id
stringlengths
33
34
doc_url
stringlengths
33
34
title
stringlengths
10
211
publication_date
timestamp[s]
update_date
timestamp[s]
authors
stringlengths
5
20.1k
category_primary
stringclasses
78 values
category_all
stringlengths
5
138
abstract
stringlengths
123
3.09k
embeddings
listlengths
768
768
http://arxiv.org/abs/2407.00407v1
http://arxiv.org/pdf/2407.00407v1
SHADE: Semantic Hypernym Annotator for Domain-specific Entities -- DnD Domain Use Case
2024-06-29T11:21:40
2024-06-29T11:21:40
Akila Peiris, Nisansa de Silva
cs.CE
cs.CE, cs.CL
Manual data annotation is an important NLP task but one that takes considerable amount of resources and effort. In spite of the costs, labeling and categorizing entities is essential for NLP tasks such as semantic evaluation. Even though annotation can be done by non-experts in most cases, due to the fact that this req...
[ -0.530375421, -0.5285475254, 0.1601629108, 0.765894413, -0.0562141091, 0.1449781358, -0.1699619889, -0.669003129, 0.5684549809, 1.0533310175, -0.793761611, -0.9486752748, -0.2863430381, 0.4691395462, -0.2375213355, 1.2467778921, -0.3023647964, -0.5471509099, -0.2972345948, -0.3...
http://arxiv.org/abs/2407.00402v1
http://arxiv.org/pdf/2407.00402v1
Is It Really Long Context if All You Need Is Retrieval? Towards Genuinely Difficult Long Context NLP
2024-06-29T11:09:47
2024-06-29T11:09:47
Omer Goldman, Alon Jacovi, Aviv Slobodkin, Aviya Maimon, Ido Dagan, Reut Tsarfaty
cs.CL
cs.CL, cs.AI
Improvements in language models' capabilities have pushed their applications towards longer contexts, making long-context evaluation and development an active research area. However, many disparate use-cases are grouped together under the umbrella term of "long-context", defined simply by the total length of the model'...
[ -0.4261831939, -0.8956898451, 0.8642889261, 0.4603164792, -0.4345272481, 0.1087634712, -0.6459499002, -0.7305788994, 0.1856112182, 0.298091501, -0.31791538, -0.0077288547, -0.75928092, 0.3038070202, -0.6153059006, 0.8959131837, -0.1221834868, -0.3906239867, -0.3838150799, -0.47...
http://arxiv.org/abs/2407.00396v1
http://arxiv.org/pdf/2407.00396v1
A Study on Effect of Reference Knowledge Choice in Generating Technical Content Relevant to SAPPhIRE Model Using Large Language Model
2024-06-29T10:46:01
2024-06-29T10:46:01
Kausik Bhattacharya, Anubhab Majumder, Amaresh Chakrabarti
cs.CL
cs.CL, cs.AI
Representation of systems using the SAPPhIRE model of causality can be an inspirational stimulus in design. However, creating a SAPPhIRE model of a technical or a natural system requires sourcing technical knowledge from multiple technical documents regarding how the system works. This research investigates how to gene...
[ -0.3856601417, -0.8508597612, 0.6615054011, 0.2339722067, -0.4464017749, -0.2742822468, -0.1888784617, -0.5355609655, 0.4282794297, 0.5683755279, -0.405239284, -0.3858810365, -0.1402432621, 0.0683895722, -0.4707641602, 0.7123535872, -0.0077906679, 0.1822674572, 0.0110845305, 0....
http://arxiv.org/abs/2407.00390v1
http://arxiv.org/pdf/2407.00390v1
Advancing Process Verification for Large Language Models via Tree-Based Preference Learning
2024-06-29T10:09:49
2024-06-29T10:09:49
Mingqian He, Yongliang Shen, Wenqi Zhang, Zeqi Tan, Weiming Lu
cs.CL
cs.CL
Large Language Models (LLMs) have demonstrated remarkable potential in handling complex reasoning tasks by generating step-by-step rationales.Some methods have proven effective in boosting accuracy by introducing extra verifiers to assess these paths. However, existing verifiers, typically trained on binary-labeled rea...
[ -0.1404267251, -0.7642815709, 0.8061933517, -0.1555777639, -0.157294631, 0.2526028156, -0.321856916, -0.0682645142, -0.3116451204, 0.45664078, -0.4091957808, -0.6147741675, -0.5695640445, 0.071356535, 0.0484999716, 0.8253999352, -0.1883194298, -0.1633527577, -0.3123868108, -0.0...
http://arxiv.org/abs/2407.00379v1
http://arxiv.org/pdf/2407.00379v1
GraphArena: Benchmarking Large Language Models on Graph Computational Problems
2024-06-29T09:19:23
2024-06-29T09:19:23
Jianheng Tang, Qifan Zhang, Yuhan Li, Jia Li
cs.AI
cs.AI, cs.CL
The "arms race" of Large Language Models (LLMs) demands novel, challenging, and diverse benchmarks to faithfully examine their progresses. We introduce GraphArena, a benchmarking tool designed to evaluate LLMs on graph computational problems using million-scale real-world graphs from diverse scenarios such as knowledge...
[ -0.3075500727, -0.84767735, 0.7038660049, 0.2347865999, 0.0245022476, 0.573489964, -0.3602790833, -0.5788908005, 0.1956716329, 0.5657525659, -0.2804252207, -0.6690578461, -0.8333226442, 0.1608007103, -0.4894354939, 0.4904835522, -0.2045107186, 0.1295318305, -0.2125887722, -0.32...
http://arxiv.org/abs/2407.00377v1
http://arxiv.org/pdf/2407.00377v1
The Factuality Tax of Diversity-Intervened Text-to-Image Generation: Benchmark and Fact-Augmented Intervention
2024-06-29T09:09:42
2024-06-29T09:09:42
Yixin Wan, Di Wu, Haoran Wang, Kai-Wei Chang
cs.CL
cs.CL, cs.AI, cs.CV, cs.CY
Prompt-based "diversity interventions" are commonly adopted to improve the diversity of Text-to-Image (T2I) models depicting individuals with various racial or gender traits. However, will this strategy result in nonfactual demographic distribution, especially when generating real historical figures? In this work, we p...
[ -0.3907117844, -0.3469836414, 0.3066047728, 0.2752485275, 0.0082350653, 0.1369840205, -0.135267511, -0.1424660236, 0.0964031145, 0.4438645244, -0.4932962954, -0.2024430931, -0.5520970821, 0.0381105617, 0.1221280769, 0.9493938684, 0.1014682204, 0.063119255, -0.1696381718, -0.244...
http://arxiv.org/abs/2407.00369v1
http://arxiv.org/pdf/2407.00369v1
How to Train Your Fact Verifier: Knowledge Transfer with Multimodal Open Models
2024-06-29T08:39:07
2024-06-29T08:39:07
Jaeyoung Lee, Ximing Lu, Jack Hessel, Faeze Brahman, Youngjae Yu, Yonatan Bisk, Yejin Choi, Saadia Gabriel
cs.CL
cs.CL
Given the growing influx of misinformation across news and social media, there is a critical need for systems that can provide effective real-time verification of news claims. Large language or multimodal model based verification has been proposed to scale up online policing mechanisms for mitigating spread of false an...
[ -0.349957794, -0.7933532596, 0.6726104617, 0.1591146737, -0.1057507247, 0.0095182462, -0.2058349848, -0.5558418036, -0.3380679786, 0.3883228302, -0.4178247154, -0.1821302176, -0.8951233029, -0.0581671335, -0.5457399487, 0.9538925886, -0.327045083, 0.0754411221, -0.527581811, -0...
http://arxiv.org/abs/2407.00365v1
http://arxiv.org/pdf/2407.00365v1
Financial Knowledge Large Language Model
2024-06-29T08:26:49
2024-06-29T08:26:49
Cehao Yang, Chengjin Xu, Yiyan Qi
cs.CL
cs.CL
Artificial intelligence is making significant strides in the finance industry, revolutionizing how data is processed and interpreted. Among these technologies, large language models (LLMs) have demonstrated substantial potential to transform financial services by automating complex tasks, enhancing customer service, an...
[ -0.1990537792, -1.0255438089, 0.4239054918, 0.0773914903, -0.0897099227, 0.4897297621, -0.2751813531, -0.9136043787, 0.0204408672, 0.67962569, -0.3927769661, -0.4290526509, -0.3148411214, -0.1836373657, -0.6063767672, 0.8545022607, 0.0812488794, 0.2260665596, -0.2756551802, 0.3...
http://arxiv.org/abs/2407.00361v1
http://arxiv.org/pdf/2407.00361v1
From RAG to RICHES: Retrieval Interlaced with Sequence Generation
2024-06-29T08:16:58
2024-06-29T08:16:58
Palak Jain, Livio Baldini Soares, Tom Kwiatkowski
cs.CL
cs.CL, cs.AI
We present RICHES, a novel approach that interleaves retrieval with sequence generation tasks. RICHES offers an alternative to conventional RAG systems by eliminating the need for separate retriever and generator. It retrieves documents by directly decoding their contents, constrained on the corpus. Unifying retrieval ...
[ -0.2045614868, -0.2572936416, 0.2002073824, 0.2533922493, -0.3474173546, 0.5205960274, 0.0901391581, -0.3066115379, 0.4788369536, 0.7621689439, -0.313013196, 0.1366158426, -0.6466991901, 0.0651448742, -0.6027908921, 1.1191686392, 0.5359376073, 0.012858537, -0.1600807756, 0.0002...
http://arxiv.org/abs/2407.00342v1
http://arxiv.org/pdf/2407.00342v1
Korean Aspect-Based Sentiment Analysis via Implicit-Feature Alignment with Corpus Filtering
2024-06-29T07:01:51
2024-06-29T07:01:51
Kibeom Nam
cs.CL
cs.CL, cs.AI
Investigations into Aspect-Based Sentiment Analysis (ABSA) for Korean restaurant reviews are notably lacking in the existing literature. Our research proposes an intuitive and effective framework for ABSA in low-resource languages such as Korean. It optimizes prediction labels by integrating translated benchmark and un...
[ -0.4528357089, -0.4600450695, 0.7553081512, 0.2815397382, -0.5848488212, 0.0807968304, -0.0054154, -0.6207954884, 0.6204956174, 0.7091913819, -0.4050704241, -0.9368708134, -0.1406867951, 0.2115594894, 0.1024843529, 1.1710431576, -0.244623512, 0.4494368136, -0.315725714, -0.5042...
http://arxiv.org/abs/2407.00341v1
http://arxiv.org/pdf/2407.00341v1
Iterative Data Augmentation with Large Language Models for Aspect-based Sentiment Analysis
2024-06-29T07:00:37
2024-06-29T07:00:37
Haiyun Li, Qihuang Zhong, Ke Zhu, Juhua Liu, Bo Du, Dacheng Tao
cs.CL
cs.CL
Aspect-based Sentiment Analysis (ABSA) is an important sentiment analysis task, which aims to determine the sentiment polarity towards an aspect in a sentence. Due to the expensive and limited labeled data, data augmentation (DA) has become the standard for improving the performance of ABSA. However, current DA methods...
[ -0.6003405452, -0.8327926397, 0.5330635905, 0.3759767413, -0.6329163313, 0.2525608242, -0.2847346067, -0.4981221259, 0.466835022, 0.6113353968, -0.5533192754, -0.8424733281, -0.4232245088, 0.0774331018, -0.5201708078, 1.0370591879, -0.0489942133, -0.0097385049, -0.4054850936, -...
http://arxiv.org/abs/2407.01626v1
http://arxiv.org/pdf/2407.01626v1
SPARKLE: Enhancing SPARQL Generation with Direct KG Integration in Decoding
2024-06-29T06:43:11
2024-06-29T06:43:11
Jaebok Lee, Hyeonjeong Shin
cs.IR
cs.IR, cs.AI, cs.CL
Existing KBQA methods have traditionally relied on multi-stage methodologies, involving tasks such as entity linking, subgraph retrieval and query structure generation. However, multi-stage approaches are dependent on the accuracy of preceding steps, leading to cascading errors and increased inference time. Although a ...
[ -0.3438768089, -0.6899399161, 0.0104171298, 0.489523083, 0.0856505707, 0.2579814196, 0.1480975896, -0.3501415849, 0.0559373274, 0.720058322, -0.6000103354, -0.4220043719, -0.1848201901, 0.0096254312, -0.3239678741, 0.8678367734, 0.4330536127, -0.0084737232, -0.2149538845, 0.062...
http://arxiv.org/abs/2407.00322v1
http://arxiv.org/pdf/2407.00322v1
LLM-Generated Natural Language Meets Scaling Laws: New Explorations and Data Augmentation Methods
2024-06-29T05:40:17
2024-06-29T05:40:17
Zhenhua Wang, Guang Xu, Ming Ren
cs.CL
cs.CL
With the ascent of large language models (LLM), natural language processing has witnessed enhancements, such as LLM-based data augmentation. Nonetheless, prior research harbors two primary concerns: firstly, a lack of contemplation regarding whether the natural language generated by LLM (LLMNL) truly aligns with human ...
