publication_date
stringlengths 20
20
| doc_url
stringlengths 33
34
| id
stringlengths 33
34
| authors
stringlengths 5
20.1k
| update_date
stringlengths 20
20
| category_all
stringlengths 5
138
| abstract
stringlengths 123
3.09k
| category_primary
stringclasses 78
values | title
stringlengths 10
211
|
|---|---|---|---|---|---|---|---|---|
2024-04-22T13:56:07Z | http://arxiv.org/pdf/2404.14183v1 | http://arxiv.org/abs/2404.14183v1 | Yuxia Wang, Jonibek Mansurov, Petar Ivanov, Jinyan Su, Artem Shelmanov, Akim Tsvigun, Osama Mohammed Afzal, Tarek Mahmoud, Giovanni Puccetti, Thomas Arnold, Chenxi Whitehouse, Alham Fikri Aji, Nizar Habash, Iryna Gurevych, Preslav Nakov | 2024-04-22T13:56:07Z | cs.CL | We present the results and the main findings of SemEval-2024 Task 8:
Multigenerator, Multidomain, and Multilingual Machine-Generated Text Detection.
The task featured three subtasks. Subtask A is a binary classification task
determining whether a text is written by a human or generated by a machine.
This subtask has two tracks: a monolingual track focused solely on English
texts and a multilingual track. Subtask B is to detect the exact source of a
text, discerning whether it is written by a human or generated by a specific
LLM. Subtask C aims to identify the changing point within a text, at which the
authorship transitions from human to machine. The task attracted a large number
of participants: subtask A monolingual (126), subtask A multilingual (59),
subtask B (70), and subtask C (30). In this paper, we present the task, analyze
the results, and discuss the system submissions and the methods they used. For
all subtasks, the best systems used LLMs. | cs.CL | SemEval-2024 Task 8: Multidomain, Multimodel and Multilingual Machine-Generated Text Detection |
2024-04-22T12:21:12Z | http://arxiv.org/pdf/2404.14122v1 | http://arxiv.org/abs/2404.14122v1 | Dawei Zhu, Pinzhen Chen, Miaoran Zhang, Barry Haddow, Xiaoyu Shen, Dietrich Klakow | 2024-04-22T12:21:12Z | cs.CL | Traditionally, success in multilingual machine translation can be attributed
to three key factors in training data: large volume, diverse translation
directions, and high quality. In the current practice of fine-tuning large
language models (LLMs) for translation, we revisit the importance of all these
factors. We find that LLMs display strong translation capability after being
fine-tuned on as few as 32 training instances, and that fine-tuning on a single
translation direction effectively enables LLMs to translate in multiple
directions. However, the choice of direction is critical: fine-tuning LLMs with
English on the target side can lead to task misinterpretation, which hinders
translations into non-English languages. A similar problem arises when noise is
introduced into the target side of parallel data, especially when the target
language is well-represented in the LLM's pre-training. In contrast, noise in
an under-represented language has a less pronounced effect. Our findings
suggest that attaining successful alignment hinges on teaching the model to
maintain a "superficial" focus, thereby avoiding the learning of erroneous
biases beyond translation. | cs.CL | Fine-Tuning Large Language Models to Translate: Will a Touch of Noisy Data in Misaligned Languages Suffice? |
2024-04-22T12:06:54Z | http://arxiv.org/pdf/2404.14465v1 | http://arxiv.org/abs/2404.14465v1 | Dimitris Asimopoulos, Ilias Siniosoglou, Vasileios Argyriou, Thomai Karamitsou, Eleftherios Fountoukidis, Sotirios K. Goudos, Ioannis D. Moscholios, Konstantinos E. Psannis, Panagiotis Sarigiannidis | 2024-04-22T12:06:54Z | cs.CL, cs.AI, cs.IR | In the realm of data privacy, the ability to effectively anonymise text is
paramount. With the proliferation of deep learning and, in particular,
transformer architectures, there is a burgeoning interest in leveraging these
advanced models for text anonymisation tasks. This paper presents a
comprehensive benchmarking study comparing the performance of transformer-based
models and Large Language Models(LLM) against traditional architectures for
text anonymisation. Utilising the CoNLL-2003 dataset, known for its robustness
and diversity, we evaluate several models. Our results showcase the strengths
and weaknesses of each approach, offering a clear perspective on the efficacy
of modern versus traditional methods. Notably, while modern models exhibit
advanced capabilities in capturing con textual nuances, certain traditional
architectures still keep high performance. This work aims to guide researchers
in selecting the most suitable model for their anonymisation needs, while also
shedding light on potential paths for future advancements in the field. | cs.CL | Benchmarking Advanced Text Anonymisation Methods: A Comparative Study on Novel and Traditional Approaches |
2024-04-22T10:16:02Z | http://arxiv.org/pdf/2404.14057v2 | http://arxiv.org/abs/2404.14057v2 | Shir Lissak, Yaakov Ophir, Refael Tikochinski, Anat Brunstein Klomek, Itay Sisso, Eyal Fruchter, Roi Reichart | 2024-04-26T11:50:25Z | cs.CL | Background: Recent advancements in Artificial Intelligence (AI) contributed
significantly to suicide assessment, however, our theoretical understanding of
this complex behavior is still limited. Objective: This study aimed to harness
AI methodologies to uncover hidden risk factors that trigger or aggravate
suicide behaviors. Method: The primary dataset included 228,052 Facebook
postings by 1,006 users who completed the gold-standard Columbia Suicide
Severity Rating Scale. This dataset was analyzed using a bottom-up research
pipeline without a-priory hypotheses and its findings were validated using a
top-down analysis of a new dataset. This secondary dataset included responses
by 1,062 participants to the same suicide scale as well as to well-validated
scales measuring depression and boredom. Results: An almost fully automated,
AI-guided research pipeline resulted in four Facebook topics that predicted the
risk of suicide, of which the strongest predictor was boredom. A comprehensive
literature review using APA PsycInfo revealed that boredom is rarely perceived
as a unique risk factor of suicide. A complementing top-down path analysis of
the secondary dataset uncovered an indirect relationship between boredom and
suicide, which was mediated by depression. An equivalent mediated relationship
was observed in the primary Facebook dataset as well. However, here, a direct
relationship between boredom and suicide risk was also observed. Conclusions:
Integrating AI methods allowed the discovery of an under-researched risk factor
of suicide. The study signals boredom as a maladaptive 'ingredient' that might
trigger suicide behaviors, regardless of depression. Further studies are
recommended to direct clinicians' attention to this burdening, and sometimes
existential experience. | cs.CL | Bored to Death: Artificial Intelligence Research Reveals the Role of Boredom in Suicide Behavior |
2024-04-22T10:06:21Z | http://arxiv.org/pdf/2404.14052v1 | http://arxiv.org/abs/2404.14052v1 | Kun Sun, Rong Wang | 2024-04-22T10:06:21Z | cs.CL, stat.ME | Data-driven approaches have revolutionized scientific research. Machine
learning and statistical analysis are commonly utilized in this type of
research. Despite their widespread use, these methodologies differ
significantly in their techniques and objectives. Few studies have utilized a
consistent dataset to demonstrate these differences within the social sciences,
particularly in language and cognitive sciences. This study leverages the
Buckeye Speech Corpus to illustrate how both machine learning and statistical
analysis are applied in data-driven research to obtain distinct insights. This
study significantly enhances our understanding of the diverse approaches
employed in data-driven strategies. | cs.CL | Differential contributions of machine learning and statistical analysis to language and cognitive sciences |
2024-04-22T09:56:59Z | http://arxiv.org/pdf/2404.14043v1 | http://arxiv.org/abs/2404.14043v1 | Keheng Wang, Feiyu Duan, Peiguang Li, Sirui Wang, Xunliang Cai | 2024-04-22T09:56:59Z | cs.CL | Retrieval-Augmented Generation (RAG) demonstrates great value in alleviating
outdated knowledge or hallucination by supplying LLMs with updated and relevant
knowledge. However, there are still several difficulties for RAG in
understanding complex multi-hop query and retrieving relevant documents, which
require LLMs to perform reasoning and retrieve step by step. Inspired by
human's reasoning process in which they gradually search for the required
information, it is natural to ask whether the LLMs could notice the missing
information in each reasoning step. In this work, we first experimentally
verified the ability of LLMs to extract information as well as to know the
missing. Based on the above discovery, we propose a Missing Information Guided
Retrieve-Extraction-Solving paradigm (MIGRES), where we leverage the
identification of missing information to generate a targeted query that steers
the subsequent knowledge retrieval. Besides, we design a sentence-level
re-ranking filtering approach to filter the irrelevant content out from
document, along with the information extraction capability of LLMs to extract
useful information from cleaned-up documents, which in turn to bolster the
overall efficacy of RAG. Extensive experiments conducted on multiple public
datasets reveal the superiority of the proposed MIGRES method, and analytical
experiments demonstrate the effectiveness of our proposed modules. | cs.CL | LLMs Know What They Need: Leveraging a Missing Information Guided Framework to Empower Retrieval-Augmented Generation |
2024-04-22T09:40:07Z | http://arxiv.org/pdf/2404.14024v1 | http://arxiv.org/abs/2404.14024v1 | Alexandre Bittar, Philip N. Garner | 2024-04-22T09:40:07Z | cs.CL, q-bio.NC | Understanding cognitive processes in the brain demands sophisticated models
capable of replicating neural dynamics at large scales. We present a
physiologically inspired speech recognition architecture, compatible and
scalable with deep learning frameworks, and demonstrate that end-to-end
gradient descent training leads to the emergence of neural oscillations in the
central spiking neural network. Significant cross-frequency couplings,
indicative of these oscillations, are measured within and across network layers
during speech processing, whereas no such interactions are observed when
handling background noise inputs. Furthermore, our findings highlight the
crucial inhibitory role of feedback mechanisms, such as spike frequency
adaptation and recurrent connections, in regulating and synchronising neural
activity to improve recognition performance. Overall, on top of developing our
understanding of synchronisation phenomena notably observed in the human
auditory pathway, our architecture exhibits dynamic and efficient information
processing, with relevance to neuromorphic technology. | cs.CL | Exploring neural oscillations during speech perception via surrogate gradient spiking neural networks |
2024-04-22T09:25:05Z | http://arxiv.org/pdf/2404.14464v1 | http://arxiv.org/abs/2404.14464v1 | Li Jiapeng, Liu Runze, Li Yabo, Zhou Tong, Li Mingling, Chen Xiang | 2024-04-22T09:25:05Z | cs.CL, cs.AI, cs.IR | Multi-hop question answering is a knowledge-intensive complex problem. Large
Language Models (LLMs) use their Chain of Thoughts (CoT) capability to reason
complex problems step by step, and retrieval-augmentation can effectively
alleviate factual errors caused by outdated and unknown knowledge in LLMs.
Recent works have introduced retrieval-augmentation in the CoT reasoning to
solve multi-hop question answering. However, these chain methods have the
following problems: 1) Retrieved irrelevant paragraphs may mislead the
reasoning; 2) An error in the chain structure may lead to a cascade of errors.
In this paper, we propose a dynamic retrieval framework called Tree of
Reviews (ToR), where the root node is the question, and the other nodes are
paragraphs from retrieval, extending different reasoning paths from the root
node to other nodes. Our framework dynamically decides to initiate a new
search, reject, or accept based on the paragraphs on the reasoning paths.
Compared to related work, we introduce a tree structure to handle each
retrieved paragraph separately, alleviating the misleading effect of irrelevant
paragraphs on the reasoning path; the diversity of reasoning path extension
reduces the impact of a single reasoning error on the whole. We conducted
experiments on three different multi-hop question answering datasets. The
results show that compared to the baseline methods, ToR achieves
state-of-the-art performance in both retrieval and response generation. In
addition, we propose two tree-based search optimization strategies, pruning and
effective expansion, to reduce time overhead and increase the diversity of path
extension. We will release our code. | cs.CL | Tree of Reviews: A Tree-based Dynamic Iterative Retrieval Framework for Multi-hop Question Answering |
2024-04-22T09:07:50Z | http://arxiv.org/pdf/2404.14463v1 | http://arxiv.org/abs/2404.14463v1 | Sergio Burdisso, Ernesto Reyes-Ramírez, Esaú Villatoro-Tello, Fernando Sánchez-Vega, Pastor López-Monroy, Petr Motlicek | 2024-04-22T09:07:50Z | cs.CL, cs.AI, cs.CY, cs.LG | Automatic depression detection from conversational data has gained
significant interest in recent years. The DAIC-WOZ dataset, interviews
conducted by a human-controlled virtual agent, has been widely used for this
task. Recent studies have reported enhanced performance when incorporating
interviewer's prompts into the model. In this work, we hypothesize that this
improvement might be mainly due to a bias present in these prompts, rather than
the proposed architectures and methods. Through ablation experiments and
qualitative analysis, we discover that models using interviewer's prompts learn
to focus on a specific region of the interviews, where questions about past
experiences with mental health issues are asked, and use them as discriminative
shortcuts to detect depressed participants. In contrast, models using
participant responses gather evidence from across the entire interview.
Finally, to highlight the magnitude of this bias, we achieve a 0.90 F1 score by
intentionally exploiting it, the highest result reported to date on this
dataset using only textual information. Our findings underline the need for
caution when incorporating interviewers' prompts into models, as they may
inadvertently learn to exploit targeted prompts, rather than learning to
characterize the language and behavior that are genuinely indicative of the
patient's mental health condition. | cs.CL | DAIC-WOZ: On the Validity of Using the Therapist's prompts in Automatic Depression Detection from Clinical Interviews |
2024-04-22T08:47:27Z | http://arxiv.org/pdf/2404.13985v2 | http://arxiv.org/abs/2404.13985v2 | Xiaoxia Cheng, Zeqi Tan, Wei Xue, Weiming Lu | 2024-05-24T07:28:49Z | cs.CL | Improving the reasoning capabilities of large language models (LLMs) has
attracted considerable interest. Recent approaches primarily focus on improving
the reasoning process to yield a more precise final answer. However, in
scenarios involving contextually aware reasoning, these methods neglect the
importance of first identifying logical relationships from the context before
proceeding with the reasoning. This oversight could lead to a superficial
understanding and interaction with the context, potentially undermining the
quality and reliability of the reasoning outcomes. In this paper, we propose an
information re-organization (InfoRE) method before proceeding with the
reasoning to enhance the reasoning ability of LLMs. Our re-organization method
involves initially extracting logical relationships from the contextual
content, such as documents or paragraphs, and subsequently pruning redundant
content to minimize noise. Then, we utilize the re-organized information in the
reasoning process. This enables LLMs to deeply understand the contextual
content by clearly perceiving these logical relationships, while also ensuring
high-quality responses by eliminating potential noise. To demonstrate the
effectiveness of our approach in improving the reasoning ability, we conduct
experiments using Llama2-70B, GPT-3.5, and GPT-4 on various contextually aware
multi-hop reasoning tasks. Using only a zero-shot setting, our method achieves
an average absolute improvement of 4% across all tasks, highlighting its
potential to improve the reasoning performance of LLMs. Our source code is
available at https://github.com/hustcxx/InfoRE. | cs.CL | Information Re-Organization Improves Reasoning in Large Language Models |
2024-04-22T08:28:13Z | http://arxiv.org/pdf/2404.15154v1 | http://arxiv.org/abs/2404.15154v1 | Kyomin Hwang, Suyoung Kim, JunHoo Lee, Nojun Kwak | 2024-04-22T08:28:13Z | cs.CL, cs.AI | Large Models (LMs) have heightened expectations for the potential of general
AI as they are akin to human intelligence. This paper shows that recent large
models such as Stable Diffusion and DALL-E3 also share the vulnerability of
human intelligence, namely the "white bear phenomenon". We investigate the
causes of the white bear phenomenon by analyzing their representation space.
