) based on Rotary Position Embedding is currently a topic of considerable interest. The mainstream approach to addressing extrapolation with LLMs involves modifying RoPE by replacing 10000, the rotary base of $\theta_n={10000}^{-2n/d}$ in the original RoPE, wit...
http://arxiv.org/pdf/2310.05209v1
2023-10-08T15:50:36
2023-10-08T15:50:36
{ "title": [ -0.017181396, 0.0037593842, -0.014717102, -0.027328491, -0.01676941, 0.00061511993, -0.008529663, -0.027648926, -0.050964355, 0.05230713, -0.011642456, 0.0067329407, -0.008514404, -0.020843506, -0.033355713, -0.020385742, 0.011703491, ...
http://arxiv.org/abs/2310.05203v1
A Comparative Study of Voice Conversion Models with Large-Scale Speech and Singing Data: The T13 Systems for the Singing Voice Conversion Challenge 2023
Ryuichi Yamamoto, Reo Yoneyama, Lester Phillip Violeta, Wen-Chin Huang, Tomoki Toda
eess.AS, cs.CL, cs.LG, cs.SD, eess.SP
This paper presents our systems (denoted as T13) for the singing voice conversion challenge (SVCC) 2023. For both in-domain and cross-domain English singing voice conversion (SVC) tasks (Task 1 and Task 2), we adopt a recognition-synthesis approach with self-supervised learning-based representation. To achieve data-eff...
http://arxiv.org/pdf/2310.05203v1
2023-10-08T15:30:44
2023-10-08T15:30:44
{ "title": [ -0.011016846, -0.0038642883, -0.053466797, -0.053344727, -0.009002686, -0.018508911, -0.056427002, 0.0085372925, 0.008720398, 0.047943115, -0.022750854, -0.005657196, 0.02571106, -0.028182983, -0.010749817, -0.0025901794, 0.007293701, ...
http://arxiv.org/abs/2310.05199v4
Loose lips sink ships: Mitigating Length Bias in Reinforcement Learning from Human Feedback
Wei Shen, Rui Zheng, Wenyu Zhan, Jun Zhao, Shihan Dou, Tao Gui, Qi Zhang, Xuanjing Huang
cs.CL
Reinforcement learning from human feedback serves as a crucial bridge, aligning large language models with human and societal values. This alignment requires a vast corpus of human feedback to learn a reward model, which is subsequently used to finetune language models. However, we have identified that the reward model...
http://arxiv.org/pdf/2310.05199v4
2023-11-06T10:28:16
2023-10-08T15:14:39
{ "title": [ -0.009254456, 0.0029506683, -0.017425537, 0.007171631, -0.0037670135, -0.02999878, 0.020507812, -0.045715332, -0.04675293, 0.007358551, -0.02949524, 0.022415161, -0.055847168, -0.009239197, -0.023483276, -0.046844482, -0.008354187, 0...
http://arxiv.org/abs/2310.05191v1
FABRIC: Automated Scoring and Feedback Generation for Essays
Jieun Han, Haneul Yoo, Junho Myung, Minsun Kim, Hyunseung Lim, Yoonsu Kim, Tak Yeon Lee, Hwajung Hong, Juho Kim, So-Yeon Ahn, Alice Oh
cs.CL
Automated essay scoring (AES) provides a useful tool for students and instructors in writing classes by generating essay scores in real-time. However, previous AES models do not provide more specific rubric-based scores nor feedback on how to improve the essays, which can be even more important than the overall scores ...
http://arxiv.org/pdf/2310.05191v1
2023-10-08T15:00:04
2023-10-08T15:00:04
{ "title": [ -0.023132324, -0.00157547, -0.026245117, -0.016784668, -0.0039405823, 0.0032901764, -0.003019333, -0.035339355, 0.015975952, 0.013084412, -0.03186035, 0.013900757, -0.023391724, -0.0010976791, -0.06323242, -0.017028809, -0.010307312, ...
http://arxiv.org/abs/2310.05189v2
Factuality Challenges in the Era of Large Language Models
Isabelle Augenstein, Timothy Baldwin, Meeyoung Cha, Tanmoy Chakraborty, Giovanni Luca Ciampaglia, David Corney, Renee DiResta, Emilio Ferrara, Scott Hale, Alon Halevy, Eduard Hovy, Heng Ji, Filippo Menczer, Ruben Miguez, Preslav Nakov, Dietram Scheufele, Shivam Sharma, Giovanni Zagni
cs.CL, cs.AI, cs.LG
The emergence of tools based on Large Language Models (LLMs), such as OpenAI's ChatGPT, Microsoft's Bing Chat, and Google's Bard, has garnered immense public attention. These incredibly useful, natural-sounding tools mark significant advances in natural language generation, yet they exhibit a propensity to generate fal...
http://arxiv.org/pdf/2310.05189v2
2023-10-10T03:34:46
2023-10-08T14:55:02
{ "title": [ -0.010566711, -0.010765076, -0.041290283, -0.07348633, -0.012840271, -0.017486572, -0.046417236, 0.0022239685, 0.011550903, 0.034240723, -0.00969696, -0.003528595, 0.034240723, -0.020187378, -0.03137207, 0.03692627, -0.015853882, 0.0...
http://arxiv.org/abs/2310.05185v2
Text2NKG: Fine-Grained N-ary Relation Extraction for N-ary relational Knowledge Graph Construction
Haoran Luo, Haihong E, Yuhao Yang, Tianyu Yao, Yikai Guo, Zichen Tang, Wentai Zhang, Kaiyang Wan, Shiyao Peng, Meina Song, Wei Lin
cs.AI, cs.CL
Beyond traditional binary relational facts, n-ary relational knowledge graphs (NKGs) are comprised of n-ary relational facts containing more than two entities, which are closer to real-world facts with broader applications. However, the construction of NKGs still significantly relies on manual labor, and n-ary relation...
http://arxiv.org/pdf/2310.05185v2
2023-10-12T18:26:46
2023-10-08T14:47:13
{ "title": [ -0.035980225, -0.020004272, -0.088134766, -0.08557129, 0.0040130615, 0.0149383545, -0.06591797, 0.025817871, 0.010818481, 0.025024414, -0.011062622, 0.019485474, -0.014892578, -0.014526367, -0.056915283, 0.0005903244, -0.0001860857, ...
http://arxiv.org/abs/2310.05178v1
Optimizing Large Language Models to Expedite the Development of Smart Contracts
Nii Osae Osae Dade, Margaret Lartey-Quaye, Emmanuel Teye-Kofi Odonkor, Paul Ammah
cs.SE, cs.AI, cs.CL
Programming has always been at the heart of technological innovation in the 21st century. With the advent of blockchain technologies and the proliferation of web3 paradigms of decentralised applications, smart contracts have been very instrumental in enabling developers to build applications that reside on decentralise...
http://arxiv.org/pdf/2310.05178v1
2023-10-08T14:29:33
2023-10-08T14:29:33
{ "title": [ -0.056152344, -0.01449585, -0.064086914, -0.0574646, -0.0013179779, 0.0023021698, -0.058502197, 0.018676758, -0.009513855, 0.026123047, -0.031280518, -0.0003311634, -0.0029735565, -0.0038719177, -0.03427124, -0.0065841675, -0.0035057068,...
http://arxiv.org/abs/2310.05177v1
Do Large Language Models Know about Facts?