[ -0.3401651978, -0.8422507644, 0.6899889708, 0.0058080759, -0.2078597248, -0.1370877475, -0.5574917793, -0.6564424634, -0.0134308301, 0.5117374659, -0.4745800197, -0.2535297871, -0.6526940465, 0.2000285089, -0.2547249198, 1.1505820751, 0.1769574732, -0.1062258482, -0.0796502009, ...
http://arxiv.org/abs/2407.00320v1
http://arxiv.org/pdf/2407.00320v1
LiteSearch: Efficacious Tree Search for LLM
2024-06-29T05:14:04
2024-06-29T05:14:04
Ante Wang, Linfeng Song, Ye Tian, Baolin Peng, Dian Yu, Haitao Mi, Jinsong Su, Dong Yu
cs.CL
cs.CL, cs.AI, cs.LG
Recent research suggests that tree search algorithms (e.g. Monte Carlo Tree Search) can dramatically boost LLM performance on complex mathematical reasoning tasks. However, they often require more than 10 times the computational resources of greedy decoding due to wasteful search strategies, making them difficult to be...
[ -0.1887157112, -0.6918367147, 0.5935990214, 0.0945390165, -0.4557612836, 0.6319175363, 0.1568795145, -0.4747287929, 0.0208049063, 0.259907335, -0.8986737132, -0.4227674901, -0.4394864142, -0.0099300649, -0.2042234093, 0.3205630481, -0.0103380801, -0.1064726263, 0.0663066506, 0....
http://arxiv.org/abs/2407.00263v1
http://arxiv.org/pdf/2407.00263v1
From Local Concepts to Universals: Evaluating the Multicultural Understanding of Vision-Language Models
2024-06-28T23:28:28
2024-06-28T23:28:28
Mehar Bhatia, Sahithya Ravi, Aditya Chinchure, Eunjeong Hwang, Vered Shwartz
cs.CL
cs.CL, cs.AI, cs.CV
Despite recent advancements in vision-language models, their performance remains suboptimal on images from non-western cultures due to underrepresentation in training datasets. Various benchmarks have been proposed to test models' cultural inclusivity, but they have limited coverage of cultures and do not adequately as...
[ -0.6795905828, -0.1407295614, 0.0829972401, 0.0674752519, -0.1897555888, 0.1250342876, -0.3567504883, -0.7546774149, -0.1193757877, 0.3309024572, -0.2945100963, -0.4866107702, -0.5426476002, 0.1327214539, 0.0667941719, 0.7384659648, -0.1452284902, 0.2895279825, -0.3141114116, -...
http://arxiv.org/abs/2407.00256v1
http://arxiv.org/pdf/2407.00256v1
One Prompt is not Enough: Automated Construction of a Mixture-of-Expert Prompts
2024-06-28T23:05:08
2024-06-28T23:05:08
Ruochen Wang, Sohyun An, Minhao Cheng, Tianyi Zhou, Sung Ju Hwang, Cho-Jui Hsieh
cs.AI
cs.AI, cs.CL, cs.LG, stat.ML, 68T01
Large Language Models (LLMs) exhibit strong generalization capabilities to novel tasks when prompted with language instructions and in-context demos. Since this ability sensitively depends on the quality of prompts, various methods have been explored to automate the instruction design. While these methods demonstrated ...
[ -0.5910390615, -0.9744223356, 0.7240334153, 0.3358186483, -0.1972964704, 0.1878728867, -0.2000126839, 0.1146309823, 0.3358914554, 0.3363606036, -0.6872843504, -0.5249596238, -0.6048403382, 0.0967027023, -0.231647104, 0.669218123, -0.1669880152, -0.1303491741, -0.3288054168, 0.0...
http://arxiv.org/abs/2407.00250v1
http://arxiv.org/pdf/2407.00250v1
Mind the Gap: Analyzing Lacunae with Transformer-Based Transcription
2024-06-28T22:52:39
2024-06-28T22:52:39
Jaydeep Borkar, David A. Smith
cs.CV
cs.CV, cs.CL
Historical documents frequently suffer from damage and inconsistencies, including missing or illegible text resulting from issues such as holes, ink problems, and storage damage. These missing portions or gaps are referred to as lacunae. In this study, we employ transformer-based optical character recognition (OCR) mod...
[ -0.245868668, -0.6995880008, 0.6327978969, -0.4467039406, 0.0367836319, 0.0240719412, 0.3425245583, -0.6616017222, 0.3635809422, 0.912319243, -0.7961966395, -0.1414268017, -0.5485801101, 0.211658746, -0.3510347605, 0.743142128, -0.1763804853, -0.0680154189, 0.1776358634, -0.074...
http://arxiv.org/abs/2407.00248v1
http://arxiv.org/pdf/2407.00248v1
DiffuseDef: Improved Robustness to Adversarial Attacks
2024-06-28T22:36:17
2024-06-28T22:36:17
Zhenhao Li, Marek Rei, Lucia Specia
cs.CL
cs.CL
Pretrained language models have significantly advanced performance across various natural language processing tasks. However, adversarial attacks continue to pose a critical challenge to system built using these models, as they can be exploited with carefully crafted adversarial texts. Inspired by the ability of diffus...
[ -0.4649100006, -1.2137688398, 0.2997947335, -0.018554071, 0.0198698062, 0.2532583773, -0.2453383356, -0.1811561733, -0.2138543576, 0.4812868536, -0.3027981818, -0.5115690827, -1.1793912649, -0.1417719275, -0.8313951492, 0.9416069984, 0.2174574435, 0.1173733026, -0.0174186323, -...
http://arxiv.org/abs/2407.00242v1
http://arxiv.org/pdf/2407.00242v1
EHRmonize: A Framework for Medical Concept Abstraction from Electronic Health Records using Large Language Models
2024-06-28T21:39:20
2024-06-28T21:39:20
João Matos, Jack Gallifant, Jian Pei, A. Ian Wong
cs.CL
cs.CL
Electronic health records (EHRs) contain vast amounts of complex data, but harmonizing and processing this information remains a challenging and costly task requiring significant clinical expertise. While large language models (LLMs) have shown promise in various healthcare applications, their potential for abstracting...
[ -0.038544517, -0.9506825805, 0.9942871332, -0.12947236, -0.2999189496, -0.2155092359, -0.3937033713, -0.5213770866, 0.1554507017, 0.3273172081, -0.3878784478, -0.7902246118, -0.6696728468, 0.2464921623, -0.217497766, 1.0005383492, -0.2225008309, 0.0711178556, -0.2439059615, 0.0...
http://arxiv.org/abs/2407.00219v1
http://arxiv.org/pdf/2407.00219v1
Evaluating Human Alignment and Model Faithfulness of LLM Rationale
2024-06-28T20:06:30
2024-06-28T20:06:30
Mohsen Fayyaz, Fan Yin, Jiao Sun, Nanyun Peng
cs.CL
cs.CL, cs.AI
We study how well large language models (LLMs) explain their generations with rationales -- a set of tokens extracted from the input texts that reflect the decision process of LLMs. We examine LLM rationales extracted with two methods: 1) attribution-based methods that use attention or gradients to locate important tok...
[ -0.2097644955, -0.6581833363, 0.9103560448, 0.1084671691, -0.1595924646, -0.1454733759, -0.3221784234, -0.2764572501, -0.244693622, 0.434338659, -0.9002410173, -0.1741640121, -0.2522474527, 0.2637667954, -0.2802333236, 0.5666814446, -0.1589654535, 0.4632425308, -0.49765414, -0....
http://arxiv.org/abs/2407.00211v1
http://arxiv.org/pdf/2407.00211v1
Detection and Measurement of Syntactic Templates in Generated Text
2024-06-28T19:34:23
2024-06-28T19:34:23
Chantal Shaib, Yanai Elazar, Junyi Jessy Li, Byron C. Wallace
cs.CL
cs.CL
Recent work on evaluating the diversity of text generated by LLMs has focused on word-level features. Here we offer an analysis of syntactic features to characterize general repetition in models, beyond frequent n-grams. Specifically, we define syntactic templates and show that models tend to produce templated text in ...
[ -0.1755803376, -0.6422778368, 0.6706773639, -0.0428726114, -0.4962661564, -0.0686419383, -0.1374864131, -0.1536769718, 0.3802250922, 0.5907884836, -0.4360382557, -0.7461092472, -0.5372264981, 0.4123013318, -0.922490716, 0.9947141409, -0.2692828774, 0.1873201281, -0.383225739, -...
http://arxiv.org/abs/2407.00191v1
http://arxiv.org/pdf/2407.00191v1
MetaKP: On-Demand Keyphrase Generation
2024-06-28T19:02:59
2024-06-28T19:02:59
Di Wu, Xiaoxian Shen, Kai-Wei Chang
cs.CL
cs.CL
Traditional keyphrase prediction methods predict a single set of keyphrases per document, failing to cater to the diverse needs of users and downstream applications. To bridge the gap, we introduce on-demand keyphrase generation, a novel paradigm that requires keyphrases that conform to specific high-level goals or int...
[ -0.0049925721, -0.3981222808, 0.5794472098, 0.317073375, -0.2900639474, 0.0383612923, -0.0842666999, -0.0966257676, 0.3483062088, 0.7460689545, -0.6321733594, -0.5652459264, -0.5288524032, 0.370601058, -0.290204674, 0.8100059628, 0.15274854, -0.2334885299, -0.05467638, -0.00272...
http://arxiv.org/abs/2407.00167v1
http://arxiv.org/pdf/2407.00167v1
Can GPT-4 Help Detect Quit Vaping Intentions? An Exploration of Automatic Data Annotation Approach
2024-06-28T18:06:48
2024-06-28T18:06:48
Sai Krishna Revanth Vuruma, Dezhi Wu, Saborny Sen Gupta, Lucas Aust, Valerie Lookingbill, Wyatt Bellamy, Yang Ren, Erin Kasson, Li-Shiun Chen, Patricia Cavazos-Rehg, Dian Hu, Ming Huang
cs.CL
cs.CL, cs.AI, cs.ET, cs.HC, cs.SI
In recent years, the United States has witnessed a significant surge in the popularity of vaping or e-cigarette use, leading to a notable rise in cases of e-cigarette and vaping use-associated lung injury (EVALI) that caused hospitalizations and fatalities during the EVALI outbreak in 2019, highlighting the urgency to ...
[ -0.2770230174, -1.0206807852, 1.1687413454, -0.0941127241, -0.28037256, -0.2697667778, -0.1443644315, -0.3342641294, 0.3447209895, 0.3519612551, -0.4405963421, -0.6518403292, -0.4746566117, 0.1894759983, -0.2670343816, 1.1298840046, -0.3757420778, -0.299739182, 0.0708361715, 0....
http://arxiv.org/abs/2406.20098v1
http://arxiv.org/pdf/2406.20098v1
Web2Code: A Large-scale Webpage-to-Code Dataset and Evaluation Framework for Multimodal LLMs
2024-06-28T17:59:46
2024-06-28T17:59:46
Sukmin Yun, Haokun Lin, Rusiru Thushara, Mohammad Qazim Bhat, Yongxin Wang, Zutao Jiang, Mingkai Deng, Jinhong Wang, Tianhua Tao, Junbo Li, Haonan Li, Preslav Nakov, Timothy Baldwin, Zhengzhong Liu, Eric P. Xing, Xiaodan Liang, Zhiqiang Shen
cs.CV
cs.CV, cs.AI, cs.CL
Multimodal large language models (MLLMs) have shown impressive success across modalities such as image, video, and audio in a variety of understanding and generation tasks. However, current MLLMs are surprisingly poor at understanding webpage screenshots and generating their corresponding HTML code. To address this pro...
[ -0.4126002789, -0.4248456955, 0.1190844402, 0.3703280091, -0.2106480598, -0.0762380287, -0.5255057216, -0.1397948712, -0.5721955895, 0.2967363596, -0.5026885867, -0.6051462889, -0.3821427226, 0.0771602392, -0.5381037593, 0.7169160843, 0.0387706533, 0.2896876633, -0.5735808611, ...
http://arxiv.org/abs/2406.20095v1
http://arxiv.org/pdf/2406.20095v1
LLaRA: Supercharging Robot Learning Data for Vision-Language Policy
2024-06-28T17:59:12
2024-06-28T17:59:12
Xiang Li, Cristina Mata, Jongwoo Park, Kumara Kahatapitiya, Yoo Sung Jang, Jinghuan Shang, Kanchana Ranasinghe, Ryan Burgert, Mu Cai, Yong Jae Lee, Michael S. Ryoo
cs.RO
cs.RO, cs.AI, cs.CL, cs.CV, cs.LG
Large Language Models (LLMs) equipped with extensive world knowledge and strong reasoning skills can tackle diverse tasks across domains, often by posing them as conversation-style instruction-response pairs. In this paper, we propose LLaRA: Large Language and Robotics Assistant, a framework which formulates robot acti...