Based on this analysis, we propose a simple prompt-based attack method, which
generates figures prohibited by the LM provider's policy. To counter these
attacks, we introduce prompt-based defense strategies inspired by cognitive
therapy techniques, successfully mitigating attacks by up to 48.22\%. | cs.CL | Do not think pink elephant! |
2024-04-22T08:16:07Z | http://arxiv.org/pdf/2404.13968v2 | http://arxiv.org/abs/2404.13968v2 | Zichuan Liu, Zefan Wang, Linjie Xu, Jinyu Wang, Lei Song, Tianchun Wang, Chunlin Chen, Wei Cheng, Jiang Bian | 2024-05-16T13:26:57Z | cs.CL, cs.AI, cs.CR | The advent of large language models (LLMs) has revolutionized the field of
natural language processing, yet they might be attacked to produce harmful
content. Despite efforts to ethically align LLMs, these are often fragile and
can be circumvented by jailbreaking attacks through optimized or manual
adversarial prompts. To address this, we introduce the Information Bottleneck
Protector (IBProtector), a defense mechanism grounded in the information
bottleneck principle, and we modify the objective to avoid trivial solutions.
The IBProtector selectively compresses and perturbs prompts, facilitated by a
lightweight and trainable extractor, preserving only essential information for
the target LLMs to respond with the expected answer. Moreover, we further
consider a situation where the gradient is not visible to be compatible with
any LLM. Our empirical evaluations show that IBProtector outperforms current
defense methods in mitigating jailbreak attempts, without overly affecting
response quality or inference speed. Its effectiveness and adaptability across
various attack methods and target LLMs underscore the potential of IBProtector
as a novel, transferable defense that bolsters the security of LLMs without
requiring modifications to the underlying models. | cs.CL | Protecting Your LLMs with Information Bottleneck |
2024-04-22T08:00:51Z | http://arxiv.org/pdf/2404.13957v1 | http://arxiv.org/abs/2404.13957v1 | Man Tik Ng, Hui Tung Tse, Jen-tse Huang, Jingjing Li, Wenxuan Wang, Michael R. Lyu | 2024-04-22T08:00:51Z | cs.CL | The role-play ability of Large Language Models (LLMs) has emerged as a
popular research direction. However, existing studies focus on imitating
well-known public figures or fictional characters, overlooking the potential
for simulating ordinary individuals. Such an oversight limits the potential for
advancements in digital human clones and non-player characters in video games.
To bridge this gap, we introduce ECHO, an evaluative framework inspired by the
Turing test. This framework engages the acquaintances of the target individuals
to distinguish between human and machine-generated responses. Notably, our
framework focuses on emulating average individuals rather than historical or
fictional figures, presenting a unique advantage to apply the Turing Test. We
evaluated three role-playing LLMs using ECHO, with GPT-3.5 and GPT-4 serving as
foundational models, alongside the online application GPTs from OpenAI. Our
results demonstrate that GPT-4 more effectively deceives human evaluators, and
GPTs achieves a leading success rate of 48.3%. Furthermore, we investigated
whether LLMs could discern between human-generated and machine-generated texts.
While GPT-4 can identify differences, it could not determine which texts were
human-produced. Our code and results of reproducing the role-playing LLMs are
made publicly available via https://github.com/CUHK-ARISE/ECHO. | cs.CL | How Well Can LLMs Echo Us? Evaluating AI Chatbots' Role-Play Ability with ECHO |
2024-04-22T07:49:36Z | http://arxiv.org/pdf/2404.13948v1 | http://arxiv.org/abs/2404.13948v1 | Sukmin Cho, Soyeong Jeong, Jeongyeon Seo, Taeho Hwang, Jong C. Park | 2024-04-22T07:49:36Z | cs.CL | The robustness of recent Large Language Models (LLMs) has become increasingly
crucial as their applicability expands across various domains and real-world
applications. Retrieval-Augmented Generation (RAG) is a promising solution for
addressing the limitations of LLMs, yet existing studies on the robustness of
RAG often overlook the interconnected relationships between RAG components or
the potential threats prevalent in real-world databases, such as minor textual
errors. In this work, we investigate two underexplored aspects when assessing
the robustness of RAG: 1) vulnerability to noisy documents through low-level
perturbations and 2) a holistic evaluation of RAG robustness. Furthermore, we
introduce a novel attack method, the Genetic Attack on RAG (\textit{GARAG}),
which targets these aspects. Specifically, GARAG is designed to reveal
vulnerabilities within each component and test the overall system functionality
against noisy documents. We validate RAG robustness by applying our
\textit{GARAG} to standard QA datasets, incorporating diverse retrievers and
LLMs. The experimental results show that GARAG consistently achieves high
attack success rates. Also, it significantly devastates the performance of each
component and their synergy, highlighting the substantial risk that minor
textual inaccuracies pose in disrupting RAG systems in the real world. | cs.CL | Typos that Broke the RAG's Back: Genetic Attack on RAG Pipeline by Simulating Documents in the Wild via Low-level Perturbations |
2024-04-22T07:32:03Z | http://arxiv.org/pdf/2404.13940v2 | http://arxiv.org/abs/2404.13940v2 | Jiayin Wang, Fengran Mo, Weizhi Ma, Peijie Sun, Min Zhang, Jian-Yun Nie | 2024-04-23T01:58:24Z | cs.CL | Large Language Models (LLMs) are essential tools to collaborate with users on
different tasks. Evaluating their performance to serve users' needs in
real-world scenarios is important. While many benchmarks have been created,
they mainly focus on specific predefined model abilities. Few have covered the
intended utilization of LLMs by real users. To address this oversight, we
propose benchmarking LLMs from a user perspective in both dataset construction
and evaluation designs. We first collect 1846 real-world use cases with 15 LLMs
from a user study with 712 participants from 23 countries. These self-reported
cases form the User Reported Scenarios(URS) dataset with a categorization of 7
user intents. Secondly, on this authentic multi-cultural dataset, we benchmark
10 LLM services on their efficacy in satisfying user needs. Thirdly, we show
that our benchmark scores align well with user-reported experience in LLM
interactions across diverse intents, both of which emphasize the overlook of
subjective scenarios. In conclusion, our study proposes to benchmark LLMs from
a user-centric perspective, aiming to facilitate evaluations that better
reflect real user needs. The benchmark dataset and code are available at
https://github.com/Alice1998/URS. | cs.CL | A User-Centric Benchmark for Evaluating Large Language Models |
2024-04-22T07:21:27Z | http://arxiv.org/pdf/2404.16068v1 | http://arxiv.org/abs/2404.16068v1 | Yifan Jiang, Filip Ilievski, Kaixin Ma | 2024-04-22T07:21:27Z | cs.AI, cs.CL, cs.LG | While vertical thinking relies on logical and commonsense reasoning, lateral
thinking requires systems to defy commonsense associations and overwrite them
through unconventional thinking. Lateral thinking has been shown to be
challenging for current models but has received little attention. A recent
benchmark, BRAINTEASER, aims to evaluate current models' lateral thinking
ability in a zero-shot setting. In this paper, we split the original benchmark
to also support fine-tuning setting and present SemEval Task 9:
BRAIN-TEASER(S), the first task at this competition designed to test the
system's reasoning and lateral thinking ability. As a popular task,
BRAINTEASER(S)'s two subtasks receive 483 team submissions from 182
participants during the competition. This paper provides a fine-grained system
analysis of the competition results, together with a reflection on what this
means for the ability of the systems to reason laterally. We hope that the
BRAINTEASER(S) subtasks and findings in this paper can stimulate future work on
lateral thinking and robust reasoning by computational models. | cs.AI | SemEval-2024 Task 9: BRAINTEASER: A Novel Task Defying Common Sense |
2024-04-22T07:03:44Z | http://arxiv.org/pdf/2404.13925v1 | http://arxiv.org/abs/2404.13925v1 | Boning Zhang, Chengxi Li, Kai Fan | 2024-04-22T07:03:44Z | cs.CL | Large language models (LLMs) have been explored in a variety of reasoning
tasks including solving of mathematical problems. Each math dataset typically
includes its own specially designed evaluation script, which, while suitable
for its intended use, lacks generalizability across different datasets.
Consequently, updates and adaptations to these evaluation tools tend to occur
without being systematically reported, leading to inconsistencies and obstacles
to fair comparison across studies. To bridge this gap, we introduce a
comprehensive mathematical evaluation toolkit that not only utilizes a python
computer algebra system (CAS) for its numerical accuracy, but also integrates
an optional LLM, known for its considerable natural language processing
capabilities. To validate the effectiveness of our toolkit, we manually
annotated two distinct datasets. Our experiments demonstrate that the toolkit
yields more robust evaluation results compared to prior works, even without an
LLM. Furthermore, when an LLM is incorporated, there is a notable enhancement.
The code for our method will be made available at
\url{https://github.com/MARIO-Math-Reasoning/math_evaluation}. | cs.CL | MARIO Eval: Evaluate Your Math LLM with your Math LLM--A mathematical dataset evaluation toolkit |
2024-04-22T06:57:43Z | http://arxiv.org/pdf/2404.13919v1 | http://arxiv.org/abs/2404.13919v1 | Yukyung Lee, Soonwon Ka, Bokyung Son, Pilsung Kang, Jaewook Kang | 2024-04-22T06:57:43Z | cs.CL, cs.AI, cs.HC | Large Language Models (LLMs) have significantly impacted the writing process,
enabling collaborative content creation and enhancing productivity. However,
generating high-quality, user-aligned text remains challenging. In this paper,
we propose Writing Path, a framework that uses explicit outlines to guide LLMs
in generating goal-oriented, high-quality pieces of writing. Our approach draws
inspiration from structured writing planning and reasoning paths, focusing on
capturing and reflecting user intentions throughout the writing process. We
construct a diverse dataset from unstructured blog posts to benchmark writing
performance and introduce a comprehensive evaluation framework assessing the
quality of outlines and generated texts. Our evaluations with GPT-3.5-turbo,
GPT-4, and HyperCLOVA X demonstrate that the Writing Path approach
significantly enhances text quality according to both LLMs and human
evaluations. This study highlights the potential of integrating
writing-specific techniques into LLMs to enhance their ability to meet the
diverse writing needs of users. | cs.CL | Navigating the Path of Writing: Outline-guided Text Generation with Large Language Models |
2024-04-22T06:33:28Z | http://arxiv.org/pdf/2404.13906v2 | http://arxiv.org/abs/2404.13906v2 | Yu-Xiang Lin, Wei-Yun Ma | 2024-05-14T08:05:00Z | cs.CL, cs.AI | The goal of product copywriting is to capture the interest of potential
buyers by emphasizing the features of products through text descriptions. As
e-commerce platforms offer a wide range of services, it's becoming essential to
dynamically adjust the styles of these auto-generated descriptions. Typical
approaches to copywriting generation often rely solely on specified product
attributes, which may result in dull and repetitive content. To tackle this
issue, we propose to generate copywriting based on customer reviews, as they
provide firsthand practical experiences with products, offering a richer source
of information than just product attributes. We have developed a
sequence-to-sequence framework, enhanced with reinforcement learning, to
produce copywriting that is attractive, authentic, and rich in information. Our
framework outperforms all existing baseline and zero-shot large language
models, including LLaMA-2-chat-7B and GPT-3.5, in terms of both attractiveness
and faithfulness. Furthermore, this work features the use of LLMs for
aspect-based summaries collection and argument allure assessment. Experiments
demonstrate the effectiveness of using LLMs for marketing domain corpus
construction. The code and the dataset is publicly available at:
https://github.com/YuXiangLin1234/Copywriting-Generation. | cs.CL | Generating Attractive and Authentic Copywriting from Customer Reviews |
2024-04-22T06:30:05Z | http://arxiv.org/pdf/2404.15155v1 | http://arxiv.org/abs/2404.15155v1 | Yubin Kim, Chanwoo Park, Hyewon Jeong, Yik Siu Chan, Xuhai Xu, Daniel McDuff, Cynthia Breazeal, Hae Won Park | 2024-04-22T06:30:05Z | cs.CL, cs.AI, cs.LG | Foundation models have become invaluable in advancing the medical field.
Despite their promise, the strategic deployment of LLMs for effective utility
in complex medical tasks remains an open question. Our novel framework, Medical
Decision-making Agents (MDAgents) aims to address this gap by automatically
assigning the effective collaboration structure for LLMs. Assigned solo or
group collaboration structure is tailored to the complexity of the medical task
at hand, emulating real-world medical decision making processes. We evaluate
our framework and baseline methods with state-of-the-art LLMs across a suite of
challenging medical benchmarks: MedQA, MedMCQA, PubMedQA, DDXPlus, PMC-VQA,
Path-VQA, and MedVidQA, achieving the best performance in 5 out of 7 benchmarks
that require an understanding of multi-modal medical reasoning. Ablation
studies reveal that MDAgents excels in adapting the number of collaborating
agents to optimize efficiency and accuracy, showcasing its robustness in
diverse scenarios. We also explore the dynamics of group consensus, offering
insights into how collaborative agents could behave in complex clinical team
dynamics. Our code can be found at https://github.com/mitmedialab/MDAgents. | cs.CL | Adaptive Collaboration Strategy for LLMs in Medical Decision Making |
2024-04-22T06:18:37Z | http://arxiv.org/pdf/2404.13899v1 | http://arxiv.org/abs/2404.13899v1 | Yihang Wu, Xiao Cao, Kaixin Li, Zitan Chen, Haonan Wang, Lei Meng, Zhiyong Huang | 2024-04-22T06:18:37Z | cs.CL, cs.AI, cs.MM | In text-to-image generation tasks, the advancements of diffusion models have
facilitated the fidelity of generated results. However, these models encounter
challenges when processing text prompts containing multiple entities and
attributes. The uneven distribution of attention results in the issues of
entity leakage and attribute misalignment. Training from scratch to address
this issue requires numerous labeled data and is resource-consuming. Motivated
by this, we propose an attribution-focusing mechanism, a training-free
phase-wise mechanism by modulation of attention for diffusion model. One of our
core ideas is to guide the model to concentrate on the corresponding syntactic
components of the prompt at distinct timesteps. To achieve this, we incorporate
a temperature control mechanism within the early phases of the self-attention
modules to mitigate entity leakage issues. An object-focused masking scheme and
a phase-wise dynamic weight control mechanism are integrated into the
cross-attention modules, enabling the model to discern the affiliation of
semantic information between entities more effectively. The experimental
results in various alignment scenarios demonstrate that our model attain better
image-text alignment with minimal additional computational cost. | cs.CL | Towards Better Text-to-Image Generation Alignment via Attention Modulation |
2024-04-22T05:12:52Z | http://arxiv.org/pdf/2404.13885v1 | http://arxiv.org/abs/2404.13885v1 | Qingyang Wu, Ying Xu, Tingsong Xiao, Yunze Xiao, Yitong Li, Tianyang Wang, Yichi Zhang, Shanghai Zhong, Yuwei Zhang, Wei Lu, Yifan Yang | 2024-04-22T05:12:52Z | cs.CY, cs.AI, cs.CL, cs.LG | Large Language Models (LLMs) have emerged as potent tools for advancing the
United Nations' Sustainable Development Goals (SDGs). However, the attitudinal
disparities between LLMs and humans towards these goals can pose significant
challenges. This study conducts a comprehensive review and analysis of the
existing literature on the attitudes of LLMs towards the 17 SDGs, emphasizing
the comparison between their attitudes and support for each goal and those of
humans. We examine the potential disparities, primarily focusing on aspects
such as understanding and emotions, cultural and regional differences, task
objective variations, and factors considered in the decision-making process.