Xuming Hu, Junzhe Chen, Xiaochuan Li, Yufei Guo, Lijie Wen, Philip S. Yu, Zhijiang Guo
cs.CL
Large language models (LLMs) have recently driven striking performance improvements across a range of natural language processing tasks. The factual knowledge acquired during pretraining and instruction tuning can be useful in various downstream tasks, such as question answering, and language generation. Unlike convent...
http://arxiv.org/pdf/2310.05177v1
2023-10-08T14:26:55
2023-10-08T14:26:55
{ "title": [ -0.023666382, -0.0042915344, -0.02986145, -0.056793213, -0.053619385, -0.008354187, -0.066345215, 0.0004930496, -0.0002887249, 0.02558899, -0.038513184, -0.009498596, 0.013381958, -0.023834229, -0.014427185, -0.0020828247, -0.0067863464,...
http://arxiv.org/abs/2310.05165v1
On the Zero-Shot Generalization of Machine-Generated Text Detectors
Xiao Pu, Jingyu Zhang, Xiaochuang Han, Yulia Tsvetkov, Tianxing He
cs.CL
The rampant proliferation of large language models, fluent enough to generate text indistinguishable from human-written language, gives unprecedented importance to the detection of machine-generated text. This work is motivated by an important research question: How will the detectors of machine-generated text perform ...
http://arxiv.org/pdf/2310.05165v1
2023-10-08T13:49:51
2023-10-08T13:49:51
{ "title": [ -0.0049705505, 0.0020008087, -0.013336182, -0.024917603, 0.0061569214, -0.005870819, -0.0004143715, -0.014183044, 0.023025513, 0.0501709, -0.023040771, 0.008811951, -0.006000519, 0.003768921, -0.07354736, 0.015197754, -0.01876831, 0....
http://arxiv.org/abs/2310.05163v3
An Investigation of LLMs' Inefficacy in Understanding Converse Relations
Chengwen Qi, Bowen Li, Binyuan Hui, Bailin Wang, Jinyang Li, Jinwang Wu, Yuanjun Laili
cs.CL
Large Language Models (LLMs) have achieved remarkable success in many formal language oriented tasks, such as structural data-to-text and semantic parsing. However current benchmarks mostly follow the data distribution of the pre-training data of LLMs. Therefore, a natural question rises that do LLMs really understand ...
http://arxiv.org/pdf/2310.05163v3
2023-11-13T07:24:14
2023-10-08T13:45:05
{ "title": [ -0.011833191, -0.002462387, -0.016052246, -0.011711121, -0.0345459, -0.016815186, 0.0008907318, -0.027801514, 0.0082092285, 0.020477295, -0.039398193, -0.025253296, -0.0016202927, -0.025787354, -0.029571533, -0.024902344, -0.025741577, ...
http://arxiv.org/abs/2310.05161v2
Recurrent Neural Language Models as Probabilistic Finite-state Automata
Anej Svete, Ryan Cotterell
cs.CL, cs.CC, cs.LG
Studying language models (LMs) in terms of well-understood formalisms allows us to precisely characterize their abilities and limitations. Previous work has investigated the representational capacity of recurrent neural network (RNN) LMs in terms of their capacity to recognize unweighted formal languages. However, LMs ...
http://arxiv.org/pdf/2310.05161v2
2023-10-19T11:49:10
2023-10-08T13:36:05
{ "title": [ -0.029891968, 0.015365601, -0.03479004, -0.08459473, 0.023620605, 0.021636963, -0.061279297, 0.0088272095, -0.004261017, 0.020980835, 0.0027236938, 0.004135132, -0.022079468, -0.018981934, -0.04156494, 0.004447937, -0.0137786865, 0.0...
http://arxiv.org/abs/2310.05157v1
MenatQA: A New Dataset for Testing the Temporal Comprehension and Reasoning Abilities of Large Language Models
Yifan Wei, Yisong Su, Huanhuan Ma, Xiaoyan Yu, Fangyu Lei, Yuanzhe Zhang, Jun Zhao, Kang Liu
cs.CL, cs.AI
Large language models (LLMs) have shown nearly saturated performance on many natural language processing (NLP) tasks. As a result, it is natural for people to believe that LLMs have also mastered abilities such as time understanding and reasoning. However, research on the temporal sensitivity of LLMs has been insuffici...
http://arxiv.org/pdf/2310.05157v1
2023-10-08T13:19:52
2023-10-08T13:19:52
{ "title": [ -0.022918701, -0.025985718, -0.056671143, -0.0725708, -0.010856628, -0.018447876, -0.06304932, 0.021972656, 0.02784729, 0.008094788, -0.032562256, -0.009277344, 0.0063476562, -0.023345947, -0.053527832, -0.0020179749, -0.016082764, 0...
http://arxiv.org/abs/2310.05155v1
Toolink: Linking Toolkit Creation and Using through Chain-of-Solving on Open-Source Model
Cheng Qian, Chenyan Xiong, Zhenghao Liu, Zhiyuan Liu
cs.CL, cs.AI
Large Language Models (LLMs) have demonstrated remarkable progress in utilizing tools, but their closed-source nature and high inference costs pose limitations on their adaptability, necessitating a valid method that leverages smaller, open-sourced models. In this paper, we introduce Toolink, a comprehensive framework ...
http://arxiv.org/pdf/2310.05155v1
2023-10-08T13:07:42
2023-10-08T13:07:42
{ "title": [ -0.038208008, -0.009742737, -0.045166016, -0.024795532, -0.026657104, 0.0067825317, -0.04345703, -0.019989014, -0.03640747, 0.028427124, -0.012374878, 0.01600647, 0.014633179, -0.022659302, 0.009307861, -0.008110046, -0.0024223328, -...
http://arxiv.org/abs/2310.05150v1
From Data to Dialogue: Leveraging the Structure of Knowledge Graphs for Conversational Exploratory Search
Phillip Schneider, Nils Rehtanz, Kristiina Jokinen, Florian Matthes
cs.CL, cs.IR
Exploratory search is an open-ended information retrieval process that aims at discovering knowledge about a topic or domain rather than searching for a specific answer or piece of information. Conversational interfaces are particularly suitable for supporting exploratory search, allowing users to refine queries and ex...
http://arxiv.org/pdf/2310.05150v1
2023-10-08T12:52:09
2023-10-08T12:52:09
{ "title": [ -0.013793945, -0.022781372, -0.051696777, -0.051208496, -0.0037174225, -0.014755249, -0.036071777, 0.011413574, 0.021530151, 0.019729614, -0.027008057, 0.009185791, -0.03753662, -0.012573242, -0.05810547, -0.02796936, -0.013130188, 0...
http://arxiv.org/abs/2310.05149v1
Retrieval-Generation Synergy Augmented Large Language Models
Zhangyin Feng, Xiaocheng Feng, Dezhi Zhao, Maojin Yang, Bing Qin
cs.CL
Large language models augmented with task-relevant documents have demonstrated impressive performance on knowledge-intensive tasks. However, regarding how to obtain effective documents, the existing methods are mainly divided into two categories. One is to retrieve from an external knowledge base, and the other is to u...
http://arxiv.org/pdf/2310.05149v1
2023-10-08T12:50:57
2023-10-08T12:50:57
{ "title": [ -0.009536743, -0.0072669983, -0.02986145, -0.07324219, -0.011016846, 0.0124435425, -0.066101074, 0.008781433, 0.015853882, 0.06335449, -0.009178162, 0.00051641464, 0.012039185, -0.008056641, -0.061798096, 0.02722168, -0.013885498, 0....