[ -0.0078341505, -0.6955738664, 0.6090714931, -0.2028282732, 0.032342758, 0.1525054425, -0.1283980608, -0.1838590354, 0.6665604711, 0.9792294502, -0.6969590187, -0.3886430264, -0.4855886698, -0.0498812012, -0.426243186, 0.7720205188, 0.1277687401, 0.1168110222, -0.0912136137, -0....
http://arxiv.org/abs/2406.20094v1
http://arxiv.org/pdf/2406.20094v1
Scaling Synthetic Data Creation with 1,000,000,000 Personas
2024-06-28T17:59:01
2024-06-28T17:59:01
Xin Chan, Xiaoyang Wang, Dian Yu, Haitao Mi, Dong Yu
cs.CL
cs.CL, cs.LG
We propose a novel persona-driven data synthesis methodology that leverages various perspectives within a large language model (LLM) to create diverse synthetic data. To fully exploit this methodology at scale, we introduce Persona Hub -- a collection of 1 billion diverse personas automatically curated from web data. T...
[ -0.6946753263, -0.7320794463, 0.7547993064, 0.601916492, 0.4585291445, 0.2782869637, 0.0776696652, -0.2927174568, 0.3863556385, 0.6143229008, -0.867921114, -0.4043411613, 0.0184396561, 0.1348640174, -0.3542422354, 0.8667187095, 0.4020848274, 0.0857641473, -0.4024116695, 0.06087...
http://arxiv.org/abs/2406.20087v1
http://arxiv.org/pdf/2406.20087v1
ProgressGym: Alignment with a Millennium of Moral Progress
2024-06-28T17:55:24
2024-06-28T17:55:24
Tianyi Qiu, Yang Zhang, Xuchuan Huang, Jasmine Xinze Li, Jiaming Ji, Yaodong Yang
cs.LG
cs.LG, cs.AI, cs.CL, cs.CY, cs.HC
Frontier AI systems, including large language models (LLMs), hold increasing influence over the epistemology of human users. Such influence can reinforce prevailing societal values, potentially contributing to the lock-in of misguided moral beliefs and, consequently, the perpetuation of problematic moral practices on a...
[ 0.0859305263, -0.5705635548, 0.6793205738, 0.0841904655, 0.1751241088, 0.562589705, 0.0720075071, -0.3030786216, 0.0229034517, 0.491445154, -0.7914457917, -0.4199841917, -0.6284319162, -0.3605127931, 0.0385769196, 1.0257009268, 0.0094040185, 0.0003185522, -0.2131952196, -0.2263...
http://arxiv.org/abs/2406.20086v1
http://arxiv.org/pdf/2406.20086v1
Token Erasure as a Footprint of Implicit Vocabulary Items in LLMs
2024-06-28T17:54:47
2024-06-28T17:54:47
Sheridan Feucht, David Atkinson, Byron Wallace, David Bau
cs.CL
cs.CL, cs.LG, I.2.7
LLMs process text as sequences of tokens that roughly correspond to words, where less common words are represented by multiple tokens. However, individual tokens are often semantically unrelated to the meanings of the words/concepts they comprise. For example, Llama-2-7b's tokenizer splits the word "northeastern" into ...
[ -0.3797490001, -0.8815509081, 0.5021921992, -0.1077764034, -0.5249354243, 0.4320823252, 0.0986924991, -0.6477326155, 0.3542830944, 0.3453064859, -0.501758337, -0.5907111764, -0.7359321117, 0.1681817472, -0.6679397225, 0.9568408132, -0.1049004793, -0.1472025216, 0.1190630496, -0...
http://arxiv.org/abs/2406.20079v1
http://arxiv.org/pdf/2406.20079v1
Molecular Facts: Desiderata for Decontextualization in LLM Fact Verification
2024-06-28T17:43:48
2024-06-28T17:43:48
Anisha Gunjal, Greg Durrett
cs.CL
cs.CL, cs.AI
Automatic factuality verification of large language model (LLM) generations is becoming more and more widely used to combat hallucinations. A major point of tension in the literature is the granularity of this fact-checking: larger chunks of text are hard to fact-check, but more atomic facts like propositions may lack ...
[ -0.299059093, -1.072740078, 1.0109586716, -0.2046029121, -0.4277898967, -0.11855492, -0.1491708457, -0.5589125752, 0.2092115432, 0.3318251073, -0.3188097179, -0.6673899293, -0.6188434362, 0.3913282156, -0.6858055592, 0.7619459629, 0.2176003158, -0.3638617694, -0.4782520533, -0....
http://arxiv.org/abs/2406.20060v1
http://arxiv.org/pdf/2406.20060v1
Applying RLAIF for Code Generation with API-usage in Lightweight LLMs
2024-06-28T17:16:03
2024-06-28T17:16:03
Sujan Dutta, Sayantan Mahinder, Raviteja Anantha, Bortik Bandyopadhyay
cs.CL
cs.CL
Reinforcement Learning from AI Feedback (RLAIF) has demonstrated significant potential across various domains, including mitigating harm in LLM outputs, enhancing text summarization, and mathematical reasoning. This paper introduces an RLAIF framework for improving the code generation abilities of lightweight (<1B para...
[ -0.2162793279, -0.7348615527, 0.2528093159, 0.2979615033, -0.3883237839, -0.0666756704, -0.322583735, -0.4640190899, 0.1885572821, 0.3499567509, -0.7561818957, -0.5957519412, -0.5251787901, 0.1170145646, -0.0911078155, 1.1697449684, -0.0706286877, -0.1211474612, -0.2587923408, ...
http://arxiv.org/abs/2406.20054v1
http://arxiv.org/pdf/2406.20054v1
To Word Senses and Beyond: Inducing Concepts with Contextualized Language Models
2024-06-28T17:07:06
2024-06-28T17:07:06
Bastien Liétard, Pascal Denis, Mikaella Keller
cs.CL
cs.CL
Polysemy and synonymy are two crucial interrelated facets of lexical ambiguity. While both phenomena have been studied extensively in NLP, leading to dedicated systems, they are often been considered independently. While many tasks dealing with polysemy (e.g. Word Sense Disambiguiation or Induction) highlight the role ...
[ -0.3992277384, -0.5809077621, 0.5709566474, -0.1328356564, -0.4484730065, -0.0440839194, -0.3473490477, -0.3750242591, 0.3450784385, -0.0910470933, -0.2484495938, -0.3515529037, -0.6549407244, 0.0653204918, -0.5786670446, 0.7402541041, 0.0620720722, -0.0984100327, -0.6131843328, ...
http://arxiv.org/abs/2406.20053v1
http://arxiv.org/pdf/2406.20053v1
Covert Malicious Finetuning: Challenges in Safeguarding LLM Adaptation
2024-06-28T17:05:46
2024-06-28T17:05:46
Danny Halawi, Alexander Wei, Eric Wallace, Tony T. Wang, Nika Haghtalab, Jacob Steinhardt
cs.CR
cs.CR, cs.AI, cs.CL, cs.LG
Black-box finetuning is an emerging interface for adapting state-of-the-art language models to user needs. However, such access may also let malicious actors undermine model safety. To demonstrate the challenge of defending finetuning interfaces, we introduce covert malicious finetuning, a method to compromise model sa...
[ -0.1341727823, -1.0009230375, 0.3847390711, -0.2978529036, -0.2005336881, -0.0210977457, 0.0915724859, -0.4816833735, -0.6173506975, 0.9346382022, -0.6947898269, -0.2936579287, -0.5297370553, -0.172425285, -0.5358013511, 1.4665788412, 0.1695574969, -0.1030148566, 0.1059160829, ...
http://arxiv.org/abs/2406.20052v1
http://arxiv.org/pdf/2406.20052v1
Understanding and Mitigating Language Confusion in LLMs
2024-06-28T17:03:51
2024-06-28T17:03:51
Kelly Marchisio, Wei-Yin Ko, Alexandre Bérard, Théo Dehaze, Sebastian Ruder
cs.CL
cs.CL
We investigate a surprising limitation of LLMs: their inability to consistently generate text in a user's desired language. We create the Language Confusion Benchmark (LCB) to evaluate such failures, covering 15 typologically diverse languages with existing and newly-created English and multilingual prompts. We evaluat...
[ -0.195063144, -0.910461545, 0.3909439147, 0.5376868248, -0.1009563878, 0.417424202, -0.2284558117, -0.4709812403, -0.3419286311, 0.321077317, -0.6745494008, -0.2610925138, -0.5553387403, 0.3617014885, -0.2216846049, 0.8366870284, -0.2800337374, 0.3399476707, -0.4359011948, -0.1...
http://arxiv.org/abs/2407.00146v1
http://arxiv.org/pdf/2407.00146v1
The Qiyas Benchmark: Measuring ChatGPT Mathematical and Language Understanding in Arabic
2024-06-28T16:34:31
2024-06-28T16:34:31
Shahad Al-Khalifa, Hend Al-Khalifa
cs.CL
cs.CL, cs.AI
Despite the growing importance of Arabic as a global language, there is a notable lack of language models pre-trained exclusively on Arabic data. This shortage has led to limited benchmarks available for assessing language model performance in Arabic. To address this gap, we introduce two novel benchmarks designed to e...
[ -0.2036259919, -0.5131343603, 0.5172065496, 0.2948811948, -0.2906148136, 0.3103429973, 0.1272809058, -0.4615114033, -0.4095687568, 0.1859523058, -0.402774483, -0.6071412563, -0.5243420005, -0.1830660105, -0.4381803572, 0.9529906511, -0.495348841, 0.2266560644, -0.01813576, -0.4...
http://arxiv.org/abs/2406.20038v1
http://arxiv.org/pdf/2406.20038v1
BioMNER: A Dataset for Biomedical Method Entity Recognition
2024-06-28T16:34:24
2024-06-28T16:34:24
Chen Tang, Bohao Yang, Kun Zhao, Bo Lv, Chenghao Xiao, Frank Guerin, Chenghua Lin
cs.CL
cs.CL
Named entity recognition (NER) stands as a fundamental and pivotal task within the realm of Natural Language Processing. Particularly within the domain of Biomedical Method NER, this task presents notable challenges, stemming from the continual influx of domain-specific terminologies in scholarly literature. Current re...
[ -0.3022367954, -0.7782464623, 0.5370687246, -0.1313243061, -0.3021890223, 0.2297796607, -0.1268715262, -0.8041727543, 0.4065192342, 0.6325674057, -0.3472685218, -0.7219507694, -0.4738623202, 0.6571552753, -0.0262316745, 1.1747750044, 0.0734007955, 0.488610357, -0.3124966919, -0...
http://arxiv.org/abs/2406.20030v1
http://arxiv.org/pdf/2406.20030v1
LEMoE: Advanced Mixture of Experts Adaptor for Lifelong Model Editing of Large Language Models
2024-06-28T16:17:41
2024-06-28T16:17:41
Renzhi Wang, Piji Li
cs.CL
cs.CL
Large language models (LLMs) require continual knowledge updates to stay abreast of the ever-changing world facts, prompting the formulation of lifelong model editing task. While recent years have witnessed the development of various techniques for single and batch editing, these methods either fail to apply or perform...
[ -0.185295105, -1.1261627674, 0.511348784, 0.2352142036, -0.1378787756, -0.0546209477, -0.3333765864, -0.2228628993, 0.1658542603, 0.5776475668, -0.4432106018, -0.4342967272, -0.6148171425, -0.1540189981, -0.0467118621, 0.4993073344, 0.0059710601, 0.1439680159, -0.2093810737, -0...
http://arxiv.org/abs/2406.20015v1
http://arxiv.org/pdf/2406.20015v1
ToolBeHonest: A Multi-level Hallucination Diagnostic Benchmark for Tool-Augmented Large Language Models
2024-06-28T16:03:30
2024-06-28T16:03:30
Yuxiang Zhang, Jing Chen, Junjie Wang, Yaxin Liu, Cheng Yang, Chufan Shi, Xinyu Zhu, Zihao Lin, Hanwen Wan, Yujiu Yang, Tetsuya Sakai, Tian Feng, Hayato Yamana
cs.CL
cs.CL, cs.AI
Tool-augmented large language models (LLMs) are rapidly being integrated into real-world applications. Due to the lack of benchmarks, the community still needs to fully understand the hallucination issues within these models. To address this challenge, we introduce a comprehensive diagnostic benchmark, ToolBH. Specific...
[ -0.4947895408, -0.8710588217, 0.6180759072, 0.4052726328, -0.3579827547, 0.250007689, -0.2596136928, -0.5954637527, 0.0579793192, 0.5524241924, -0.4967213273, -0.6549143195, -0.6080498099, 0.066871509, -0.6151689887, 0.5977041125, 0.069038786, 0.1223815158, -0.157039538, 0.1041...
http://arxiv.org/abs/2406.19999v1
http://arxiv.org/pdf/2406.19999v1
The SIFo Benchmark: Investigating the Sequential Instruction Following Ability of Large Language Models
2024-06-28T15:34:26
2024-06-28T15:34:26
Xinyi Chen, Baohao Liao, Jirui Qi, Panagiotis Eustratiadis, Christof Monz, Arianna Bisazza, Maarten de Rijke
cs.CL
cs.CL
Following multiple instructions is a crucial ability for large language models (LLMs). Evaluating this ability comes with significant challenges: (i) limited coherence between multiple instructions, (ii) positional bias where the order of instructions affects model performance, and (iii) a lack of objectively verifiabl...