These disparities arise from the underrepresentation and imbalance in LLM
training data, historical biases, quality issues, lack of contextual
understanding, and skewed ethical values reflected. The study also investigates
the risks and harms that may arise from neglecting the attitudes of LLMs
towards the SDGs, including the exacerbation of social inequalities, racial
discrimination, environmental destruction, and resource wastage. To address
these challenges, we propose strategies and recommendations to guide and
regulate the application of LLMs, ensuring their alignment with the principles
and goals of the SDGs, and therefore creating a more just, inclusive, and
sustainable future. | cs.CY | Surveying Attitudinal Alignment Between Large Language Models Vs. Humans Towards 17 Sustainable Development Goals |
2024-04-22T05:08:53Z | http://arxiv.org/pdf/2404.14461v2 | http://arxiv.org/abs/2404.14461v2 | Javier Rando, Francesco Croce, Kryštof Mitka, Stepan Shabalin, Maksym Andriushchenko, Nicolas Flammarion, Florian Tramèr | 2024-06-06T12:45:52Z | cs.CL, cs.AI, cs.CR, cs.LG | Large language models are aligned to be safe, preventing users from
generating harmful content like misinformation or instructions for illegal
activities. However, previous work has shown that the alignment process is
vulnerable to poisoning attacks. Adversaries can manipulate the safety training
data to inject backdoors that act like a universal sudo command: adding the
backdoor string to any prompt enables harmful responses from models that,
otherwise, behave safely. Our competition, co-located at IEEE SaTML 2024,
challenged participants to find universal backdoors in several large language
models. This report summarizes the key findings and promising ideas for future
research. | cs.CL | Competition Report: Finding Universal Jailbreak Backdoors in Aligned LLMs |
2024-04-22T04:49:22Z | http://arxiv.org/pdf/2404.13874v2 | http://arxiv.org/abs/2404.13874v2 | Haoyi Qiu, Wenbo Hu, Zi-Yi Dou, Nanyun Peng | 2024-06-06T02:53:37Z | cs.CL, cs.CV | Large Vision-Language Models (LVLMs) suffer from hallucination issues,
wherein the models generate plausible-sounding but factually incorrect outputs,
undermining their reliability. A comprehensive quantitative evaluation is
necessary to identify and understand the extent of hallucinations in these
models. However, existing benchmarks are often limited in scope, focusing
mainly on object hallucinations. Furthermore, current evaluation methods
struggle to effectively address the subtle semantic distinctions between model
outputs and reference data, as well as the balance between hallucination and
informativeness. To address these issues, we introduce a multi-dimensional
benchmark covering objects, attributes, and relations, with challenging images
selected based on associative biases. Moreover, we propose a large language
model (LLM)-based two-stage evaluation framework that generalizes the popular
CHAIR metric and incorporates both faithfulness and coverage into the
evaluation. Experiments on 10 established LVLMs demonstrate that our evaluation
metric is more comprehensive and better correlated with humans than existing
work when evaluating on our challenging human-annotated benchmark dataset. Our
work also highlights the critical balance between faithfulness and coverage of
model outputs, and encourages future works to address hallucinations in LVLMs
while keeping their outputs informative. | cs.CL | VALOR-EVAL: Holistic Coverage and Faithfulness Evaluation of Large Vision-Language Models |
2024-04-22T04:30:36Z | http://arxiv.org/pdf/2404.13865v1 | http://arxiv.org/abs/2404.13865v1 | Avinash Anand, Kritarth Prasad, Ujjwal Goel, Mohit Gupta, Naman Lal, Astha Verma, Rajiv Ratn Shah | 2024-04-22T04:30:36Z | cs.CL | Citation text plays a pivotal role in elucidating the connection between
scientific documents, demanding an in-depth comprehension of the cited paper.
Constructing citations is often time-consuming, requiring researchers to delve
into extensive literature and grapple with articulating relevant content. To
address this challenge, the field of citation text generation (CTG) has
emerged. However, while earlier methods have primarily centered on creating
single-sentence citations, practical scenarios frequently necessitate citing
multiple papers within a single paragraph. To bridge this gap, we propose a
method that leverages Large Language Models (LLMs) to generate multi-citation
sentences. Our approach involves a single source paper and a collection of
target papers, culminating in a coherent paragraph containing multi-sentence
citation text. Furthermore, we introduce a curated dataset named MCG-S2ORC,
composed of English-language academic research papers in Computer Science,
showcasing multiple citation instances. In our experiments, we evaluate three
LLMs LLaMA, Alpaca, and Vicuna to ascertain the most effective model for this
endeavor. Additionally, we exhibit enhanced performance by integrating
knowledge graphs from target papers into the prompts for generating citation
text. This research underscores the potential of harnessing LLMs for citation
generation, opening a compelling avenue for exploring the intricate connections
between scientific documents. | cs.CL | Context-Enhanced Language Models for Generating Multi-Paper Citations |
2024-04-22T03:47:00Z | http://arxiv.org/pdf/2404.13855v1 | http://arxiv.org/abs/2404.13855v1 | Sunit Bhattacharya, Ondřej Bojar | 2024-04-22T03:47:00Z | cs.CL | Multilingualism in Large Language Models (LLMs) is an yet under-explored
area. In this paper, we conduct an in-depth analysis of the multilingual
capabilities of a family of a Large Language Model, examining its architecture,
activation patterns, and processing mechanisms across languages. We introduce
novel metrics to probe the model's multilingual behaviour at different layers
and shed light on the impact of architectural choices on multilingual
processing.
Our findings reveal different patterns of multilinugal processing in the
sublayers of Feed-Forward Networks of the models. Furthermore, we uncover the
phenomenon of "over-layerization" in certain model configurations, where
increasing layer depth without corresponding adjustments to other parameters
may degrade model performance. Through comparisons within and across languages,
we demonstrate the interplay between model architecture, layer depth, and
multilingual processing capabilities of LLMs trained on multiple languages. | cs.CL | Understanding the role of FFNs in driving multilingual behaviour in LLMs |
2024-04-22T03:05:32Z | http://arxiv.org/pdf/2404.13847v1 | http://arxiv.org/abs/2404.13847v1 | Mingjie Ma, Zhihuan Yu, Yichao Ma, Guohui Li | 2024-04-22T03:05:32Z | cs.CV, cs.CL | Visual Commonsense Reasoning (VCR) is a cognitive task, challenging models to
answer visual questions requiring human commonsense, and to provide rationales
explaining why the answers are correct. With emergence of Large Language Models
(LLMs), it is natural and imperative to explore their applicability to VCR.
However, VCR task demands more external knowledge to tackle its challenging
questions, necessitating special designs to activate LLMs' commonsense
reasoning abilities. Also, most existing Multimodal LLMs adopted an abstraction
of entire input image, which makes it difficult to comprehend VCR's unique
co-reference tags between image regions and text, posing challenges for
fine-grained alignment. To address these issues, we propose EventLens that
leverages Event-Aware Pretraining and Cross-modal Linking and EnhanceS VCR.
First, by emulating the cognitive process of human reasoning, an Event-Aware
Pretraining auxiliary task is introduced to better activate LLM's global
comprehension of intricate scenarios. Second, during fine-tuning, we further
utilize reference tags to bridge RoI features with texts, while preserving both
modality semantics. Finally, we use instruct-style prompts to narrow the gap
between pretraining and fine-tuning, and task-specific adapters to better
integrate LLM's inherent knowledge with new commonsense. Experimental results
show the effectiveness of our proposed auxiliary task and fine-grained linking
strategy. | cs.CV | EventLens: Leveraging Event-Aware Pretraining and Cross-modal Linking Enhances Visual Commonsense Reasoning |
2024-04-22T03:05:19Z | http://arxiv.org/pdf/2404.13846v2 | http://arxiv.org/abs/2404.13846v2 | Tetsuro Morimura, Mitsuki Sakamoto, Yuu Jinnai, Kenshi Abe, Kaito Ariu | 2024-04-23T11:56:45Z | cs.LG, cs.AI, cs.CL | Reinforcement learning from human feedback (RLHF) plays a crucial role in
aligning language models with human preferences. While the significance of
dataset quality is generally recognized, explicit investigations into its
impact within the RLHF framework, to our knowledge, have been limited. This
paper addresses the issue of text quality within the preference dataset by
focusing on Direct Preference Optimization (DPO), an increasingly adopted
reward-model-free RLHF method. We confirm that text quality significantly
influences the performance of models optimized with DPO more than those
optimized with reward-model-based RLHF. Building on this new insight, we
propose an extension of DPO, termed filtered direct preference optimization
(fDPO). fDPO uses a trained reward model to monitor the quality of texts within
the preference dataset during DPO training. Samples of lower quality are
discarded based on comparisons with texts generated by the model being
optimized, resulting in a more accurate dataset. Experimental results
demonstrate that fDPO enhances the final model performance. Our code is
available at https://github.com/CyberAgentAILab/filtered-dpo. | cs.LG | Filtered Direct Preference Optimization |
2024-04-22T02:15:52Z | http://arxiv.org/pdf/2404.15159v2 | http://arxiv.org/abs/2404.15159v2 | Dengchun Li, Yingzi Ma, Naizheng Wang, Zhengmao Ye, Zhiyuan Cheng, Yinghao Tang, Yan Zhang, Lei Duan, Jie Zuo, Cal Yang, Mingjie Tang | 2024-05-23T05:16:33Z | cs.CL, cs.AI | Fine-tuning Large Language Models (LLMs) is a common practice to adapt
pre-trained models for specific applications. While methods like LoRA have
effectively addressed GPU memory constraints during fine-tuning, their
performance often falls short, especially in multi-task scenarios. In contrast,
Mixture-of-Expert (MoE) models, such as Mixtral 8x7B, demonstrate remarkable
performance in multi-task learning scenarios while maintaining a reduced
parameter count. However, the resource requirements of these MoEs remain
challenging, particularly for consumer-grade GPUs with less than 24GB memory.
To tackle these challenges, we propose MixLoRA, an approach to construct a
resource-efficient sparse MoE model based on LoRA. MixLoRA inserts multiple
LoRA-based experts within the feed-forward network block of a frozen
pre-trained dense model and employs a commonly used top-k router. Unlike other
LoRA-based MoE methods, MixLoRA enhances model performance by utilizing
independent attention-layer LoRA adapters. Additionally, an auxiliary load
balance loss is employed to address the imbalance problem of the router. Our
evaluations show that MixLoRA improves about 9% accuracy compared to
state-of-the-art PEFT methods in multi-task learning scenarios. We also propose
a new high-throughput framework to alleviate the computation and memory
bottlenecks during the training and inference of MOE models. This framework
reduces GPU memory consumption by 40% and token computation latency by 30%
during both training and inference. | cs.CL | MixLoRA: Enhancing Large Language Models Fine-Tuning with LoRA-based Mixture of Experts |
2024-04-22T01:22:23Z | http://arxiv.org/pdf/2404.13813v1 | http://arxiv.org/abs/2404.13813v1 | Maxim Enis, Mark Hopkins | 2024-04-22T01:22:23Z | cs.CL, cs.AI | We show that Claude 3 Opus, a large language model (LLM) released by
Anthropic in March 2024, exhibits stronger machine translation competence than
other LLMs. Though we find evidence of data contamination with Claude on
FLORES-200, we curate new benchmarks that corroborate the effectiveness of
Claude for low-resource machine translation into English. We find that Claude
has remarkable \textit{resource efficiency} -- the degree to which the quality
of the translation model depends on a language pair's resource level. Finally,
we show that advancements in LLM translation can be compressed into traditional
neural machine translation (NMT) models. Using Claude to generate synthetic
data, we demonstrate that knowledge distillation advances the state-of-the-art
in Yoruba-English translation, meeting or surpassing strong baselines like
NLLB-54B and Google Translate. | cs.CL | From LLM to NMT: Advancing Low-Resource Machine Translation with Claude |
2024-04-22T00:07:55Z | http://arxiv.org/pdf/2404.15157v1 | http://arxiv.org/abs/2404.15157v1 | Si Chen, Feiyang Kang, Ning Yu, Ruoxi Jia | 2024-04-22T00:07:55Z | cs.CL, cs.AI | Fact tracing seeks to identify specific training examples that serve as the
knowledge source for a given query. Existing approaches to fact tracing rely on
assessing the similarity between each training sample and the query along a
certain dimension, such as lexical similarity, gradient, or embedding space.
However, these methods fall short of effectively distinguishing between samples
that are merely relevant and those that actually provide supportive evidence
for the information sought by the query. This limitation often results in
suboptimal effectiveness. Moreover, these approaches necessitate the
examination of the similarity of individual training points for each query,
imposing significant computational demands and creating a substantial barrier
for practical applications. This paper introduces FASTTRACK, a novel approach
that harnesses the capabilities of Large Language Models (LLMs) to validate
supportive evidence for queries and at the same time clusters the training
database towards a reduced extent for LLMs to trace facts. Our experiments show
that FASTTRACK substantially outperforms existing methods in both accuracy and
efficiency, achieving more than 100\% improvement in F1 score over the
state-of-the-art methods while being X33 faster than \texttt{TracIn}. | cs.CL | FASTTRACK: Fast and Accurate Fact Tracing for LLMs |
2024-04-21T23:14:02Z | http://arxiv.org/pdf/2404.13793v1 | http://arxiv.org/abs/2404.13793v1 | Mustafa Erolcan Er, Murathan Kurfalı, Deniz Zeyrek | 2024-04-21T23:14:02Z | cs.CL, I.2.7 | In this work, we introduce a lightweight discourse connective detection
system. Employing gradient boosting trained on straightforward, low-complexity
features, this proposed approach sidesteps the computational demands of the
current approaches that rely on deep neural networks. Considering its
simplicity, our approach achieves competitive results while offering
significant gains in terms of time even on CPU. Furthermore, the stable
performance across two unrelated languages suggests the robustness of our
system in the multilingual scenario. The model is designed to support the
annotation of discourse relations, particularly in scenarios with limited
resources, while minimizing performance loss. | cs.CL | Lightweight Connective Detection Using Gradient Boosting |
2024-04-21T23:03:47Z | http://arxiv.org/pdf/2404.13792v1 | http://arxiv.org/abs/2404.13792v1 | Donghuo Zeng, Roberto S. Legaspi, Yuewen Sun, Xinshuai Dong, Kazushi Ikeda, Peter Spirtes, kun Zhang | 2024-04-21T23:03:47Z | cs.MM, cs.AI, cs.CL, cs.HC | Customizing persuasive conversations related to the outcome of interest for
specific users achieves better persuasion results. However, existing persuasive
conversation systems rely on persuasive strategies and encounter challenges in
dynamically adjusting dialogues to suit the evolving states of individual users
during interactions. This limitation restricts the system's ability to deliver
flexible or dynamic conversations and achieve suboptimal persuasion outcomes.