http://arxiv.org/abs/2310.05146v1
Large Language Model (LLM) as a System of Multiple Expert Agents: An Approach to solve the Abstraction and Reasoning Corpus (ARC) Challenge
John Chong Min Tan, Mehul Motani
cs.AI, cs.CL
We attempt to solve the Abstraction and Reasoning Corpus (ARC) Challenge using Large Language Models (LLMs) as a system of multiple expert agents. Using the flexibility of LLMs to be prompted to do various novel tasks using zero-shot, few-shot, context-grounded prompting, we explore the feasibility of using LLMs to sol...
http://arxiv.org/pdf/2310.05146v1
2023-10-08T12:37:28
2023-10-08T12:37:28
{ "title": [ -0.037963867, -0.022445679, -0.06945801, -0.06536865, -0.019515991, -0.008110046, -0.06085205, 0.04522705, -0.00017940998, 0.027175903, -0.025772095, -0.0057373047, 0.013343811, -0.027633667, -0.024002075, -0.0074920654, -0.024734497, ...
http://arxiv.org/abs/2310.05140v1
Harnessing the Power of Large Language Models for Empathetic Response Generation: Empirical Investigations and Improvements
Yushan Qian, Wei-Nan Zhang, Ting Liu
cs.CL, cs.AI
Empathetic dialogue is an indispensable part of building harmonious social relationships and contributes to the development of a helpful AI. Previous approaches are mainly based on fine small-scale language models. With the advent of ChatGPT, the application effect of large language models (LLMs) in this field has attr...
http://arxiv.org/pdf/2310.05140v1
2023-10-08T12:21:24
2023-10-08T12:21:24
{ "title": [ -0.009513855, -0.01979065, -0.02923584, -0.047302246, -0.0073165894, -0.013267517, -0.042175293, 0.002828598, 0.009140015, 0.013404846, -0.003320694, -0.013412476, 0.007457733, -0.038879395, -0.038909912, -0.002986908, -0.0068130493, ...
http://arxiv.org/abs/2310.05135v1
Are Emily and Greg Still More Employable than Lakisha and Jamal? Investigating Algorithmic Hiring Bias in the Era of ChatGPT
Akshaj Kumar Veldanda, Fabian Grob, Shailja Thakur, Hammond Pearce, Benjamin Tan, Ramesh Karri, Siddharth Garg
cs.CL, cs.AI, cs.LG
Large Language Models (LLMs) such as GPT-3.5, Bard, and Claude exhibit applicability across numerous tasks. One domain of interest is their use in algorithmic hiring, specifically in matching resumes with job categories. Yet, this introduces issues of bias on protected attributes like gender, race and maternity status....
http://arxiv.org/pdf/2310.05135v1
2023-10-08T12:08:48
2023-10-08T12:08:48
{ "title": [ 0.0025844574, -0.017486572, -0.024795532, 0.044281006, -0.020614624, -0.03643799, 0.042236328, -0.0236969, -0.009262085, 0.0131073, -0.042297363, 0.027954102, -0.0071487427, 0.0046043396, -0.023361206, -0.03564453, -0.02166748, 0.037...
http://arxiv.org/abs/2310.05130v1
Fast-DetectGPT: Efficient Zero-Shot Detection of Machine-Generated Text via Conditional Probability Curvature
Guangsheng Bao, Yanbin Zhao, Zhiyang Teng, Linyi Yang, Yue Zhang
cs.CL
Large language models (LLMs) have shown the ability to produce fluent and cogent content, presenting both productivity opportunities and societal risks. To build trustworthy AI systems, it is imperative to distinguish between machine-generated and human-authored content. The leading zero-shot detector, DetectGPT, showc...
http://arxiv.org/pdf/2310.05130v1
2023-10-08T11:41:28
2023-10-08T11:41:28
{ "title": [ -0.00051259995, -0.0036945343, -0.020507812, -0.04171753, 0.0020103455, -0.0134887695, -0.013191223, -0.018356323, 0.0041007996, 0.03062439, -0.014808655, 0.0284729, 0.0129776, -0.01864624, -0.029693604, 0.013595581, -0.0101623535, 0...
http://arxiv.org/abs/2310.05128v2
Instances and Labels: Hierarchy-aware Joint Supervised Contrastive Learning for Hierarchical Multi-Label Text Classification
Simon Chi Lok U, Jie He, Víctor Gutiérrez-Basulto, Jeff Z. Pan
cs.CL, cs.AI, cs.LG
Hierarchical multi-label text classification (HMTC) aims at utilizing a label hierarchy in multi-label classification. Recent approaches to HMTC deal with the problem of imposing an over-constrained premise on the output space by using contrastive learning on generated samples in a semi-supervised manner to bring text ...
http://arxiv.org/pdf/2310.05128v2
2023-10-14T15:49:09
2023-10-08T11:36:45
{ "title": [ 0.006298065, 0.00655365, -0.030731201, -0.062438965, 0.01210022, -0.008277893, -0.012680054, -0.025817871, 0.011352539, 0.023742676, -0.024505615, 0.040008545, 0.0031929016, -0.012878418, -0.036224365, -0.0019836426, -0.037994385, 0....
http://arxiv.org/abs/2310.05991v3
Enhancing Document-level Event Argument Extraction with Contextual Clues and Role Relevance
Wanlong Liu, Shaohuan Cheng, Dingyi Zeng, Hong Qu
cs.CL
Document-level event argument extraction poses new challenges of long input and cross-sentence inference compared to its sentence-level counterpart. However, most prior works focus on capturing the relations between candidate arguments and the event trigger in each event, ignoring two crucial points: a) non-argument co...
http://arxiv.org/pdf/2310.05991v3
2023-10-20T02:20:20
2023-10-08T11:29:10
{ "title": [ 0.028961182, 0.0023288727, -0.030670166, -0.109680176, 0.021575928, 0.0009937286, -0.05493164, -0.0023727417, -0.007675171, 0.013710022, 0.015296936, 0.0056915283, -0.0031261444, -0.028518677, -0.022232056, 0.008872986, -0.015914917, ...
http://arxiv.org/abs/2310.05116v1
CARLG: Leveraging Contextual Clues and Role Correlations for Improving Document-level Event Argument Extraction
Wanlong Liu, Wenyu Chen, Dingyi Zeng, Li Zhou, Hong Qu
cs.CL, cs.IR
Document-level event argument extraction (EAE) is a crucial but challenging subtask in information extraction. Most existing approaches focus on the interaction between arguments and event triggers, ignoring two critical points: the information of contextual clues and the semantic correlations among argument roles. In ...
http://arxiv.org/pdf/2310.05116v1
2023-10-08T11:09:16
2023-10-08T11:09:16
{ "title": [ 0.030654907, 0.00078725815, -0.029968262, -0.10449219, 0.02470398, 0.0014228821, -0.042266846, -0.004901886, -0.015960693, 0.015640259, 0.021896362, 0.021896362, -0.008613586, -0.021575928, -0.022857666, 0.0063667297, -0.022079468, 0...
http://arxiv.org/abs/2310.05115v1
Breaking Down Word Semantics from Pre-trained Language Models through Layer-wise Dimension Selection
Nayoung Choi
cs.CL, cs.AI
Contextual word embeddings obtained from pre-trained language model (PLM) have proven effective for various natural language processing tasks at the word level. However, interpreting the hidden aspects within embeddings, such as syntax and semantics, remains challenging. Disentangled representation learning has emerged...
