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/2310.11237v1
Watermarking LLMs with Weight Quantization
Linyang Li, Botian Jiang, Pengyu Wang, Ke Ren, Hang Yan, Xipeng Qiu
cs.CL
Abuse of large language models reveals high risks as large language models are being deployed at an astonishing speed. It is important to protect the model weights to avoid malicious usage that violates licenses of open-source large language models. This paper proposes a novel watermarking strategy that plants watermar...
http://arxiv.org/pdf/2310.11237v1
2023-10-17T13:06:59
2023-10-17T13:06:59
{ "title": [ -0.04220581, 0.010696411, -0.05239868, -0.013008118, -0.022125244, -0.016998291, -0.03717041, -0.022964478, -0.04421997, 0.055541992, 0.016433716, 0.010101318, -0.010818481, -0.040283203, -0.024765015, -0.004436493, 0.0036201477, 0.0...
http://arxiv.org/abs/2310.11227v1
RealBehavior: A Framework for Faithfully Characterizing Foundation Models' Human-like Behavior Mechanisms
Enyu Zhou, Rui Zheng, Zhiheng Xi, Songyang Gao, Xiaoran Fan, Zichu Fei, Jingting Ye, Tao Gui, Qi Zhang, Xuanjing Huang
cs.CL, cs.AI
Reports of human-like behaviors in foundation models are growing, with psychological theories providing enduring tools to investigate these behaviors. However, current research tends to directly apply these human-oriented tools without verifying the faithfulness of their outcomes. In this paper, we introduce a framewor...
http://arxiv.org/pdf/2310.11227v1
2023-10-17T12:58:17
2023-10-17T12:58:17
{ "title": [ -0.02154541, 0.0074768066, -0.027786255, -0.049346924, 0.015541077, -0.006904602, -0.019363403, -0.037109375, -0.007637024, 0.024475098, 0.005302429, 0.0231781, -0.00051021576, -0.029083252, -0.019561768, 0.022720337, -0.019058228, 0...
http://arxiv.org/abs/2310.11220v1
KG-GPT: A General Framework for Reasoning on Knowledge Graphs Using Large Language Models
Jiho Kim, Yeonsu Kwon, Yohan Jo, Edward Choi
cs.CL
While large language models (LLMs) have made considerable advancements in understanding and generating unstructured text, their application in structured data remains underexplored. Particularly, using LLMs for complex reasoning tasks on knowledge graphs (KGs) remains largely untouched. To address this, we propose KG-G...
http://arxiv.org/pdf/2310.11220v1
2023-10-17T12:51:35
2023-10-17T12:51:35
{ "title": [ -0.038238525, -0.030960083, -0.05758667, -0.097961426, -0.007873535, 0.027770996, -0.091918945, 0.018310547, 0.020629883, 0.012130737, -0.016326904, -0.008583069, 0.0012550354, -0.021820068, -0.025253296, 0.0015745163, -0.0041389465, ...
http://arxiv.org/abs/2310.11207v1
Can Large Language Models Explain Themselves? A Study of LLM-Generated Self-Explanations
Shiyuan Huang, Siddarth Mamidanna, Shreedhar Jangam, Yilun Zhou, Leilani H. Gilpin
cs.CL, cs.LG
Large language models (LLMs) such as ChatGPT have demonstrated superior performance on a variety of natural language processing (NLP) tasks including sentiment analysis, mathematical reasoning and summarization. Furthermore, since these models are instruction-tuned on human conversations to produce "helpful" responses,...
http://arxiv.org/pdf/2310.11207v1
2023-10-17T12:34:32
2023-10-17T12:34:32
{ "title": [ -0.057739258, -0.017440796, -0.03765869, -0.058563232, -0.04071045, 0.013183594, -0.07684326, 0.016799927, -0.01902771, 0.017242432, -0.0029792786, 0.01737976, -0.030349731, -0.02671814, -0.035614014, 0.004108429, -0.014328003, 0.031...
http://arxiv.org/abs/2310.11191v2
Medical Text Simplification: Optimizing for Readability with Unlikelihood Training and Reranked Beam Search Decoding
Lorenzo Jaime Yu Flores, Heyuan Huang, Kejian Shi, Sophie Chheang, Arman Cohan
cs.CL, cs.AI
Text simplification has emerged as an increasingly useful application of AI for bridging the communication gap in specialized fields such as medicine, where the lexicon is often dominated by technical jargon and complex constructs. Despite notable progress, methods in medical simplification sometimes result in the gene...
http://arxiv.org/pdf/2310.11191v2
2023-10-26T02:54:41
2023-10-17T12:14:03
{ "title": [ -0.048217773, 0.004600525, -0.034332275, -0.048919678, -0.011474609, 0.026397705, -0.06768799, -0.021896362, -0.020141602, 0.058776855, 0.0075149536, 0.01852417, 0.028961182, -0.0126571655, -0.041625977, -0.0034980774, -0.008399963, ...
http://arxiv.org/abs/2310.11166v2
ViSoBERT: A Pre-Trained Language Model for Vietnamese Social Media Text Processing
Quoc-Nam Nguyen, Thang Chau Phan, Duc-Vu Nguyen, Kiet Van Nguyen
cs.CL
English and Chinese, known as resource-rich languages, have witnessed the strong development of transformer-based language models for natural language processing tasks. Although Vietnam has approximately 100M people speaking Vietnamese, several pre-trained models, e.g., PhoBERT, ViBERT, and vELECTRA, performed well on ...
http://arxiv.org/pdf/2310.11166v2
2023-10-28T16:22:19
2023-10-17T11:34:50
{ "title": [ -0.0007677078, -0.010574341, -0.0005507469, -0.045166016, -0.019760132, 0.009483337, -0.044525146, -0.017089844, -0.021606445, 0.055603027, -0.03274536, 0.026412964, 0.004901886, 0.007522583, -0.058929443, -0.0015296936, -0.014961243, ...
http://arxiv.org/abs/2310.11163v1
IMTLab: An Open-Source Platform for Building, Evaluating, and Diagnosing Interactive Machine Translation Systems
Xu Huang, Zhirui Zhang, Ruize Gao, Yichao Du, Lemao Liu, Gouping Huang, Shuming Shi, Jiajun Chen, Shujian Huang
cs.CL
We present IMTLab, an open-source end-to-end interactive machine translation (IMT) system platform that enables researchers to quickly build IMT systems with state-of-the-art models, perform an end-to-end evaluation, and diagnose the weakness of systems. IMTLab treats the whole interactive translation process as a task...
http://arxiv.org/pdf/2310.11163v1
2023-10-17T11:29:04
2023-10-17T11:29:04
{ "title": [ -0.011474609, -0.008880615, -0.0045433044, -0.029464722, -0.042175293, 0.008132935, -0.05682373, -0.032470703, -0.009567261, 0.048858643, -0.015396118, 0.0038776398, 0.05114746, -0.024490356, -0.013259888, 0.011062622, -0.0022068024, ...
http://arxiv.org/abs/2310.11158v1
Probing the Creativity of Large Language Models: Can models produce divergent semantic association?
Honghua Chen, Nai Ding
cs.CL, cs.LG
Large language models possess remarkable capacity for processing language, but it remains unclear whether these models can further generate creative content. The present study aims to investigate the creative thinking of large language models through a cognitive perspective. We utilize the divergent association task (D...
http://arxiv.org/pdf/2310.11158v1
2023-10-17T11:23:32
2023-10-17T11:23:32
{ "title": [ -0.041259766, -0.002117157, -0.037322998, -0.03375244, -0.0071640015, -0.0031394958, -0.04928589, -0.01134491, -0.002161026, 0.03552246, -0.007575989, -0.00082588196, 0.012420654, -0.016494751, -0.038360596, 0.010353088, -0.029663086, ...
http://arxiv.org/abs/2310.11146v1
The Quo Vadis of the Relationship between Language and Large Language Models
Evelina Leivada, Vittoria Dentella, Elliot Murphy
cs.CL
In the field of Artificial (General) Intelligence (AI), the several recent advancements in Natural language processing (NLP) activities relying on Large Language Models (LLMs) have come to encourage the adoption of LLMs as scientific models of language. While the terminology employed for the characterization of LLMs fa...
http://arxiv.org/pdf/2310.11146v1
2023-10-17T10:54:24
2023-10-17T10:54:24
{ "title": [ -0.019866943, -0.0104904175, -0.059173584, -0.054260254, -0.02166748, -0.03274536, -0.025482178, 0.01525116, -0.004047394, 0.024963379, -0.051849365, -0.01737976, 0.018325806, -0.031402588, -0.027633667, -0.006137848, 0.003042221, 0....
http://arxiv.org/abs/2310.11141v1
Long-form Simultaneous Speech Translation: Thesis Proposal
Peter Polák
cs.CL, cs.AI, cs.SD, eess.AS
Simultaneous speech translation (SST) aims to provide real-time translation of spoken language, even before the speaker finishes their sentence. Traditionally, SST has been addressed primarily by cascaded systems that decompose the task into subtasks, including speech recognition, segmentation, and machine translation....
http://arxiv.org/pdf/2310.11141v1
2023-10-17T10:44:05
2023-10-17T10:44:05
{ "title": [ -0.012718201, 0.013580322, -0.023406982, -0.060638428, -0.005290985, -0.0047683716, -0.032287598, 0.0046310425, -0.05126953, 0.064697266, -0.0044059753, 0.0039863586, 0.03930664, -0.03555298, -0.017364502, -0.005882263, -0.0065193176, ...
http://arxiv.org/abs/2310.13012v2
H2O Open Ecosystem for State-of-the-art Large Language Models
Arno Candel, Jon McKinney, Philipp Singer, Pascal Pfeiffer, Maximilian Jeblick, Chun Ming Lee, Marcos V. Conde
cs.CL, cs.AI
Large Language Models (LLMs) represent a revolution in AI. However, they also pose many significant risks, such as the presence of biased, private, copyrighted or harmful text. For this reason we need open, transparent and safe solutions. We introduce a complete open-source ecosystem for developing and testing LLMs. Th...
http://arxiv.org/pdf/2310.13012v2
2023-10-23T10:24:55
2023-10-17T09:40:58
{ "title": [ -0.038726807, -0.010231018, -0.058776855, -0.09185791, 0.0013093948, 0.011978149, -0.078125, 0.036743164, -0.0011281967, 0.03955078, -0.006450653, -0.014892578, 0.024307251, -0.02204895, -0.040893555, 0.01689148, -0.009559631, 0.0458...
http://arxiv.org/abs/2310.11097v4
Experimenting AI Technologies for Disinformation Combat: the IDMO Project
Lorenzo Canale, Alberto Messina
cs.CL
The Italian Digital Media Observatory (IDMO) project, part of a European initiative, focuses on countering disinformation and fake news. This report outlines contributions from Rai-CRITS to the project, including: (i) the creation of novel datasets for testing technologies (ii) development of an automatic model for cat...
http://arxiv.org/pdf/2310.11097v4
2023-11-08T09:57:39
2023-10-17T09:27:43
{ "title": [ -0.0011014938, 0.0071525574, -0.04348755, -0.03277588, 0.0014390945, -0.026779175, 0.007820129, -0.027908325, -0.03579712, 0.0209198, -0.01687622, -0.0024719238, 0.029083252, -0.044799805, 0.0054626465, -0.026168823, -0.033172607, 0....
http://arxiv.org/abs/2310.11085v1
In-Context Few-Shot Relation Extraction via Pre-Trained Language Models
Yilmazcan Ozyurt, Stefan Feuerriegel, Ce Zhang
cs.CL, cs.AI, cs.LG
Relation extraction aims at inferring structured human knowledge from textual documents. State-of-the-art methods based on language models commonly have two limitations: (1) they require named entities to be either given as input or infer them, which introduces additional noise, and (2) they require human annotations o...