In this paper, we present a novel approach that tracks a user's latent
personality dimensions (LPDs) during ongoing persuasion conversation and
generates tailored counterfactual utterances based on these LPDs to optimize
the overall persuasion outcome. In particular, our proposed method leverages a
Bi-directional Generative Adversarial Network (BiCoGAN) in tandem with a
Dialogue-based Personality Prediction Regression (DPPR) model to generate
counterfactual data. This enables the system to formulate alternative
persuasive utterances that are more suited to the user. Subsequently, we
utilize the D3QN model to learn policies for optimized selection of system
utterances on counterfactual data. Experimental results we obtained from using
the PersuasionForGood dataset demonstrate the superiority of our approach over
the existing method, BiCoGAN. The cumulative rewards and Q-values produced by
our method surpass ground truth benchmarks, showcasing the efficacy of
employing counterfactual reasoning and LPDs to optimize reinforcement learning
policy in online interactions. | cs.MM | Counterfactual Reasoning Using Predicted Latent Personality Dimensions for Optimizing Persuasion Outcome |
2024-04-21T22:18:13Z | http://arxiv.org/pdf/2404.16873v1 | http://arxiv.org/abs/2404.16873v1 | Anselm Paulus, Arman Zharmagambetov, Chuan Guo, Brandon Amos, Yuandong Tian | 2024-04-21T22:18:13Z | cs.CR, cs.AI, cs.CL, cs.LG | While recently Large Language Models (LLMs) have achieved remarkable
successes, they are vulnerable to certain jailbreaking attacks that lead to
generation of inappropriate or harmful content. Manual red-teaming requires
finding adversarial prompts that cause such jailbreaking, e.g. by appending a
suffix to a given instruction, which is inefficient and time-consuming. On the
other hand, automatic adversarial prompt generation often leads to semantically
meaningless attacks that can easily be detected by perplexity-based filters,
may require gradient information from the TargetLLM, or do not scale well due
to time-consuming discrete optimization processes over the token space. In this
paper, we present a novel method that uses another LLM, called the AdvPrompter,
to generate human-readable adversarial prompts in seconds, $\sim800\times$
faster than existing optimization-based approaches. We train the AdvPrompter
using a novel algorithm that does not require access to the gradients of the
TargetLLM. This process alternates between two steps: (1) generating
high-quality target adversarial suffixes by optimizing the AdvPrompter
predictions, and (2) low-rank fine-tuning of the AdvPrompter with the generated
adversarial suffixes. The trained AdvPrompter generates suffixes that veil the
input instruction without changing its meaning, such that the TargetLLM is
lured to give a harmful response. Experimental results on popular open source
TargetLLMs show state-of-the-art results on the AdvBench dataset, that also
transfer to closed-source black-box LLM APIs. Further, we demonstrate that by
fine-tuning on a synthetic dataset generated by AdvPrompter, LLMs can be made
more robust against jailbreaking attacks while maintaining performance, i.e.
high MMLU scores. | cs.CR | AdvPrompter: Fast Adaptive Adversarial Prompting for LLMs |
2024-04-21T21:30:17Z | http://arxiv.org/pdf/2404.13784v1 | http://arxiv.org/abs/2404.13784v1 | Ali Naseh, Katherine Thai, Mohit Iyyer, Amir Houmansadr | 2024-04-21T21:30:17Z | cs.CR, cs.CL, cs.CV | With the digital imagery landscape rapidly evolving, image stocks and
AI-generated image marketplaces have become central to visual media.
Traditional stock images now exist alongside innovative platforms that trade in
prompts for AI-generated visuals, driven by sophisticated APIs like DALL-E 3
and Midjourney. This paper studies the possibility of employing multi-modal
models with enhanced visual understanding to mimic the outputs of these
platforms, introducing an original attack strategy. Our method leverages
fine-tuned CLIP models, a multi-label classifier, and the descriptive
capabilities of GPT-4V to create prompts that generate images similar to those
available in marketplaces and from premium stock image providers, yet at a
markedly lower expense. In presenting this strategy, we aim to spotlight a new
class of economic and security considerations within the realm of digital
imagery. Our findings, supported by both automated metrics and human
assessment, reveal that comparable visual content can be produced for a
fraction of the prevailing market prices ($0.23 - $0.27 per image), emphasizing
the need for awareness and strategic discussions about the integrity of digital
media in an increasingly AI-integrated landscape. Our work also contributes to
the field by assembling a dataset consisting of approximately 19 million
prompt-image pairs generated by the popular Midjourney platform, which we plan
to release publicly. | cs.CR | Iteratively Prompting Multimodal LLMs to Reproduce Natural and AI-Generated Images |
2024-04-21T21:22:28Z | http://arxiv.org/pdf/2404.13781v1 | http://arxiv.org/abs/2404.13781v1 | Alireza Salemi, Hamed Zamani | 2024-04-21T21:22:28Z | cs.CL, cs.IR | Evaluating retrieval-augmented generation (RAG) presents challenges,
particularly for retrieval models within these systems. Traditional end-to-end
evaluation methods are computationally expensive. Furthermore, evaluation of
the retrieval model's performance based on query-document relevance labels
shows a small correlation with the RAG system's downstream performance. We
propose a novel evaluation approach, eRAG, where each document in the retrieval
list is individually utilized by the large language model within the RAG
system. The output generated for each document is then evaluated based on the
downstream task ground truth labels. In this manner, the downstream performance
for each document serves as its relevance label. We employ various downstream
task metrics to obtain document-level annotations and aggregate them using
set-based or ranking metrics. Extensive experiments on a wide range of datasets
demonstrate that eRAG achieves a higher correlation with downstream RAG
performance compared to baseline methods, with improvements in Kendall's $\tau$
correlation ranging from 0.168 to 0.494. Additionally, eRAG offers significant
computational advantages, improving runtime and consuming up to 50 times less
GPU memory than end-to-end evaluation. | cs.CL | Evaluating Retrieval Quality in Retrieval-Augmented Generation |
2024-04-21T21:19:36Z | http://arxiv.org/pdf/2404.13779v1 | http://arxiv.org/abs/2404.13779v1 | Ziqing Guo | 2024-04-21T21:19:36Z | cs.CL, cs.LG | Biomedical literature is a rapidly expanding field of science and technology.
Classification of biomedical texts is an essential part of biomedicine
research, especially in the field of biology. This work proposes the fine-tuned
DistilBERT, a methodology-specific, pre-trained generative classification
language model for mining biomedicine texts. The model has proven its
effectiveness in linguistic understanding capabilities and has reduced the size
of BERT models by 40\% but by 60\% faster. The main objective of this project
is to improve the model and assess the performance of the model compared to the
non-fine-tuned model. We used DistilBert as a support model and pre-trained on
a corpus of 32,000 abstracts and complete text articles; our results were
impressive and surpassed those of traditional literature classification methods
by using RNN or LSTM. Our aim is to integrate this highly specialised and
specific model into different research industries. | cs.CL | Automated Text Mining of Experimental Methodologies from Biomedical Literature |
2024-04-21T20:21:24Z | http://arxiv.org/pdf/2404.13764v1 | http://arxiv.org/abs/2404.13764v1 | Li Siyan, Teresa Shao, Zhou Yu, Julia Hirschberg | 2024-04-21T20:21:24Z | cs.CL | Existing English-teaching chatbots rarely incorporate empathy explicitly in
their feedback, but empathetic feedback could help keep students engaged and
reduce learner anxiety. Toward this end, we propose the task of negative
emotion detection via audio, for recognizing empathetic feedback opportunities
in language learning. We then build the first spoken English-teaching chatbot
with adaptive, empathetic feedback. This feedback is synthesized through
automatic prompt optimization of ChatGPT and is evaluated with English
learners. We demonstrate the effectiveness of our system through a preliminary
user study. | cs.CL | Using Adaptive Empathetic Responses for Teaching English |
2024-04-21T20:16:35Z | http://arxiv.org/pdf/2404.13760v1 | http://arxiv.org/abs/2404.13760v1 | Elisa Bassignana, Viggo Unmack Gascou, Frida Nøhr Laustsen, Gustav Kristensen, Marie Haahr Petersen, Rob van der Goot, Barbara Plank | 2024-04-21T20:16:35Z | cs.CL | Current language models require a lot of training data to obtain high
performance. For Relation Classification (RC), many datasets are
domain-specific, so combining datasets to obtain better performance is
non-trivial. We explore a multi-domain training setup for RC, and attempt to
improve performance by encoding domain information. Our proposed models improve
> 2 Macro-F1 against the baseline setup, and our analysis reveals that not all
the labels benefit the same: The classes which occupy a similar space across
domains (i.e., their interpretation is close across them, for example
"physical") benefit the least, while domain-dependent relations (e.g.,
"part-of'') improve the most when encoding domain information. | cs.CL | How to Encode Domain Information in Relation Classification |
2024-04-21T19:24:15Z | http://arxiv.org/pdf/2404.13752v2 | http://arxiv.org/abs/2404.13752v2 | Yihao Zhang, Zeming Wei, Jun Sun, Meng Sun | 2024-05-23T13:06:59Z | cs.LG, cs.AI, cs.CL, cs.CR, math.OC | Since the development of Large Language Models (LLMs) has achieved remarkable
success, understanding and controlling their internal complex mechanisms has
become an urgent problem. Recent research has attempted to interpret their
behaviors through the lens of inner representation. However, developing
practical and efficient methods for applying these representations for general
and flexible model editing remains challenging. In this work, we explore how to
use representation engineering methods to guide the editing of LLMs by
deploying a representation sensor as an oracle. We first identify the
importance of a robust and reliable sensor during editing, then propose an
Adversarial Representation Engineering (ARE) framework to provide a unified and
interpretable approach for conceptual model editing without compromising
baseline performance. Experiments on multiple model editing paradigms
demonstrate the effectiveness of ARE in various settings. Code and data are
available at
https://github.com/Zhang-Yihao/Adversarial-Representation-Engineering. | cs.LG | Towards General Conceptual Model Editing via Adversarial Representation Engineering |
2024-04-21T19:20:42Z | http://arxiv.org/pdf/2404.13751v1 | http://arxiv.org/abs/2404.13751v1 | Kevin Scaria, Abyn Scaria, Ben Scaria | 2024-04-21T19:20:42Z | cs.CL | Aspect Based Sentiment Analysis (ABSA) tasks involve the extraction of
fine-grained sentiment tuples from sentences, aiming to discern the author's
opinions. Conventional methodologies predominantly rely on supervised
approaches; however, the efficacy of such methods diminishes in low-resource
domains lacking labeled datasets since they often lack the ability to
generalize across domains. To address this challenge, we propose a simple and
novel unsupervised approach to extract opinion terms and the corresponding
sentiment polarity for aspect terms in a sentence. Our experimental
evaluations, conducted on four benchmark datasets, demonstrate compelling
performance to extract the aspect oriented opinion words as well as assigning
sentiment polarity. Additionally, unsupervised approaches for opinion word
mining have not been explored and our work establishes a benchmark for the
same. | cs.CL | Embarrassingly Simple Unsupervised Aspect Based Sentiment Tuple Extraction |
2024-04-21T17:20:19Z | http://arxiv.org/pdf/2404.13721v1 | http://arxiv.org/abs/2404.13721v1 | Arnulf Hagen | 2024-04-21T17:20:19Z | cs.AI, cs.CL | The thesis develops a view of design in a concept formation framework and
outlines a language to describe both the object of the design and the process
of designing. The unknown object at the outset of the design work may be seen
as an unknown concept that the designer is to define. Throughout the process,
she develops a description of this object by relating it to known concepts. The
search stops when the designer is satisfied that the design specification is
complete enough to satisfy the requirements from it once built. It is then a
collection of propositions that all contribute towards defining the design
object - a collection of sentences describing relationships between the object
and known concepts. Also, the design process itself may be described by
relating known concepts - by organizing known abilities into particular
patterns of activation, or mobilization. In view of the demands posed to a
language to use in this concept formation process, the framework of a Design
Process Language (DPL) is developed. The basis for the language are linguistic
categories that act as classes of relations used to combine concepts,
containing relations used for describing process and object within the same
general system, with some relations being process specific, others being object
specific, and with the bulk being used both for process and object description.
Another outcome is the distinction of modal relations, or relations describing
futurity, possibility, willingness, hypothetical events, and the like. The
design process almost always includes aspects such as these, and it is thus
necessary for a language facilitating design process description to support
such relationships to be constructed. The DPL is argued to be a foundation
whereupon to build a language that can be used for enabling computers to be
more useful - act more intelligently - in the design process. | cs.AI | The Framework of a Design Process Language |
2024-04-21T13:31:16Z | http://arxiv.org/pdf/2404.13660v1 | http://arxiv.org/abs/2404.13660v1 | Narek Maloyan, Ekansh Verma, Bulat Nutfullin, Bislan Ashinov | 2024-04-21T13:31:16Z | cs.CL | Large Language Models (LLMs) have demonstrated remarkable capabilities in
various domains, but their vulnerability to trojan or backdoor attacks poses
significant security risks. This paper explores the challenges and insights
gained from the Trojan Detection Competition 2023 (TDC2023), which focused on
identifying and evaluating trojan attacks on LLMs. We investigate the
difficulty of distinguishing between intended and unintended triggers, as well
as the feasibility of reverse engineering trojans in real-world scenarios. Our
comparative analysis of various trojan detection methods reveals that achieving
high Recall scores is significantly more challenging than obtaining high
Reverse-Engineering Attack Success Rate (REASR) scores. The top-performing
methods in the competition achieved Recall scores around 0.16, comparable to a
simple baseline of randomly sampling sentences from a distribution similar to
the given training prefixes. This finding raises questions about the
detectability and recoverability of trojans inserted into the model, given only
the harmful targets. Despite the inability to fully solve the problem, the
competition has led to interesting observations about the viability of trojan
detection and improved techniques for optimizing LLM input prompts. The
phenomenon of unintended triggers and the difficulty in distinguishing them
from intended triggers highlights the need for further research into the
robustness and interpretability of LLMs. The TDC2023 has provided valuable
insights into the challenges and opportunities associated with trojan detection
in LLMs, laying the groundwork for future research in this area to ensure their
safety and reliability in real-world applications. | cs.CL | Trojan Detection in Large Language Models: Insights from The Trojan Detection Challenge |
2024-04-21T12:41:02Z | http://arxiv.org/pdf/2404.13645v1 | http://arxiv.org/abs/2404.13645v1 | Feiqi Cao, Caren Han, Hyunsuk Chung | 2024-04-21T12:41:02Z | cs.CL | In this work, we propose a novel tree-based explanation technique, PEACH
(Pretrained-embedding Explanation Across Contextual and Hierarchical
Structure), that can explain how text-based documents are classified by using
any pretrained contextual embeddings in a tree-based human-interpretable
manner. Note that PEACH can adopt any contextual embeddings of the PLMs as a
training input for the decision tree. Using the proposed PEACH, we perform a
comprehensive analysis of several contextual embeddings on nine different NLP
text classification benchmarks. This analysis demonstrates the flexibility of
the model by applying several PLM contextual embeddings, its attribute
selections, scaling, and clustering methods. Furthermore, we show the utility
of explanations by visualising the feature selection and important trend of
text classification via human-interpretable word-cloud-based trees, which
clearly identify model mistakes and assist in dataset debugging. Besides
interpretability, PEACH outperforms or is similar to those from pretrained
models. | cs.CL | PEACH: Pretrained-embedding Explanation Across Contextual and Hierarchical Structure |
2024-04-21T12:15:44Z | http://arxiv.org/pdf/2404.13633v1 | http://arxiv.org/abs/2404.13633v1 | Samuel Rhys Cox | 2024-04-21T12:15:44Z | cs.HC, cs.CL | The ubiquity of smartphones has led to an increase in on demand healthcare
being supplied. For example, people can share their illness-related experiences
with others similar to themselves, and healthcare experts can offer advice for
better treatment and care for remediable, terminal and mental illnesses. As
well as this human-to-human communication, there has been an increased use of
human-to-computer digital health messaging, such as chatbots. These can prove
advantageous as they offer synchronous and anonymous feedback without the need
for a human conversational partner. However, there are many subtleties involved
in human conversation that a computer agent may not properly exhibit. For
example, there are various conversational styles, etiquettes, politeness
strategies or empathic responses that need to be chosen appropriately for the
conversation. Encouragingly, computers are social actors (CASA) posits that
people apply the same social norms to computers as they would do to people. On
from this, previous studies have focused on applying conversational strategies
to computer agents to make them embody more favourable human characteristics.
However, if a computer agent fails in this regard it can lead to negative
reactions from users. Therefore, in this dissertation we describe a series of
studies we carried out to lead to more effective human-to-computer digital
health messaging.
In our first study, we use the crowd [...]
Our second study investigates the effect of a health chatbot's conversational
style [...]
In our final study, we investigate the format used by a chatbot when [...]