[ -0.05065706, -0.4098867178, 0.7079063058, 0.5851570368, 0.1198927388, 0.0933633, -0.0202927887, -0.3956198394, -0.4159326255, 0.7400007248, -0.6260660887, -0.5428302288, -0.8412648439, 0.3576915562, -0.4440494776, 0.7901489139, -0.2157093436, 0.1407868117, -0.2302388996, -0.091...
http://arxiv.org/abs/2406.19995v1
http://arxiv.org/pdf/2406.19995v1
Single Parent Family: A Spectrum of Family Members from a Single Pre-Trained Foundation Model
2024-06-28T15:27:57
2024-06-28T15:27:57
Habib Hajimolahoseini, Mohammad Hassanpour, Foozhan Ataiefard, Boxing Chen, Yang Liu
cs.CL
cs.CL, cs.AI, cs.LG
This paper introduces a novel method of Progressive Low Rank Decomposition (PLRD) tailored for the compression of large language models. Our approach leverages a pre-trained model, which is then incrementally decompressed to smaller sizes using progressively lower ranks. This method allows for significant reductions in...
[ -0.5231897235, -0.5549325943, -0.053934563, 0.0438290983, -0.2705177367, 0.4308332503, -0.4332859814, 0.1787554473, 0.0154062547, 0.5559815764, -0.7196780443, -0.427998215, -0.6497001648, -0.0838026255, -0.2837101519, 0.8135674, 0.2523333728, -0.0689341277, -0.1459095031, -0.16...
http://arxiv.org/abs/2406.19967v1
http://arxiv.org/pdf/2406.19967v1
Into the Unknown: Generating Geospatial Descriptions for New Environments
2024-06-28T14:56:21
2024-06-28T14:56:21
Tzuf Paz-Argaman, John Palowitch, Sayali Kulkarni, Reut Tsarfaty, Jason Baldridge
cs.CL
cs.CL, cs.AI
Similar to vision-and-language navigation (VLN) tasks that focus on bridging the gap between vision and language for embodied navigation, the new Rendezvous (RVS) task requires reasoning over allocentric spatial relationships (independent of the observer's viewpoint) using non-sequential navigation instructions and map...
[ -0.6022315025, -0.7313144803, 1.0329319239, 0.0092973597, -0.0333968215, -0.1194578335, -0.5101830959, -0.6176307201, 0.1315037459, 0.539239347, -0.75807935, -0.7612394691, -0.290789336, -0.0988743454, -0.3709385395, 0.6916823983, -0.190421313, -0.4295124114, -0.4496944547, 0.0...
http://arxiv.org/abs/2406.19966v1
http://arxiv.org/pdf/2406.19966v1
Simulating Financial Market via Large Language Model based Agents
2024-06-28T14:54:12
2024-06-28T14:54:12
Shen Gao, Yuntao Wen, Minghang Zhu, Jianing Wei, Yuhan Cheng, Qunzi Zhang, Shuo Shang
cs.CL
cs.CL
Most economic theories typically assume that financial market participants are fully rational individuals and use mathematical models to simulate human behavior in financial markets. However, human behavior is often not entirely rational and is challenging to predict accurately with mathematical models. In this paper, ...
[ -0.0482632928, -1.0987126827, 0.164688617, 0.3686380684, -0.0264842939, 0.3015168011, -0.1884266585, -0.5243904591, 0.5900945067, 0.4846665561, -0.3305070996, -0.2315566093, -0.082793951, -0.0136541435, -0.5205939412, 0.478212446, 0.0071980329, 0.2074071765, 0.1635289192, 0.326...
http://arxiv.org/abs/2406.19954v1
http://arxiv.org/pdf/2406.19954v1
BESTOW: Efficient and Streamable Speech Language Model with the Best of Two Worlds in GPT and T5
2024-06-28T14:40:03
2024-06-28T14:40:03
Zhehuai Chen, He Huang, Oleksii Hrinchuk, Krishna C. Puvvada, Nithin Rao Koluguri, Piotr Żelasko, Jagadeesh Balam, Boris Ginsburg
cs.CL
cs.CL, cs.HC, cs.SD, eess.AS, 68T10, I.2.7
Incorporating speech understanding capabilities into pretrained large-language models has become a vital research direction (SpeechLLM). The previous architectures can be categorized as: i) GPT-style, prepend speech prompts to the text prompts as a sequence of LLM inputs like a decoder-only model; ii) T5-style, introdu...
[ -0.036980778, -0.5511693358, 0.3758074343, 0.3227292597, -0.3438417614, 0.1613076478, -0.330783844, -0.5949793458, 0.1137766168, 0.5280594826, -0.5436440706, -0.4557972848, -0.6062541604, -0.1677287072, -0.36573717, 0.7951079607, 0.1101664156, -0.0805578306, -0.1436733752, -0.1...
http://arxiv.org/abs/2406.19951v1
http://arxiv.org/pdf/2406.19951v1
Mining Reasons For And Against Vaccination From Unstructured Data Using Nichesourcing and AI Data Augmentation
2024-06-28T14:36:31
2024-06-28T14:36:31
Damián Ariel Furman, Juan Junqueras, Z. Burçe Gümüslü, Edgar Altszyler, Joaquin Navajas, Ophelia Deroy, Justin Sulik
cs.CL
cs.CL
We present Reasons For and Against Vaccination (RFAV), a dataset for predicting reasons for and against vaccination, and scientific authorities used to justify them, annotated through nichesourcing and augmented using GPT4 and GPT3.5-Turbo. We show how it is possible to mine these reasons in non-structured text, under ...
[ -0.0556536615, -0.5517722368, 0.4114234447, -0.3282644153, -0.3276346326, -0.2398836911, 0.0431969799, -0.4431908429, 0.1086589918, 0.3756191134, -0.3786034286, -0.5913075209, -0.402784586, 0.2922378778, -0.425701201, 1.3324828148, 0.1883154213, -0.0684264749, -0.6660725474, -0...
http://arxiv.org/abs/2406.19949v1
http://arxiv.org/pdf/2406.19949v1
Calibrating LLMs with Preference Optimization on Thought Trees for Generating Rationale in Science Question Scoring
2024-06-28T14:33:05
2024-06-28T14:33:05
Jiazheng Li, Hainiu Xu, Zhaoyue Sun, Yuxiang Zhou, David West, Cesare Aloisi, Yulan He
cs.CL
cs.CL
Generating rationales that justify scoring decisions has been a promising way to facilitate explainability in automated scoring systems. However, existing methods do not match the accuracy of classifier-based methods. Plus, the generated rationales often contain hallucinated information. To address these issues, we pro...
[ -0.3842498064, -0.6036087871, 0.8601791859, -0.0715750828, -0.3458983898, 0.0378810428, -0.2740579545, -0.2600618005, 0.264701575, 0.372549355, -0.5863285661, -0.4025275409, -0.077407077, 0.1781475395, 0.087960124, 0.9699587822, -0.0583576337, 0.2264748365, -0.6171466708, -0.04...
http://arxiv.org/abs/2406.19934v1
http://arxiv.org/pdf/2406.19934v1
From the Least to the Most: Building a Plug-and-Play Visual Reasoner via Data Synthesis
2024-06-28T14:04:10
2024-06-28T14:04:10
Chuanqi Cheng, Jian Guan, Wei Wu, Rui Yan
cs.CL
cs.CL, cs.AI
We explore multi-step reasoning in vision-language models (VLMs). The problem is challenging, as reasoning data consisting of multiple steps of visual and language processing are barely available. To overcome the challenge, we first introduce a least-to-most visual reasoning paradigm, which interleaves steps of decompo...
[ -0.7104043365, -0.7486436963, 0.9903303981, -0.1107080504, -0.2053027153, -0.1455603391, 0.0789694488, -0.3065402508, -0.2878215909, 0.640481174, -0.6611457467, -0.5818446875, -0.0887973234, 0.0815753341, -0.5080236197, 0.4899757802, -0.1452589482, -0.0188062321, -0.5114811659, ...
http://arxiv.org/abs/2406.19928v1
http://arxiv.org/pdf/2406.19928v1
Interactive Topic Models with Optimal Transport
2024-06-28T13:57:27
2024-06-28T13:57:27
Garima Dhanania, Sheshera Mysore, Chau Minh Pham, Mohit Iyyer, Hamed Zamani, Andrew McCallum
cs.CL
cs.CL, cs.HC, cs.IR
Topic models are widely used to analyze document collections. While they are valuable for discovering latent topics in a corpus when analysts are unfamiliar with the corpus, analysts also commonly start with an understanding of the content present in a corpus. This may be through categories obtained from an initial pas...
[ -0.539526403, -1.0684254169, 0.9122747779, -0.2247208059, -0.2549183667, 0.1196154803, -0.109644793, -0.0996615514, 0.3473870754, 0.5773890615, -0.4687483013, -0.5762220621, -0.5048141479, -0.2414932549, -0.6429908276, 1.0998995304, -0.1838323623, -0.2395182848, -0.4162316322, ...
http://arxiv.org/abs/2406.19898v1
http://arxiv.org/pdf/2406.19898v1
Paraphrase Types Elicit Prompt Engineering Capabilities
2024-06-28T13:06:31
2024-06-28T13:06:31
Jan Philip Wahle, Terry Ruas, Yang Xu, Bela Gipp
cs.CL
cs.CL
Much of the success of modern language models depends on finding a suitable prompt to instruct the model. Until now, it has been largely unknown how variations in the linguistic expression of prompts affect these models. This study systematically and empirically evaluates which linguistic features influence models thro...
[ -0.2256752849, -0.7429916263, 0.8898122907, 0.6747484803, 0.0795467198, -0.0670725852, 0.0250372533, -0.2128180265, -0.0629366189, 0.461203903, -0.2769128084, -0.2104185224, -0.5381984115, 0.7580075264, -0.2413494885, 0.8756211996, -0.1816099584, -0.1306821555, -0.0323386639, -...
http://arxiv.org/abs/2406.19892v1
http://arxiv.org/pdf/2406.19892v1
Untangling the Unrestricted Web: Automatic Identification of Multilingual Registers
2024-06-28T13:00:30
2024-06-28T13:00:30
Erik Henriksson, Amanda Myntti, Anni Eskelinen, Selcen Erten-Johansson, Saara Hellström, Veronika Laippala
cs.CL
cs.CL
This article explores deep learning models for the automatic identification of registers - text varieties such as news reports and discussion forums - in web-based datasets across 16 languages. Web register (or genre) identification would provide a robust solution for understanding the content of web-scale datasets, wh...
[ -0.4046032429, -0.3985894918, 0.1066320613, -0.0342060104, -0.2832091451, 0.4212519228, -0.5451291203, -0.8573048711, -0.2126219273, 0.2546804249, -0.4231213033, -0.5819487572, -0.616538465, 0.1597191393, -0.5498052239, 0.8136150837, 0.069156535, 0.016601285, -0.1855757535, -0....
http://arxiv.org/abs/2406.19884v1
http://arxiv.org/pdf/2406.19884v1
Investigating the Timescales of Language Processing with EEG and Language Models
2024-06-28T12:49:27
2024-06-28T12:49:27
Davide Turco, Conor Houghton
cs.CL
cs.CL, q-bio.NC
This study explores the temporal dynamics of language processing by examining the alignment between word representations from a pre-trained transformer-based language model, and EEG data. Using a Temporal Response Function (TRF) model, we investigate how neural activity corresponds to model representations across diffe...
[ -0.188111335, -1.2874873877, 0.7255941033, 0.4316550791, -0.0444979444, 0.0754544362, -0.0932081714, -0.6831698418, 0.2579210401, 0.1969576031, -0.5278174877, -0.1252264678, -0.3606410921, 0.0215205103, -0.5894153714, 0.2424865067, -0.0282529257, -0.024478063, -0.0506643169, -0...
http://arxiv.org/abs/2406.19874v1
http://arxiv.org/pdf/2406.19874v1
Detecting Subtle Differences between Human and Model Languages Using Spectrum of Relative Likelihood
2024-06-28T12:28:52
2024-06-28T12:28:52
Yang Xu, Yu Wang, Hao An, Zhichen Liu, Yongyuan Li
cs.CL
cs.CL, cs.AI, I.2.7
Human and model-generated texts can be distinguished by examining the magnitude of likelihood in language. However, it is becoming increasingly difficult as language model's capabilities of generating human-like texts keep evolving. This study provides a new perspective by using the relative likelihood values instead o...