http://arxiv.org/pdf/2310.05115v1
2023-10-08T11:07:19
2023-10-08T11:07:19
{ "title": [ -0.02684021, -0.013282776, -0.031585693, -0.063964844, -0.002981186, 0.0010404587, -0.052886963, -0.0020198822, -0.014595032, 0.03527832, -0.0019273758, -0.00017523766, 0.016998291, -0.022903442, -0.031921387, 0.0075263977, 0.0025196075,...
http://arxiv.org/abs/2310.05103v1
Zero-Shot Detection of Machine-Generated Codes
Xianjun Yang, Kexun Zhang, Haifeng Chen, Linda Petzold, William Yang Wang, Wei Cheng
cs.CL, cs.AI, cs.CR, cs.LG
This work proposes a training-free approach for the detection of LLMs-generated codes, mitigating the risks associated with their indiscriminate usage. To the best of our knowledge, our research is the first to investigate zero-shot detection techniques applied to code generated by advanced black-box LLMs like ChatGPT....
http://arxiv.org/pdf/2310.05103v1
2023-10-08T10:08:21
2023-10-08T10:08:21
{ "title": [ -0.011222839, -0.0008029938, -0.023880005, -0.021728516, -0.0003244877, -0.023147583, 0.0054016113, -0.010475159, 0.008659363, 0.044555664, 0.00907135, 0.008338928, 0.0030670166, -0.024627686, -0.04425049, 0.029449463, -0.004711151, ...
http://arxiv.org/abs/2310.05095v1
How Reliable Are AI-Generated-Text Detectors? An Assessment Framework Using Evasive Soft Prompts
Tharindu Kumarage, Paras Sheth, Raha Moraffah, Joshua Garland, Huan Liu
cs.CL, cs.AI
In recent years, there has been a rapid proliferation of AI-generated text, primarily driven by the release of powerful pre-trained language models (PLMs). To address the issue of misuse associated with AI-generated text, various high-performing detectors have been developed, including the OpenAI detector and the Stanf...
http://arxiv.org/pdf/2310.05095v1
2023-10-08T09:53:46
2023-10-08T09:53:46
{ "title": [ -0.021835327, -0.003074646, -0.010543823, -0.0022697449, -0.0063323975, -0.008766174, 0.01725769, -0.03201294, -0.006061554, 0.022125244, -0.03555298, 0.0017929077, -0.0060920715, -0.0077552795, -0.05029297, -0.021835327, -0.025604248, ...
http://arxiv.org/abs/2310.05092v1
Benchmarking Large Language Models with Augmented Instructions for Fine-grained Information Extraction
Jun Gao, Huan Zhao, Yice Zhang, Wei Wang, Changlong Yu, Ruifeng Xu
cs.CL
Information Extraction (IE) is an essential task in Natural Language Processing. Traditional methods have relied on coarse-grained extraction with simple instructions. However, with the emergence of Large Language Models (LLMs), there is a need to adapt IE techniques to leverage the capabilities of these models. This p...
http://arxiv.org/pdf/2310.05092v1
2023-10-08T09:41:18
2023-10-08T09:41:18
{ "title": [ -0.041931152, -0.011489868, -0.036346436, -0.064453125, -0.0046195984, 0.015014648, -0.06762695, 0.007255554, 0.0057640076, 0.03857422, -0.02658081, 0.019485474, -0.005039215, 0.0031528473, -0.058532715, 0.0073013306, -0.010826111, 0...
http://arxiv.org/abs/2310.05083v1
FLatS: Principled Out-of-Distribution Detection with Feature-Based Likelihood Ratio Score
Haowei Lin, Yuntian Gu
cs.LG, cs.AI, cs.CL
Detecting out-of-distribution (OOD) instances is crucial for NLP models in practical applications. Although numerous OOD detection methods exist, most of them are empirical. Backed by theoretical analysis, this paper advocates for the measurement of the "OOD-ness" of a test case $\boldsymbol{x}$ through the likelihood ...
http://arxiv.org/pdf/2310.05083v1
2023-10-08T09:16:46
2023-10-08T09:16:46
{ "title": [ -0.006515503, 0.0047187805, -0.022216797, -0.035614014, 0.022216797, -0.0074043274, -0.0070152283, -0.029693604, 0.0025844574, 0.0023441315, -0.032928467, 0.012123108, -0.02520752, -0.038146973, -0.029586792, -0.028961182, -0.018371582, ...
http://arxiv.org/abs/2310.18324v1
"A Nova Eletricidade: Aplicações, Riscos e Tendências da IA Moderna -- "The New Electricity": Applications, Risks, and Trends in Current AI
Ana L. C. Bazzan, Anderson R. Tavares, André G. Pereira, Cláudio R. Jung, Jacob Scharcanski, Joel Luis Carbonera, Luís C. Lamb, Mariana Recamonde-Mendoza, Thiago L. T. da Silveira, Viviane Moreira
cs.AI, cs.CL, cs.CY, cs.LG, 68, I.2
The thought-provoking analogy between AI and electricity, made by computer scientist and entrepreneur Andrew Ng, summarizes the deep transformation that recent advances in Artificial Intelligence (AI) have triggered in the world. This chapter presents an overview of the ever-evolving landscape of AI, written in Portugu...
http://arxiv.org/pdf/2310.18324v1
2023-10-08T09:01:18
2023-10-08T09:01:18
{ "title": [ -0.0021877289, -0.0013637543, 0.01889038, 0.0121536255, -0.030288696, 0.013046265, 0.016433716, -0.023864746, -0.04586792, 0.061584473, -0.012588501, 0.03237915, 0.01701355, -0.017852783, 0.008811951, -0.028533936, 0.023849487, -0.02...
http://arxiv.org/abs/2310.05074v3
DialCoT Meets PPO: Decomposing and Exploring Reasoning Paths in Smaller Language Models
Chengcheng Han, Xiaowei Du, Che Zhang, Yixin Lian, Xiang Li, Ming Gao, Baoyuan Wang
cs.CL, cs.AI
Chain-of-Thought (CoT) prompting has proven to be effective in enhancing the reasoning capabilities of Large Language Models (LLMs) with at least 100 billion parameters. However, it is ineffective or even detrimental when applied to reasoning tasks in Smaller Language Models (SLMs) with less than 10 billion parameters....
http://arxiv.org/pdf/2310.05074v3
2023-10-23T09:38:01
2023-10-08T08:52:13
{ "title": [ -0.04309082, -0.04055786, -0.046691895, -0.09069824, -0.029418945, 0.025421143, -0.08392334, 0.0335083, -0.00944519, 0.023010254, -0.017807007, 0.008270264, 0.014442444, -0.019760132, -0.024383545, -0.004711151, -0.026489258, 0.02087...
http://arxiv.org/abs/2310.05073v1
Enhancing Argument Structure Extraction with Efficient Leverage of Contextual Information
Yun Luo, Zhen Yang, Fandong Meng, Yingjie Li, Jie Zhou, Yue Zhang
cs.CL
Argument structure extraction (ASE) aims to identify the discourse structure of arguments within documents. Previous research has demonstrated that contextual information is crucial for developing an effective ASE model. However, we observe that merely concatenating sentences in a contextual window does not fully utili...