In summary, we have researched how to create more effective digital health
interventions starting from generating health messages, to choosing an
appropriate formality of messaging, and finally to formatting messages which
reference a user's previous utterances. | cs.HC | Incorporating Different Verbal Cues to Improve Text-Based Computer-Delivered Health Messaging |
2024-04-21T12:06:05Z | http://arxiv.org/pdf/2404.13630v2 | http://arxiv.org/abs/2404.13630v2 | Keqin Li, Armando Zhu, Peng Zhao, Jintong Song, Jiabei Liu | 2024-05-03T13:07:18Z | cs.SE, cs.AI, cs.CL, cs.LG | This study explores the application of deep learning technologies in software
development processes, particularly in automating code reviews, error
prediction, and test generation to enhance code quality and development
efficiency. Through a series of empirical studies, experimental groups using
deep learning tools and control groups using traditional methods were compared
in terms of code error rates and project completion times. The results
demonstrated significant improvements in the experimental group, validating the
effectiveness of deep learning technologies. The research also discusses
potential optimization points, methodologies, and technical challenges of deep
learning in software development, as well as how to integrate these
technologies into existing software development workflows. | cs.SE | Utilizing Deep Learning to Optimize Software Development Processes |
2024-04-21T11:59:53Z | http://arxiv.org/pdf/2404.13628v1 | http://arxiv.org/abs/2404.13628v1 | Xun Wu, Shaohan Huang, Furu Wei | 2024-04-21T11:59:53Z | cs.CL, cs.LG, cs.MM | LoRA has gained widespread acceptance in the fine-tuning of large pre-trained
models to cater to a diverse array of downstream tasks, showcasing notable
effectiveness and efficiency, thereby solidifying its position as one of the
most prevalent fine-tuning techniques. Due to the modular nature of LoRA's
plug-and-play plugins, researchers have delved into the amalgamation of
multiple LoRAs to empower models to excel across various downstream tasks.
Nonetheless, extant approaches for LoRA fusion grapple with inherent
challenges. Direct arithmetic merging may result in the loss of the original
pre-trained model's generative capabilities or the distinct identity of LoRAs,
thereby yielding suboptimal outcomes. On the other hand, Reference tuning-based
fusion exhibits limitations concerning the requisite flexibility for the
effective combination of multiple LoRAs. In response to these challenges, this
paper introduces the Mixture of LoRA Experts (MoLE) approach, which harnesses
hierarchical control and unfettered branch selection. The MoLE approach not
only achieves superior LoRA fusion performance in comparison to direct
arithmetic merging but also retains the crucial flexibility for combining LoRAs
effectively. Extensive experimental evaluations conducted in both the Natural
Language Processing (NLP) and Vision & Language (V&L) domains substantiate the
efficacy of MoLE. | cs.CL | Mixture of LoRA Experts |
2024-04-21T11:51:13Z | http://arxiv.org/pdf/2404.13627v1 | http://arxiv.org/abs/2404.13627v1 | Chunkit Chan, Cheng Jiayang, Yauwai Yim, Zheye Deng, Wei Fan, Haoran Li, Xin Liu, Hongming Zhang, Weiqi Wang, Yangqiu Song | 2024-04-21T11:51:13Z | cs.CL, cs.AI | Large Language Models (LLMs) have sparked substantial interest and debate
concerning their potential emergence of Theory of Mind (ToM) ability. Theory of
mind evaluations currently focuses on testing models using machine-generated
data or game settings prone to shortcuts and spurious correlations, which lacks
evaluation of machine ToM ability in real-world human interaction scenarios.
This poses a pressing demand to develop new real-world scenario benchmarks. We
introduce NegotiationToM, a new benchmark designed to stress-test machine ToM
in real-world negotiation surrounding covered multi-dimensional mental states
(i.e., desires, beliefs, and intentions). Our benchmark builds upon the
Belief-Desire-Intention (BDI) agent modeling theory and conducts the necessary
empirical experiments to evaluate large language models. Our findings
demonstrate that NegotiationToM is challenging for state-of-the-art LLMs, as
they consistently perform significantly worse than humans, even when employing
the chain-of-thought (CoT) method. | cs.CL | NegotiationToM: A Benchmark for Stress-testing Machine Theory of Mind on Negotiation Surrounding |
2024-04-21T10:49:41Z | http://arxiv.org/pdf/2404.13613v1 | http://arxiv.org/abs/2404.13613v1 | Shai Meital, Lior Rokach, Roman Vainshtein, Nir Grinberg | 2024-04-21T10:49:41Z | cs.CL, cs.LG | Multi-participant discussions tend to unfold in a tree structure rather than
a chain structure. Branching may occur for multiple reasons -- from the
asynchronous nature of online platforms to a conscious decision by an
interlocutor to disengage with part of the conversation. Predicting branching
and understanding the reasons for creating new branches is important for many
downstream tasks such as summarization and thread disentanglement and may help
develop online spaces that encourage users to engage in online discussions in
more meaningful ways. In this work, we define the novel task of branch
prediction and propose GLOBS (Global Branching Score) -- a deep neural network
model for predicting branching. GLOBS is evaluated on three large discussion
forums from Reddit, achieving significant improvements over an array of
competitive baselines and demonstrating better transferability. We affirm that
structural, temporal, and linguistic features contribute to GLOBS success and
find that branching is associated with a greater number of conversation
participants and tends to occur in earlier levels of the conversation tree. We
publicly release GLOBS and our implementation of all baseline models to allow
reproducibility and promote further research on this important task. | cs.CL | The Branch Not Taken: Predicting Branching in Online Conversations |
2024-04-21T10:41:04Z | http://arxiv.org/pdf/2404.13611v2 | http://arxiv.org/abs/2404.13611v2 | Cai Chen, Runzhong Zhang, Jianjun Gao, Kejun Wu, Kim-Hui Yap, Yi Wang | 2024-06-01T06:56:16Z | cs.CV, cs.CL | Temporal sentence grounding involves the retrieval of a video moment with a
natural language query. Many existing works directly incorporate the given
video and temporally localized query for temporal grounding, overlooking the
inherent domain gap between different modalities. In this paper, we utilize
pseudo-query features containing extensive temporally global textual knowledge
sourced from the same video-query pair, to enhance the bridging of domain gaps
and attain a heightened level of similarity between multi-modal features.
Specifically, we propose a Pseudo-query Intermediary Network (PIN) to achieve
an improved alignment of visual and comprehensive pseudo-query features within
the feature space through contrastive learning. Subsequently, we utilize
learnable prompts to encapsulate the knowledge of pseudo-queries, propagating
them into the textual encoder and multi-modal fusion module, further enhancing
the feature alignment between visual and language for better temporal
grounding. Extensive experiments conducted on the Charades-STA and
ActivityNet-Captions datasets demonstrate the effectiveness of our method. | cs.CV | Video sentence grounding with temporally global textual knowledge |
2024-04-21T09:42:05Z | http://arxiv.org/pdf/2404.13599v2 | http://arxiv.org/abs/2404.13599v2 | Zhijun Xu, Siyu Yuan, Lingjie Chen, Deqing Yang | 2024-06-16T11:31:18Z | cs.CL | Puns play a vital role in academic research due to their distinct structure
and clear definition, which aid in the comprehensive analysis of linguistic
humor. However, the understanding of puns in large language models (LLMs) has
not been thoroughly examined, limiting their use in creative writing and humor
creation. In this paper, we leverage three popular tasks, i.e., pun
recognition, explanation and generation to systematically evaluate the
capabilities of LLMs in pun understanding. In addition to adopting the
automated evaluation metrics from prior research, we introduce new evaluation
methods and metrics that are better suited to the in-context learning paradigm
of LLMs. These new metrics offer a more rigorous assessment of an LLM's ability
to understand puns and align more closely with human cognition than previous
metrics. Our findings reveal the "lazy pun generation" pattern and identify the
primary challenges LLMs encounter in understanding puns. | cs.CL | "A good pun is its own reword": Can Large Language Models Understand Puns? |
2024-04-21T09:20:16Z | http://arxiv.org/pdf/2404.14454v2 | http://arxiv.org/abs/2404.14454v2 | Yousef Khan, Ahmed Abdeen Hamed | 2024-06-02T22:42:00Z | cs.CL, cs.AI, I.2; I.2.1 | Addressing the global challenge of breast cancer, this research explores the
fusion of generative AI, focusing on ChatGPT 3.5 turbo model, and the
intricacies of breast cancer risk assessment. The research aims to evaluate
ChatGPT's reasoning capabilities, emphasizing its potential to process rules
and provide explanations for screening recommendations. The study seeks to
bridge the technology gap between intelligent machines and clinicians by
demonstrating ChatGPT's unique proficiency in natural language reasoning. The
methodology employs a supervised prompt-engineering approach to enforce
detailed explanations for ChatGPT's recommendations. Synthetic use cases,
generated algorithmically, serve as the testing ground for the encoded rules,
evaluating the model's processing prowess. Findings highlight ChatGPT's
promising capacity in processing rules comparable to Expert System Shells, with
a focus on natural language reasoning. The research introduces the concept of
reinforcement explainability, showcasing its potential in elucidating outcomes
and facilitating user-friendly interfaces for breast cancer risk assessment. | cs.CL | Reinforcement of Explainability of ChatGPT Prompts by Embedding Breast Cancer Self-Screening Rules into AI Responses |
2024-04-21T08:10:20Z | http://arxiv.org/pdf/2404.15190v1 | http://arxiv.org/abs/2404.15190v1 | Suyeon Shin, Sujin jeon, Junghyun Kim, Gi-Cheon Kang, Byoung-Tak Zhang | 2024-04-21T08:10:20Z | cs.AI, cs.CL, cs.CV, cs.RO, 68T01 (Primary) 68T40, 68T50, 68T45 (Secondary) | Embodied Instruction Following (EIF) is the task of executing natural
language instructions by navigating and interacting with objects in 3D
environments. One of the primary challenges in EIF is compositional task
planning, which is often addressed with supervised or in-context learning with
labeled data. To this end, we introduce the Socratic Planner, the first
zero-shot planning method that infers without the need for any training data.
Socratic Planner first decomposes the instructions into substructural
information of the task through self-questioning and answering, translating it
into a high-level plan, i.e., a sequence of subgoals. Subgoals are executed
sequentially, with our visually grounded re-planning mechanism adjusting plans
dynamically through a dense visual feedback. We also introduce an evaluation
metric of high-level plans, RelaxedHLP, for a more comprehensive evaluation.
Experiments demonstrate the effectiveness of the Socratic Planner, achieving
competitive performance on both zero-shot and few-shot task planning in the
ALFRED benchmark, particularly excelling in tasks requiring higher-dimensional
inference. Additionally, a precise adjustments in the plan were achieved by
incorporating environmental visual information. | cs.AI | Socratic Planner: Inquiry-Based Zero-Shot Planning for Embodied Instruction Following |
2024-04-21T07:34:44Z | http://arxiv.org/pdf/2404.13565v2 | http://arxiv.org/abs/2404.13565v2 | Panfeng Li, Qikai Yang, Xieming Geng, Wenjing Zhou, Zhicheng Ding, Yi Nian | 2024-05-21T02:38:35Z | cs.CV, cs.AI, cs.CL | This study explores innovative methods for improving Visual Question
Answering (VQA) using Generative Adversarial Networks (GANs), autoencoders, and
attention mechanisms. Leveraging a balanced VQA dataset, we investigate three
distinct strategies. Firstly, GAN-based approaches aim to generate answer
embeddings conditioned on image and question inputs, showing potential but
struggling with more complex tasks. Secondly, autoencoder-based techniques
focus on learning optimal embeddings for questions and images, achieving
comparable results with GAN due to better ability on complex questions. Lastly,
attention mechanisms, incorporating Multimodal Compact Bilinear pooling (MCB),
address language priors and attention modeling, albeit with a
complexity-performance trade-off. This study underscores the challenges and
opportunities in VQA and suggests avenues for future research, including
alternative GAN formulations and attentional mechanisms. | cs.CV | Exploring Diverse Methods in Visual Question Answering |
2024-04-21T07:03:55Z | http://arxiv.org/pdf/2404.13556v1 | http://arxiv.org/abs/2404.13556v1 | Kelong Mao, Chenlong Deng, Haonan Chen, Fengran Mo, Zheng Liu, Tetsuya Sakai, Zhicheng Dou | 2024-04-21T07:03:55Z | cs.IR, cs.CL | Conversational search requires accurate interpretation of user intent from
complex multi-turn contexts. This paper presents ChatRetriever, which inherits
the strong generalization capability of large language models to robustly
represent complex conversational sessions for dense retrieval. To achieve this,
we propose a simple and effective dual-learning approach that adapts LLM for
retrieval via contrastive learning while enhancing the complex session
understanding through masked instruction tuning on high-quality conversational
instruction tuning data. Extensive experiments on five conversational search
benchmarks demonstrate that ChatRetriever substantially outperforms existing
conversational dense retrievers, achieving state-of-the-art performance on par
with LLM-based rewriting approaches. Furthermore, ChatRetriever exhibits
superior robustness in handling diverse conversational contexts. Our work
highlights the potential of adapting LLMs for retrieval with complex inputs
like conversational search sessions and proposes an effective approach to
advance this research direction. | cs.IR | ChatRetriever: Adapting Large Language Models for Generalized and Robust Conversational Dense Retrieval |
2024-04-21T06:03:43Z | http://arxiv.org/pdf/2404.13547v1 | http://arxiv.org/abs/2404.13547v1 | Mao-Siang Chen, An-Zi Yen | 2024-04-21T06:03:43Z | cs.CL | To optimize the preparation process for educators in academic lectures and
associated question-and-answer sessions, this paper presents E-QGen, a lecture
abstract-based question generation system. Given a lecture abstract, E-QGen
generates potential student inquiries. The questions suggested by our system
are expected to not only facilitate teachers in preparing answers in advance
but also enable them to supply additional resources when necessary. | cs.CL | E-QGen: Educational Lecture Abstract-based Question Generation System |
2024-04-21T04:55:13Z | http://arxiv.org/pdf/2404.13530v1 | http://arxiv.org/abs/2404.13530v1 | Aviral Agrawal, Carlos Mateo Samudio Lezcano, Iqui Balam Heredia-Marin, Prabhdeep Singh Sethi | 2024-04-21T04:55:13Z | cs.CV, cs.CL, cs.LG | Video-based Question Answering (Video QA) is a challenging task and becomes
even more intricate when addressing Socially Intelligent Question Answering
(SIQA). SIQA requires context understanding, temporal reasoning, and the
integration of multimodal information, but in addition, it requires processing
nuanced human behavior. Furthermore, the complexities involved are exacerbated
by the dominance of the primary modality (text) over the others. Thus, there is
a need to help the task's secondary modalities to work in tandem with the
primary modality. In this work, we introduce a cross-modal alignment and
subsequent representation fusion approach that achieves state-of-the-art
results (82.06\% accuracy) on the Social IQ 2.0 dataset for SIQA. Our approach
exhibits an improved ability to leverage the video modality by using the audio
modality as a bridge with the language modality. This leads to enhanced
performance by reducing the prevalent issue of language overfitting and
resultant video modality bypassing encountered by current existing techniques.
Our code and models are publicly available at
https://github.com/sts-vlcc/sts-vlcc | cs.CV | Listen Then See: Video Alignment with Speaker Attention |
2024-04-21T03:52:46Z | http://arxiv.org/pdf/2404.14453v1 | http://arxiv.org/abs/2404.14453v1 | Xiping Liu, Zhao Tan | 2024-04-21T03:52:46Z | cs.CL, cs.AI, cs.DB | The conversion of natural language queries into SQL queries, known as
Text-to-SQL, is a critical yet challenging task. This paper introduces EPI-SQL,
a novel methodological framework leveraging Large Language Models (LLMs) to
enhance the performance of Text-to-SQL tasks. EPI-SQL operates through a
four-step process. Initially, the method involves gathering instances from the
Spider dataset on which LLMs are prone to failure. These instances are then
utilized to generate general error-prevention instructions (EPIs).