http://arxiv.org/pdf/2310.11085v1
2023-10-17T09:10:27
2023-10-17T09:10:27
{ "title": [ 0.015914917, -0.01687622, -0.03125, -0.0847168, -0.0026130676, 0.00039696693, -0.0463562, 0.012283325, 0.025894165, 0.016601562, -0.013786316, -0.0036945343, -0.0055389404, -0.011680603, -0.048095703, -0.00010514259, 0.0065574646, 0....
http://arxiv.org/abs/2310.11081v1
Understanding writing style in social media with a supervised contrastively pre-trained transformer
Javier Huertas-Tato, Alejandro Martin, David Camacho
cs.CL, cs.SI
Online Social Networks serve as fertile ground for harmful behavior, ranging from hate speech to the dissemination of disinformation. Malicious actors now have unprecedented freedom to misbehave, leading to severe societal unrest and dire consequences, as exemplified by events such as the Capitol assault during the US ...
http://arxiv.org/pdf/2310.11081v1
2023-10-17T09:01:17
2023-10-17T09:01:17
{ "title": [ -0.004600525, 0.01626587, -0.03262329, -0.028121948, 0.0051231384, -0.014785767, -0.030044556, -0.025421143, -0.019226074, 0.033599854, -0.03274536, 0.039489746, 0.019592285, 0.0058631897, -0.009941101, 0.0055885315, -0.0087509155, 0...
http://arxiv.org/abs/2310.11079v1
Learning from Red Teaming: Gender Bias Provocation and Mitigation in Large Language Models
Hsuan Su, Cheng-Chu Cheng, Hua Farn, Shachi H Kumar, Saurav Sahay, Shang-Tse Chen, Hung-yi Lee
cs.CL, cs.AI
Recently, researchers have made considerable improvements in dialogue systems with the progress of large language models (LLMs) such as ChatGPT and GPT-4. These LLM-based chatbots encode the potential biases while retaining disparities that can harm humans during interactions. The traditional biases investigation metho...
http://arxiv.org/pdf/2310.11079v1
2023-10-17T08:56:04
2023-10-17T08:56:04
{ "title": [ -0.049438477, -0.0061035156, -0.056243896, -0.04373169, -0.009864807, -0.019363403, -0.049957275, -0.0027770996, 0.0015773773, 0.03326416, -0.018051147, 0.0005517006, 0.009521484, -0.0067863464, -0.045318604, -0.0112838745, -0.027648926,...
http://arxiv.org/abs/2310.11069v4
VoxArabica: A Robust Dialect-Aware Arabic Speech Recognition System
Abdul Waheed, Bashar Talafha, Peter Sullivan, AbdelRahim Elmadany, Muhammad Abdul-Mageed
cs.CL, cs.SD, eess.AS
Arabic is a complex language with many varieties and dialects spoken by over 450 millions all around the world. Due to the linguistic diversity and variations, it is challenging to build a robust and generalized ASR system for Arabic. In this work, we address this gap by developing and demoing a system, dubbed VoxArabi...
http://arxiv.org/pdf/2310.11069v4
2023-10-27T13:32:15
2023-10-17T08:33:02
{ "title": [ -0.017501831, -0.006729126, -0.04562378, -0.0736084, 0.009986877, 0.0063667297, -0.050872803, 0.024383545, 0.0018339157, 0.040527344, -0.005241394, -0.0107040405, -0.009414673, -0.027328491, -0.070251465, 0.0012922287, 0.014808655, 0...
http://arxiv.org/abs/2310.11053v2
Denevil: Towards Deciphering and Navigating the Ethical Values of Large Language Models via Instruction Learning
Shitong Duan, Xiaoyuan Yi, Peng Zhang, Tun Lu, Xing Xie, Ning Gu
cs.CL, cs.AI, cs.CY
Large Language Models (LLMs) have made unprecedented breakthroughs, yet their increasing integration into everyday life might raise societal risks due to generated unethical content. Despite extensive study on specific issues like bias, the intrinsic values of LLMs remain largely unexplored from a moral philosophy pers...
http://arxiv.org/pdf/2310.11053v2
2023-10-30T02:30:35
2023-10-17T07:42:40
{ "title": [ -0.032592773, -0.0054130554, -0.046417236, -0.018859863, -0.0072135925, -0.024307251, -0.021560669, -0.013839722, -0.009414673, 0.019470215, -0.030059814, 0.0018444061, 0.0012950897, -0.027999878, -0.021835327, -0.020126343, -0.024368286...
http://arxiv.org/abs/2310.11049v1
Nonet at SemEval-2023 Task 6: Methodologies for Legal Evaluation
Shubham Kumar Nigam, Aniket Deroy, Noel Shallum, Ayush Kumar Mishra, Anup Roy, Shubham Kumar Mishra, Arnab Bhattacharya, Saptarshi Ghosh, Kripabandhu Ghosh
cs.CL, cs.AI, cs.IR, cs.LG
This paper describes our submission to the SemEval-2023 for Task 6 on LegalEval: Understanding Legal Texts. Our submission concentrated on three subtasks: Legal Named Entity Recognition (L-NER) for Task-B, Legal Judgment Prediction (LJP) for Task-C1, and Court Judgment Prediction with Explanation (CJPE) for Task-C2. We...
http://arxiv.org/pdf/2310.11049v1
2023-10-17T07:35:11
2023-10-17T07:35:11
{ "title": [ -0.012634277, -0.013298035, -0.025863647, -0.04232788, 0.0019664764, -0.009994507, -0.03665161, -0.030670166, -0.009109497, 0.012969971, -0.009147644, 0.017578125, -0.02935791, -0.030181885, -0.01965332, 0.0072517395, -0.028564453, 0...
http://arxiv.org/abs/2310.11035v1
Lyricist-Singer Entropy Affects Lyric-Lyricist Classification Performance
Mitsuki Morita, Masato Kikuchi, Tadachika Ozono
cs.SD, cs.CL, eess.AS
Although lyrics represent an essential component of music, few music information processing studies have been conducted on the characteristics of lyricists. Because these characteristics may be valuable for musical applications, such as recommendations, they warrant further study. We considered a potential method that ...
http://arxiv.org/pdf/2310.11035v1
2023-10-17T07:02:26
2023-10-17T07:02:26
{ "title": [ 0.01663208, 0.018096924, -0.07952881, -0.01914978, -0.010719299, -0.043884277, -0.0028800964, 0.008880615, -0.0058250427, 0.053497314, -0.05795288, -0.005874634, 0.030456543, -0.017623901, -0.03048706, -0.010551453, -0.003944397, 0.0...
http://arxiv.org/abs/2310.11026v1
Exploring Automatic Evaluation Methods based on a Decoder-based LLM for Text Generation
Tomohito Kasahara, Daisuke Kawahara
cs.CL
Automatic evaluation of text generation is essential for improving the accuracy of generation tasks. In light of the current trend towards increasingly larger decoder-based language models, we investigate automatic evaluation methods based on such models for text generation. This paper compares various methods, includi...
http://arxiv.org/pdf/2310.11026v1
2023-10-17T06:53:00
2023-10-17T06:53:00
{ "title": [ -0.07336426, -0.007587433, -0.039001465, -0.022491455, -0.036865234, 0.015281677, -0.0904541, -0.015808105, -0.0024738312, 0.037353516, -0.012390137, 0.018127441, -0.00088739395, -0.026168823, -0.0340271, 0.0011024475, -0.039489746, ...
http://arxiv.org/abs/2310.11016v1
Reading Order Matters: Information Extraction from Visually-rich Documents by Token Path Prediction
Chong Zhang, Ya Guo, Yi Tu, Huan Chen, Jinyang Tang, Huijia Zhu, Qi Zhang, Tao Gui
cs.CL
Recent advances in multimodal pre-trained models have significantly improved information extraction from visually-rich documents (VrDs), in which named entity recognition (NER) is treated as a sequence-labeling task of predicting the BIO entity tags for tokens, following the typical setting of NLP. However, BIO-tagging...
http://arxiv.org/pdf/2310.11016v1
2023-10-17T06:08:55
2023-10-17T06:08:55
{ "title": [ -0.028289795, -0.018798828, -0.035186768, -0.057647705, -0.020462036, 0.0026931763, -0.045806885, -0.01083374, 0.006542206, 0.022033691, -0.025054932, 0.012062073, -0.010696411, -0.0029182434, -0.052581787, -0.01235199, -0.0058784485, ...
http://arxiv.org/abs/2310.11003v1
Correction Focused Language Model Training for Speech Recognition
Yingyi Ma, Zhe Liu, Ozlem Kalinli
cs.CL, cs.LG, cs.SD, eess.AS
Language models (LMs) have been commonly adopted to boost the performance of automatic speech recognition (ASR) particularly in domain adaptation tasks. Conventional way of LM training treats all the words in corpora equally, resulting in suboptimal improvements in ASR performance. In this work, we introduce a novel co...
http://arxiv.org/pdf/2310.11003v1
2023-10-17T05:10:39
2023-10-17T05:10:39
{ "title": [ -0.01890564, -0.00083112717, -0.036468506, -0.04260254, -0.017623901, -0.007484436, -0.05545044, 0.010726929, -0.027130127, 0.047332764, -0.028457642, -0.008705139, 0.013343811, -0.030303955, -0.033355713, -0.008918762, -0.027755737, ...
http://arxiv.org/abs/2310.10981v1
Instructive Dialogue Summarization with Query Aggregations
Bin Wang, Zhengyuan Liu, Nancy F. Chen
cs.CL
Conventional dialogue summarization methods directly generate summaries and do not consider user's specific interests. This poses challenges in cases where the users are more focused on particular topics or aspects. With the advancement of instruction-finetuned language models, we introduce instruction-tuning to dialog...
http://arxiv.org/pdf/2310.10981v1
2023-10-17T04:03:00
2023-10-17T04:03:00
{ "title": [ 0.014595032, -0.0060424805, -0.0385437, -0.12054443, 0.014167786, 0.009719849, -0.060180664, 0.032318115, 0.01461792, 0.057281494, -0.008399963, 0.007965088, 0.004951477, -0.009880066, -0.07476807, 0.006969452, -0.0071868896, 0.02447...
http://arxiv.org/abs/2310.10967v1
EXMODD: An EXplanatory Multimodal Open-Domain Dialogue dataset
Hang Yin, Pinren Lu, Ziang Li, Bin Sun, Kan Li
cs.CL, cs.AI, cs.HC
The need for high-quality data has been a key issue hindering the research of dialogue tasks. Recent studies try to build datasets through manual, web crawling, and large pre-trained models. However, man-made data is expensive and data collected from the internet often includes generic responses, meaningless statements...
http://arxiv.org/pdf/2310.10967v1
2023-10-17T03:28:29
2023-10-17T03:28:29
{ "title": [ -0.013687134, -0.00011318922, -0.053131104, -0.08288574, 0.012634277, -0.014549255, -0.05831909, 0.014442444, 0.00026249886, 0.020507812, -0.0035514832, 0.008163452, -0.014099121, -0.022003174, -0.028549194, 0.001291275, -0.02772522, ...
http://arxiv.org/abs/2310.10962v1
Semantic-Aware Contrastive Sentence Representation Learning with Large Language Models
Huiming Wang, Liying Cheng, Zhaodonghui Li, De Wen Soh, Lidong Bing
cs.CL
Contrastive learning has been proven to be effective in learning better sentence representations. However, to train a contrastive learning model, large numbers of labeled sentences are required to construct positive and negative pairs explicitly, such as those in natural language inference (NLI) datasets. Unfortunately...
http://arxiv.org/pdf/2310.10962v1
2023-10-17T03:21:43
2023-10-17T03:21:43
{ "title": [ 0.0050849915, 0.004463196, -0.015853882, -0.054779053, 0.0021705627, 0.00094795227, -0.037597656, -0.0118637085, 0.030395508, 0.033569336, -0.0335083, 0.00856781, 0.008079529, -0.0029621124, -0.0501709, 0.011520386, -0.008911133, 0.0...