http://arxiv.org/pdf/2310.05073v1
2023-10-08T08:47:10
2023-10-08T08:47:10
{ "title": [ 0.027923584, 0.009025574, -0.030334473, -0.10131836, 0.01701355, 0.005039215, -0.050628662, -0.0095825195, -0.013374329, 0.019622803, 0.005504608, 0.011741638, -0.025604248, -0.021759033, -0.033325195, -0.00058317184, -0.013191223, 0...
http://arxiv.org/abs/2310.05069v2
Unleashing the Multilingual Encoder Potential: Boosting Zero-Shot Performance via Probability Calibration
Ercong Nie, Helmut Schmid, Hinrich Schütze
cs.CL
Pretrained multilingual encoder models can directly perform zero-shot multilingual tasks or linguistic probing by reformulating the input examples into cloze-style prompts. This is accomplished by predicting the probabilities of the label words at the masked token position, without requiring any updates to the model pa...
http://arxiv.org/pdf/2310.05069v2
2023-10-19T15:58:05
2023-10-08T08:31:05
{ "title": [ -0.0007200241, -0.008712769, -0.02885437, -0.0496521, -0.013511658, -0.014892578, -0.035980225, -0.008575439, 0.01675415, 0.047729492, 0.009437561, -0.0011358261, 0.007637024, -0.034423828, -0.03189087, 0.030044556, 0.0066452026, 0.0...
http://arxiv.org/abs/2310.05066v2
Guideline Learning for In-context Information Extraction
Chaoxu Pang, Yixuan Cao, Qiang Ding, Ping Luo
cs.CL, cs.LG
Large language models (LLMs) can perform a new task by merely conditioning on task instructions and a few input-output examples, without optimizing any parameters. This is called In-Context Learning (ICL). In-context Information Extraction (IE) has recently garnered attention in the research community. However, the per...
http://arxiv.org/pdf/2310.05066v2
2023-10-21T10:21:48
2023-10-08T08:25:16
{ "title": [ -0.030563354, -0.005203247, -0.06329346, -0.06958008, 0.007587433, -0.0015239716, -0.041137695, 0.011131287, -0.0056266785, 0.017364502, -0.035827637, 0.02128601, -0.022857666, -0.01285553, -0.034606934, -0.017318726, -0.018615723, 0...
http://arxiv.org/abs/2310.05064v1
sign.mt: Real-Time Multilingual Sign Language Translation Application
Amit Moryossef
cs.CL
This demo paper presents sign.mt, an open-source application pioneering real-time multilingual bi-directional translation between spoken and signed languages. Harnessing state-of-the-art open-source models, this tool aims to address the communication divide between the hearing and the deaf, facilitating seamless transl...
http://arxiv.org/pdf/2310.05064v1
2023-10-08T08:14:03
2023-10-08T08:14:03
{ "title": [ -0.046325684, -0.014122009, -0.025115967, -0.04168701, -0.0047683716, 0.012809753, -0.043518066, -0.016067505, 0.0066604614, 0.04067993, -0.014953613, -0.0038871765, 0.028793335, -0.033203125, -0.038360596, 0.00447464, 0.0007853508, ...
http://arxiv.org/abs/2310.05057v3
BRAINTEASER: Lateral Thinking Puzzles for Large Language Models
Yifan Jiang, Filip Ilievski, Kaixin Ma, Zhivar Sourati
cs.CL
The success of language models has inspired the NLP community to attend to tasks that require implicit and complex reasoning, relying on human-like commonsense mechanisms. While such vertical thinking tasks have been relatively popular, lateral thinking puzzles have received little attention. To bridge this gap, we dev...
http://arxiv.org/pdf/2310.05057v3
2023-11-09T19:45:13
2023-10-08T07:46:01
{ "title": [ -0.046691895, -0.023071289, -0.027770996, -0.019897461, -0.009857178, 0.00038933754, -0.024414062, -0.0022068024, -0.001991272, 0.018310547, -0.015899658, 0.0060806274, -0.0032291412, -0.01436615, -0.034484863, 0.004425049, -0.009742737,...
http://arxiv.org/abs/2310.05046v1
Harnessing the Power of ChatGPT in Fake News: An In-Depth Exploration in Generation, Detection and Explanation
Yue Huang, Lichao Sun
cs.CL
The rampant spread of fake news has adversely affected society, resulting in extensive research on curbing its spread. As a notable milestone in large language models (LLMs), ChatGPT has gained significant attention due to its exceptional natural language processing capabilities. In this study, we present a thorough ex...
http://arxiv.org/pdf/2310.05046v1
2023-10-08T07:01:07
2023-10-08T07:01:07
{ "title": [ -0.0101623535, -0.008201599, -0.019622803, -0.028671265, 0.0012960434, 0.0023269653, -0.007297516, -0.028152466, -0.056549072, 0.027862549, -0.0061149597, 0.034118652, -0.013626099, -0.019836426, -0.006225586, -0.008666992, -0.040802002,...
http://arxiv.org/abs/2310.05036v3
AvalonBench: Evaluating LLMs Playing the Game of Avalon
Jonathan Light, Min Cai, Sheng Shen, Ziniu Hu
cs.AI, cs.CL
In this paper, we explore the potential of Large Language Models (LLMs) Agents in playing the strategic social deduction game, Resistance Avalon. Players in Avalon are challenged not only to make informed decisions based on dynamically evolving game phases, but also to engage in discussions where they must deceive, ded...
http://arxiv.org/pdf/2310.05036v3
2023-11-08T16:01:32
2023-10-08T06:37:08
{ "title": [ -0.022872925, -0.0027828217, -0.041107178, 0.0074386597, -0.03579712, -0.012283325, -0.040924072, -0.031204224, -0.009590149, 0.012069702, -0.05441284, 0.02658081, -0.017730713, -0.028762817, 0.0069503784, -0.04598999, -0.011749268, ...
http://arxiv.org/abs/2310.05035v2
Self-Convinced Prompting: Few-Shot Question Answering with Repeated Introspection
Haodi Zhang, Min Cai, Xinhe Zhang, Chen Jason Zhang, Rui Mao, Kaishun Wu
cs.CL, cs.AI
While large language models (LLMs) such as ChatGPT and PaLM have demonstrated remarkable performance in various language understanding and generation tasks, their capabilities in complex reasoning and intricate knowledge utilization still fall short of human-level proficiency. Recent studies have established the effect...
http://arxiv.org/pdf/2310.05035v2
2023-10-10T15:03:35
2023-10-08T06:36:26
{ "title": [ 0.011909485, -0.016494751, -0.015686035, -0.043121338, -0.008308411, -0.01272583, -0.00466156, -0.005153656, 0.005683899, 0.020828247, -0.017242432, 0.024032593, -0.023132324, -0.025390625, -0.043640137, 0.004245758, 0.011787415, -0....
http://arxiv.org/abs/2310.05030v2
Counter Turing Test CT^2: AI-Generated Text Detection is Not as Easy as You May Think -- Introducing AI Detectability Index
Megha Chakraborty, S. M Towhidul Islam Tonmoy, S M Mehedi Zaman, Krish Sharma, Niyar R Barman, Chandan Gupta, Shreya Gautam, Tanay Kumar, Vinija Jain, Aman Chadha, Amit P. Sheth, Amitava Das
cs.CL, cs.AI
With the rise of prolific ChatGPT, the risk and consequences of AI-generated text has increased alarmingly. To address the inevitable question of ownership attribution for AI-generated artifacts, the US Copyright Office released a statement stating that 'If a work's traditional elements of authorship were produced by a...