[ -0.506947875, -0.9169764519, 0.6769904494, -0.2049833387, -0.096862413, 0.238345027, -0.3614927828, -0.8383321166, 0.0889634341, 0.4302839637, -0.4543944299, -0.4551801383, -0.3180409074, 0.084316276, -0.4610476792, 0.6050832868, 0.242851302, -0.1505008936, 0.1736578345, -0.127...
http://arxiv.org/abs/2406.19853v1
http://arxiv.org/pdf/2406.19853v1
YuLan: An Open-source Large Language Model
2024-06-28T11:52:53
2024-06-28T11:52:53
Yutao Zhu, Kun Zhou, Kelong Mao, Wentong Chen, Yiding Sun, Zhipeng Chen, Qian Cao, Yihan Wu, Yushuo Chen, Feng Wang, Lei Zhang, Junyi Li, Xiaolei Wang, Lei Wang, Beichen Zhang, Zican Dong, Xiaoxue Cheng, Yuhan Chen, Xinyu Tang, Yupeng Hou, Qiangqiang Ren, Xincheng Pang, Shufang Xie, Wayne Xin Zhao, Zhicheng Dou, Jiaxin...
cs.CL
cs.CL, cs.AI
Large language models (LLMs) have become the foundation of many applications, leveraging their extensive capabilities in processing and understanding natural language. While many open-source LLMs have been released with technical reports, the lack of training details hinders further research and development. This paper...
[ -0.0824172571, -0.6268384457, 0.4213737845, 0.0481243916, -0.1847153455, -0.0609306581, -0.4999572337, -0.477704078, 0.2880646586, 0.6589056253, -0.2659112513, -0.4061401784, -0.2344815731, -0.1432285905, -0.2700957656, 1.0605162382, 0.0907967538, 0.0466090441, 0.3764007986, -0...
http://arxiv.org/abs/2406.19840v1
http://arxiv.org/pdf/2406.19840v1
AnomaLLMy -- Detecting anomalous tokens in black-box LLMs through low-confidence single-token predictions
2024-06-28T11:28:44
2024-06-28T11:28:44
Waligóra Witold
cs.CL
cs.CL, cs.AI
This paper introduces AnomaLLMy, a novel technique for the automatic detection of anomalous tokens in black-box Large Language Models (LLMs) with API-only access. Utilizing low-confidence single-token predictions as a cost-effective indicator, AnomaLLMy identifies irregularities in model behavior, addressing the issue ...
[ -0.0810568854, -0.875459969, 0.5582386851, 0.1755743325, -0.0894415304, 0.2465010583, 0.1144737452, -0.3062659502, 0.3939180374, 0.7831476331, -0.6975728273, -0.817648828, -0.6937877536, 0.2155941427, -0.799276948, 0.4082040489, 0.1691375077, -0.1679030508, -0.2610885203, -0.01...
http://arxiv.org/abs/2406.19820v1
http://arxiv.org/pdf/2406.19820v1
BeamAggR: Beam Aggregation Reasoning over Multi-source Knowledge for Multi-hop Question Answering
2024-06-28T10:53:48
2024-06-28T10:53:48
Zheng Chu, Jingchang Chen, Qianglong Chen, Haotian Wang, Kun Zhu, Xiyuan Du, Weijiang Yu, Ming Liu, Bing Qin
cs.CL
cs.CL, cs.AI
Large language models (LLMs) have demonstrated strong reasoning capabilities. Nevertheless, they still suffer from factual errors when tackling knowledge-intensive tasks. Retrieval-augmented reasoning represents a promising approach. However, significant challenges still persist, including inaccurate and insufficient r...
[ -0.420923382, -1.1662745476, 0.9573814273, 0.1680403799, 0.0366506539, 0.4927895963, 0.2287600338, -0.3012241721, -0.0225167535, 0.2943302989, -0.2420100421, -0.4772390127, -0.6476097703, -0.1361202747, -0.1454418451, 0.5334106684, 0.0053057391, 0.046456296, -0.5310708284, -0.2...
http://arxiv.org/abs/2407.00134v1
http://arxiv.org/pdf/2407.00134v1
A Simple Attention-Based Mechanism for Bimodal Emotion Classification
2024-06-28T10:43:02
2024-06-28T10:43:02
Mazen Elabd, Sardar Jaf
cs.LG
cs.LG, cs.AI, cs.CL
Big data contain rich information for machine learning algorithms to utilize when learning important features during classification tasks. Human beings express their emotion using certain words, speech (tone, pitch, speed) or facial expression. Artificial Intelligence approach to emotion classification are largely base...
[ -0.3684341013, -0.6379048824, 0.2774282992, 0.4981923997, -0.1545216888, 0.0497239679, -0.3838177025, -0.4964033067, -0.081396699, 0.0741849691, -0.4202027619, -0.3795281053, -0.4323536158, -0.122139141, -0.0548733659, 0.8079385161, -0.0133303003, 0.241919592, -0.1161890626, -0...
http://arxiv.org/abs/2406.19803v1
http://arxiv.org/pdf/2406.19803v1
Scalable and Domain-General Abstractive Proposition Segmentation
2024-06-28T10:24:31
2024-06-28T10:24:31
Mohammad Javad Hosseini, Yang Gao, Tim Baumgärtner, Alex Fabrikant, Reinald Kim Amplayo
cs.CL
cs.CL
Segmenting text into fine-grained units of meaning is important to a wide range of NLP applications. The default approach of segmenting text into sentences is often insufficient, especially since sentences are usually complex enough to include multiple units of meaning that merit separate treatment in the downstream ta...
[ -0.3388991356, -0.8252557516, 0.7001099586, 0.2123760283, -0.456469208, -0.0483738147, -0.0706938431, -0.4364225566, 0.0785018802, 0.5386307836, -0.2923656702, -0.473131448, -0.5922113657, 0.1378200203, -0.4406704605, 0.9700929523, 0.0838945881, 0.0103323227, -0.5604965091, -0....
http://arxiv.org/abs/2406.19783v1
http://arxiv.org/pdf/2406.19783v1
NLPerturbator: Studying the Robustness of Code LLMs to Natural Language Variations
2024-06-28T09:39:33
2024-06-28T09:39:33
Junkai Chen, Zhenhao Li, Xing Hu, Xin Xia
cs.SE
cs.SE, cs.CL
Large language models (LLMs) achieve promising results in code generation based on a given natural language description. They have been integrated into open-source projects and commercial products to facilitate daily coding activities. The natural language description in the prompt is crucial for LLMs to comprehend use...
[ -0.6429966688, -1.1320098639, 0.220827058, 0.4921558201, -0.1952807307, 0.1034067422, -0.2852716148, -0.3201429546, 0.105535306, 0.628326416, -0.5452213287, -0.4618072808, -0.4460015297, 0.7029525042, -0.3814408183, 0.957719326, 0.0163991973, -0.2195061147, -0.2035463601, 0.219...
http://arxiv.org/abs/2406.19774v1
http://arxiv.org/pdf/2406.19774v1
Direct Preference Knowledge Distillation for Large Language Models
2024-06-28T09:23:40
2024-06-28T09:23:40
Yixing Li, Yuxian Gu, Li Dong, Dequan Wang, Yu Cheng, Furu Wei
cs.CL
cs.CL
In the field of large language models (LLMs), Knowledge Distillation (KD) is a critical technique for transferring capabilities from teacher models to student models. However, existing KD methods face limitations and challenges in distillation of LLMs, including efficiency and insufficient measurement capabilities of t...
[ -0.3407057226, -0.4404237568, 0.9480926394, 0.1490387022, -0.4579737186, 0.11430718, -0.6041375399, 0.1853193492, -0.1880892217, 0.3980720043, -0.3343929946, -0.5424882174, -0.9030621052, 0.0037996334, -0.2262460291, 0.9292529225, -0.2477949262, 0.1936476231, -0.3617744744, 0.2...
http://arxiv.org/abs/2406.19764v1
http://arxiv.org/pdf/2406.19764v1
Belief Revision: The Adaptability of Large Language Models Reasoning
2024-06-28T09:09:36
2024-06-28T09:09:36
Bryan Wilie, Samuel Cahyawijaya, Etsuko Ishii, Junxian He, Pascale Fung
cs.CL
cs.CL
The capability to reason from text is crucial for real-world NLP applications. Real-world scenarios often involve incomplete or evolving data. In response, individuals update their beliefs and understandings accordingly. However, most existing evaluations assume that language models (LMs) operate with consistent inform...
[ -0.4222872555, -0.5766450167, 0.8462521434, 0.0508455262, -0.0578900166, -0.2645668983, -0.3818856776, -0.4564682543, -0.0139651094, 0.5689839125, -0.6971364021, -0.2277885228, -0.4626850784, 0.2675710618, -0.10994073, 0.7097845078, -0.1071065888, -0.0526997484, -0.0549492314, ...
http://arxiv.org/abs/2406.19760v1
http://arxiv.org/pdf/2406.19760v1
Learning Interpretable Legal Case Retrieval via Knowledge-Guided Case Reformulation
2024-06-28T08:59:45
2024-06-28T08:59:45
Chenlong Deng, Kelong Mao, Zhicheng Dou
cs.IR
cs.IR, cs.CL
Legal case retrieval for sourcing similar cases is critical in upholding judicial fairness. Different from general web search, legal case retrieval involves processing lengthy, complex, and highly specialized legal documents. Existing methods in this domain often overlook the incorporation of legal expert knowledge, wh...
[ -0.0366675779, -0.7101771235, 1.1325314045, -0.2080609202, -0.4712901711, -0.1463858336, -0.2268592119, -0.3435174525, 0.202954188, 1.0085117817, -0.1901593506, -0.7614532709, -0.5650732517, -0.0930740386, -0.5460785627, 1.066129446, 0.1007372141, 0.2772680223, -0.130102396, -0...
http://arxiv.org/abs/2406.19759v1
http://arxiv.org/pdf/2406.19759v1
Breaking the Script Barrier in Multilingual Pre-Trained Language Models with Transliteration-Based Post-Training Alignment
2024-06-28T08:59:24
2024-06-28T08:59:24
Orgest Xhelili, Yihong Liu, Hinrich Schütze
cs.CL
cs.CL
Multilingual pre-trained models (mPLMs) have shown impressive performance on cross-lingual transfer tasks. However, the transfer performance is often hindered when a low-resource target language is written in a different script than the high-resource source language, even though the two languages may be related or shar...
[ -0.1201932505, 0.0683339238, 0.3516430855, 0.1035532728, -0.2728025019, 0.4032234251, -0.2448504418, -0.2745788097, 0.1629561782, 0.3293561339, -0.6614829302, -0.3867125809, -0.7274461985, 0.1021063998, 0.0060005574, 0.8368098736, -0.530684948, 0.4846522808, 0.2128733844, -0.23...
http://arxiv.org/abs/2406.19736v1
http://arxiv.org/pdf/2406.19736v1
MM-Instruct: Generated Visual Instructions for Large Multimodal Model Alignment
2024-06-28T08:25:27
2024-06-28T08:25:27
Jihao Liu, Xin Huang, Jinliang Zheng, Boxiao Liu, Jia Wang, Osamu Yoshie, Yu Liu, Hongsheng Li
cs.CV
cs.CV, cs.AI, cs.CL, cs.LG
This paper introduces MM-Instruct, a large-scale dataset of diverse and high-quality visual instruction data designed to enhance the instruction-following capabilities of large multimodal models (LMMs). While existing visual instruction datasets often focus on question-answering, they struggle to generalize to broader ...
[ -0.2671428323, -0.8213998079, 0.4282616973, -0.1004944742, -0.1363350898, 0.1110796556, -0.0499553718, -0.004300314, -0.0151974522, 0.5712806582, -0.9371930957, -0.6743268371, -0.4829587638, -0.2091493905, -0.1601987928, 0.8011721373, -0.1048050374, 0.3508986235, -0.6370321512, ...
http://arxiv.org/abs/2406.19731v1
http://arxiv.org/pdf/2406.19731v1
Message du troisi{è}me type : irruption d'un tiers dans un dialogue en ligne
2024-06-28T08:20:32
2024-06-28T08:20:32
Ludovic Tanguy, Céline Poudat, Lydia-Mai Ho-Dac
cs.CL
cs.CL
Our study focuses on Wikipedia talk pages, from a global perspective analyzing contributors' behaviors in online interactions. Using a corpus comprising all Wikipedia talk pages in French, totaling more than 300,000 discussion threads, we examine how discussions with more than two participants (multiparty conversation)...
[ -0.7813003063, -0.8161631823, 0.5327888131, 0.4973872304, -0.046809379, -0.0208244994, -0.0471752286, -0.7669833899, 0.803897202, 0.339224577, -0.4888570309, -0.0538086258, -0.5497005582, 0.2724878788, -0.7577619553, 1.2908415794, -0.0956271887, 0.1533869505, -0.1202883497, -0....
http://arxiv.org/abs/2406.19729v1
http://arxiv.org/pdf/2406.19729v1
Le sens de la famille : analyse du vocabulaire de la parent{é} par les plongements de mots
2024-06-28T08:19:36
2024-06-28T08:19:36
Ludovic Tanguy, Cécile Fabre, Nabil Hathout, Lydia-Mai Ho-Dac
cs.CL
cs.CL
In this study, we propose a corpus analysis of an area of the French lexicon that is both dense and highly structured: the vocabulary of family relationships. Starting with a lexicon of 25 nouns designating the main relationships (son, cousin, mother, grandfather, sister-in-law etc.), we examine how these terms are pos...