http://arxiv.org/abs/2310.10956v1
Computing the optimal keyboard through a geometric analysis of the English language
Jules Deschamps, Quentin Hubert, Lucas Ryckelynck
cs.CL, cs.HC
In the context of a group project for the course COMSW4995 002 - Geometric Data Analysis, we bring our attention to the design of fast-typing keyboards. Leveraging some geometric tools in an optimization framework allowed us to propose novel keyboard layouts that offer a faster typing.
http://arxiv.org/pdf/2310.10956v1
2023-10-17T03:05:42
2023-10-17T03:05:42
{ "title": [ -0.025146484, -0.011169434, 0.007881165, 0.019073486, -0.05682373, -0.013130188, -0.015640259, -0.034729004, -0.023742676, 0.04272461, -0.008468628, -0.011039734, 0.029663086, -0.040252686, -0.0031471252, 0.0015392303, 0.0095825195, ...
http://arxiv.org/abs/2310.10955v1
A State-Vector Framework for Dataset Effects
Esmat Sahak, Zining Zhu, Frank Rudzicz
cs.CL, cs.AI, cs.LG
The impressive success of recent deep neural network (DNN)-based systems is significantly influenced by the high-quality datasets used in training. However, the effects of the datasets, especially how they interact with each other, remain underexplored. We propose a state-vector framework to enable rigorous studies in ...
http://arxiv.org/pdf/2310.10955v1
2023-10-17T03:05:06
2023-10-17T03:05:06
{ "title": [ 0.02720642, 0.0039405823, -0.025497437, -0.079956055, 0.016784668, -0.019622803, -0.033477783, -0.025970459, 0.010719299, 0.027038574, -0.042785645, 0.011528015, -0.041534424, -0.004146576, -0.06149292, -0.012306213, 0.014183044, 0.0...
http://arxiv.org/abs/2310.10944v1
TEQ: Trainable Equivalent Transformation for Quantization of LLMs
Wenhua Cheng, Yiyang Cai, Kaokao Lv, Haihao Shen
cs.CL
As large language models (LLMs) become more prevalent, there is a growing need for new and improved quantization methods that can meet the computationalast layer demands of these modern architectures while maintaining the accuracy. In this paper, we present TEQ, a trainable equivalent transformation that preserves the ...
http://arxiv.org/pdf/2310.10944v1
2023-10-17T02:42:34
2023-10-17T02:42:34
{ "title": [ -0.07342529, -0.0033893585, -0.047729492, -0.01763916, -0.02217102, 0.0031528473, -0.057739258, -0.011459351, -0.0064048767, 0.050689697, 0.0069084167, -0.00014173985, 0.015335083, -0.031402588, -0.029418945, 0.015960693, 0.0021190643, ...
http://arxiv.org/abs/2310.10941v1
MASON-NLP at eRisk 2023: Deep Learning-Based Detection of Depression Symptoms from Social Media Texts
Fardin Ahsan Sakib, Ahnaf Atef Choudhury, Ozlem Uzuner
cs.CL, cs.LG
Depression is a mental health disorder that has a profound impact on people's lives. Recent research suggests that signs of depression can be detected in the way individuals communicate, both through spoken words and written texts. In particular, social media posts are a rich and convenient text source that we may exam...
http://arxiv.org/pdf/2310.10941v1
2023-10-17T02:34:34
2023-10-17T02:34:34
{ "title": [ 0.008369446, -0.008125305, -0.009994507, -0.027648926, -0.005153656, -0.0132751465, -0.008705139, -0.012496948, 0.006931305, 0.02722168, -0.04034424, 0.017028809, 0.04244995, -0.02017212, -0.019165039, 0.0048675537, -0.011955261, 0.0...
http://arxiv.org/abs/2310.10935v1
Intent Detection and Slot Filling for Home Assistants: Dataset and Analysis for Bangla and Sylheti
Fardin Ahsan Sakib, A H M Rezaul Karim, Saadat Hasan Khan, Md Mushfiqur Rahman
cs.CL, cs.LG
As voice assistants cement their place in our technologically advanced society, there remains a need to cater to the diverse linguistic landscape, including colloquial forms of low-resource languages. Our study introduces the first-ever comprehensive dataset for intent detection and slot filling in formal Bangla, collo...
http://arxiv.org/pdf/2310.10935v1
2023-10-17T02:12:12
2023-10-17T02:12:12
{ "title": [ -0.009979248, 0.017929077, -0.04727173, -0.0029125214, 0.016082764, -0.03149414, 0.0020313263, -0.028457642, -0.026031494, 0.049316406, -0.024353027, -0.0049057007, 0.03665161, -0.028259277, -0.014350891, -0.0061912537, -0.01222229, ...
http://arxiv.org/abs/2310.10930v1
Enhanced Transformer Architecture for Natural Language Processing
Woohyeon Moon, Taeyoung Kim, Bumgeun Park, Dongsoo Har
cs.CL, cs.AI
Transformer is a state-of-the-art model in the field of natural language processing (NLP). Current NLP models primarily increase the number of transformers to improve processing performance. However, this technique requires a lot of training resources such as computing capacity. In this paper, a novel structure of Tran...
http://arxiv.org/pdf/2310.10930v1
2023-10-17T01:59:07
2023-10-17T01:59:07
{ "title": [ -0.013397217, 0.001209259, -0.039886475, -0.059265137, 0.017028809, -0.0023117065, -0.03164673, 0.0013074875, 0.0005455017, 0.039978027, -0.041412354, -0.0014104843, 0.030303955, -0.0037822723, -0.037109375, 0.0141067505, -0.010894775, ...
http://arxiv.org/abs/2310.10922v1
Spatial HuBERT: Self-supervised Spatial Speech Representation Learning for a Single Talker from Multi-channel Audio
Antoni Dimitriadis, Siqi Pan, Vidhyasaharan Sethu, Beena Ahmed
cs.CL, cs.SD, eess.AS
Self-supervised learning has been used to leverage unlabelled data, improving accuracy and generalisation of speech systems through the training of representation models. While many recent works have sought to produce effective representations across a variety of acoustic domains, languages, modalities and even simulta...
http://arxiv.org/pdf/2310.10922v1
2023-10-17T01:31:59
2023-10-17T01:31:59
{ "title": [ 0.0095825195, 0.0055770874, -0.041259766, -0.055633545, -0.016159058, -0.0021648407, -0.047454834, 0.0115356445, -0.029891968, 0.059753418, -0.015052795, 0.008049011, 0.02999878, -0.023117065, -0.028366089, -0.0036830902, -0.008979797, ...
http://arxiv.org/abs/2310.13011v1
Compositional preference models for aligning LMs
Dongyoung Go, Tomasz Korbak, Germán Kruszewski, Jos Rozen, Marc Dymetman
cs.CL, cs.LG
As language models (LMs) become more capable, it is increasingly important to align them with human preferences. However, the dominant paradigm for training Preference Models (PMs) for that purpose suffers from fundamental limitations, such as lack of transparency and scalability, along with susceptibility to overfitti...
http://arxiv.org/pdf/2310.13011v1
2023-10-17T01:31:59
2023-10-17T01:31:59
{ "title": [ -0.032409668, 0.00932312, -0.046936035, -0.048461914, -0.005859375, 0.0079956055, -0.07373047, -0.024719238, -0.0035629272, 0.033843994, -0.018081665, -0.022964478, -0.0034294128, -0.028137207, -0.04135132, -0.012886047, -0.02659607, ...
http://arxiv.org/abs/2310.10920v1
NuclearQA: A Human-Made Benchmark for Language Models for the Nuclear Domain
Anurag Acharya, Sai Munikoti, Aaron Hellinger, Sara Smith, Sridevi Wagle, Sameera Horawalavithana
cs.CL, cs.AI, I.2.7
As LLMs have become increasingly popular, they have been used in almost every field. But as the application for LLMs expands from generic fields to narrow, focused science domains, there exists an ever-increasing gap in ways to evaluate their efficacy in those fields. For the benchmarks that do exist, a lot of them foc...
http://arxiv.org/pdf/2310.10920v1
2023-10-17T01:27:20
2023-10-17T01:27:20
{ "title": [ -0.0473938, -0.007129669, -0.03366089, -0.04348755, -0.008598328, 0.004299164, -0.06994629, -0.009803772, -0.003786087, 0.033111572, 0.02130127, -0.012649536, 0.010231018, -0.038330078, -0.066711426, 0.024261475, -0.0287323, 0.044067...
http://arxiv.org/abs/2310.10903v2
Emergent AI-Assisted Discourse: Case Study of a Second Language Writer Authoring with ChatGPT
Sharin Jacob, Tamara Tate, Mark Warschauer
cs.CL
The rapid proliferation of ChatGPT has incited debates regarding its impact on human writing. Amid concerns about declining writing standards, this study investigates the role of ChatGPT in facilitating academic writing, especially among language learners. Using a case study approach, this study examines the experience...
http://arxiv.org/pdf/2310.10903v2
2023-10-23T17:31:26
2023-10-17T00:22:10
{ "title": [ -0.029144287, -0.016159058, -0.0031089783, 0.003873825, -0.036956787, 0.0107040405, -0.0049095154, -0.019226074, -0.018432617, 0.030670166, 0.0015993118, 0.02822876, -0.022354126, -0.014205933, -0.041015625, 0.006477356, -0.03579712, ...
http://arxiv.org/abs/2310.10873v1
IDEAL: Influence-Driven Selective Annotations Empower In-Context Learners in Large Language Models
Shaokun Zhang, Xiaobo Xia, Zhaoqing Wang, Ling-Hao Chen, Jiale Liu, Qingyun Wu, Tongliang Liu
cs.CL
In-context learning is a promising paradigm that utilizes in-context examples as prompts for the predictions of large language models. These prompts are crucial for achieving strong performance. However, since the prompts need to be sampled from a large volume of annotated examples, finding the right prompt may result ...
http://arxiv.org/pdf/2310.10873v1
2023-10-16T22:53:54
2023-10-16T22:53:54
{ "title": [ 0.017578125, -0.010131836, -0.028289795, -0.093688965, -0.0064582825, -0.0067253113, -0.030593872, 0.017532349, -0.0067749023, 0.018295288, -0.023727417, 0.009147644, 0.008201599, -0.02281189, -0.013336182, -0.011138916, -0.019256592, ...
http://arxiv.org/abs/2310.10865v1
Will the Prince Get True Love's Kiss? On the Model Sensitivity to Gender Perturbation over Fairytale Texts
Christina Chance, Da Yin, Dakuo Wang, Kai-Wei Chang
cs.CL
Recent studies show that traditional fairytales are rife with harmful gender biases. To help mitigate these gender biases in fairytales, this work aims to assess learned biases of language models by evaluating their robustness against gender perturbations. Specifically, we focus on Question Answering (QA) tasks in fair...
http://arxiv.org/pdf/2310.10865v1
2023-10-16T22:25:09
2023-10-16T22:25:09
{ "title": [ 0.02218628, 0.015205383, -0.023269653, 0.031585693, 0.00774765, -0.049804688, 0.045318604, -0.038604736, -0.004573822, 0.017562866, -0.005657196, -0.0019741058, 0.024993896, -0.044677734, -0.020858765, -0.011405945, -0.01285553, 0.01...
http://arxiv.org/abs/2310.10845v1
CoTFormer: More Tokens With Attention Make Up For Less Depth
Amirkeivan Mohtashami, Matteo Pagliardini, Martin Jaggi
cs.CL, cs.LG
The race to continually develop ever larger and deeper foundational models is underway. However, techniques like the Chain-of-Thought (CoT) method continue to play a pivotal role in achieving optimal downstream performance. In this work, we establish an approximate parallel between using chain-of-thought and employing ...