http://arxiv.org/pdf/2310.05030v2
2023-10-24T00:36:29
2023-10-08T06:20:36
{ "title": [ -0.040008545, 0.0054092407, 0.020217896, 0.070007324, -0.030578613, 0.010917664, 0.025924683, -0.07269287, 0.010650635, 0.019973755, -0.06439209, 0.0043563843, -0.0027236938, -0.018066406, -0.022338867, -0.057617188, -0.0079956055, 0...
http://arxiv.org/abs/2310.05029v1
Walking Down the Memory Maze: Beyond Context Limit through Interactive Reading
Howard Chen, Ramakanth Pasunuru, Jason Weston, Asli Celikyilmaz
cs.CL
Large language models (LLMs) have advanced in large strides due to the effectiveness of the self-attention mechanism that processes and compares all tokens at once. However, this mechanism comes with a fundamental issue -- the predetermined context window is bound to be limited. Despite attempts to extend the context w...
http://arxiv.org/pdf/2310.05029v1
2023-10-08T06:18:14
2023-10-08T06:18:14
{ "title": [ -0.005870819, -0.007648468, -0.021133423, -0.023986816, -0.0015058517, -0.015670776, -0.0041122437, -0.029754639, -0.0077323914, 0.03668213, -0.015350342, 0.0095825195, -0.030288696, -0.0096206665, -0.0690918, -0.01499176, 0.0071907043, ...
http://arxiv.org/abs/2310.05028v3
Revisiting Large Language Models as Zero-shot Relation Extractors
Guozheng Li, Peng Wang, Wenjun Ke
cs.AI, cs.CL
Relation extraction (RE) consistently involves a certain degree of labeled or unlabeled data even if under zero-shot setting. Recent studies have shown that large language models (LLMs) transfer well to new tasks out-of-the-box simply given a natural language prompt, which provides the possibility of extracting relatio...
http://arxiv.org/pdf/2310.05028v3
2023-10-11T06:16:30
2023-10-08T06:17:39
{ "title": [ -0.008857727, -0.01637268, -0.043151855, -0.08215332, -0.003396988, 0.0038890839, -0.062469482, 0.020126343, 0.03845215, 0.014831543, -0.004043579, -0.0040740967, -0.0138168335, -0.007068634, -0.059020996, 0.016830444, -0.0059547424, ...
http://arxiv.org/abs/2310.05025v1
Synslator: An Interactive Machine Translation Tool with Online Learning
Jiayi Wang, Ke Wang, Fengming Zhou, Chengyu Wang, Zhiyong Fu, Zeyu Feng, Yu Zhao, Yuqi Zhang
cs.CL
Interactive machine translation (IMT) has emerged as a progression of the computer-aided translation paradigm, where the machine translation system and the human translator collaborate to produce high-quality translations. This paper introduces Synslator, a user-friendly computer-aided translation (CAT) tool that not o...
http://arxiv.org/pdf/2310.05025v1
2023-10-08T06:05:55
2023-10-08T06:05:55
{ "title": [ -0.035980225, -0.01184082, -0.01309967, -0.037200928, -0.023071289, 0.013496399, -0.03842163, -0.022506714, -0.0029239655, 0.035949707, -0.0047073364, 0.008758545, 0.0017595291, -0.025497437, -0.038879395, 0.0056419373, 0.007583618, ...
http://arxiv.org/abs/2310.10672v1
Hybrid Quantum-Classical Machine Learning for Sentiment Analysis
Abu Kaisar Mohammad Masum, Anshul Maurya, Dhruthi Sridhar Murthy, Pratibha, Naveed Mahmud
cs.CL, cs.LG
The collaboration between quantum computing and classical machine learning offers potential advantages in natural language processing, particularly in the sentiment analysis of human emotions and opinions expressed in large-scale datasets. In this work, we propose a methodology for sentiment analysis using hybrid quant...
http://arxiv.org/pdf/2310.10672v1
2023-10-08T05:45:22
2023-10-08T05:45:22
{ "title": [ -0.014083862, 0.007774353, -0.0154953, 0.00049209595, 0.0024547577, -0.019592285, 0.024337769, -0.03237915, -0.0067634583, 0.045959473, 0.012550354, -0.012573242, -0.0009975433, -0.025604248, -0.064697266, -0.003993988, -0.024841309, ...
http://arxiv.org/abs/2310.05009v1
WikiIns: A High-Quality Dataset for Controlled Text Editing by Natural Language Instruction
Xiang Chen, Zheng Li, Xiaojun Wan
cs.CL
Text editing, i.e., the process of modifying or manipulating text, is a crucial step in human writing process. In this paper, we study the problem of controlled text editing by natural language instruction. According to a given instruction that conveys the edit intention and necessary information, an original draft tex...
http://arxiv.org/pdf/2310.05009v1
2023-10-08T04:46:39
2023-10-08T04:46:39
{ "title": [ -0.045776367, -0.019729614, -0.030853271, -0.019302368, -0.010116577, 0.00007683039, -0.015472412, -0.014312744, -0.012374878, 0.025863647, -0.0087890625, 0.0064582825, 0.009155273, -0.023025513, -0.02230835, -0.008613586, -0.028442383, ...
http://arxiv.org/abs/2310.05007v1
MinPrompt: Graph-based Minimal Prompt Data Augmentation for Few-shot Question Answering
Xiusi Chen, Jyun-Yu Jiang, Wei-Cheng Chang, Cho-Jui Hsieh, Hsiang-Fu Yu, Wei Wang
cs.CL
Few-shot question answering (QA) aims at achieving satisfactory results on machine question answering when only a few training samples are available. Recent advances mostly rely on the power of pre-trained large language models (LLMs) and fine-tuning in specific settings. Although the pre-training stage has already equ...
http://arxiv.org/pdf/2310.05007v1
2023-10-08T04:44:36
2023-10-08T04:44:36
{ "title": [ 0.014625549, -0.02418518, -0.045410156, -0.06915283, -0.008895874, -0.0038204193, -0.02116394, 0.03704834, 0.026245117, 0.011230469, -0.03353882, 0.014144897, -0.020492554, -0.010261536, -0.052459717, -0.013305664, 0.010467529, 0.004...
http://arxiv.org/abs/2310.05002v1
Self-Knowledge Guided Retrieval Augmentation for Large Language Models
Yile Wang, Peng Li, Maosong Sun, Yang Liu
cs.CL
Large language models (LLMs) have shown superior performance without task-specific fine-tuning. Despite the success, the knowledge stored in the parameters of LLMs could still be incomplete and difficult to update due to the computational costs. As complementary, retrieval-based methods can offer non-parametric world k...
http://arxiv.org/pdf/2310.05002v1
2023-10-08T04:22:33
2023-10-08T04:22:33
{ "title": [ 0.0007529259, -0.00762558, -0.053222656, -0.07165527, -0.0020561218, -0.0074272156, -0.039520264, 0.022918701, 0.022506714, 0.05154419, -0.014373779, 0.0019397736, 0.0234375, -0.005935669, -0.054382324, 0.02909851, 0.00982666, 0.0163...
http://arxiv.org/abs/2310.04994v1
Distantly-Supervised Joint Entity and Relation Extraction with Noise-Robust Learning
Yufei Li, Xiao Yu, Yanghong Guo, Yanchi Liu, Haifeng Chen, Cong Liu
cs.CL, cs.AI, cs.LG
Joint entity and relation extraction is a process that identifies entity pairs and their relations using a single model. We focus on the problem of training these models on distantly-labeled data, which is generated by aligning entity mentions in a text corpus with their corresponding entity and relation types in a kno...