[ -0.5783603191, -0.6263800859, 0.2190634012, 0.4117914438, -0.015486774, -0.0372683369, 0.11031349, -0.0730391666, 0.4232236147, 0.3992402554, -0.3948345184, -0.1506674439, -0.7817195654, 0.2016411573, -0.3038857281, 0.6713960171, -0.1327625066, -0.111735411, -0.0462958291, -0.5...
http://arxiv.org/abs/2406.19712v1
http://arxiv.org/pdf/2406.19712v1
Uncertainty Quantification in Large Language Models Through Convex Hull Analysis
2024-06-28T07:47:34
2024-06-28T07:47:34
Ferhat Ozgur Catak, Murat Kuzlu
cs.AI
cs.AI, cs.CL
Uncertainty quantification approaches have been more critical in large language models (LLMs), particularly high-risk applications requiring reliable outputs. However, traditional methods for uncertainty quantification, such as probabilistic models and ensemble techniques, face challenges when applied to the complex an...
[ -0.513001442, -1.5498044491, 0.8257203102, 0.2809129357, -0.2013531476, -0.0949424505, -0.3717650175, -0.1050427482, 0.1242177933, -0.0065258499, -0.3834001422, -0.6880958676, -0.6743109226, -0.1211486086, -0.0773141384, 0.3433456719, 0.0170542467, 0.1166050285, -0.1555151343, ...
http://arxiv.org/abs/2406.19674v1
http://arxiv.org/pdf/2406.19674v1
Less is More: Accurate Speech Recognition & Translation without Web-Scale Data
2024-06-28T06:22:23
2024-06-28T06:22:23
Krishna C. Puvvada, Piotr Żelasko, He Huang, Oleksii Hrinchuk, Nithin Rao Koluguri, Kunal Dhawan, Somshubra Majumdar, Elena Rastorgueva, Zhehuai Chen, Vitaly Lavrukhin, Jagadeesh Balam, Boris Ginsburg
cs.CL
cs.CL, cs.LG, cs.SD, eess.AS
Recent advances in speech recognition and translation rely on hundreds of thousands of hours of Internet speech data. We argue that state-of-the art accuracy can be reached without relying on web-scale data. Canary - multilingual ASR and speech translation model, outperforms current state-of-the-art models - Whisper, O...
[ -0.2203813344, -0.3852702379, 0.1295791715, 0.1589633524, -0.0662766472, 0.0156876519, -0.1785585433, -0.7605012655, 0.1191319302, 0.2168671489, -0.5315590501, -0.2530270219, -0.6628367305, 0.079041101, -0.8120577335, 0.5640937686, 0.401827246, 0.3059968948, -0.0246896558, -0.1...
http://arxiv.org/abs/2406.19650v1
http://arxiv.org/pdf/2406.19650v1
DECOR: Improving Coherence in L2 English Writing with a Novel Benchmark for Incoherence Detection, Reasoning, and Rewriting
2024-06-28T04:38:54
2024-06-28T04:38:54
Xuanming Zhang, Anthony Diaz, Zixun Chen, Qingyang Wu, Kun Qian, Erik Voss, Zhou Yu
cs.CL
cs.CL
Coherence in writing, an aspect that second-language (L2) English learners often struggle with, is crucial in assessing L2 English writing. Existing automated writing evaluation systems primarily use basic surface linguistic features to detect coherence in writing. However, little effort has been made to correct the de...
[ -0.3240197003, -0.7683856487, 0.3368673921, 0.1710564345, -0.049061805, 0.2817984819, -0.2382568121, -0.7651689649, 0.2983316481, 0.1710488796, -0.458584249, -0.2328320146, -0.4872061908, 0.3134238124, -0.1457717866, 0.6874874234, -0.3552566171, -0.0204515457, -0.7014642358, -0...
http://arxiv.org/abs/2406.19648v1
http://arxiv.org/pdf/2406.19648v1
Designing and Evaluating Multi-Chatbot Interface for Human-AI Communication: Preliminary Findings from a Persuasion Task
2024-06-28T04:33:41
2024-06-28T04:33:41
Sion Yoon, Tae Eun Kim, Yoo Jung Oh
cs.HC
cs.HC, cs.AI, cs.CL
The dynamics of human-AI communication have been reshaped by language models such as ChatGPT. However, extant research has primarily focused on dyadic communication, leaving much to be explored regarding the dynamics of human-AI communication in group settings. The availability of multiple language model chatbots prese...
[ -0.4467735589, -0.7839599848, 0.2180032879, 0.2169076353, -0.2792926133, 0.3689554632, -0.2405842245, -0.5543775558, 0.1379711926, 0.0750322416, -0.6492617726, -0.019006284, -0.5203102231, 0.134833321, -0.1734930128, 1.0925841331, 0.4152938426, 0.260954082, -0.2172061652, -0.06...
http://arxiv.org/abs/2406.19643v1
http://arxiv.org/pdf/2406.19643v1
Unlocking Varied Perspectives: A Persona-Based Multi-Agent Framework with Debate-Driven Text Planning for Argument Generation
2024-06-28T04:21:20
2024-06-28T04:21:20
Zhe Hu, Hou Pong Chan, Jing Li, Yu Yin
cs.CL
cs.CL, cs.AI
Writing persuasive arguments is a challenging task for both humans and machines. It entails incorporating high-level beliefs from various perspectives on the topic, along with deliberate reasoning and planning to construct a coherent narrative. Current language models often generate surface tokens autoregressively, lac...
[ -0.6930878162, -0.5987663269, 0.8138468266, 0.2424824685, 0.0056768111, 0.2888530195, 0.1844407171, 0.0402989425, 0.2305372506, 0.6603063941, -0.7825362682, -0.3751943111, -0.2090015411, 0.1781162024, 0.1977904588, 1.0581718683, -0.119388856, -0.1204975024, -0.2658283114, -0.31...
http://arxiv.org/abs/2406.19642v1
http://arxiv.org/pdf/2406.19642v1
IDT: Dual-Task Adversarial Attacks for Privacy Protection
2024-06-28T04:14:35
2024-06-28T04:14:35
Pedro Faustini, Shakila Mahjabin Tonni, Annabelle McIver, Qiongkai Xu, Mark Dras
cs.CL
cs.CL, cs.CR, cs.LG
Natural language processing (NLP) models may leak private information in different ways, including membership inference, reconstruction or attribute inference attacks. Sensitive information may not be explicit in the text, but hidden in underlying writing characteristics. Methods to protect privacy can involve using re...
[ -0.2413528264, -0.882729888, 0.1196557283, 0.206150189, -0.1555518806, 0.3086087406, -0.1326987147, -0.4453470111, -0.2505066395, 0.6807975173, -0.5662062168, -0.4000751078, -0.8937205076, -0.0501254909, -0.5780422091, 1.3089020252, 0.2137897164, -0.1466649175, -0.0617148615, -...
http://arxiv.org/abs/2406.19598v1
http://arxiv.org/pdf/2406.19598v1
Mixture of In-Context Experts Enhance LLMs' Long Context Awareness
2024-06-28T01:46:41
2024-06-28T01:46:41
Hongzhan Lin, Ang Lv, Yuhan Chen, Chen Zhu, Yang Song, Hengshu Zhu, Rui Yan
cs.CL
cs.CL
Many studies have revealed that large language models (LLMs) exhibit uneven awareness of different contextual positions.Their limited context awareness can lead to overlooking critical information and subsequent task failures. While several approaches have been proposed to enhance LLMs' context awareness, achieving bot...
[ -0.5445863008, -0.5253634453, 0.4895606637, 0.2463314235, -0.3606783152, 0.2233168185, -0.3377449811, -0.388653785, 0.2941249311, 0.3310904801, -0.319085598, -0.5045212507, -0.5010355711, -0.0551972874, -0.2519824207, 0.8237846494, -0.025369361, -0.1399134994, -0.2675053477, -0...
http://arxiv.org/abs/2406.19593v1
http://arxiv.org/pdf/2406.19593v1
SK-VQA: Synthetic Knowledge Generation at Scale for Training Context-Augmented Multimodal LLMs
2024-06-28T01:14:43
2024-06-28T01:14:43
Xin Su, Man Luo, Kris W Pan, Tien Pei Chou, Vasudev Lal, Phillip Howard
cs.CL
cs.CL, cs.CV
Synthetic data generation has gained significant attention recently for its utility in training large vision and language models. However, the application of synthetic data to the training of multimodal context-augmented generation systems has been relatively unexplored. This gap in existing work is important because e...
[ -0.5877443552, -0.4449463487, 0.6220257878, 0.0021850704, -0.1924984008, 0.2110498399, -0.1075700447, -0.4022342563, 0.1260933131, 0.5991278887, -0.7686533928, -0.012719579, -0.1700150818, 0.2172077149, -0.3044763803, 0.7370516658, 0.091495797, 0.3554244339, -0.6745983958, -0.3...
http://arxiv.org/abs/2406.19578v1
http://arxiv.org/pdf/2406.19578v1
PathAlign: A vision-language model for whole slide images in histopathology
2024-06-27T23:43:36
2024-06-27T23:43:36
Faruk Ahmed, Andrew Sellergren, Lin Yang, Shawn Xu, Boris Babenko, Abbi Ward, Niels Olson, Arash Mohtashamian, Yossi Matias, Greg S. Corrado, Quang Duong, Dale R. Webster, Shravya Shetty, Daniel Golden, Yun Liu, David F. Steiner, Ellery Wulczyn
cs.CV
cs.CV, cs.AI, cs.CL, cs.LG
Microscopic interpretation of histopathology images underlies many important diagnostic and treatment decisions. While advances in vision-language modeling raise new opportunities for analysis of such images, the gigapixel-scale size of whole slide images (WSIs) introduces unique challenges. Additionally, pathology rep...
[ 0.0636096746, -0.2519106269, 0.8191940784, 0.1255326718, -0.6018443704, -0.0735631585, -0.0886728689, -0.434568733, 0.5155949593, 0.6026813984, -0.2302839756, -0.5911412239, -0.7305087447, 0.1128618494, -0.2883316576, 0.9983676076, 0.0517430715, 0.3244157434, -0.313691318, -0.2...
http://arxiv.org/abs/2406.19564v1
http://arxiv.org/pdf/2406.19564v1
Voices Unheard: NLP Resources and Models for Yorùbá Regional Dialects
2024-06-27T22:38:04
2024-06-27T22:38:04
Orevaoghene Ahia, Anuoluwapo Aremu, Diana Abagyan, Hila Gonen, David Ifeoluwa Adelani, Daud Abolade, Noah A. Smith, Yulia Tsvetkov
cs.CL
cs.CL
Yor\`ub\'a an African language with roughly 47 million speakers encompasses a continuum with several dialects. Recent efforts to develop NLP technologies for African languages have focused on their standard dialects, resulting in disparities for dialects and varieties for which there are little to no resources or tools...
[ -0.6673585773, -0.6711573601, 0.1298196614, 0.1890816838, -0.1403449029, 0.1421262622, -0.2529540658, -0.5451335311, 0.3898304403, 0.8253393769, -0.5721207261, -0.2419085354, -0.4487274289, 0.2724126577, -0.4501020312, 0.7690122128, 0.7605822086, -0.0728480816, 0.1308858693, -0...
http://arxiv.org/abs/2406.19552v1
http://arxiv.org/pdf/2406.19552v1
Rethinking harmless refusals when fine-tuning foundation models
2024-06-27T22:08:22
2024-06-27T22:08:22
Florin Pop, Judd Rosenblatt, Diogo Schwerz de Lucena, Michael Vaiana
cs.CL
cs.CL, cs.AI, cs.LG
In this paper, we investigate the degree to which fine-tuning in Large Language Models (LLMs) effectively mitigates versus merely conceals undesirable behavior. Through the lens of semi-realistic role-playing exercises designed to elicit such behaviors, we explore the response dynamics of LLMs post fine-tuning interven...
[ -0.5323575735, -1.1566114426, 0.5235360861, -0.0123284031, -0.2477667779, 0.0417574309, -0.2130339444, -0.3983316123, 0.0131863952, 0.9580917954, -0.9164805412, -0.0676273406, -0.3557564616, -0.0394084789, -0.5329958797, 1.1694625616, 0.0850138441, -0.1190937236, 0.0213650633, ...
http://arxiv.org/abs/2406.19545v1
http://arxiv.org/pdf/2406.19545v1
Leveraging Machine-Generated Rationales to Facilitate Social Meaning Detection in Conversations
2024-06-27T21:47:42
2024-06-27T21:47:42
Ritam Dutt, Zhen Wu, Kelly Shi, Divyanshu Sheth, Prakhar Gupta, Carolyn Penstein Rose
cs.CL
cs.CL, cs.AI
We present a generalizable classification approach that leverages Large Language Models (LLMs) to facilitate the detection of implicitly encoded social meaning in conversations. We design a multi-faceted prompt to extract a textual explanation of the reasoning that connects visible cues to underlying social meanings. T...