http://arxiv.org/pdf/2310.10845v1
2023-10-16T21:37:34
2023-10-16T21:37:34
{ "title": [ 0.014434814, -0.024032593, -0.04373169, -0.039764404, -0.021820068, -0.017089844, -0.019515991, 0.0007185936, -0.037353516, 0.018081665, -0.018096924, -0.011306763, 0.0058250427, -0.041137695, -0.0074882507, -0.052581787, 0.000023901463,...
http://arxiv.org/abs/2310.10844v1
Survey of Vulnerabilities in Large Language Models Revealed by Adversarial Attacks
Erfan Shayegani, Md Abdullah Al Mamun, Yu Fu, Pedram Zaree, Yue Dong, Nael Abu-Ghazaleh
cs.CL, cs.CR, cs.LG
Large Language Models (LLMs) are swiftly advancing in architecture and capability, and as they integrate more deeply into complex systems, the urgency to scrutinize their security properties grows. This paper surveys research in the emerging interdisciplinary field of adversarial attacks on LLMs, a subfield of trustwor...
http://arxiv.org/pdf/2310.10844v1
2023-10-16T21:37:24
2023-10-16T21:37:24
{ "title": [ -0.016601562, -0.0029449463, -0.031921387, -0.059539795, 0.011886597, -0.004962921, -0.032836914, 0.00844574, -0.029830933, 0.029022217, 0.008460999, 0.011688232, -0.011268616, -0.016357422, -0.034210205, -0.00086069107, -0.014564514, ...
http://arxiv.org/abs/2310.10830v1
Fake News in Sheep's Clothing: Robust Fake News Detection Against LLM-Empowered Style Attacks
Jiaying Wu, Bryan Hooi
cs.CL
It is commonly perceived that online fake news and reliable news exhibit stark differences in writing styles, such as the use of sensationalist versus objective language. However, we emphasize that style-related features can also be exploited for style-based attacks. Notably, the rise of powerful Large Language Models ...
http://arxiv.org/pdf/2310.10830v1
2023-10-16T21:05:12
2023-10-16T21:05:12
{ "title": [ -0.029220581, 0.018295288, -0.054260254, -0.019104004, 0.01927185, -0.014434814, -0.014968872, -0.022018433, -0.056518555, 0.021896362, -0.003917694, 0.051513672, -0.010643005, -0.02961731, -0.0033416748, -0.00982666, -0.030517578, 0...
http://arxiv.org/abs/2310.10803v1
SD-HuBERT: Self-Distillation Induces Syllabic Organization in HuBERT
Cheol Jun Cho, Abdelrahman Mohamed, Shang-Wen Li, Alan W Black, Gopala K. Anumanchipalli
cs.CL, eess.AS
Data-driven unit discovery in self-supervised learning (SSL) of speech has embarked on a new era of spoken language processing. Yet, the discovered units often remain in phonetic space, limiting the utility of SSL representations. Here, we demonstrate that a syllabic organization emerges in learning sentence-level repr...
http://arxiv.org/pdf/2310.10803v1
2023-10-16T20:05:36
2023-10-16T20:05:36
{ "title": [ -0.013755798, 0.021102905, -0.068603516, -0.047546387, -0.02029419, -0.009986877, -0.09106445, -0.004940033, -0.047027588, 0.053253174, 0.0019369125, -0.008834839, 0.027862549, -0.047058105, 0.012008667, -0.025787354, -0.015991211, 0...
http://arxiv.org/abs/2310.10788v1
Self-Supervised Models of Speech Infer Universal Articulatory Kinematics
Cheol Jun Cho, Abdelrahman Mohamed, Alan W Black, Gopala K. Anumanchipalli
eess.AS, cs.CL
Self-Supervised Learning (SSL) based models of speech have shown remarkable performance on a range of downstream tasks. These state-of-the-art models have remained blackboxes, but many recent studies have begun "probing" models like HuBERT, to correlate their internal representations to different aspects of speech. In ...
http://arxiv.org/pdf/2310.10788v1
2023-10-16T19:50:01
2023-10-16T19:50:01
{ "title": [ -0.018218994, 0.010627747, -0.04748535, -0.0569458, -0.025390625, -0.011810303, -0.06536865, 0.016082764, -0.052246094, 0.054382324, -0.00084877014, 0.019424438, 0.010055542, -0.03564453, -0.018203735, 0.0043182373, 0.012969971, 0.05...
http://arxiv.org/abs/2310.10781v1
BanglaNLP at BLP-2023 Task 1: Benchmarking different Transformer Models for Violence Inciting Text Detection in Bengali
Saumajit Saha, Albert Nanda
cs.CL
This paper presents the system that we have developed while solving this shared task on violence inciting text detection in Bangla. We explain both the traditional and the recent approaches that we have used to make our models learn. Our proposed system helps to classify if the given text contains any threat. We studie...
http://arxiv.org/pdf/2310.10781v1
2023-10-16T19:35:04
2023-10-16T19:35:04
{ "title": [ 0.008506775, -0.007675171, -0.013999939, -0.02003479, -0.014205933, -0.01272583, -0.008674622, -0.0057792664, -0.006526947, 0.032470703, -0.033996582, 0.0357666, 0.025802612, -0.022613525, 0.00374794, 0.00026202202, -0.032684326, 0.0...
http://arxiv.org/abs/2310.10765v3
BiomedJourney: Counterfactual Biomedical Image Generation by Instruction-Learning from Multimodal Patient Journeys
Yu Gu, Jianwei Yang, Naoto Usuyama, Chunyuan Li, Sheng Zhang, Matthew P. Lungren, Jianfeng Gao, Hoifung Poon
cs.CV, cs.AI, cs.CL
Rapid progress has been made in instruction-learning for image editing with natural-language instruction, as exemplified by InstructPix2Pix. In biomedicine, such methods can be applied to counterfactual image generation, which helps differentiate causal structure from spurious correlation and facilitate robust image in...
http://arxiv.org/pdf/2310.10765v3
2023-10-21T02:59:47
2023-10-16T18:59:31
{ "title": [ 0.012504578, -0.0023288727, -0.02746582, -0.019546509, -0.009559631, -0.02305603, -0.00061655045, -0.036315918, 0.0124435425, 0.021621704, -0.021774292, 0.015670776, -0.02999878, -0.032928467, -0.03250122, -0.019943237, 0.0053520203, ...
http://arxiv.org/abs/2310.10760v1
Towards reducing hallucination in extracting information from financial reports using Large Language Models
Bhaskarjit Sarmah, Tianjie Zhu, Dhagash Mehta, Stefano Pasquali
cs.CL, q-fin.PM, q-fin.ST, stat.AP
For a financial analyst, the question and answer (Q\&A) segment of the company financial report is a crucial piece of information for various analysis and investment decisions. However, extracting valuable insights from the Q\&A section has posed considerable challenges as the conventional methods such as detailed read...
http://arxiv.org/pdf/2310.10760v1
2023-10-16T18:45:38
2023-10-16T18:45:38
{ "title": [ -0.035858154, -0.003419876, -0.060058594, -0.051239014, -0.013511658, 0.0009813309, -0.06262207, -0.0017957687, 0.01638794, 0.038909912, -0.03378296, -0.007827759, 0.0024909973, 0.009346008, -0.07159424, -0.007534027, -0.011245728, 0...
http://arxiv.org/abs/2310.10735v1
Building Persona Consistent Dialogue Agents with Offline Reinforcement Learning
Ryan Shea, Zhou Yu
cs.CL
Maintaining a consistent persona is a key quality for any open domain dialogue system. Current state-of-the-art systems do this by training agents with supervised learning or online reinforcement learning (RL). However, systems trained with supervised learning often lack consistency as they are never punished for utter...
http://arxiv.org/pdf/2310.10735v1
2023-10-16T18:05:54
2023-10-16T18:05:54
{ "title": [ -0.0151901245, -0.008911133, -0.026489258, -0.040039062, 0.0037441254, 0.0027332306, -0.013809204, -0.0000718236, -0.016830444, 0.016418457, 0.0017318726, -0.008613586, -0.013252258, -0.029708862, -0.027801514, -0.020721436, -0.0362854, ...
http://arxiv.org/abs/2310.10648v1
Step-by-Step Remediation of Students' Mathematical Mistakes
Rose E. Wang, Qingyang Zhang, Carly Robinson, Susanna Loeb, Dorottya Demszky
cs.CL, cs.AI
Scaling high-quality tutoring is a major challenge in education. Because of the growing demand, many platforms employ novice tutors who, unlike professional educators, struggle to effectively address student mistakes and thus fail to seize prime learning opportunities for students. In this paper, we explore the potenti...
http://arxiv.org/pdf/2310.10648v1
2023-10-16T17:59:50
2023-10-16T17:59:50
{ "title": [ 0.048950195, 0.0045280457, -0.0016412735, -0.023071289, -0.022033691, -0.015945435, 0.0090789795, -0.017196655, -0.020812988, 0.04421997, -0.034057617, -0.002779007, 0.033355713, -0.029418945, 0.02406311, -0.019577026, 0.030151367, -...
http://arxiv.org/abs/2310.10645v1
Interactive Task Planning with Language Models
Boyi Li, Philipp Wu, Pieter Abbeel, Jitendra Malik
cs.RO, cs.AI, cs.CL, cs.HC
An interactive robot framework accomplishes long-horizon task planning and can easily generalize to new goals or distinct tasks, even during execution. However, most traditional methods require predefined module design, which makes it hard to generalize to different goals. Recent large language model based approaches c...
http://arxiv.org/pdf/2310.10645v1
2023-10-16T17:59:12
2023-10-16T17:59:12
{ "title": [ -0.04321289, -0.01625061, -0.032226562, -0.0625, -0.0131073, 0.022491455, -0.070373535, 0.00422287, -0.014961243, 0.029876709, -0.009033203, -0.010047913, -0.02027893, -0.012435913, -0.029434204, -0.011161804, 0.0036697388, 0.0109634...
http://arxiv.org/abs/2310.10638v3
In-Context Pretraining: Language Modeling Beyond Document Boundaries
Weijia Shi, Sewon Min, Maria Lomeli, Chunting Zhou, Margaret Li, Xi Victoria Lin, Noah A. Smith, Luke Zettlemoyer, Scott Yih, Mike Lewis
cs.CL, cs.AI, cs.LG
Large language models (LMs) are currently trained to predict tokens given document prefixes, enabling them to directly perform long-form generation and prompting-style tasks which can be reduced to document completion. Existing pretraining pipelines train LMs by concatenating random sets of short documents to create in...
http://arxiv.org/pdf/2310.10638v3
2023-10-20T23:44:56
2023-10-16T17:57:12
{ "title": [ -0.004016876, -0.006160736, -0.028015137, -0.07092285, 0.0033988953, -0.0008983612, -0.047821045, 0.00000423193, 0.005531311, 0.03564453, -0.023101807, 0.017990112, 0.01184845, -0.0152282715, -0.035217285, 0.0085372925, -0.0023860931, ...
http://arxiv.org/abs/2310.10637v1
"Mistakes Help Us Grow": Facilitating and Evaluating Growth Mindset Supportive Language in Classrooms
Kunal Handa, Margaret Clapper, Jessica Boyle, Rose E Wang, Diyi Yang, David S Yeager, Dorottya Demszky
cs.CL
Teachers' growth mindset supportive language (GMSL)--rhetoric emphasizing that one's skills can be improved over time--has been shown to significantly reduce disparities in academic achievement and enhance students' learning outcomes. Although teachers espouse growth mindset principles, most find it difficult to adopt ...
http://arxiv.org/pdf/2310.10637v1
2023-10-16T17:56:07
2023-10-16T17:56:07
{ "title": [ -0.030380249, -0.023452759, -0.0725708, -0.02168274, -0.017456055, -0.035247803, -0.041503906, 0.0178833, -0.011184692, 0.0032596588, -0.013214111, 0.026504517, -0.013999939, -0.02961731, -0.0070610046, -0.022323608, -0.013191223, 0....