http://arxiv.org/pdf/2310.04994v1
2023-10-08T03:42:15
2023-10-08T03:42:15
{ "title": [ 0.0061683655, 0.022979736, -0.06463623, -0.089660645, 0.04626465, -0.016220093, -0.040863037, -0.0051002502, 0.025054932, 0.01197052, -0.046539307, 0.023895264, 0.014472961, -0.0013093948, -0.019638062, -0.0014533997, -0.014587402, 0...
http://arxiv.org/abs/2310.04982v1
Comparative Analysis of Transfer Learning in Deep Learning Text-to-Speech Models on a Few-Shot, Low-Resource, Customized Dataset
Ze Liu
cs.SD, cs.AI, cs.CL, eess.AS
Text-to-Speech (TTS) synthesis using deep learning relies on voice quality. Modern TTS models are advanced, but they need large amount of data. Given the growing computational complexity of these models and the scarcity of large, high-quality datasets, this research focuses on transfer learning, especially on few-shot,...
http://arxiv.org/pdf/2310.04982v1
2023-10-08T03:08:25
2023-10-08T03:08:25
{ "title": [ -0.0009498596, 0.0044517517, -0.035095215, -0.04748535, 0.0013208389, -0.01739502, -0.01537323, 0.008422852, -0.023620605, 0.038360596, -0.023803711, 0.032928467, 0.0071792603, -0.011116028, -0.011764526, -0.006931305, -0.010223389, ...
http://arxiv.org/abs/2310.04978v1
TopicAdapt- An Inter-Corpora Topics Adaptation Approach
Pritom Saha Akash, Trisha Das, Kevin Chen-Chuan Chang
cs.CL, cs.LG
Topic models are popular statistical tools for detecting latent semantic topics in a text corpus. They have been utilized in various applications across different fields. However, traditional topic models have some limitations, including insensitivity to user guidance, sensitivity to the amount and quality of data, and...
http://arxiv.org/pdf/2310.04978v1
2023-10-08T02:56:44
2023-10-08T02:56:44
{ "title": [ -0.013328552, 0.008926392, -0.057373047, -0.058624268, 0.021377563, -0.0038585663, -0.035827637, -0.017684937, 0.030807495, 0.028823853, -0.033111572, 0.03387451, -0.0032100677, 0.0015335083, -0.025756836, 0.014099121, -0.008071899, ...
http://arxiv.org/abs/2310.04965v1
MULTISCRIPT: Multimodal Script Learning for Supporting Open Domain Everyday Tasks
Jingyuan Qi, Minqian Liu, Ying Shen, Zhiyang Xu, Lifu Huang
cs.CL, cs.AI, cs.LG
Automatically generating scripts (i.e. sequences of key steps described in text) from video demonstrations and reasoning about the subsequent steps are crucial to the modern AI virtual assistants to guide humans to complete everyday tasks, especially unfamiliar ones. However, current methods for generative script learn...
http://arxiv.org/pdf/2310.04965v1
2023-10-08T01:51:17
2023-10-08T01:51:17
{ "title": [ -0.022857666, -0.010292053, -0.05996704, -0.05947876, -0.0060195923, -0.01626587, -0.051605225, 0.01386261, -0.01146698, 0.016204834, -0.015342712, 0.020309448, -0.018066406, -0.024414062, -0.010864258, -0.016311646, 0.0055160522, 0....
http://arxiv.org/abs/2310.04960v1
Exploring the Usage of Chinese Pinyin in Pretraining
Baojun Wang, Kun Xu, Lifeng Shang
cs.CL
Unlike alphabetic languages, Chinese spelling and pronunciation are different. Both characters and pinyin take an important role in Chinese language understanding. In Chinese NLP tasks, we almost adopt characters or words as model input, and few works study how to use pinyin. However, pinyin is essential in many scenar...
http://arxiv.org/pdf/2310.04960v1
2023-10-08T01:26:44
2023-10-08T01:26:44
{ "title": [ 0.009887695, 0.0050582886, -0.020812988, 0.034851074, -0.02507019, -0.028060913, 0.033599854, -0.031311035, -0.04171753, 0.046844482, 0.0023288727, -0.017837524, 0.025741577, -0.046203613, 0.027160645, -0.026229858, 0.008369446, 0.02...
http://arxiv.org/abs/2310.04959v1
Towards Better Chain-of-Thought Prompting Strategies: A Survey
Zihan Yu, Liang He, Zhen Wu, Xinyu Dai, Jiajun Chen
cs.CL
Chain-of-Thought (CoT), a step-wise and coherent reasoning chain, shows its impressive strength when used as a prompting strategy for large language models (LLM). Recent years, the prominent effect of CoT prompting has attracted emerging research. However, there still lacks of a systematic summary about key factors of ...
http://arxiv.org/pdf/2310.04959v1
2023-10-08T01:16:55
2023-10-08T01:16:55
{ "title": [ 0.023345947, -0.017150879, -0.009971619, -0.03866577, 0.0051651, -0.015777588, -0.0021343231, -0.024780273, -0.00705719, 0.0077171326, -0.022354126, 0.014709473, -0.013618469, -0.031433105, -0.018829346, 0.00080013275, -0.0024414062, ...
http://arxiv.org/abs/2310.04949v1
Domain Knowledge Graph Construction Via A Simple Checker
Yueling Zeng, Li-C. Wang
cs.CL
With the availability of large language models, there is a growing interest for semiconductor chip design companies to leverage the technologies. For those companies, deployment of a new methodology must include two important considerations: confidentiality and scalability. In this context, this work tackles the proble...
http://arxiv.org/pdf/2310.04949v1
2023-10-08T00:09:31
2023-10-08T00:09:31
{ "title": [ -0.014389038, -0.007926941, -0.072265625, -0.10510254, 0.017929077, -0.008659363, -0.07116699, 0.010978699, -0.012069702, 0.013023376, -0.009643555, -0.00015044212, -0.008041382, -0.028564453, -0.023742676, -0.0032806396, 0.0004608631, ...
http://arxiv.org/abs/2310.04948v2
TEMPO: Prompt-based Generative Pre-trained Transformer for Time Series Forecasting
Defu Cao, Furong Jia, Sercan O Arik, Tomas Pfister, Yixiang Zheng, Wen Ye, Yan Liu
cs.LG, cs.CL
The past decade has witnessed significant advances in time series modeling with deep learning. While achieving state-of-the-art results, the best-performing architectures vary highly across applications and domains. Meanwhile, for natural language processing, the Generative Pre-trained Transformer (GPT) has demonstrate...
http://arxiv.org/pdf/2310.04948v2
2023-10-12T02:43:13
2023-10-08T00:02:25
{ "title": [ 0.014190674, -0.012252808, -0.04714966, -0.09234619, 0.017227173, -0.0048599243, -0.03817749, -0.0013074875, 0.006275177, 0.03741455, -0.027954102, 0.01826477, -0.0090789795, -0.00548172, -0.06756592, -0.0039711, 0.013580322, 0.03053...
http://arxiv.org/abs/2310.04945v1
Balancing Specialized and General Skills in LLMs: The Impact of Modern Tuning and Data Strategy
Zheng Zhang, Chen Zheng, Da Tang, Ke Sun, Yukun Ma, Yingtong Bu, Xun Zhou, Liang Zhao
cs.CL, cs.AI
This paper introduces a multifaceted methodology for fine-tuning and evaluating large language models (LLMs) for specialized monetization tasks. The goal is to balance general language proficiency with domain-specific skills. The methodology has three main components: 1) Carefully blending in-domain and general-purpose...