[ -0.2933736444, -1.0718715191, 0.9560813904, 0.2727404833, -0.5053287745, 0.0698266551, -0.1832686663, -0.5386065841, 0.1696796119, 0.4647045135, -0.5217887163, -0.4588222802, -0.9244151115, 0.1255755723, -0.3015980721, 0.8094910979, 0.08911217, -0.0941815376, -0.255934, -0.2567...
http://arxiv.org/abs/2406.19543v1
http://arxiv.org/pdf/2406.19543v1
Demarked: A Strategy for Enhanced Abusive Speech Moderation through Counterspeech, Detoxification, and Message Management
2024-06-27T21:45:33
2024-06-27T21:45:33
Seid Muhie Yimam, Daryna Dementieva, Tim Fischer, Daniil Moskovskiy, Naquee Rizwan, Punyajoy Saha, Sarthak Roy, Martin Semmann, Alexander Panchenko, Chris Biemann, Animesh Mukherjee
cs.CL
cs.CL, cs.SI
Despite regulations imposed by nations and social media platforms, such as recent EU regulations targeting digital violence, abusive content persists as a significant challenge. Existing approaches primarily rely on binary solutions, such as outright blocking or banning, yet fail to address the complex nature of abusiv...
[ -0.5650568604, -1.1122608185, 0.2788814604, 0.1481579989, -0.5185551643, 0.5134856105, -0.0293317046, -0.38593328, 0.3772847652, 0.6452037692, -0.6347091198, -0.5224186778, -1.0349649191, 0.0699399486, -0.595289886, 1.0572293997, 0.1491872519, 0.2020236552, -0.1787960976, -0.03...
http://arxiv.org/abs/2406.19538v1
http://arxiv.org/pdf/2406.19538v1
Context Matters: An Empirical Study of the Impact of Contextual Information in Temporal Question Answering Systems
2024-06-27T21:31:30
2024-06-27T21:31:30
Dan Schumacher, Fatemeh Haji, Tara Grey, Niharika Bandlamudi, Nupoor Karnik, Gagana Uday Kumar, Jason Cho-Yu Chiang, Paul Rad, Nishant Vishwamitra, Anthony Rios
cs.CL
cs.CL
Large language models (LLMs) often struggle with temporal reasoning, crucial for tasks like historical event analysis and time-sensitive information retrieval. Despite advancements, state-of-the-art models falter in handling temporal information, especially when faced with irrelevant or noisy contexts. This paper addre...
[ -0.1914124042, -0.8794987202, 0.8888842463, 0.2208375633, -0.1393753886, -0.1941245794, 0.173607111, -0.4514296651, -0.344029367, -0.2384869456, -0.4733388424, -0.1599689275, -0.4239829481, 0.0161591526, -0.3350732625, 0.6328620315, 0.0453533754, 0.1994817555, -0.2829975486, -0...
http://arxiv.org/abs/2406.19537v1
http://arxiv.org/pdf/2406.19537v1
Handling Ontology Gaps in Semantic Parsing
2024-06-27T21:21:22
2024-06-27T21:21:22
Andrea Bacciu, Marco Damonte, Marco Basaldella, Emilio Monti
cs.CL
cs.CL, cs.AI
The majority of Neural Semantic Parsing (NSP) models are developed with the assumption that there are no concepts outside the ones such models can represent with their target symbols (closed-world assumption). This assumption leads to generate hallucinated outputs rather than admitting their lack of knowledge. Hallucin...
[ -0.679263413, -0.8325948119, 0.8193513155, 0.2681349218, 0.1015605405, -0.3402319551, 0.1907422692, -0.5010288358, 0.1634855568, 0.8970276713, -0.9241143465, -0.5724152923, -0.4429461956, 0.0846131369, -0.301515907, 0.3124619126, 0.3654007316, -0.1256619394, -0.4532674253, 0.12...
http://arxiv.org/abs/2406.19526v1
http://arxiv.org/pdf/2406.19526v1
TocBERT: Medical Document Structure Extraction Using Bidirectional Transformers
2024-06-27T20:56:57
2024-06-27T20:56:57
Majd Saleh, Sarra Baghdadi, Stéphane Paquelet
cs.CL
cs.CL, cs.IR, cs.LG
Text segmentation holds paramount importance in the field of Natural Language Processing (NLP). It plays an important role in several NLP downstream tasks like information retrieval and document summarization. In this work, we propose a new solution, namely TocBERT, for segmenting texts using bidirectional transformers...
[ -0.0736257955, -0.5628584623, 0.6487780809, 0.3003487289, -0.6077564955, 0.1214461327, -0.3018250167, -0.2178578377, 0.1725057662, 0.6556328535, -0.4067922235, -0.8650434613, -0.7842133641, 0.3149756789, -0.7170379758, 1.2698762417, -0.0526766367, 0.0353176966, -0.4640039504, -...
http://arxiv.org/abs/2406.19512v1
http://arxiv.org/pdf/2406.19512v1
Captioning Visualizations with Large Language Models (CVLLM): A Tutorial
2024-06-27T20:18:18
2024-06-27T20:18:18
Giuseppe Carenini, Jordon Johnson, Ali Salamatian
cs.CL
cs.CL, cs.AI, cs.HC
Automatically captioning visualizations is not new, but recent advances in large language models(LLMs) open exciting new possibilities. In this tutorial, after providing a brief review of Information Visualization (InfoVis) principles and past work in captioning, we introduce neural models and the transformer architect...
[ -0.245046109, -0.4558204412, 0.5111766458, 0.2543974817, -0.1689131409, 0.2355943322, -0.1562982351, -0.1268372685, 0.1196067408, 0.8239210248, -0.5840086937, -0.6897692084, -0.5576505661, -0.1859559864, -0.7218430042, 0.915096581, -0.1604414731, 0.5798859596, -0.664996624, -0....
http://arxiv.org/abs/2406.19504v1
http://arxiv.org/pdf/2406.19504v1
Are Generative Language Models Multicultural? A Study on Hausa Culture and Emotions using ChatGPT
2024-06-27T19:42:13
2024-06-27T19:42:13
Ibrahim Said Ahmad, Shiran Dudy, Resmi Ramachandranpillai, Kenneth Church
cs.CL
cs.CL
Large Language Models (LLMs), such as ChatGPT, are widely used to generate content for various purposes and audiences. However, these models may not reflect the cultural and emotional diversity of their users, especially for low-resource languages. In this paper, we investigate how ChatGPT represents Hausa's culture an...
[ -0.58549124, -0.6148859859, -0.1985052675, 0.3771602511, -0.3388352692, 0.3467919827, -0.5904890895, -0.5397210717, 0.2719349265, 0.0487790368, -0.5166518688, -0.3294277191, -0.3764699697, 0.2854251266, -0.1748208702, 0.9122460485, 0.0571531206, -0.2036626637, 0.0805073157, -0....
http://arxiv.org/abs/2406.19502v1
http://arxiv.org/pdf/2406.19502v1
Investigating How Large Language Models Leverage Internal Knowledge to Perform Complex Reasoning
2024-06-27T19:29:36
2024-06-27T19:29:36
Miyoung Ko, Sue Hyun Park, Joonsuk Park, Minjoon Seo
cs.CL
cs.CL, cs.AI
Despite significant advancements, there is a limited understanding of how large language models (LLMs) utilize knowledge for reasoning. To address this, we propose a method that deconstructs complex real-world questions into a graph, representing each question as a node with parent nodes of background knowledge needed ...
[ -0.4541739821, -0.9914886355, 0.9956120849, -0.2348611802, 0.0554814599, 0.3975107968, -0.1585639864, -0.4517645538, -0.3808560669, 0.6328647733, -0.6118972898, -0.4698210359, -0.518243134, -0.0722365379, -0.2010823786, 0.5673573613, -0.151277706, 0.0611765049, -0.2360660881, 0...
http://arxiv.org/abs/2406.19501v1
http://arxiv.org/pdf/2406.19501v1
Monitoring Latent World States in Language Models with Propositional Probes
2024-06-27T19:28:43
2024-06-27T19:28:43
Jiahai Feng, Stuart Russell, Jacob Steinhardt
cs.CL
cs.CL, cs.LG
Language models are susceptible to bias, sycophancy, backdoors, and other tendencies that lead to unfaithful responses to the input context. Interpreting internal states of language models could help monitor and correct unfaithful behavior. We hypothesize that language models represent their input contexts in a latent ...
[ 0.0772322193, -0.6982329488, 0.8090898991, 0.0133390492, -0.1329388618, -0.2290594578, 0.0026026277, -0.1782999784, -0.1042618826, 0.5512052774, -0.4910458326, -0.4170169234, -0.3436128199, 0.0515124798, -0.5830106735, 0.9482215643, -0.0227383804, 0.0719766766, -0.2701151669, 0...
http://arxiv.org/abs/2406.19500v1
http://arxiv.org/pdf/2406.19500v1
Knowledge acquisition for dialogue agents using reinforcement learning on graph representations
2024-06-27T19:28:42
2024-06-27T19:28:42
Selene Baez Santamaria, Shihan Wang, Piek Vossen
cs.AI
cs.AI, cs.CL
We develop an artificial agent motivated to augment its knowledge base beyond its initial training. The agent actively participates in dialogues with other agents, strategically acquiring new information. The agent models its knowledge as an RDF knowledge graph, integrating new beliefs acquired through conversation. Re...
[ -0.4858523905, -0.9094485044, 0.8129300475, -0.1098890081, 0.0601257645, 0.4972669482, 0.0197745152, -0.5356190205, 0.0857356265, 0.6985345483, -0.4815935791, -0.5029613376, -0.5793366432, 0.0661478266, 0.0417162888, 0.953074038, 0.046649579, -0.1524907202, -0.1779041141, 0.066...
http://arxiv.org/abs/2406.19497v1
http://arxiv.org/pdf/2406.19497v1
Inclusivity in Large Language Models: Personality Traits and Gender Bias in Scientific Abstracts
2024-06-27T19:26:11
2024-06-27T19:26:11
Naseela Pervez, Alexander J. Titus
cs.CL
cs.CL, cs.AI
Large language models (LLMs) are increasingly utilized to assist in scientific and academic writing, helping authors enhance the coherence of their articles. Previous studies have highlighted stereotypes and biases present in LLM outputs, emphasizing the need to evaluate these models for their alignment with human narr...
[ -0.3414131701, -0.3208912611, 0.7288342714, 0.4179031551, 0.1737459451, 0.1535258591, -0.2858733535, -0.6582283974, 0.3268783391, 0.3702171445, -0.4375327229, -0.5320704579, -0.7757984996, 0.715446353, -0.6737023592, 1.0661759377, -0.0737841129, 0.3832054436, -0.4298838079, -0....
http://arxiv.org/abs/2406.19493v1
http://arxiv.org/pdf/2406.19493v1
Development and Evaluation of a Retrieval-Augmented Generation Tool for Creating SAPPhIRE Models of Artificial Systems
2024-06-27T19:20:09
2024-06-27T19:20:09
Anubhab Majumder, Kausik Bhattacharya, Amaresh Chakrabarti
cs.CL
cs.CL, cs.AI
Representing systems using the SAPPhIRE causality model is found useful in supporting design-by-analogy. However, creating a SAPPhIRE model of artificial or biological systems is an effort-intensive process that requires human experts to source technical knowledge from multiple technical documents regarding how the sys...
[ -0.1112629771, -0.61791116, 0.7512149215, 0.3170712292, -0.1879056841, 0.0720440447, -0.0655356199, -0.3202012777, 0.4177069664, 0.3035984039, -0.3398428559, -0.1906581074, -0.2115973234, 0.1325370073, -0.649418354, 0.871552527, 0.0811513141, 0.1696240753, -0.0170441363, 0.3585...
http://arxiv.org/abs/2406.19486v1
http://arxiv.org/pdf/2406.19486v1
LoPT: Low-Rank Prompt Tuning for Parameter Efficient Language Models
2024-06-27T19:02:41
2024-06-27T19:02:41
Shouchang Guo, Sonam Damani, Keng-hao Chang
cs.CL
cs.CL, cs.AI, cs.ET, cs.LG, eess.SP
In prompt tuning, a prefix or suffix text is added to the prompt, and the embeddings (soft prompts) or token indices (hard prompts) of the prefix/suffix are optimized to gain more control over language models for specific tasks. This approach eliminates the need for hand-crafted prompt engineering or explicit model fin...
[ -0.4653531611, -0.982018888, 0.5060248971, 0.1372089088, -0.4215906262, -0.2097508907, -0.2881796062, -0.0096938098, -0.0190959889, 0.3751750886, -0.763473928, -0.3655635417, -0.4494721591, 0.2257787138, 0.1564597338, 0.8311577439, 0.2195783705, -0.0489182211, 0.0070558884, 0.0...
http://arxiv.org/abs/2406.19482v1
http://arxiv.org/pdf/2406.19482v1
xTower: A Multilingual LLM for Explaining and Correcting Translation Errors
2024-06-27T18:51:46
2024-06-27T18:51:46
Marcos Treviso, Nuno M. Guerreiro, Sweta Agrawal, Ricardo Rei, José Pombal, Tania Vaz, Helena Wu, Beatriz Silva, Daan van Stigt, André F. T. Martins
cs.CL
cs.CL
While machine translation (MT) systems are achieving increasingly strong performance on benchmarks, they often produce translations with errors and anomalies. Understanding these errors can potentially help improve the translation quality and user experience. This paper introduces xTower, an open large language model (...