http://arxiv.org/abs/2310.10634v1
OpenAgents: An Open Platform for Language Agents in the Wild
Tianbao Xie, Fan Zhou, Zhoujun Cheng, Peng Shi, Luoxuan Weng, Yitao Liu, Toh Jing Hua, Junning Zhao, Qian Liu, Che Liu, Leo Z. Liu, Yiheng Xu, Hongjin Su, Dongchan Shin, Caiming Xiong, Tao Yu
cs.CL, cs.AI
Language agents show potential in being capable of utilizing natural language for varied and intricate tasks in diverse environments, particularly when built upon large language models (LLMs). Current language agent frameworks aim to facilitate the construction of proof-of-concept language agents while neglecting the n...
http://arxiv.org/pdf/2310.10634v1
2023-10-16T17:54:53
2023-10-16T17:54:53
{ "title": [ -0.032806396, -0.019134521, -0.029586792, -0.057769775, 0.008621216, 0.009101868, -0.03152466, 0.0010299683, -0.020614624, 0.04232788, -0.014526367, 0.013252258, 0.0135650635, -0.009178162, -0.033966064, 0.00041604042, -0.015014648, ...
http://arxiv.org/abs/2310.10632v1
BioPlanner: Automatic Evaluation of LLMs on Protocol Planning in Biology
Odhran O'Donoghue, Aleksandar Shtedritski, John Ginger, Ralph Abboud, Ali Essa Ghareeb, Justin Booth, Samuel G Rodriques
cs.CL, cs.AI, cs.RO
The ability to automatically generate accurate protocols for scientific experiments would represent a major step towards the automation of science. Large Language Models (LLMs) have impressive capabilities on a wide range of tasks, such as question answering and the generation of coherent text and code. However, LLMs c...
http://arxiv.org/pdf/2310.10632v1
2023-10-16T17:54:20
2023-10-16T17:54:20
{ "title": [ -0.079589844, -0.021713257, -0.052490234, -0.020736694, -0.036071777, 0.010353088, -0.064086914, -0.0066490173, -0.025604248, -0.0064582825, -0.02760315, 0.019378662, -0.008804321, -0.05130005, 0.025405884, -0.019348145, -0.006969452, ...
http://arxiv.org/abs/2310.10631v1
Llemma: An Open Language Model For Mathematics
Zhangir Azerbayev, Hailey Schoelkopf, Keiran Paster, Marco Dos Santos, Stephen McAleer, Albert Q. Jiang, Jia Deng, Stella Biderman, Sean Welleck
cs.CL, cs.AI, cs.LO
We present Llemma, a large language model for mathematics. We continue pretraining Code Llama on the Proof-Pile-2, a mixture of scientific papers, web data containing mathematics, and mathematical code, yielding Llemma. On the MATH benchmark Llemma outperforms all known open base models, as well as the unreleased Miner...
http://arxiv.org/pdf/2310.10631v1
2023-10-16T17:54:07
2023-10-16T17:54:07
{ "title": [ -0.038391113, -0.00090932846, -0.045074463, -0.05834961, 0.02067566, 0.013641357, -0.033081055, -0.017791748, -0.0015850067, 0.032714844, 0.014831543, -0.0051956177, -0.0044670105, -0.02658081, -0.035064697, 0.01600647, -0.0072669983, ...
http://arxiv.org/abs/2310.10628v1
Data Contamination Through the Lens of Time
Manley Roberts, Himanshu Thakur, Christine Herlihy, Colin White, Samuel Dooley
cs.CL
Recent claims about the impressive abilities of large language models (LLMs) are often supported by evaluating publicly available benchmarks. Since LLMs train on wide swaths of the internet, this practice raises concerns of data contamination, i.e., evaluating on examples that are explicitly or implicitly included in t...
http://arxiv.org/pdf/2310.10628v1
2023-10-16T17:51:29
2023-10-16T17:51:29
{ "title": [ 0.068847656, 0.006931305, -0.035217285, -0.07519531, -0.004863739, -0.043670654, -0.004207611, 0.020568848, -0.011528015, 0.034454346, -0.06903076, -0.022247314, 0.0041122437, -0.017333984, -0.020111084, -0.035247803, 0.011100769, 0....
http://arxiv.org/abs/2310.10627v1
Factored Verification: Detecting and Reducing Hallucination in Summaries of Academic Papers
Charlie George, Andreas Stuhlmüller
cs.CL, cs.AI
Hallucination plagues even frontier LLMs--but how bad is it really for summarizing academic papers? We evaluate Factored Verification, a simple automated method for detecting hallucinations in abstractive summaries. This method sets a new SotA on hallucination detection in the summarization task of the HaluEval benchma...
http://arxiv.org/pdf/2310.10627v1
2023-10-16T17:51:17
2023-10-16T17:51:17
{ "title": [ 0.026870728, 0.016677856, -0.02279663, -0.057373047, 0.015052795, -0.004486084, -0.02859497, -0.03717041, -0.01737976, 0.046875, -0.00048828125, 0.030456543, 0.019180298, -0.0033798218, -0.061035156, 0.010787964, -0.029785156, 0.0615...
http://arxiv.org/abs/2310.10623v1
Generating Summaries with Controllable Readability Levels
Leonardo F. R. Ribeiro, Mohit Bansal, Markus Dreyer
cs.CL, cs.AI, cs.LG
Readability refers to how easily a reader can understand a written text. Several factors affect the readability level, such as the complexity of the text, its subject matter, and the reader's background knowledge. Generating summaries based on different readability levels is critical for enabling knowledge consumption ...
http://arxiv.org/pdf/2310.10623v1
2023-10-16T17:46:26
2023-10-16T17:46:26
{ "title": [ 0.0005965233, 0.007331848, -0.03112793, -0.078308105, -0.0025844574, 0.018157959, -0.06628418, -0.026687622, 0.012298584, 0.03555298, -0.019897461, 0.0063591003, -0.0135269165, 0.009178162, -0.05505371, 0.008346558, -0.011550903, 0.0...
http://arxiv.org/abs/2310.10605v1
ForceGen: End-to-end de novo protein generation based on nonlinear mechanical unfolding responses using a protein language diffusion model
Bo Ni, David L. Kaplan, Markus J. Buehler
cond-mat.mtrl-sci, cond-mat.mes-hall, cs.CL, cs.LG, q-bio.BM
Through evolution, nature has presented a set of remarkable protein materials, including elastins, silks, keratins and collagens with superior mechanical performances that play crucial roles in mechanobiology. However, going beyond natural designs to discover proteins that meet specified mechanical properties remains c...
http://arxiv.org/pdf/2310.10605v1
2023-10-16T17:31:34
2023-10-16T17:31:34
{ "title": [ -0.07110596, -0.018814087, -0.05130005, -0.05218506, -0.016830444, 0.013557434, -0.053741455, 0.02798462, -0.010391235, 0.015457153, 0.018127441, 0.022583008, 0.012413025, -0.040405273, 0.017105103, 0.020446777, 0.008171082, 0.005828...
http://arxiv.org/abs/2310.10594v1
Motion2Language, Unsupervised learning of synchronized semantic motion segmentation
Karim Radouane, Andon Tchechmedjiev, Sylvie Ranwez, Julien Lagarde
cs.CV, cs.CL
In this paper, we investigate building a sequence to sequence architecture for motion to language translation and synchronization. The aim is to translate motion capture inputs into English natural-language descriptions, such that the descriptions are generated synchronously with the actions performed, enabling semanti...
http://arxiv.org/pdf/2310.10594v1
2023-10-16T17:16:32
2023-10-16T17:16:32
{ "title": [ -0.022415161, 0.012458801, -0.068481445, -0.010826111, -0.018966675, -0.03164673, -0.009109497, -0.0071029663, -0.026306152, 0.051452637, -0.024261475, 0.028427124, 0.021865845, -0.022644043, -0.006690979, -0.0047836304, 0.03338623, ...
http://arxiv.org/abs/2310.10590v1
Mastering the Task of Open Information Extraction with Large Language Models and Consistent Reasoning Environment
Ji Qi, Kaixuan Ji, Xiaozhi Wang, Jifan Yu, Kaisheng Zeng, Lei Hou, Juanzi Li, Bin Xu
cs.CL
Open Information Extraction (OIE) aims to extract objective structured knowledge from natural texts, which has attracted growing attention to build dedicated models with human experience. As the large language models (LLMs) have exhibited remarkable in-context learning capabilities, a question arises as to whether the ...
http://arxiv.org/pdf/2310.10590v1
2023-10-16T17:11:42
2023-10-16T17:11:42
{ "title": [ -0.0390625, -0.0043678284, -0.04660034, -0.07043457, 0.005027771, 0.010269165, -0.05606079, -0.0007534027, -0.0158844, 0.024536133, -0.026046753, -0.0041618347, -0.0031337738, -0.02079773, -0.020965576, -0.011154175, -0.0069885254, 0...
http://arxiv.org/abs/2310.10586v1
BiLL-VTG: Bridging Large Language Models and Lightweight Visual Tools for Video-based Texts Generation
Ji Qi, Kaixuan Ji, Jifan Yu, Duokang Wang, Bin Xu, Lei Hou, Juanzi Li
cs.CV, cs.CL
Building models that generate textual responses to user instructions for videos is a practical and challenging topic, as it requires both vision understanding and knowledge reasoning. Compared to language and image modalities, training efficiency remains a serious problem as existing studies train models on massive spa...
http://arxiv.org/pdf/2310.10586v1
2023-10-16T17:05:56
2023-10-16T17:05:56
{ "title": [ -0.032714844, -0.01550293, -0.039123535, -0.062683105, -0.015670776, 0.010437012, -0.06628418, -0.005088806, 0.008590698, 0.04699707, -0.028198242, 0.013442993, 0.01713562, 0.0017690659, -0.056030273, -0.0022220612, -0.010467529, 0.0...
http://arxiv.org/abs/2310.10583v1
Who Are All The Stochastic Parrots Imitating? They Should Tell Us!
Sagi Shaier, Lawrence E. Hunter, Katharina von der Wense
cs.CL, cs.LG
Both standalone language models (LMs) as well as LMs within downstream-task systems have been shown to generate statements which are factually untrue. This problem is especially severe for low-resource languages, where training data is scarce and of worse quality than for high-resource languages. In this opinion piece,...
http://arxiv.org/pdf/2310.10583v1
2023-10-16T16:57:55
2023-10-16T16:57:55
{ "title": [ -0.02293396, -0.026245117, -0.031799316, -0.010292053, -0.025939941, 0.013626099, -0.014007568, 0.0056152344, -0.042144775, 0.0011510849, 0.018203735, 0.02381897, -0.05429077, -0.048217773, 0.0076942444, -0.029251099, 0.01953125, 0.0...
http://arxiv.org/abs/2310.10571v1
Emerging Challenges in Personalized Medicine: Assessing Demographic Effects on Biomedical Question Answering Systems
Sagi Shaier, Kevin Bennett, Lawrence Hunter, Katharina von der Wense
cs.CL, cs.LG
State-of-the-art question answering (QA) models exhibit a variety of social biases (e.g., with respect to sex or race), generally explained by similar issues in their training data. However, what has been overlooked so far is that in the critical domain of biomedicine, any unjustified change in model output due to pati...
http://arxiv.org/pdf/2310.10571v1
2023-10-16T16:45:52
2023-10-16T16:45:52
{ "title": [ 0.009933472, 0.0073661804, -0.029846191, -0.06439209, 0.020843506, 0.00039672852, -0.016738892, 0.009971619, -0.014038086, 0.028411865, -0.00016784668, -0.002840042, -0.005569458, -0.023101807, -0.042510986, 0.005924225, 0.002626419, ...
http://arxiv.org/abs/2310.10570v1
On Position Bias in Summarization with Large Language Models
Mathieu Ravaut, Shafiq Joty, Aixin Sun, Nancy F. Chen
cs.CL
Large language models (LLMs) excel in zero-shot abstractive summarization tasks, delivering fluent and pertinent summaries. Recent advancements have extended their capabilities to handle long-input contexts, surpassing token limits of 32k or more. However, in the realm of multi-document question answering, language mod...