Subsequently, LLMs craft contextualized EPIs tailored to the specific context
of the current task. Finally, these context-specific EPIs are incorporated into
the prompt used for SQL generation. EPI-SQL is distinguished in that it
provides task-specific guidance, enabling the model to circumvent potential
errors for the task at hand. Notably, the methodology rivals the performance of
advanced few-shot methods despite being a zero-shot approach. An empirical
assessment using the Spider benchmark reveals that EPI-SQL achieves an
execution accuracy of 85.1\%, underscoring its effectiveness in generating
accurate SQL queries through LLMs. The findings indicate a promising direction
for future research, i.e. enhancing instructions with task-specific and
contextualized rules, for boosting LLMs' performance in NLP tasks. | cs.CL | EPI-SQL: Enhancing Text-to-SQL Translation with Error-Prevention Instructions |
2024-04-21T02:26:15Z | http://arxiv.org/pdf/2404.13506v2 | http://arxiv.org/abs/2404.13506v2 | Charith Chandra Sai Balne, Sreyoshi Bhaduri, Tamoghna Roy, Vinija Jain, Aman Chadha | 2024-04-23T21:28:26Z | cs.LG, cs.AI, cs.CL | The rise of deep learning has marked significant progress in fields such as
computer vision, natural language processing, and medical imaging, primarily
through the adaptation of pre-trained models for specific tasks. Traditional
fine-tuning methods, involving adjustments to all parameters, face challenges
due to high computational and memory demands. This has led to the development
of Parameter Efficient Fine-Tuning (PEFT) techniques, which selectively update
parameters to balance computational efficiency with performance. This review
examines PEFT approaches, offering a detailed comparison of various strategies
highlighting applications across different domains, including text generation,
medical imaging, protein modeling, and speech synthesis. By assessing the
effectiveness of PEFT methods in reducing computational load, speeding up
training, and lowering memory usage, this paper contributes to making deep
learning more accessible and adaptable, facilitating its wider application and
encouraging innovation in model optimization. Ultimately, the paper aims to
contribute towards insights into PEFT's evolving landscape, guiding researchers
and practitioners in overcoming the limitations of conventional fine-tuning
approaches. | cs.LG | Parameter Efficient Fine Tuning: A Comprehensive Analysis Across Applications |
2024-04-21T02:15:59Z | http://arxiv.org/pdf/2404.13504v1 | http://arxiv.org/abs/2404.13504v1 | Tao Feng, Lizhen Qu, Zhuang Li, Haolan Zhan, Yuncheng Hua, Gholamreza Haffari | 2024-04-21T02:15:59Z | cs.CL | Machine learning models have made incredible progress, but they still
struggle when applied to examples from unseen domains. This study focuses on a
specific problem of domain generalization, where a model is trained on one
source domain and tested on multiple target domains that are unseen during
training. We propose IMO: Invariant features Masks for Out-of-Distribution text
classification, to achieve OOD generalization by learning invariant features.
During training, IMO would learn sparse mask layers to remove irrelevant
features for prediction, where the remaining features keep invariant.
Additionally, IMO has an attention module at the token level to focus on tokens
that are useful for prediction. Our comprehensive experiments show that IMO
substantially outperforms strong baselines in terms of various evaluation
metrics and settings. | cs.CL | IMO: Greedy Layer-Wise Sparse Representation Learning for Out-of-Distribution Text Classification with Pre-trained Models |
2024-04-20T20:48:42Z | http://arxiv.org/pdf/2404.13465v1 | http://arxiv.org/abs/2404.13465v1 | Alexander Shan, John Bauer, Riley Carlson, Christopher Manning | 2024-04-20T20:48:42Z | cs.CL, cs.LG | The vast majority of the popular English named entity recognition (NER)
datasets contain American or British English data, despite the existence of
many global varieties of English. As such, it is unclear whether they
generalize for analyzing use of English globally. To test this, we build a
newswire dataset, the Worldwide English NER Dataset, to analyze NER model
performance on low-resource English variants from around the world. We test
widely used NER toolkits and transformer models, including models using the
pre-trained contextual models RoBERTa and ELECTRA, on three datasets: a
commonly used British English newswire dataset, CoNLL 2003, a more American
focused dataset OntoNotes, and our global dataset. All models trained on the
CoNLL or OntoNotes datasets experienced significant performance drops-over 10
F1 in some cases-when tested on the Worldwide English dataset. Upon examination
of region-specific errors, we observe the greatest performance drops for
Oceania and Africa, while Asia and the Middle East had comparatively strong
performance. Lastly, we find that a combined model trained on the Worldwide
dataset and either CoNLL or OntoNotes lost only 1-2 F1 on both test sets. | cs.CL | Do "English" Named Entity Recognizers Work Well on Global Englishes? |
2024-04-20T18:22:49Z | http://arxiv.org/pdf/2404.13439v1 | http://arxiv.org/abs/2404.13439v1 | Sefika Efeoglu, Adrian Paschke | 2024-04-20T18:22:49Z | cs.CL | Information resources such as newspapers have produced unstructured text data
in various languages related to the corona outbreak since December 2019.
Analyzing these unstructured texts is time-consuming without representing them
in a structured format; therefore, representing them in a structured format is
crucial. An information extraction pipeline with essential tasks -- named
entity tagging and relation extraction -- to accomplish this goal might be
applied to these texts. This study proposes a data annotation pipeline to
generate training data from corona news articles, including generic and
domain-specific entities. Named entity recognition models are trained on this
annotated corpus and then evaluated on test sentences manually annotated by
domain experts evaluating the performance of a trained model. The code base and
demonstration are available at https://github.com/sefeoglu/coronanews-ner.git. | cs.CL | Fine-Grained Named Entities for Corona News |
2024-04-20T16:48:18Z | http://arxiv.org/pdf/2404.14449v1 | http://arxiv.org/abs/2404.14449v1 | Mohammad Al-Ramahi, Izzat Alsmadi, Abdullah Wahbeh | 2024-04-20T16:48:18Z | cs.CL, cs.LG | The wealth of information available through the Internet and social media is
unprecedented. Within computing fields, websites such as Stack Overflow are
considered important sources for users seeking solutions to their computing and
programming issues. However, like other social media platforms, Stack Overflow
contains a mixture of relevant and irrelevant information. In this paper, we
evaluated neural network models to predict the quality of questions on Stack
Overflow, as an example of Question Answering (QA) communities. Our results
demonstrate the effectiveness of neural network models compared to baseline
machine learning models, achieving an accuracy of 80%. Furthermore, our
findings indicate that the number of layers in the neural network model can
significantly impact its performance. | cs.CL | Predicting Question Quality on StackOverflow with Neural Networks |
2024-04-20T15:04:25Z | http://arxiv.org/pdf/2404.13402v1 | http://arxiv.org/abs/2404.13402v1 | Jiongliang Lin, Yiwen Guo, Hao Chen | 2024-04-20T15:04:25Z | cs.CR, cs.AI, cs.CL | Intrusion detection is a long standing and crucial problem in security. A
system capable of detecting intrusions automatically is on great demand in
enterprise security solutions. Existing solutions rely heavily on hand-crafted
rules designed by security operators, which suffer from high false negative
rates and poor generalization ability to new, zero-day attacks at scale. AI and
machine learning offer promising solutions to address the issues, by inspecting
abnormal user behaviors intelligently and automatically from data. However,
existing learning-based intrusion detection systems in the literature are
mostly designed for small data, and they lack the ability to leverage the power
of big data in cloud environments. In this paper, we target at this problem and
introduce an intrusion detection system which incorporates large-scale
pre-training, so as to train a large language model based on tens of millions
of command lines for AI-based intrusion detection. Experiments performed on 30
million training samples and 10 million test samples verify the effectiveness
of our solution. | cs.CR | Intrusion Detection at Scale with the Assistance of a Command-line Language Model |
2024-04-20T14:42:43Z | http://arxiv.org/pdf/2404.13397v1 | http://arxiv.org/abs/2404.13397v1 | Sefika Efeoglu, Adrian Paschke | 2024-04-20T14:42:43Z | cs.CL, cs.AI | Information Extraction (IE) is a transformative process that converts
unstructured text data into a structured format by employing entity and
relation extraction (RE) methodologies. The identification of the relation
between a pair of entities plays a crucial role within this framework. Despite
the existence of various techniques for relation extraction, their efficacy
heavily relies on access to labeled data and substantial computational
resources. In addressing these challenges, Large Language Models (LLMs) emerge
as promising solutions; however, they might return hallucinating responses due
to their own training data. To overcome these limitations, Retrieved-Augmented
Generation-based Relation Extraction (RAG4RE) in this work is proposed,
offering a pathway to enhance the performance of relation extraction tasks.
This work evaluated the effectiveness of our RAG4RE approach utilizing
different LLMs. Through the utilization of established benchmarks, such as
TACRED, TACREV, Re-TACRED, and SemEval RE datasets, our aim is to
comprehensively evaluate the efficacy of our RAG4RE approach. In particularly,
we leverage prominent LLMs including Flan T5, Llama2, and Mistral in our
investigation. The results of our study demonstrate that our RAG4RE approach
surpasses performance of traditional RE approaches based solely on LLMs,
particularly evident in the TACRED dataset and its variations. Furthermore, our
approach exhibits remarkable performance compared to previous RE methodologies
across both TACRED and TACREV datasets, underscoring its efficacy and potential
for advancing RE tasks in natural language processing. | cs.CL | Retrieval-Augmented Generation-based Relation Extraction |
2024-04-20T14:20:24Z | http://arxiv.org/pdf/2404.13390v1 | http://arxiv.org/abs/2404.13390v1 | Jianxiang Zang, Hui Liu | 2024-04-20T14:20:24Z | cs.CL | The robustness of Transformer-based Natural Language Inference encoders is
frequently compromised as they tend to rely more on dataset biases than on the
intended task-relevant features. Recent studies have attempted to mitigate this
by reducing the weight of biased samples during the training process. However,
these debiasing methods primarily focus on identifying which samples are biased
without explicitly determining the biased components within each case. This
limitation restricts those methods' capability in out-of-distribution
inference. To address this issue, we aim to train models to adopt the logic
humans use in explaining causality. We propose a simple, comprehensive, and
interpretable method: Explanation based Bias Decoupling Regularization
(EBD-Reg). EBD-Reg employs human explanations as criteria, guiding the encoder
to establish a tripartite parallel supervision of Distinguishing, Decoupling
and Aligning. This method enables encoders to identify and focus on keywords
that represent the task-relevant features during inference, while discarding
the residual elements acting as biases. Empirical evidence underscores that
EBD-Reg effectively guides various Transformer-based encoders to decouple
biases through a human-centric lens, significantly surpassing other methods in
terms of out-of-distribution inference capabilities. | cs.CL | Explanation based Bias Decoupling Regularization for Natural Language Inference |
2024-04-20T13:15:27Z | http://arxiv.org/pdf/2404.13370v1 | http://arxiv.org/abs/2404.13370v1 | Zihao Yue, Yepeng Zhang, Ziheng Wang, Qin Jin | 2024-04-20T13:15:27Z | cs.CV, cs.CL, cs.MM | Automatic movie narration targets at creating video-aligned plot descriptions
to assist visually impaired audiences. It differs from standard video
captioning in that it requires not only describing key visual details but also
inferring the plots developed across multiple movie shots, thus posing unique
and ongoing challenges. To advance the development of automatic movie narrating
systems, we first revisit the limitations of existing datasets and develop a
large-scale, bilingual movie narration dataset, Movie101v2. Second, taking into
account the essential difficulties in achieving applicable movie narration, we
break the long-term goal into three progressive stages and tentatively focus on
the initial stages featuring understanding within individual clips. We also
introduce a new narration assessment to align with our staged task goals.