[ 0.0845445171, -0.4683046639, 0.3994119167, 0.4106849432, -0.1728445292, -0.0816226006, -0.1898221821, -0.3090023994, 0.0896120146, 0.3915023208, -0.5530022383, -0.5575610399, -0.7307703495, 0.116382964, -0.5243203044, 0.9852640033, -0.6225008368, 0.12632972, 0.219507888, -0.274...
http://arxiv.org/abs/2406.19478v1
http://arxiv.org/pdf/2406.19478v1
Sparse Regression for Machine Translation
2024-06-27T18:43:51
2024-06-27T18:43:51
Ergun Biçici
cs.CL
cs.CL, cs.AI, G.3; I.2.7
We use transductive regression techniques to learn mappings between source and target features of given parallel corpora and use these mappings to generate machine translation outputs. We show the effectiveness of $L_1$ regularized regression (\textit{lasso}) to learn the mappings between sparsely observed feature sets...
[ -0.0328126736, -0.7212315798, 0.5116245747, 0.3198904395, -0.203316167, 0.0148001537, -0.3962534964, -0.3310483992, 0.1914101988, 0.5106889606, -0.4916546941, -0.4717359245, -0.8242682219, 0.4530525208, -0.4965531826, 1.2248278856, -0.1772749573, 0.4560705125, -0.0816509426, -0...
http://arxiv.org/abs/2406.19470v1
http://arxiv.org/pdf/2406.19470v1
Changing Answer Order Can Decrease MMLU Accuracy
2024-06-27T18:21:32
2024-06-27T18:21:32
Vipul Gupta, David Pantoja, Candace Ross, Adina Williams, Megan Ung
cs.CL
cs.CL
As large language models (LLMs) have grown in prevalence, particular benchmarks have become essential for the evaluation of these models and for understanding model capabilities. Most commonly, we use test accuracy averaged across multiple subtasks in order to rank models on leaderboards, to determine which model is be...
[ -0.2954673469, -0.947476387, 0.5038996935, -0.070323728, -0.0024321829, 0.0656881332, 0.0682344884, -0.5779307485, 0.0929799452, 0.1544857919, -0.5695306063, -0.2530087829, -0.8355882764, 0.1886087507, -0.3715248406, 0.8702311516, -0.0461361334, 0.227846235, -0.4515379071, -0.7...
http://arxiv.org/abs/2406.19465v1
http://arxiv.org/pdf/2406.19465v1
Can Large Language Models Generate High-quality Patent Claims?
2024-06-27T18:07:40
2024-06-27T18:07:40
Lekang Jiang, Caiqi Zhang, Pascal A Scherz, Stephan Goetz
cs.CL
cs.CL
Large language models (LLMs) have shown exceptional performance across various text generation tasks but remain under-explored in the patent domain, which offers highly structured and precise language. This paper constructs a dataset to investigate the performance of current LLMs in patent claim generation. Our results...
[ -0.0679422542, -0.4489767551, 0.6672207117, 0.2097385675, -0.0851293132, 0.24017106, -0.2464645356, -0.7652289271, 0.1360560954, 0.5654148459, 0.1207076684, -0.4270220995, -0.6346586347, 0.1847249269, -0.3122494519, 0.699398756, -0.010897289, -0.243407771, -0.1959509552, 0.3288...
http://arxiv.org/abs/2406.19388v1
http://arxiv.org/pdf/2406.19388v1
Taming Data and Transformers for Audio Generation
2024-06-27T17:58:54
2024-06-27T17:58:54
Moayed Haji-Ali, Willi Menapace, Aliaksandr Siarohin, Guha Balakrishnan, Sergey Tulyakov, Vicente Ordonez
cs.SD
cs.SD, cs.CL, cs.CV, cs.MM, eess.AS
Generating ambient sounds and effects is a challenging problem due to data scarcity and often insufficient caption quality, making it difficult to employ large-scale generative models for the task. In this work, we tackle the problem by introducing two new models. First, we propose AutoCap, a high-quality and efficient...
[ -0.7766627073, -0.6043565869, 0.2714280784, 0.2571496964, -0.1227630302, 0.3463197649, 0.0220418032, -0.4583424032, 0.0695786402, 0.6033875346, -0.6378999949, -0.3406114876, -0.465367347, 0.1203668192, -0.8320067525, 0.6623582244, 0.1211219132, -0.3536193371, -0.0444468632, 0.0...
http://arxiv.org/abs/2406.19384v1
http://arxiv.org/pdf/2406.19384v1
The Remarkable Robustness of LLMs: Stages of Inference?
2024-06-27T17:57:03
2024-06-27T17:57:03
Vedang Lad, Wes Gurnee, Max Tegmark
cs.LG
cs.LG, cs.AI, cs.CL
We demonstrate and investigate the remarkable robustness of Large Language Models by deleting and swapping adjacent layers. We find that deleting and swapping interventions retain 72-95\% of the original model's prediction accuracy without fine-tuning, whereas models with more layers exhibit more robustness. Based on t...
[ -0.2278255969, -0.8587782383, 0.7738869786, 0.0304621775, -0.0015166418, 0.0795516819, -0.0806326717, -0.4364953041, -0.2081277668, 0.553371489, -0.5462569594, -0.4619857967, -0.4880237877, -0.1407040656, -0.7381148338, 0.7993941307, 0.2090148181, 0.1673013568, 0.0746666268, 0....
http://arxiv.org/abs/2406.19371v1
http://arxiv.org/pdf/2406.19371v1
Suri: Multi-constraint Instruction Following for Long-form Text Generation
2024-06-27T17:50:35
2024-06-27T17:50:35
Chau Minh Pham, Simeng Sun, Mohit Iyyer
cs.CL
cs.CL
Existing research on instruction following largely focuses on tasks with simple instructions and short responses. In this work, we explore multi-constraint instruction following for generating long-form text. We create Suri, a dataset with 20K human-written long-form texts paired with LLM-generated backtranslated instr...
[ -0.0611308329, -0.487687856, 0.4604957998, 0.1793876737, -0.0002255726, -0.1345138103, -0.2603928745, -0.4971957505, -0.1129774228, 1.0259346962, -1.0569701195, -0.6260743737, -0.2386939824, 0.2730541825, -0.0602019541, 1.0448330641, -0.2502327859, -0.0853617415, -0.2900972962, ...
http://arxiv.org/abs/2407.00121v1
http://arxiv.org/pdf/2407.00121v1
Granite-Function Calling Model: Introducing Function Calling Abilities via Multi-task Learning of Granular Tasks
2024-06-27T17:47:26
2024-06-27T17:47:26
Ibrahim Abdelaziz, Kinjal Basu, Mayank Agarwal, Sadhana Kumaravel, Matthew Stallone, Rameswar Panda, Yara Rizk, GP Bhargav, Maxwell Crouse, Chulaka Gunasekara, Shajith Ikbal, Sachin Joshi, Hima Karanam, Vineet Kumar, Asim Munawar, Sumit Neelam, Dinesh Raghu, Udit Sharma, Adriana Meza Soria, Dheeraj Sreedhar, Praveen Ve...
cs.LG
cs.LG, cs.AI, cs.CL
Large language models (LLMs) have recently shown tremendous promise in serving as the backbone to agentic systems, as demonstrated by their performance in multi-faceted, challenging benchmarks like SWE-Bench and Agent-Bench. However, to realize the true potential of LLMs as autonomous agents, they must learn to identif...
[ -0.2937496305, -0.8631148934, 0.4893568754, 0.2617434859, -0.0517801419, 0.0189716406, -0.4139297903, -0.5085480213, -0.3284886181, 0.8083678484, -0.5936694741, -0.3517948687, -0.4442733824, -0.0357527062, -0.1669945717, 0.7993221283, 0.0723167956, 0.4168303907, -0.5838394165, ...
http://arxiv.org/abs/2406.19358v1
http://arxiv.org/pdf/2406.19358v1
The Model Arena for Cross-lingual Sentiment Analysis: A Comparative Study in the Era of Large Language Models
2024-06-27T17:38:45
2024-06-27T17:38:45
Xiliang Zhu, Shayna Gardiner, Tere Roldán, David Rossouw
cs.CL
cs.CL
Sentiment analysis serves as a pivotal component in Natural Language Processing (NLP). Advancements in multilingual pre-trained models such as XLM-R and mT5 have contributed to the increasing interest in cross-lingual sentiment analysis. The recent emergence in Large Language Models (LLM) has significantly advanced gen...
[ -0.5095820427, -0.2543129921, 0.5675985813, 0.5721489191, -0.2161262035, 0.1977135837, -0.2888497114, -0.699432373, 0.0784229115, 0.240324825, -0.2279721349, -0.4829381406, -0.6537781954, 0.4618213475, -0.1648886055, 1.0086796284, -0.0385474376, 0.2973096073, -0.0233029388, -0....
http://arxiv.org/abs/2406.19356v1
http://arxiv.org/pdf/2406.19356v1
DiVERT: Distractor Generation with Variational Errors Represented as Text for Math Multiple-choice Questions
2024-06-27T17:37:31
2024-06-27T17:37:31
Nigel Fernandez, Alexander Scarlatos, Simon Woodhead, Andrew Lan
cs.CL
cs.CL, cs.CY, cs.LG
High-quality distractors are crucial to both the assessment and pedagogical value of multiple-choice questions (MCQs), where manually crafting ones that anticipate knowledge deficiencies or misconceptions among real students is difficult. Meanwhile, automated distractor generation, even with the help of large language ...
[ -0.4846668839, -0.6488329172, 0.6466665864, 0.2390616387, -0.2291415483, 0.2112288624, 0.0102030067, 0.0426863767, -0.0570865236, 0.6058220863, -0.9636710882, 0.0793137625, -0.531717658, 0.1961255223, -0.0088953124, 0.8504316211, -0.3961399198, 0.2238416821, -0.2209044099, 0.02...
http://arxiv.org/abs/2406.19354v1
http://arxiv.org/pdf/2406.19354v1
Fundamental Problems With Model Editing: How Should Rational Belief Revision Work in LLMs?
2024-06-27T17:33:03
2024-06-27T17:33:03
Peter Hase, Thomas Hofweber, Xiang Zhou, Elias Stengel-Eskin, Mohit Bansal
cs.CL
cs.CL, cs.AI
The model editing problem concerns how language models should learn new facts about the world over time. While empirical research on model editing has drawn widespread attention, the conceptual foundations of model editing remain shaky -- perhaps unsurprisingly, since model editing is essentially belief revision, a sto...
[ -0.0927358568, -0.91730088, 0.8812292814, -0.0025517314, 0.0465980992, 0.0304614864, -0.2592016459, -0.4093419909, -0.1396788806, 0.5881155133, -0.5603974462, -0.4566047788, -0.5255980492, 0.2541123033, -0.1464077383, 0.7004337907, -0.4767677188, 0.0949769244, -0.3064751923, -0...
http://arxiv.org/abs/2406.19349v1
http://arxiv.org/pdf/2406.19349v1
IndoToxic2024: A Demographically-Enriched Dataset of Hate Speech and Toxicity Types for Indonesian Language
2024-06-27T17:26:38
2024-06-27T17:26:38
Lucky Susanto, Musa Izzanardi Wijanarko, Prasetia Anugrah Pratama, Traci Hong, Ika Idris, Alham Fikri Aji, Derry Wijaya
cs.CL
cs.CL, cs.AI
Hate speech poses a significant threat to social harmony. Over the past two years, Indonesia has seen a ten-fold increase in the online hate speech ratio, underscoring the urgent need for effective detection mechanisms. However, progress is hindered by the limited availability of labeled data for Indonesian texts. The ...
[ -0.471267134, -0.6304606795, -0.021048557, 0.3668730855, -0.1893087178, -0.131717816, -0.4090877175, -0.6541134119, 0.0977427512, 0.7577537894, -0.3301605582, -0.5680958629, -0.686216116, 0.2493277192, -0.3336386383, 1.2041809559, 0.5767021775, -0.1925511509, -0.1129033267, -0....
http://arxiv.org/abs/2406.19317v1
http://arxiv.org/pdf/2406.19317v1
Jump Starting Bandits with LLM-Generated Prior Knowledge
2024-06-27T16:52:19
2024-06-27T16:52:19
Parand A. Alamdari, Yanshuai Cao, Kevin H. Wilson
cs.LG
cs.LG, cs.AI, cs.CL
We present substantial evidence demonstrating the benefits of integrating Large Language Models (LLMs) with a Contextual Multi-Armed Bandit framework. Contextual bandits have been widely used in recommendation systems to generate personalized suggestions based on user-specific contexts. We show that LLMs, pre-trained o...
[ -0.4737771451, -0.6747410297, 0.7550473809, -0.0405894779, -0.0473137386, 0.1471441984, -0.459894985, -0.3109333217, 0.2930303216, 0.871949017, -0.5135807991, -0.6555714011, -0.5977692604, -0.2575989962, -0.1401121467, 1.2788038254, 0.20125027, -0.2161205262, -0.1249119267, 0.1...