http://arxiv.org/pdf/2310.10570v1
2023-10-16T16:45:12
2023-10-16T16:45:12
{ "title": [ -0.005176544, 0.0018081665, -0.02784729, -0.06854248, -0.008956909, 0.004047394, -0.056915283, -0.008712769, 0.017089844, 0.052703857, -0.020889282, 0.005050659, 0.005748749, 0.0076828003, -0.06329346, 0.010826111, -0.015670776, 0.03...
http://arxiv.org/abs/2310.10567v2
RegaVAE: A Retrieval-Augmented Gaussian Mixture Variational Auto-Encoder for Language Modeling
Jingcheng Deng, Liang Pang, Huawei Shen, Xueqi Cheng
cs.CL
Retrieval-augmented language models show promise in addressing issues like outdated information and hallucinations in language models (LMs). However, current research faces two main problems: 1) determining what information to retrieve, and 2) effectively combining retrieved information during generation. We argue that...
http://arxiv.org/pdf/2310.10567v2
2023-10-23T12:16:44
2023-10-16T16:42:01
{ "title": [ -0.0345459, -0.013198853, -0.034301758, -0.0335083, -0.02784729, 0.00592041, -0.060180664, 0.0085372925, 0.007537842, 0.06414795, 0.005153656, 0.018295288, 0.03543091, -0.011116028, -0.05279541, 0.0345459, -0.005065918, 0.025405884, ...
http://arxiv.org/abs/2310.10707v1
Demonstrations Are All You Need: Advancing Offensive Content Paraphrasing using In-Context Learning
Anirudh Som, Karan Sikka, Helen Gent, Ajay Divakaran, Andreas Kathol, Dimitra Vergyri
cs.CL, cs.AI
Paraphrasing of offensive content is a better alternative to content removal and helps improve civility in a communication environment. Supervised paraphrasers; however, rely heavily on large quantities of labelled data to help preserve meaning and intent. They also retain a large portion of the offensiveness of the or...
http://arxiv.org/pdf/2310.10707v1
2023-10-16T16:18:55
2023-10-16T16:18:55
{ "title": [ 0.042633057, -0.003019333, -0.031097412, -0.05899048, 0.010353088, -0.03314209, -0.0043678284, -0.019332886, -0.013809204, 0.044799805, -0.046844482, 0.01777649, 0.022338867, -0.007209778, -0.023208618, -0.027313232, -0.012779236, 0....
http://arxiv.org/abs/2310.10543v1
ViPE: Visualise Pretty-much Everything
Hassan Shahmohammadi, Adhiraj Ghosh, Hendrik P. A. Lensch
cs.CL, cs.CV
Figurative and non-literal expressions are profoundly integrated in human communication. Visualising such expressions allow us to convey our creative thoughts, and evoke nuanced emotions. Recent text-to-image models like Stable Diffusion, on the other hand, struggle to depict non-literal expressions. Recent works prima...
http://arxiv.org/pdf/2310.10543v1
2023-10-16T16:14:20
2023-10-16T16:14:20
{ "title": [ 0.03173828, -0.015686035, -0.028793335, -0.054718018, -0.013252258, -0.016357422, -0.0043029785, 0.012672424, 0.023406982, 0.046783447, -0.014228821, -0.01525116, 0.02519226, -0.034973145, -0.026565552, -0.00036978722, 0.024093628, 0...
http://arxiv.org/abs/2310.10532v1
One For All & All For One: Bypassing Hyperparameter Tuning with Model Averaging For Cross-Lingual Transfer
Fabian David Schmidt, Ivan Vulić, Goran Glavaš
cs.CL
Multilingual language models enable zero-shot cross-lingual transfer (ZS-XLT): fine-tuned on sizable source-language task data, they perform the task in target languages without labeled instances. The effectiveness of ZS-XLT hinges on the linguistic proximity between languages and the amount of pretraining data for a l...
http://arxiv.org/pdf/2310.10532v1
2023-10-16T15:50:34
2023-10-16T15:50:34
{ "title": [ -0.016601562, -0.00484848, -0.018539429, -0.04067993, -0.01146698, 0.0049934387, -0.054473877, -0.023620605, 0.010131836, 0.040985107, -0.0015792847, 0.009674072, 0.038726807, -0.045288086, -0.005039215, 0.030136108, 0.003522873, 0.0...
http://arxiv.org/abs/2310.10520v2
Semantic Parsing by Large Language Models for Intricate Updating Strategies of Zero-Shot Dialogue State Tracking
Yuxiang Wu, Guanting Dong, Weiran Xu
cs.CL, cs.AI, cs.LG
Zero-shot Dialogue State Tracking (DST) addresses the challenge of acquiring and annotating task-oriented dialogues, which can be time consuming and costly. However, DST extends beyond simple slot-filling and requires effective updating strategies for tracking dialogue state as conversations progress. In this paper, we...
http://arxiv.org/pdf/2310.10520v2
2023-10-18T06:56:29
2023-10-16T15:38:02
{ "title": [ 0.0022182465, -0.023773193, -0.023529053, -0.0814209, -0.022262573, -0.003276825, -0.072509766, 0.014709473, -0.01927185, 0.024795532, 0.013885498, 0.0058174133, -0.0024223328, -0.02999878, -0.02067566, -0.003293991, -0.000664711, 0....
http://arxiv.org/abs/2310.10501v1
NeMo Guardrails: A Toolkit for Controllable and Safe LLM Applications with Programmable Rails
Traian Rebedea, Razvan Dinu, Makesh Sreedhar, Christopher Parisien, Jonathan Cohen
cs.CL, cs.AI
NeMo Guardrails is an open-source toolkit for easily adding programmable guardrails to LLM-based conversational systems. Guardrails (or rails for short) are a specific way of controlling the output of an LLM, such as not talking about topics considered harmful, following a predefined dialogue path, using a particular l...
http://arxiv.org/pdf/2310.10501v1
2023-10-16T15:20:30
2023-10-16T15:20:30
{ "title": [ -0.051818848, -0.018447876, -0.06008911, -0.054016113, -0.0362854, 0.011795044, -0.06304932, 0.030380249, -0.034332275, -0.0016403198, -0.045410156, 0.018875122, -0.010398865, -0.028320312, 0.023971558, -0.030929565, -0.024307251, 0....
http://arxiv.org/abs/2310.10495v1
Metric Ensembles For Hallucination Detection
Grant C. Forbes, Parth Katlana, Zeydy Ortiz
cs.CL
Abstractive text summarization has garnered increased interest as of late, in part due to the proliferation of large language models (LLMs). One of the most pressing problems related to generation of abstractive summaries is the need to reduce "hallucinations," information that was not included in the document being su...
http://arxiv.org/pdf/2310.10495v1
2023-10-16T15:17:22
2023-10-16T15:17:22
{ "title": [ 0.002691269, 0.012954712, -0.010223389, -0.0036621094, 0.0034866333, -0.02053833, 0.02218628, -0.039489746, 0.011306763, 0.05239868, -0.008460999, 0.0022830963, 0.020584106, -0.012245178, -0.07165527, 0.018356323, 0.004173279, 0.0530...
http://arxiv.org/abs/2310.10492v1
UNO-DST: Leveraging Unlabelled Data in Zero-Shot Dialogue State Tracking
Chuang Li, Yan Zhang, Min-Yen Kan, Haizhou Li
cs.CL
Previous zero-shot dialogue state tracking (DST) methods only apply transfer learning, but ignore unlabelled data in the target domain. We transform zero-shot DST into few-shot DST by utilising such unlabelled data via joint and self-training methods. Our method incorporates auxiliary tasks that generate slot types as ...
http://arxiv.org/pdf/2310.10492v1
2023-10-16T15:16:16
2023-10-16T15:16:16
{ "title": [ 0.018493652, -0.008743286, -0.03933716, -0.072387695, -0.010696411, -0.021057129, -0.030044556, -0.009590149, 0.013648987, 0.012260437, -0.012245178, 0.0151901245, -0.021148682, -0.033569336, -0.022735596, 0.0039596558, -0.011993408, ...
http://arxiv.org/abs/2310.10706v2
Harnessing the Power of LLMs: Evaluating Human-AI Text Co-Creation through the Lens of News Headline Generation
Zijian Ding, Alison Smith-Renner, Wenjuan Zhang, Joel R. Tetreault, Alejandro Jaimes
cs.CL, cs.AI
To explore how humans can best leverage LLMs for writing and how interacting with these models affects feelings of ownership and trust in the writing process, we compared common human-AI interaction types (e.g., guiding system, selecting from system outputs, post-editing outputs) in the context of LLM-assisted news hea...
http://arxiv.org/pdf/2310.10706v2
2023-10-18T00:55:57
2023-10-16T15:11:01
{ "title": [ -0.055877686, -0.0032100677, -0.041412354, 0.0051841736, -0.035827637, -0.012023926, -0.031097412, -0.017211914, -0.0357666, 0.037628174, -0.040618896, 0.042755127, -0.02420044, 0.0015497208, -0.04559326, -0.026351929, -0.02861023, 0...
http://arxiv.org/abs/2310.10487v1
Type-aware Decoding via Explicitly Aggregating Event Information for Document-level Event Extraction
Gang Zhao, Yidong Shi, Shudong Lu, Xinjie Yang, Guanting Dong, Jian Xu, Xiaocheng Gong, Si Li
cs.CL, cs.AI, cs.IR, cs.LG
Document-level event extraction (DEE) faces two main challenges: arguments-scattering and multi-event. Although previous methods attempt to address these challenges, they overlook the interference of event-unrelated sentences during event detection and neglect the mutual interference of different event roles during arg...
http://arxiv.org/pdf/2310.10487v1
2023-10-16T15:10:42
2023-10-16T15:10:42
{ "title": [ -0.019119263, 0.000182271, -0.05404663, -0.107910156, 0.0076828003, 0.0070724487, -0.086242676, 0.008956909, 0.036987305, 0.037109375, 0.0023078918, -0.0037460327, 0.027557373, -0.019454956, -0.049957275, 0.032165527, -0.0020446777, ...
http://arxiv.org/abs/2310.10482v1
xCOMET: Transparent Machine Translation Evaluation through Fine-grained Error Detection
Nuno M. Guerreiro, Ricardo Rei, Daan van Stigt, Luisa Coheur, Pierre Colombo, André F. T. Martins
cs.CL
Widely used learned metrics for machine translation evaluation, such as COMET and BLEURT, estimate the quality of a translation hypothesis by providing a single sentence-level score. As such, they offer little insight into translation errors (e.g., what are the errors and what is their severity). On the other hand, gen...
http://arxiv.org/pdf/2310.10482v1
2023-10-16T15:03:14
2023-10-16T15:03:14
{ "title": [ -0.04714966, -0.015602112, -0.017211914, -0.053710938, -0.00033926964, 0.022125244, -0.056396484, -0.0064811707, 0.008605957, 0.04510498, -0.024429321, 0.008934021, 0.030319214, -0.0065193176, -0.061035156, 0.014305115, -0.02154541, ...
http://arxiv.org/abs/2310.10481v1
DemoSG: Demonstration-enhanced Schema-guided Generation for Low-resource Event Extraction
Gang Zhao, Xiaocheng Gong, Xinjie Yang, Guanting Dong, Shudong Lu, Si Li
cs.CL, cs.AI, cs.IR, cs.LG
Most current Event Extraction (EE) methods focus on the high-resource scenario, which requires a large amount of annotated data and can hardly be applied to low-resource domains. To address EE more effectively with limited resources, we propose the Demonstration-enhanced Schema-guided Generation (DemoSG) model, which b...