Third, using our new dataset, we baseline several leading large vision-language
models, including GPT-4V, and conduct in-depth investigations into the
challenges current models face for movie narration generation. Our findings
reveal that achieving applicable movie narration generation is a fascinating
goal that requires thorough research. | cs.CV | Movie101v2: Improved Movie Narration Benchmark |
2024-04-20T12:16:35Z | http://arxiv.org/pdf/2404.13364v1 | http://arxiv.org/abs/2404.13364v1 | Ruturaj Ghatage, Aditya Kulkarni, Rajlaxmi Patil, Sharvi Endait, Raviraj Joshi | 2024-04-20T12:16:35Z | cs.CL, cs.LG | Question-answering systems have revolutionized information retrieval, but
linguistic and cultural boundaries limit their widespread accessibility. This
research endeavors to bridge the gap of the absence of efficient QnA datasets
in low-resource languages by translating the English Question Answering Dataset
(SQuAD) using a robust data curation approach. We introduce MahaSQuAD, the
first-ever full SQuAD dataset for the Indic language Marathi, consisting of
118,516 training, 11,873 validation, and 11,803 test samples. We also present a
gold test set of manually verified 500 examples. Challenges in maintaining
context and handling linguistic nuances are addressed, ensuring accurate
translations. Moreover, as a QnA dataset cannot be simply converted into any
low-resource language using translation, we need a robust method to map the
answer translation to its span in the translated passage. Hence, to address
this challenge, we also present a generic approach for translating SQuAD into
any low-resource language. Thus, we offer a scalable approach to bridge
linguistic and cultural gaps present in low-resource languages, in the realm of
question-answering systems. The datasets and models are shared publicly at
https://github.com/l3cube-pune/MarathiNLP . | cs.CL | MahaSQuAD: Bridging Linguistic Divides in Marathi Question-Answering |
2024-04-20T12:08:00Z | http://arxiv.org/pdf/2404.13362v1 | http://arxiv.org/abs/2404.13362v1 | Samuael Adnew, Paul Pu Liang | 2024-04-20T12:08:00Z | cs.CL, cs.AI, cs.LG, eess.AS | Automatic Speech Recognition (ASR) can play a crucial role in enhancing the
accessibility of spoken languages worldwide. In this paper, we build a set of
ASR tools for Amharic, a language spoken by more than 50 million people
primarily in eastern Africa. Amharic is written in the Ge'ez script, a sequence
of graphemes with spacings denoting word boundaries. This makes computational
processing of Amharic challenging since the location of spacings can
significantly impact the meaning of formed sentences. We find that existing
benchmarks for Amharic ASR do not account for these spacings and only measure
individual grapheme error rates, leading to significantly inflated measurements
of in-the-wild performance. In this paper, we first release corrected
transcriptions of existing Amharic ASR test datasets, enabling the community to
accurately evaluate progress. Furthermore, we introduce a post-processing
approach using a transformer encoder-decoder architecture to organize raw ASR
outputs into a grammatically complete and semantically meaningful Amharic
sentence. Through experiments on the corrected test dataset, our model enhances
the semantic correctness of Amharic speech recognition systems, achieving a
Character Error Rate (CER) of 5.5\% and a Word Error Rate (WER) of 23.3\%. | cs.CL | Semantically Corrected Amharic Automatic Speech Recognition |
2024-04-20T11:10:37Z | http://arxiv.org/pdf/2404.13350v1 | http://arxiv.org/abs/2404.13350v1 | Maneesha U. Athukorala, Deshan K. Sumanathilaka | 2024-04-20T11:10:37Z | cs.CL | Machine Transliteration provides the ability to transliterate a basic
language into different languages in a computational way. Transliteration is an
important technical process that has caught the attention most recently. The
Sinhala transliteration has many constraints because of the insufficiency of
resources in the Sinhala language. Due to these limitations, Sinhala
Transliteration is highly complex and time-consuming. Therefore, the majority
of the Sri Lankans uses non-formal texting language named 'Singlish' to make
that process simple. This study has focused on the transliteration of the
Singlish language at the word level by reducing the complication in the
transliteration. A new approach of coding system has invented with the
rule-based approach that can map the matching Sinhala words even without the
vowels. Various typing patterns were collected by different communities for
this. The collected data have analyzed with every Sinhala character and unique
Singlish patterns related to them were generated. The system has introduced a
newly initiated numeric coding system to use with the Singlish letters by
matching with the recognized typing patterns. For the mapping process, fuzzy
logic-based implementation has used. A codified dictionary has also implemented
including unique numeric values. In this system, Each Romanized English letter
was assigned with a unique numeric code that can construct a unique pattern for
each word. The system can identify the most relevant Sinhala word that matches
with the pattern of the Singlish word or it gives the most related word
suggestions. For example, the word 'kiyanna,kianna, kynna, kynn, kiynna' have
mapped with the accurate Sinhala word "kiyanna". These results revealed that
the 'Swa Bhasha' transliteration system has the ability to enhance the Sinhala
users' experience while conducting the texting in Singlish to Sinhala. | cs.CL | Swa Bhasha: Message-Based Singlish to Sinhala Transliteration |
2024-04-20T10:41:02Z | http://arxiv.org/pdf/2404.13343v1 | http://arxiv.org/abs/2404.13343v1 | Ana-Cristina Rogoz, Radu Tudor Ionescu | 2024-04-20T10:41:02Z | cs.CL, cs.AI, cs.LG | This work explores a novel data augmentation method based on Large Language
Models (LLMs) for predicting item difficulty and response time of retired USMLE
Multiple-Choice Questions (MCQs) in the BEA 2024 Shared Task. Our approach is
based on augmenting the dataset with answers from zero-shot LLMs (Falcon,
Meditron, Mistral) and employing transformer-based models based on six
alternative feature combinations. The results suggest that predicting the
difficulty of questions is more challenging. Notably, our top performing
methods consistently include the question text, and benefit from the
variability of LLM answers, highlighting the potential of LLMs for improving
automated assessment in medical licensing exams. We make our code available
https://github.com/ana-rogoz/BEA-2024. | cs.CL | UnibucLLM: Harnessing LLMs for Automated Prediction of Item Difficulty and Response Time for Multiple-Choice Questions |
2024-04-20T08:08:28Z | http://arxiv.org/pdf/2404.14445v1 | http://arxiv.org/abs/2404.14445v1 | Yefeng Yuan, Yuhong Liu, Liang Cheng | 2024-04-20T08:08:28Z | cs.LG, cs.AI, cs.CL | The rapid advancements in generative AI and large language models (LLMs) have
opened up new avenues for producing synthetic data, particularly in the realm
of structured tabular formats, such as product reviews. Despite the potential
benefits, concerns regarding privacy leakage have surfaced, especially when
personal information is utilized in the training datasets. In addition, there
is an absence of a comprehensive evaluation framework capable of quantitatively
measuring the quality of the generated synthetic data and their utility for
downstream tasks. In response to this gap, we introduce SynEval, an open-source
evaluation framework designed to assess the fidelity, utility, and privacy
preservation of synthetically generated tabular data via a suite of diverse
evaluation metrics. We validate the efficacy of our proposed framework -
SynEval - by applying it to synthetic product review data generated by three
state-of-the-art LLMs: ChatGPT, Claude, and Llama. Our experimental findings
illuminate the trade-offs between various evaluation metrics in the context of
synthetic data generation. Furthermore, SynEval stands as a critical instrument
for researchers and practitioners engaged with synthetic tabular data,,
empowering them to judiciously determine the suitability of the generated data
for their specific applications, with an emphasis on upholding user privacy. | cs.LG | A Multi-Faceted Evaluation Framework for Assessing Synthetic Data Generated by Large Language Models |
2024-04-20T07:29:06Z | http://arxiv.org/pdf/2404.13307v1 | http://arxiv.org/abs/2404.13307v1 | Soumyadeep Roy, Aparup Khatua, Fatemeh Ghoochani, Uwe Hadler, Wolfgang Nejdl, Niloy Ganguly | 2024-04-20T07:29:06Z | cs.CL | GPT-4 demonstrates high accuracy in medical QA tasks, leading with an
accuracy of 86.70%, followed by Med-PaLM 2 at 86.50%. However, around 14% of
errors remain. Additionally, current works use GPT-4 to only predict the
correct option without providing any explanation and thus do not provide any
insight into the thinking process and reasoning used by GPT-4 or other LLMs.
Therefore, we introduce a new domain-specific error taxonomy derived from
collaboration with medical students. Our GPT-4 USMLE Error (G4UE) dataset
comprises 4153 GPT-4 correct responses and 919 incorrect responses to the
United States Medical Licensing Examination (USMLE) respectively. These
responses are quite long (258 words on average), containing detailed
explanations from GPT-4 justifying the selected option. We then launch a
large-scale annotation study using the Potato annotation platform and recruit
44 medical experts through Prolific, a well-known crowdsourcing platform. We
annotated 300 out of these 919 incorrect data points at a granular level for
different classes and created a multi-label span to identify the reasons behind
the error. In our annotated dataset, a substantial portion of GPT-4's incorrect
responses is categorized as a "Reasonable response by GPT-4," by annotators.
This sheds light on the challenge of discerning explanations that may lead to
incorrect options, even among trained medical professionals. We also provide
medical concepts and medical semantic predications extracted using the SemRep
tool for every data point. We believe that it will aid in evaluating the
ability of LLMs to answer complex medical questions. We make the resources
available at https://github.com/roysoumya/usmle-gpt4-error-taxonomy . | cs.CL | Beyond Accuracy: Investigating Error Types in GPT-4 Responses to USMLE Questions |
2024-04-20T06:49:15Z | http://arxiv.org/pdf/2404.13292v1 | http://arxiv.org/abs/2404.13292v1 | Khuyagbaatar Batsuren, Ekaterina Vylomova, Verna Dankers, Tsetsuukhei Delgerbaatar, Omri Uzan, Yuval Pinter, Gábor Bella | 2024-04-20T06:49:15Z | cs.CL, cs.AI | The popular subword tokenizers of current language models, such as Byte-Pair
Encoding (BPE), are known not to respect morpheme boundaries, which affects the
downstream performance of the models. While many improved tokenization
algorithms have been proposed, their evaluation and cross-comparison is still
an open problem. As a solution, we propose a combined intrinsic-extrinsic
evaluation framework for subword tokenization. Intrinsic evaluation is based on
our new UniMorph Labeller tool that classifies subword tokenization as either
morphological or alien. Extrinsic evaluation, in turn, is performed via the
Out-of-Vocabulary Generalization Challenge 1.0 benchmark, which consists of
three newly specified downstream text classification tasks. Our empirical
findings show that the accuracy of UniMorph Labeller is 98%, and that, in all
language models studied (including ALBERT, BERT, RoBERTa, and DeBERTa), alien
tokenization leads to poorer generalizations compared to morphological
tokenization for semantic compositionality of word meanings. | cs.CL | Evaluating Subword Tokenization: Alien Subword Composition and OOV Generalization Challenge |
2024-04-20T06:32:00Z | http://arxiv.org/pdf/2404.13289v1 | http://arxiv.org/abs/2404.13289v1 | Jingqi Kang, Tongtong Wu, Jinming Zhao, Guitao Wang, Yinwei Wei, Hao Yang, Guilin Qi, Yuan-Fang Li, Gholamreza Haffari | 2024-04-20T06:32:00Z | cs.CL, cs.MM, cs.SD, eess.AS | Speech event detection is crucial for multimedia retrieval, involving the
tagging of both semantic and acoustic events. Traditional ASR systems often
overlook the interplay between these events, focusing solely on content, even
though the interpretation of dialogue can vary with environmental context. This
paper tackles two primary challenges in speech event detection: the continual
integration of new events without forgetting previous ones, and the
disentanglement of semantic from acoustic events. We introduce a new task,
continual event detection from speech, for which we also provide two benchmark
datasets. To address the challenges of catastrophic forgetting and effective
disentanglement, we propose a novel method, 'Double Mixture.' This method
merges speech expertise with robust memory mechanisms to enhance adaptability
and prevent forgetting. Our comprehensive experiments show that this task
presents significant challenges that are not effectively addressed by current
state-of-the-art methods in either computer vision or natural language
processing. Our approach achieves the lowest rates of forgetting and the
highest levels of generalization, proving robust across various continual
learning sequences. Our code and data are available at
https://anonymous.4open.science/status/Continual-SpeechED-6461. | cs.CL | Double Mixture: Towards Continual Event Detection from Speech |
2024-04-20T04:14:28Z | http://arxiv.org/pdf/2404.14443v1 | http://arxiv.org/abs/2404.14443v1 | Kewei Yuan, Qiurong Zhao, Yang Xu, Xiao Zhang, Huansheng Ning | 2024-04-20T04:14:28Z | cs.CL, cs.AI | In view of the fact that most of the existing machine translation evaluation
algorithms only consider the lexical and syntactic information, but ignore the
deep semantic information contained in the sentence, this paper proposes a
computational method for evaluating the semantic correctness of machine
translations based on reference translations and incorporating semantic
dependencies and sentence keyword information. Use the language technology
platform developed by the Social Computing and Information Retrieval Research
Center of Harbin Institute of Technology to conduct semantic dependency
analysis and keyword analysis on sentences, and obtain semantic dependency
graphs, keywords, and weight information corresponding to keywords. It includes
all word information with semantic dependencies in the sentence and keyword
information that affects semantic information. Construct semantic association
pairs including word and dependency multi-features. The key semantics of the
sentence cannot be highlighted in the semantic information extracted through
semantic dependence, resulting in vague semantics analysis. Therefore, the
sentence keyword information is also included in the scope of machine
translation semantic evaluation. To achieve a comprehensive and in-depth
evaluation of the semantic correctness of sentences, the experimental results
show that the accuracy of the evaluation algorithm has been improved compared
with similar methods, and it can more accurately measure the semantic
correctness of machine translation. | cs.CL | Evaluation of Machine Translation Based on Semantic Dependencies and Keywords |
2024-04-20T03:16:13Z | http://arxiv.org/pdf/2404.13246v1 | http://arxiv.org/abs/2404.13246v1 | Zekai Li, Yanxia Qin, Qian Liu, Min-Yen Kan | 2024-04-20T03:16:13Z | cs.CL | We propose Iterative Facuality Refining on Informative Scientific
Question-Answering (ISQA) feedback\footnote{Code is available at
\url{https://github.com/lizekai-richard/isqa}}, a method following human
learning theories that employs model-generated feedback consisting of both
positive and negative information. Through iterative refining of summaries, it
probes for the underlying rationale of statements to enhance the factuality of
scientific summarization. ISQA does this in a fine-grained manner by asking a
summarization agent to reinforce validated statements in positive feedback and
fix incorrect ones in negative feedback. Our findings demonstrate that the ISQA
feedback mechanism significantly improves the factuality of various open-source
LLMs on the summarization task, as evaluated across multiple scientific
datasets. | cs.CL | ISQA: Informative Factuality Feedback for Scientific Summarization |
2024-04-20T02:30:21Z | http://arxiv.org/pdf/2404.13238v1 | http://arxiv.org/abs/2404.13238v1 | Feibo Jiang, Li Dong, Siwei Tu, Yubo Peng, Kezhi Wang, Kun Yang, Cunhua Pan, Dusit Niyato | 2024-04-20T02:30:21Z | cs.LG, cs.AI, cs.CL | Large Language Models (LLMs) have revolutionized natural language processing
tasks. However, their deployment in wireless networks still face challenges,
i.e., a lack of privacy and security protection mechanisms. Federated Learning
(FL) has emerged as a promising approach to address these challenges. Yet, it
suffers from issues including inefficient handling with big and heterogeneous
data, resource-intensive training, and high communication overhead. To tackle
these issues, we first compare different learning stages and their features of
LLMs in wireless networks. Next, we introduce two personalized wireless
federated fine-tuning methods with low communication overhead, i.e., (1)
Personalized Federated Instruction Tuning (PFIT), which employs reinforcement
learning to fine-tune local LLMs with diverse reward models to achieve
personalization; (2) Personalized Federated Task Tuning (PFTT), which can
leverage global adapters and local Low-Rank Adaptations (LoRA) to
collaboratively fine-tune local LLMs, where the local LoRAs can be applied to
achieve personalization without aggregation. Finally, we perform simulations to
demonstrate the effectiveness of the proposed two methods and comprehensively
discuss open issues. | cs.LG | Personalized Wireless Federated Learning for Large Language Models |
2024-04-19T22:55:23Z | http://arxiv.org/pdf/2404.13208v1 | http://arxiv.org/abs/2404.13208v1 | Eric Wallace, Kai Xiao, Reimar Leike, Lilian Weng, Johannes Heidecke, Alex Beutel | 2024-04-19T22:55:23Z | cs.CR, cs.CL, cs.LG | Today's LLMs are susceptible to prompt injections, jailbreaks, and other
attacks that allow adversaries to overwrite a model's original instructions
with their own malicious prompts. In this work, we argue that one of the
primary vulnerabilities underlying these attacks is that LLMs often consider
system prompts (e.g., text from an application developer) to be the same
priority as text from untrusted users and third parties. To address this, we
propose an instruction hierarchy that explicitly defines how models should
behave when instructions of different priorities conflict. We then propose a
data generation method to demonstrate this hierarchical instruction following
behavior, which teaches LLMs to selectively ignore lower-privileged
instructions. We apply this method to GPT-3.5, showing that it drastically
increases robustness -- even for attack types not seen during training -- while
imposing minimal degradations on standard capabilities. | cs.CR | The Instruction Hierarchy: Training LLMs to Prioritize Privileged Instructions |
2024-04-19T21:39:37Z | http://arxiv.org/pdf/2404.13192v1 | http://arxiv.org/abs/2404.13192v1 | Yuchen Zhang, Xiaoxiao Ma, Jia Wu, Jian Yang, Hao Fan | 2024-04-19T21:39:37Z | cs.CL, cs.AI | Fake news is pervasive on social media, inflicting substantial harm on public
discourse and societal well-being. We investigate the explicit structural
information and textual features of news pieces by constructing a heterogeneous
graph concerning the relations among news topics, entities, and content.
Through our study, we reveal that fake news can be effectively detected in
terms of the atypical heterogeneous subgraphs centered on them, which
encapsulate the essential semantics and intricate relations between news
elements. However, suffering from the heterogeneity, exploring such
heterogeneous subgraphs remains an open problem. To bridge the gap, this work
proposes a heterogeneous subgraph transformer (HeteroSGT) to exploit subgraphs
in our constructed heterogeneous graph. In HeteroSGT, we first employ a
pre-trained language model to derive both word-level and sentence-level
semantics. Then the random walk with restart (RWR) is applied to extract
subgraphs centered on each news, which are further fed to our proposed subgraph
Transformer to quantify the authenticity. Extensive experiments on five
real-world datasets demonstrate the superior performance of HeteroSGT over five
baselines. Further case and ablation studies validate our motivation and
demonstrate that performance improvement stems from our specially designed
components. | cs.CL | Heterogeneous Subgraph Transformer for Fake News Detection |
2024-04-19T20:14:15Z | http://arxiv.org/pdf/2404.13163v1 | http://arxiv.org/abs/2404.13163v1 | Alireza Javadian Sabet, Sarah H. Bana, Renzhe Yu, Morgan R. Frank | 2024-04-19T20:14:15Z | econ.GN, cs.CL, q-fin.EC | Higher education plays a critical role in driving an innovative economy by
equipping students with knowledge and skills demanded by the workforce. While
researchers and practitioners have developed data systems to track detailed
occupational skills, such as those established by the U.S. Department of Labor
(DOL), much less effort has been made to document skill development in higher
education at a similar granularity. Here, we fill this gap by presenting a
longitudinal dataset of skills inferred from over three million course syllabi
taught at nearly three thousand U.S. higher education institutions. To
construct this dataset, we apply natural language processing to extract from
course descriptions detailed workplace activities (DWAs) used by the DOL to
describe occupations. We then aggregate these DWAs to create skill profiles for
institutions and academic majors. Our dataset offers a large-scale
representation of college-educated workers and their role in the economy. To
showcase the utility of this dataset, we use it to 1) compare the similarity of
skills taught and skills in the workforce according to the US Bureau of Labor
Statistics, 2) estimate gender differences in acquired skills based on
enrollment data, 3) depict temporal trends in the skills taught in social
science curricula, and 4) connect college majors' skill distinctiveness to
salary differences of graduates. Overall, this dataset can enable new research
on the source of skills in the context of workforce development and provide
actionable insights for shaping the future of higher education to meet evolving
labor demands especially in the face of new technologies. | econ.GN | A national longitudinal dataset of skills taught in U.S. higher education curricula |
2024-04-19T19:34:35Z | http://arxiv.org/pdf/2404.13149v1 | http://arxiv.org/abs/2404.13149v1 | Chia-Hsuan Chang, Mary M. Lucas, Yeawon Lee, Christopher C. Yang, Grace Lu-Yao | 2024-04-19T19:34:35Z | cs.CL, cs.AI | Advances in large language models (LLMs) have encouraged their adoption in
the healthcare domain where vital clinical information is often contained in
unstructured notes. Cancer staging status is available in clinical reports, but
it requires natural language processing to extract the status from the
unstructured text. With the advance in clinical-oriented LLMs, it is promising
to extract such status without extensive efforts in training the algorithms.