http://arxiv.org/pdf/2310.04945v1
2023-10-07T23:29:00
2023-10-07T23:29:00
{ "title": [ -0.047973633, -0.010551453, -0.062347412, -0.03845215, -0.023635864, -0.02381897, -0.031799316, 0.030273438, -0.022216797, 0.027679443, -0.034942627, 0.027694702, -0.003791809, -0.021118164, -0.0032806396, -0.009033203, -0.0063056946, ...
http://arxiv.org/abs/2310.04928v2
Large Language Models Only Pass Primary School Exams in Indonesia: A Comprehensive Test on IndoMMLU
Fajri Koto, Nurul Aisyah, Haonan Li, Timothy Baldwin
cs.CL
Although large language models (LLMs) are often pre-trained on large-scale multilingual texts, their reasoning abilities and real-world knowledge are mainly evaluated based on English datasets. Assessing LLM capabilities beyond English is increasingly vital but hindered due to the lack of suitable datasets. In this wor...
http://arxiv.org/pdf/2310.04928v2
2023-10-21T17:13:05
2023-10-07T21:49:38
{ "title": [ -0.013595581, -0.030456543, -0.0013637543, -0.015487671, -0.06536865, -0.0014238358, -0.03050232, 0.004814148, -0.0038890839, 0.016342163, -0.005809784, -0.027633667, -0.005290985, -0.047851562, -0.008911133, -0.040924072, -0.029449463, ...
http://arxiv.org/abs/2310.04921v2
Crystal: Introspective Reasoners Reinforced with Self-Feedback
Jiacheng Liu, Ramakanth Pasunuru, Hannaneh Hajishirzi, Yejin Choi, Asli Celikyilmaz
cs.AI, cs.CL, cs.LG
Extensive work has shown that the performance and interpretability of commonsense reasoning can be improved via knowledge-augmented reasoning methods, where the knowledge that underpins the reasoning process is explicitly verbalized and utilized. However, existing implementations, including "chain-of-thought" and its v...
http://arxiv.org/pdf/2310.04921v2
2023-10-18T14:52:03
2023-10-07T21:23:58
{ "title": [ -0.02909851, -0.00793457, -0.06274414, -0.034362793, -0.01576233, -0.0115737915, -0.029922485, -0.01260376, -0.0132369995, 0.027786255, 0.005619049, -0.0010461807, -0.00049972534, -0.04397583, -0.032714844, -0.012077332, 0.005317688, ...
http://arxiv.org/abs/2310.04914v1
Analyzing Zero-Shot Abilities of Vision-Language Models on Video Understanding Tasks
Avinash Madasu, Anahita Bhiwandiwalla, Vasudev Lal
cs.CV, cs.AI, cs.CL
Foundational multimodal models pre-trained on large scale image-text pairs or video-text pairs or both have shown strong generalization abilities on downstream tasks. However unlike image-text models, pretraining video-text models is always not feasible due to the difficulty in collecting large-scale clean and aligned ...
http://arxiv.org/pdf/2310.04914v1
2023-10-07T20:57:54
2023-10-07T20:57:54
{ "title": [ -0.012123108, -0.010017395, -0.033050537, -0.03451538, -0.017349243, -0.026748657, -0.016815186, -0.0012397766, 0.032836914, 0.030166626, -0.017196655, -0.017318726, -0.0028057098, -0.0132369995, -0.07824707, 0.000009417534, 0.008743286,...
http://arxiv.org/abs/2310.04910v1
Faithful Knowledge Graph Explanations for Commonsense Reasoning
Weihe Zhai, Arkaitz Zubiaga, Bingquan Liu
cs.CL, cs.AI
While fusing language models (LMs) and knowledge graphs (KGs) has become common in commonsense question answering research, enabling faithful chain-of-thought explanations in these models remains an open problem. One major weakness of current KG-based explanation techniques is that they overlook the faithfulness of gen...
http://arxiv.org/pdf/2310.04910v1
2023-10-07T20:29:45
2023-10-07T20:29:45
{ "title": [ -0.0128479, -0.022735596, -0.034698486, -0.0791626, -0.02230835, 0.0009756088, -0.038482666, 0.015777588, 0.0015211105, 0.024002075, -0.0034332275, -0.000028371811, -0.011779785, -0.033050537, -0.023620605, 0.0028591156, -0.006137848, ...
http://arxiv.org/abs/2310.04889v1
GradXKG: A Universal Explain-per-use Temporal Knowledge Graph Explainer
Chenhan Yuan, Hoda Eldardiry
cs.LG, cs.CL
Temporal knowledge graphs (TKGs) have shown promise for reasoning tasks by incorporating a temporal dimension to represent how facts evolve over time. However, existing TKG reasoning (TKGR) models lack explainability due to their black-box nature. Recent work has attempted to address this through customized model archi...
http://arxiv.org/pdf/2310.04889v1
2023-10-07T18:21:35
2023-10-07T18:21:35
{ "title": [ -0.030532837, -0.03012085, -0.059051514, -0.107055664, -0.014732361, 0.0028381348, -0.0602417, 0.032073975, 0.0011386871, 0.020935059, -0.0025691986, 0.021026611, -0.0107040405, -0.018600464, -0.030670166, 0.008178711, 0.0075569153, ...
http://arxiv.org/abs/2310.04875v1
Prompt-to-OS (P2OS): Revolutionizing Operating Systems and Human-Computer Interaction with Integrated AI Generative Models
Gabriele Tolomei, Cesare Campagnano, Fabrizio Silvestri, Giovanni Trappolini
cs.LG, cs.CL, cs.CY, cs.HC, cs.OS
In this paper, we present a groundbreaking paradigm for human-computer interaction that revolutionizes the traditional notion of an operating system. Within this innovative framework, user requests issued to the machine are handled by an interconnected ecosystem of generative AI models that seamlessly integrate with ...
http://arxiv.org/pdf/2310.04875v1
2023-10-07T17:16:34
2023-10-07T17:16:34
{ "title": [ 0.018554688, -0.008049011, -0.007575989, -0.03152466, -0.01838684, -0.01499939, -0.013000488, -0.020141602, -0.005290985, 0.016799927, -0.018737793, -0.007671356, 0.013648987, -0.053619385, 0.0042037964, -0.008094788, 0.011680603, -0...
http://arxiv.org/abs/2310.04869v1
ILuvUI: Instruction-tuned LangUage-Vision modeling of UIs from Machine Conversations
Yue Jiang, Eldon Schoop, Amanda Swearngin, Jeffrey Nichols
cs.HC, cs.AI, cs.CL, cs.CV
Multimodal Vision-Language Models (VLMs) enable powerful applications from their fused understanding of images and language, but many perform poorly on UI tasks due to the lack of UI training data. In this paper, we adapt a recipe for generating paired text-image training data for VLMs to the UI domain by combining exi...
http://arxiv.org/pdf/2310.04869v1
2023-10-07T16:32:34
2023-10-07T16:32:34
{ "title": [ -0.038848877, -0.03265381, -0.024627686, -0.034240723, -0.019714355, -0.005126953, -0.045196533, 0.0016937256, 0.025878906, 0.027893066, -0.055511475, -0.009147644, 0.024520874, -0.00440979, -0.059265137, -0.014076233, -0.011657715, ...