http://arxiv.org/pdf/2310.10481v1
2023-10-16T15:02:37
2023-10-16T15:02:37
{ "title": [ -0.00894165, -0.010307312, -0.06304932, -0.115112305, 0.019439697, 0.0060272217, -0.07623291, 0.011299133, 0.022125244, 0.025527954, 0.009918213, 0.0017366409, -0.0011796951, -0.034454346, -0.036865234, 0.020339966, -0.007068634, 0.0...
http://arxiv.org/abs/2310.10480v1
G-SPEED: General SParse Efficient Editing MoDel
Haoke Zhang, Yue Wang, Juntao Li, Xiabing Zhou, Min Zhang
cs.CL
Large Language Models~(LLMs) have demonstrated incredible capabilities in understanding, generating, and manipulating languages. Through human-model interactions, LLMs can automatically understand human-issued instructions and output the expected contents, which can significantly increase working efficiency. In various...
http://arxiv.org/pdf/2310.10480v1
2023-10-16T15:01:18
2023-10-16T15:01:18
{ "title": [ -0.062347412, -0.01309967, -0.05960083, -0.034179688, -0.009895325, 0.0020198822, -0.042114258, -0.0072135925, 0.01424408, 0.039093018, 0.008735657, -0.013282776, 0.02748108, -0.031707764, -0.034088135, 0.024032593, -0.0127334595, 0....
http://arxiv.org/abs/2310.10477v2
Gaining Wisdom from Setbacks: Aligning Large Language Models via Mistake Analysis
Kai Chen, Chunwei Wang, Kuo Yang, Jianhua Han, Lanqing Hong, Fei Mi, Hang Xu, Zhengying Liu, Wenyong Huang, Zhenguo Li, Dit-Yan Yeung, Lifeng Shang, Xin Jiang, Qun Liu
cs.CL, cs.AI, cs.LG
The rapid advancement of large language models (LLMs) presents both opportunities and challenges, particularly concerning unintentional generation of harmful and toxic responses. While the traditional alignment methods strive to steer LLMs towards desired performance and shield them from malicious content, this study p...
http://arxiv.org/pdf/2310.10477v2
2023-10-20T13:50:24
2023-10-16T14:59:10
{ "title": [ -0.042388916, -0.012557983, -0.054656982, -0.049743652, -0.0038070679, 0.0029182434, -0.073913574, -0.011978149, 0.015029907, 0.030227661, -0.01776123, -0.0021972656, 0.013664246, -0.013656616, -0.046325684, 0.003534317, -0.017181396, ...
http://arxiv.org/abs/2310.10467v1
Stance Detection with Collaborative Role-Infused LLM-Based Agents
Xiaochong Lan, Chen Gao, Depeng Jin, Yong Li
cs.CL, cs.AI
Stance detection automatically detects the stance in a text towards a target, vital for content analysis in web and social media research. Despite their promising capabilities, LLMs encounter challenges when directly applied to stance detection. First, stance detection demands multi-aspect knowledge, from deciphering e...
http://arxiv.org/pdf/2310.10467v1
2023-10-16T14:46:52
2023-10-16T14:46:52
{ "title": [ -0.044799805, -0.009788513, -0.048797607, -0.04547119, -0.02406311, -0.00069761276, -0.056732178, -0.0011940002, 0.015380859, 0.022201538, -0.040283203, 0.0029945374, -0.015357971, -0.013374329, -0.034362793, -0.018875122, -0.030227661, ...
http://arxiv.org/abs/2310.10449v2
Text Summarization Using Large Language Models: A Comparative Study of MPT-7b-instruct, Falcon-7b-instruct, and OpenAI Chat-GPT Models
Lochan Basyal, Mihir Sanghvi
cs.CL, cs.AI, cs.LG
Text summarization is a critical Natural Language Processing (NLP) task with applications ranging from information retrieval to content generation. Leveraging Large Language Models (LLMs) has shown remarkable promise in enhancing summarization techniques. This paper embarks on an exploration of text summarization with ...
http://arxiv.org/pdf/2310.10449v2
2023-10-17T19:54:16
2023-10-16T14:33:02
{ "title": [ -0.019256592, -0.0030117035, -0.0037899017, -0.044036865, -0.027160645, 0.020050049, -0.055786133, -0.020004272, 0.0025558472, 0.039245605, -0.019500732, 0.028839111, 0.016296387, 0.00071668625, -0.028884888, 0.015525818, -0.014526367, ...
http://arxiv.org/abs/2310.10445v1
MechGPT, a language-based strategy for mechanics and materials modeling that connects knowledge across scales, disciplines and modalities
Markus J. Buehler
cs.CL, cond-mat.mtrl-sci
For centuries, researchers have sought out ways to connect disparate areas of knowledge. While early scholars (Galileo, da Vinci, etc.) were experts across fields, specialization has taken hold later. With the advent of Artificial Intelligence, we can now explore relationships across areas (e.g., mechanics-biology) or ...
http://arxiv.org/pdf/2310.10445v1
2023-10-16T14:29:35
2023-10-16T14:29:35
{ "title": [ -0.014549255, -0.04272461, -0.038513184, -0.014961243, -0.060150146, -0.012138367, -0.04437256, 0.008354187, 0.0049476624, 0.028549194, 0.020965576, 0.041381836, 0.022384644, -0.04333496, 0.05203247, 0.029205322, 0.014945984, -0.0426...
http://arxiv.org/abs/2310.10429v1
Exploiting User Comments for Early Detection of Fake News Prior to Users' Commenting
Qiong Nan, Qiang Sheng, Juan Cao, Yongchun Zhu, Danding Wang, Guang Yang, Jintao Li, Kai Shu
cs.CL, cs.CY, cs.SI
Both accuracy and timeliness are key factors in detecting fake news on social media. However, most existing methods encounter an accuracy-timeliness dilemma: Content-only methods guarantee timeliness but perform moderately because of limited available information, while social context-based ones generally perform bette...
http://arxiv.org/pdf/2310.10429v1
2023-10-16T14:13:38
2023-10-16T14:13:38
{ "title": [ 0.010902405, 0.005290985, -0.030899048, -0.038909912, 0.02949524, -0.013771057, 0.003824234, -0.027618408, -0.058166504, 0.03161621, -0.031204224, 0.024993896, -0.019714355, 0.000052452087, -0.03010559, -0.025390625, -0.025436401, 0....
http://arxiv.org/abs/2310.10400v1
Can Word Sense Distribution Detect Semantic Changes of Words?
Xiaohang Tang, Yi Zhou, Taichi Aida, Procheta Sen, Danushka Bollegala
cs.CL, cs.AI, cs.LG
Semantic Change Detection (SCD) of words is an important task for various NLP applications that must make time-sensitive predictions. Some words are used over time in novel ways to express new meanings, and these new meanings establish themselves as novel senses of existing words. On the other hand, Word Sense Disambig...
http://arxiv.org/pdf/2310.10400v1
2023-10-16T13:41:27
2023-10-16T13:41:27
{ "title": [ -0.03717041, 0.01864624, -0.06512451, 0.02545166, -0.023605347, -0.020965576, -0.019165039, -0.027679443, 0.010444641, 0.051116943, -0.04675293, 0.0023460388, 0.044555664, -0.008529663, -0.018569946, 0.005569458, 0.018234253, 0.03872...
http://arxiv.org/abs/2310.10397v1
$\textit{Swap and Predict}$ -- Predicting the Semantic Changes in Words across Corpora by Context Swapping
Taichi Aida, Danushka Bollegala
cs.CL
Meanings of words change over time and across domains. Detecting the semantic changes of words is an important task for various NLP applications that must make time-sensitive predictions. We consider the problem of predicting whether a given target word, $w$, changes its meaning between two different text corpora, $\ma...
http://arxiv.org/pdf/2310.10397v1
2023-10-16T13:39:44
2023-10-16T13:39:44
{ "title": [ -0.006378174, 0.012863159, -0.07067871, -0.022018433, -0.004257202, -0.024169922, -0.04623413, -0.012336731, 0.0030174255, 0.040405273, -0.05303955, 0.039276123, 0.041503906, -0.00023889542, 0.015411377, 0.0022621155, 0.0065689087, 0...
http://arxiv.org/abs/2310.10385v2
Towards a Better Understanding of Variations in Zero-Shot Neural Machine Translation Performance
Shaomu Tan, Christof Monz
cs.CL, cs.LG
Multilingual Neural Machine Translation (MNMT) facilitates knowledge sharing but often suffers from poor zero-shot (ZS) translation qualities. While prior work has explored the causes of overall low ZS performance, our work introduces a fresh perspective: the presence of high variations in ZS performance. This suggests...
http://arxiv.org/pdf/2310.10385v2
2023-10-31T10:35:17
2023-10-16T13:26:05
{ "title": [ -0.027908325, -0.010437012, -0.014892578, -0.029373169, -0.010131836, -0.005996704, -0.0234375, 0.008453369, -0.0020008087, 0.060028076, 0.01033783, 0.0061836243, 0.017807007, -0.018371582, -0.06109619, 0.03353882, 0.0003130436, 0.03...
http://arxiv.org/abs/2310.10383v1
Privacy in Large Language Models: Attacks, Defenses and Future Directions
Haoran Li, Yulin Chen, Jinglong Luo, Yan Kang, Xiaojin Zhang, Qi Hu, Chunkit Chan, Yangqiu Song
cs.CL, cs.CR
The advancement of large language models (LLMs) has significantly enhanced the ability to effectively tackle various downstream NLP tasks and unify these tasks into generative pipelines. On the one hand, powerful language models, trained on massive textual data, have brought unparalleled accessibility and usability for...
http://arxiv.org/pdf/2310.10383v1
2023-10-16T13:23:54
2023-10-16T13:23:54
{ "title": [ -0.027420044, -0.019515991, -0.046691895, -0.07421875, 0.01020813, -0.0039978027, -0.033935547, 0.032562256, -0.016357422, 0.024612427, -0.006965637, 0.012016296, 0.004234314, -0.01739502, -0.024307251, 0.004497528, -0.019119263, 0.0...
http://arxiv.org/abs/2310.10380v1
Contextual Data Augmentation for Task-Oriented Dialog Systems
Dustin Axman, Avik Ray, Shubham Garg, Jing Huang
cs.CL
Collection of annotated dialogs for training task-oriented dialog systems have been one of the key bottlenecks in improving current models. While dialog response generation has been widely studied on the agent side, it is not evident if similar generative models can be used to generate a large variety of, and often une...
http://arxiv.org/pdf/2310.10380v1
2023-10-16T13:22:34
2023-10-16T13:22:34
{ "title": [ 0.015945435, 0.004512787, -0.068725586, -0.09454346, 0.037322998, -0.018737793, -0.03201294, 0.027389526, 0.022201538, 0.057495117, -0.023452759, -0.0042381287, 0.012878418, -0.0055274963, -0.062316895, 0.002609253, -0.009010315, 0.0...
http://arxiv.org/abs/2310.10378v4
Cross-Lingual Consistency of Factual Knowledge in Multilingual Language Models
Jirui Qi, Raquel Fernández, Arianna Bisazza
cs.CL, cs.AI, cs.HC, cs.LG
Multilingual large-scale Pretrained Language Models (PLMs) have been shown to store considerable amounts of factual knowledge, but large variations are observed across languages. With the ultimate goal of ensuring that users with different language backgrounds obtain consistent feedback from the same model, we study th...
http://arxiv.org/pdf/2310.10378v4
2023-11-09T12:04:45
2023-10-16T13:19:17
{ "title": [ -0.038391113, -0.013206482, -0.047607422, -0.07470703, -0.00046348572, -0.0021800995, -0.07141113, 0.004989624, 0.0087890625, 0.052001953, -0.0060806274, -0.002439499, 0.03338623, -0.012176514, -0.0541687, 0.0206604, -0.02003479, 0.0...
http://arxiv.org/abs/2310.10358v1
Tabular Representation, Noisy Operators, and Impacts on Table Structure Understanding Tasks in LLMs
Ananya Singha, José Cambronero, Sumit Gulwani, Vu Le, Chris Parnin
cs.CL, cs.AI
Large language models (LLMs) are increasingly applied for tabular tasks using in-context learning. The prompt representation for a table may play a role in the LLMs ability to process the table. Inspired by prior work, we generate a collection of self-supervised structural tasks (e.g. navigate to a cell and row; transp...
http://arxiv.org/pdf/2310.10358v1
2023-10-16T12:51:24
2023-10-16T12:51:24
{ "title": [ -0.037139893, 0.0017938614, -0.032409668, -0.050323486, -0.013519287, 0.0043182373, -0.053253174, -0.0059165955, 0.025039673, 0.023468018, -0.04989624, -0.00074243546, 0.019012451, -0.013572693, -0.022033691, -0.017868042, 0.007041931, ...
http://arxiv.org/abs/2310.10348v1
Attribution Patching Outperforms Automated Circuit Discovery
Aaquib Syed, Can Rager, Arthur Conmy
cs.LG, cs.AI, cs.CL
Automated interpretability research has recently attracted attention as a potential research direction that could scale explanations of neural network behavior to large models. Existing automated circuit discovery work applies activation patching to identify subnetworks responsible for solving specific tasks (circuits)...
http://arxiv.org/pdf/2310.10348v1
2023-10-16T12:34:43
2023-10-16T12:34:43
{ "title": [ 0.012062073, -0.0022964478, -0.03564453, -0.023269653, -0.017028809, -0.028396606, -0.02180481, -0.033447266, -0.044525146, 0.028015137, -0.034362793, 0.022705078, -0.012748718, -0.025039673, 0.016067505, -0.010223389, 0.0033035278, ...
http://arxiv.org/abs/2310.10333v3
Legal NLP Meets MiCAR: Advancing the Analysis of Crypto White Papers
Carolina Camassa
cs.CY, cs.CL, q-fin.GN
In the rapidly evolving field of crypto assets, white papers are essential documents for investor guidance, and are now subject to unprecedented content requirements under the European Union's Markets in Crypto-Assets Regulation (MiCAR). Natural Language Processing (NLP) can serve as a powerful tool for both analyzing ...
http://arxiv.org/pdf/2310.10333v3
2023-10-25T08:18:40
2023-10-16T12:17:11
{ "title": [ -0.047821045, 0.0050697327, -0.09039307, -0.008453369, -0.014579773, -0.02519226, -0.03665161, -0.0008134842, -0.018005371, 0.030548096, -0.040161133, 0.018981934, 0.039978027, -0.022064209, -0.0018920898, -0.0030288696, -0.021148682, ...
http://arxiv.org/abs/2310.10704v1
Optimized Tokenization for Transcribed Error Correction
Tomer Wullach, Shlomo E. Chazan
cs.CL, cs.SD, eess.AS
The challenges facing speech recognition systems, such as variations in pronunciations, adverse audio conditions, and the scarcity of labeled data, emphasize the necessity for a post-processing step that corrects recurring errors. Previous research has shown the advantages of employing dedicated error correction models...
http://arxiv.org/pdf/2310.10704v1
2023-10-16T12:14:21
2023-10-16T12:14:21
{ "title": [ -0.039276123, -0.011253357, -0.04034424, -0.08013916, -0.005340576, 0.0093688965, -0.070129395, 0.008995056, 0.00605011, 0.06008911, -0.031555176, 0.0040397644, 0.00919342, -0.0016021729, -0.05697632, 0.009262085, -0.018936157, 0.032...
http://arxiv.org/abs/2310.10322v1
Untying the Reversal Curse via Bidirectional Language Model Editing
Jun-Yu Ma, Jia-Chen Gu, Zhen-Hua Ling, Quan Liu, Cong Liu
cs.CL
Recent studies have demonstrated that large language models (LLMs) store massive factual knowledge within their parameters. But existing LLMs are prone to hallucinate unintended text due to false or outdated knowledge. Since retraining LLMs is resource intensive, there has been a growing interest in the concept of mode...
http://arxiv.org/pdf/2310.10322v1
2023-10-16T12:04:13
2023-10-16T12:04:13
{ "title": [ -0.0236969, -0.012275696, -0.022247314, -0.03930664, -0.025222778, 0.011734009, -0.050994873, -0.02319336, 0.0025157928, 0.055877686, 0.00028586388, -0.033111572, 0.06378174, -0.02293396, -0.051757812, 0.021774292, -0.022262573, 0.04...
http://arxiv.org/abs/2310.10318v1
Interpreting and Exploiting Functional Specialization in Multi-Head Attention under Multi-task Learning
Chong Li, Shaonan Wang, Yunhao Zhang, Jiajun Zhang, Chengqing Zong
cs.CL, cs.AI
Transformer-based models, even though achieving super-human performance on several downstream tasks, are often regarded as a black box and used as a whole. It is still unclear what mechanisms they have learned, especially their core module: multi-head attention. Inspired by functional specialization in the human brain,...
http://arxiv.org/pdf/2310.10318v1
2023-10-16T11:55:53
2023-10-16T11:55:53
{ "title": [ 0.009460449, 0.023529053, -0.047546387, -0.060424805, 0.024398804, -0.03982544, -0.0049476624, -0.00459671, -0.023284912, 0.04800415, 0.00031495094, 0.027679443, -0.0095825195, -0.0042877197, -0.04345703, 0.0050354004, -0.02079773, 0...
http://arxiv.org/abs/2310.10310v1
Investigating Bias in Multilingual Language Models: Cross-Lingual Transfer of Debiasing Techniques
Manon Reusens, Philipp Borchert, Margot Mieskes, Jochen De Weerdt, Bart Baesens
cs.CL
This paper investigates the transferability of debiasing techniques across different languages within multilingual models. We examine the applicability of these techniques in English, French, German, and Dutch. Using multilingual BERT (mBERT), we demonstrate that cross-lingual transfer of debiasing techniques is not on...
http://arxiv.org/pdf/2310.10310v1
2023-10-16T11:43:30
2023-10-16T11:43:30
{ "title": [ -0.024276733, 0.0024280548, -0.030227661, -0.050842285, -0.0036067963, -0.00020062923, -0.0703125, -0.038970947, 0.0053100586, 0.039764404, -0.008995056, -0.00086307526, 0.03439331, -0.029434204, -0.033599854, 0.0143966675, -0.032989502,...
http://arxiv.org/abs/2310.10294v1
Key-phrase boosted unsupervised summary generation for FinTech organization
Aadit Deshpande, Shreya Goyal, Prateek Nagwanshi, Avinash Tripathy
cs.CL, cs.AI
With the recent advances in social media, the use of NLP techniques in social media data analysis has become an emerging research direction. Business organizations can particularly benefit from such an analysis of social media discourse, providing an external perspective on consumer behavior. Some of the NLP applicatio...
http://arxiv.org/pdf/2310.10294v1
2023-10-16T11:30:47
2023-10-16T11:30:47
{ "title": [ -0.029022217, 0.003692627, -0.06359863, -0.043273926, -0.01701355, 0.009742737, -0.06451416, -0.0029296875, -0.016021729, 0.05307007, -0.03753662, 0.042175293, 0.040130615, 0.0036582947, -0.019226074, 0.0075149536, -0.024154663, 0.02...
http://arxiv.org/abs/2310.10285v1
Multi-Stage Pre-training Enhanced by ChatGPT for Multi-Scenario Multi-Domain Dialogue Summarization
Weixiao Zhou, Gengyao Li, Xianfu Cheng, Xinnian Liang, Junnan Zhu, Feifei Zhai, Zhoujun Li
cs.CL
Dialogue summarization involves a wide range of scenarios and domains. However, existing methods generally only apply to specific scenarios or domains. In this study, we propose a new pre-trained model specifically designed for multi-scenario multi-domain dialogue summarization. It adopts a multi-stage pre-training str...
http://arxiv.org/pdf/2310.10285v1
2023-10-16T11:16:07
2023-10-16T11:16:07
{ "title": [ 0.03024292, -0.010307312, -0.018478394, -0.0982666, -0.012550354, 0.008384705, -0.059173584, 0.026565552, -0.034576416, 0.035980225, 0.02571106, 0.03845215, -0.0053749084, -0.026992798, -0.00818634, 0.012161255, -0.03100586, 0.043640...
http://arxiv.org/abs/2310.10266v1
Generative Calibration for In-context Learning
Zhongtao Jiang, Yuanzhe Zhang, Cao Liu, Jun Zhao, Kang Liu
cs.CL
As one of the most exciting features of large language models (LLMs), in-context learning is a mixed blessing. While it allows users to fast-prototype a task solver with only a few training examples, the performance is generally sensitive to various configurations of the prompt such as the choice or order of the traini...
http://arxiv.org/pdf/2310.10266v1
2023-10-16T10:45:02
2023-10-16T10:45:02
{ "title": [ -0.015457153, 0.0029716492, -0.049591064, -0.0413208, 0.008666992, -0.018218994, -0.025466919, -0.02142334, 0.012580872, 0.025543213, -0.053009033, 0.017868042, -0.033447266, -0.012969971, -0.037353516, -0.011726379, 0.0036258698, 0....
http://arxiv.org/abs/2310.10262v1
Enhancing Interpretability using Human Similarity Judgements to Prune Word Embeddings
Natalia Flechas Manrique, Wanqian Bao, Aurelie Herbelot, Uri Hasson
cs.CL
Interpretability methods in NLP aim to provide insights into the semantics underlying specific system architectures. Focusing on word embeddings, we present a supervised-learning method that, for a given domain (e.g., sports, professions), identifies a subset of model features that strongly improve prediction of human ...
http://arxiv.org/pdf/2310.10262v1
2023-10-16T10:38:49
2023-10-16T10:38:49
{ "title": [ -0.009346008, 0.0035247803, -0.02079773, -0.04043579, -0.00030446053, 0.0012521744, -0.03955078, -0.019851685, 0.017578125, 0.00945282, -0.03387451, -0.011329651, 0.008003235, -0.0028934479, -0.039886475, -0.013694763, -0.004627228, ...
http://arxiv.org/abs/2310.10260v1
Prediction of Arabic Legal Rulings using Large Language Models
Adel Ammar, Anis Koubaa, Bilel Benjdira, Omar Najar, Serry Sibaee
cs.CL, cs.AI, cs.LG
In the intricate field of legal studies, the analysis of court decisions is a cornerstone for the effective functioning of the judicial system. The ability to predict court outcomes helps judges during the decision-making process and equips lawyers with invaluable insights, enhancing their strategic approaches to cases...
http://arxiv.org/pdf/2310.10260v1
2023-10-16T10:37:35
2023-10-16T10:37:35
{ "title": [ -0.03781128, -0.017227173, -0.036346436, -0.07110596, -0.00843811, 0.012756348, -0.06402588, 0.0181427, 0.009346008, 0.029693604, -0.01737976, -0.015853882, 0.013374329, -0.020065308, -0.050048828, 0.0031013489, -0.0060043335, 0.0370...
http://arxiv.org/abs/2310.10226v1
Repetition In Repetition Out: Towards Understanding Neural Text Degeneration from the Data Perspective
Huayang Li, Tian Lan, Zihao Fu, Deng Cai, Lemao Liu, Nigel Collier, Taro Watanabe, Yixuan Su
cs.CL
There are a number of diverging hypotheses about the neural text degeneration problem, i.e., generating repetitive and dull loops, which makes this problem both interesting and confusing. In this work, we aim to advance our understanding by presenting a straightforward and fundamental explanation from the data perspect...
http://arxiv.org/pdf/2310.10226v1
2023-10-16T09:35:42
2023-10-16T09:35:42
{ "title": [ 0.0084991455, 0.008026123, -0.022964478, -0.056610107, 0.008979797, -0.003227234, -0.021087646, -0.00982666, 0.0077705383, 0.05795288, -0.027832031, 0.0037822723, 0.030136108, -0.00091934204, -0.07434082, 0.01625061, -0.01852417, 0.0...