Prompting approaches of the pre-trained LLMs that elicit a model's reasoning
process, such as chain-of-thought, may help to improve the trustworthiness of
the generated responses. Using self-consistency further improves model
performance, but often results in inconsistent generations across the multiple
reasoning paths. In this study, we propose an ensemble reasoning approach with
the aim of improving the consistency of the model generations. Using an open
access clinical large language model to determine the pathologic cancer stage
from real-world pathology reports, we show that the ensemble reasoning approach
is able to improve both the consistency and performance of the LLM in
determining cancer stage, thereby demonstrating the potential to use these
models in clinical or other domains where reliability and trustworthiness are
critical. | cs.CL | Beyond Self-Consistency: Ensemble Reasoning Boosts Consistency and Accuracy of LLMs in Cancer Staging |
2024-04-19T17:57:29Z | http://arxiv.org/pdf/2404.13043v1 | http://arxiv.org/abs/2404.13043v1 | Soham Gadgil, Mahtab Bigverdi | 2024-04-19T17:57:29Z | cs.CV, cs.CL, cs.LG | AI in dermatology is evolving at a rapid pace but the major limitation to
training trustworthy classifiers is the scarcity of data with ground-truth
concept level labels, which are meta-labels semantically meaningful to humans.
Foundation models like CLIP providing zero-shot capabilities can help alleviate
this challenge by leveraging vast amounts of image-caption pairs available on
the internet. CLIP can be fine-tuned using domain specific image-caption pairs
to improve classification performance. However, CLIP's pre-training data is not
well-aligned with the medical jargon that clinicians use to perform diagnoses.
The development of large language models (LLMs) in recent years has led to the
possibility of leveraging the expressive nature of these models to generate
rich text. Our goal is to use these models to generate caption text that aligns
well with both the clinical lexicon and with the natural human language used in
CLIP's pre-training data. Starting with captions used for images in PubMed
articles, we extend them by passing the raw captions through an LLM fine-tuned
on the field's several textbooks. We find that using captions generated by an
expressive fine-tuned LLM like GPT-3.5 improves downstream zero-shot concept
classification performance. | cs.CV | Data Alignment for Zero-Shot Concept Generation in Dermatology AI |
2024-04-19T17:51:52Z | http://arxiv.org/pdf/2404.13039v1 | http://arxiv.org/abs/2404.13039v1 | Tiancheng Gu, Kaicheng Yang, Dongnan Liu, Weidong Cai | 2024-04-19T17:51:52Z | cs.CV, cs.CL | Medical visual question answering (Med-VQA) aims to automate the prediction
of correct answers for medical images and questions, thereby assisting
physicians in reducing repetitive tasks and alleviating their workload.
Existing approaches primarily focus on pre-training models using additional and
comprehensive datasets, followed by fine-tuning to enhance performance in
downstream tasks. However, there is also significant value in exploring
existing models to extract clinically relevant information. In this paper, we
propose the Latent Prompt Assist model (LaPA) for medical visual question
answering. Firstly, we design a latent prompt generation module to generate the
latent prompt with the constraint of the target answer. Subsequently, we
propose a multi-modal fusion block with latent prompt fusion module that
utilizes the latent prompt to extract clinical-relevant information from
uni-modal and multi-modal features. Additionally, we introduce a prior
knowledge fusion module to integrate the relationship between diseases and
organs with the clinical-relevant information. Finally, we combine the final
integrated information with image-language cross-modal information to predict
the final answers. Experimental results on three publicly available Med-VQA
datasets demonstrate that LaPA outperforms the state-of-the-art model ARL,
achieving improvements of 1.83%, 0.63%, and 1.80% on VQA-RAD, SLAKE, and
VQA-2019, respectively. The code is publicly available at
https://github.com/GaryGuTC/LaPA_model. | cs.CV | LaPA: Latent Prompt Assist Model For Medical Visual Question Answering |
2024-04-19T17:47:02Z | http://arxiv.org/pdf/2404.13033v1 | http://arxiv.org/abs/2404.13033v1 | Biyang Guo, He Wang, Wenyilin Xiao, Hong Chen, Zhuxin Lee, Songqiao Han, Hailiang Huang | 2024-04-19T17:47:02Z | cs.CL | In the burgeoning field of Large Language Models (LLMs) like ChatGPT and
LLaMA, Prompt Engineering (PE) is renowned for boosting zero-shot or in-context
learning (ICL) through prompt modifications. Yet, the realm of the sample
design for downstream fine-tuning, crucial for task-specific LLM adaptation, is
largely unexplored. This paper introduces Sample Design Engineering (SDE), a
methodical approach to enhancing LLMs' post-tuning performance by refining
input, output, and reasoning designs. We conduct a series of in-domain (ID) and
out-of-domain (OOD) experiments to assess the impact of various design options
on LLMs' downstream performance, revealing several intriguing patterns that
hold consistently across different LLMs. Based on these insights, we propose an
integrated SDE strategy, combining the most effective options, and validate its
consistent superiority over heuristic sample designs in complex downstream
tasks like multi-aspect sentiment analysis, event extraction, and nested entity
recognition. Additionally, analyses of LLMs' inherent prompt/output perplexity,
zero-shot, and ICL abilities illustrate that good PE strategies may not always
translate to good SDE strategies. Code available at
https://github.com/beyondguo/LLM-Tuning. | cs.CL | Sample Design Engineering: An Empirical Study of What Makes Good Downstream Fine-Tuning Samples for LLMs |
2024-04-19T17:30:10Z | http://arxiv.org/pdf/2404.13020v1 | http://arxiv.org/abs/2404.13020v1 | Gregory Yauney, David Mimno | 2024-04-19T17:30:10Z | cs.CL, cs.LG | Evaluating the in-context learning classification performance of language
models poses challenges due to small dataset sizes, extensive prompt-selection
using the validation set, and intentionally difficult tasks that lead to
near-random performance. The standard random baseline -- the expected accuracy
of guessing labels uniformly at random -- is stable when the evaluation set is
used only once or when the dataset is large. We account for the common practice
of validation set reuse and existing small datasets with a stronger random
baseline: the expected maximum accuracy across multiple random classifiers.
When choosing the best prompt demonstrations across six quantized language
models applied to 16 BIG-bench Lite tasks, more than 20\% of the few-shot
results that exceed the standard baseline do not exceed this stronger random
baseline. When held-out test sets are available, this stronger baseline is also
a better predictor of held-out performance than the standard baseline, avoiding
unnecessary test set evaluations. This maximum random baseline provides an
easily calculated drop-in replacement for the standard baseline. | cs.CL | Stronger Random Baselines for In-Context Learning |
2024-04-19T17:22:51Z | http://arxiv.org/pdf/2404.13013v1 | http://arxiv.org/abs/2404.13013v1 | Chuofan Ma, Yi Jiang, Jiannan Wu, Zehuan Yuan, Xiaojuan Qi | 2024-04-19T17:22:51Z | cs.CV, cs.AI, cs.CL, cs.LG | We introduce Groma, a Multimodal Large Language Model (MLLM) with grounded
and fine-grained visual perception ability. Beyond holistic image
understanding, Groma is adept at region-level tasks such as region captioning
and visual grounding. Such capabilities are built upon a localized visual
tokenization mechanism, where an image input is decomposed into regions of
interest and subsequently encoded into region tokens. By integrating region
tokens into user instructions and model responses, we seamlessly enable Groma
to understand user-specified region inputs and ground its textual output to
images. Besides, to enhance the grounded chat ability of Groma, we curate a
visually grounded instruction dataset by leveraging the powerful GPT-4V and
visual prompting techniques. Compared with MLLMs that rely on the language
model or external module for localization, Groma consistently demonstrates
superior performances in standard referring and grounding benchmarks,
highlighting the advantages of embedding localization into image tokenization.
Project page: https://groma-mllm.github.io/. | cs.CV | Groma: Localized Visual Tokenization for Grounding Multimodal Large Language Models |
2024-04-19T16:45:50Z | http://arxiv.org/pdf/2404.12994v2 | http://arxiv.org/abs/2404.12994v2 | Clemencia Siro, Mohammad Aliannejadi, Maarten de Rijke | 2024-04-29T19:00:34Z | cs.IR, cs.CL | In ad-hoc retrieval, evaluation relies heavily on user actions, including
implicit feedback. In a conversational setting such signals are usually
unavailable due to the nature of the interactions, and, instead, the evaluation
often relies on crowdsourced evaluation labels. The role of user feedback in
annotators' assessment of turns in a conversational perception has been little
studied. We focus on how the evaluation of task-oriented dialogue systems
(TDSs), is affected by considering user feedback, explicit or implicit, as
provided through the follow-up utterance of a turn being evaluated. We explore
and compare two methodologies for assessing TDSs: one includes the user's
follow-up utterance and one without. We use both crowdworkers and large
language models (LLMs) as annotators to assess system responses across four
aspects: relevance, usefulness, interestingness, and explanation quality. Our
findings indicate that there is a distinct difference in ratings assigned by
both annotator groups in the two setups, indicating user feedback does
influence system evaluation. Workers are more susceptible to user feedback on
usefulness and interestingness compared to LLMs on interestingness and
relevance. User feedback leads to a more personalized assessment of usefulness
by workers, aligning closely with the user's explicit feedback. Additionally,
in cases of ambiguous or complex user requests, user feedback improves
agreement among crowdworkers. These findings emphasize the significance of user
feedback in refining system evaluations and suggest the potential for automated
feedback integration in future research. We publicly release the annotated data
to foster research in this area. | cs.IR | Rethinking the Evaluation of Dialogue Systems: Effects of User Feedback on Crowdworkers and LLMs |
2024-04-19T15:40:39Z | http://arxiv.org/pdf/2404.12957v1 | http://arxiv.org/abs/2404.12957v1 | Qinyuan Wu, Mohammad Aflah Khan, Soumi Das, Vedant Nanda, Bishwamittra Ghosh, Camila Kolling, Till Speicher, Laurent Bindschaedler, Krishna P. Gummadi, Evimaria Terzi | 2024-04-19T15:40:39Z | cs.CL, cs.LG | We propose an approach for estimating the latent knowledge embedded inside
large language models (LLMs). We leverage the in-context learning (ICL)
abilities of LLMs to estimate the extent to which an LLM knows the facts stored
in a knowledge base. Our knowledge estimator avoids reliability concerns with
previous prompting-based methods, is both conceptually simpler and easier to
apply, and we demonstrate that it can surface more of the latent knowledge
embedded in LLMs. We also investigate how different design choices affect the
performance of ICL-based knowledge estimation. Using the proposed estimator, we
perform a large-scale evaluation of the factual knowledge of a variety of open
source LLMs, like OPT, Pythia, Llama(2), Mistral, Gemma, etc. over a large set
of relations and facts from the Wikidata knowledge base. We observe differences
in the factual knowledge between different model families and models of
different sizes, that some relations are consistently better known than others
but that models differ in the precise facts they know, and differences in the
knowledge of base models and their finetuned counterparts. | cs.CL | Towards Reliable Latent Knowledge Estimation in LLMs: In-Context Learning vs. Prompting Based Factual Knowledge Extraction |
2024-04-19T15:08:06Z | http://arxiv.org/pdf/2404.12938v2 | http://arxiv.org/abs/2404.12938v2 | Oana Ignat, Xiaomeng Xu, Rada Mihalcea | 2024-06-19T03:34:42Z | cs.CL, cs.AI | Deceptive reviews are becoming increasingly common, especially given the
increase in performance and the prevalence of LLMs. While work to date has
addressed the development of models to differentiate between truthful and
deceptive human reviews, much less is known about the distinction between real
reviews and AI-authored fake reviews. Moreover, most of the research so far has
focused primarily on English, with very little work dedicated to other
languages. In this paper, we compile and make publicly available the MAiDE-up
dataset, consisting of 10,000 real and 10,000 AI-generated fake hotel reviews,
balanced across ten languages. Using this dataset, we conduct extensive
linguistic analyses to (1) compare the AI fake hotel reviews to real hotel
reviews, and (2) identify the factors that influence the deception detection
model performance. We explore the effectiveness of several models for deception
detection in hotel reviews across three main dimensions: sentiment, location,
and language. We find that these dimensions influence how well we can detect
AI-generated fake reviews. | cs.CL | MAiDE-up: Multilingual Deception Detection of GPT-generated Hotel Reviews |
2024-04-19T15:04:30Z | http://arxiv.org/pdf/2404.12933v2 | http://arxiv.org/abs/2404.12933v2 | Oana Ignat, Gayathri Ganesh Lakshmy, Rada Mihalcea | 2024-06-19T03:27:43Z | cs.CL, cs.AI | Inspiration is linked to various positive outcomes, such as increased
creativity, productivity, and happiness. Although inspiration has great
potential, there has been limited effort toward identifying content that is
inspiring, as opposed to just engaging or positive. Additionally, most research
has concentrated on Western data, with little attention paid to other cultures.
This work is the first to study cross-cultural inspiration through machine
learning methods. We aim to identify and analyze real and AI-generated
cross-cultural inspiring posts. To this end, we compile and make publicly
available the InspAIred dataset, which consists of 2,000 real inspiring posts,
2,000 real non-inspiring posts, and 2,000 generated inspiring posts evenly
distributed across India and the UK. The real posts are sourced from Reddit,
while the generated posts are created using the GPT-4 model. Using this
dataset, we conduct extensive computational linguistic analyses to (1) compare
inspiring content across cultures, (2) compare AI-generated inspiring posts to
real inspiring posts, and (3) determine if detection models can accurately
distinguish between inspiring content across cultures and data sources. | cs.CL | Cross-cultural Inspiration Detection and Analysis in Real and LLM-generated Social Media Data |
2024-04-19T14:05:03Z | http://arxiv.org/pdf/2404.12897v2 | http://arxiv.org/abs/2404.12897v2 | Ahmed Elshabrawy, Yongxin Huang, Iryna Gurevych, Alham Fikri Aji | 2024-04-22T10:17:07Z | cs.CL | While Large Language Models (LLMs) exhibit remarkable capabilities in
zero-shot and few-shot scenarios, they often require computationally
prohibitive sizes. Conversely, smaller Masked Language Models (MLMs) like BERT
and RoBERTa achieve state-of-the-art results through fine-tuning but struggle
with extending to few-shot and zero-shot settings due to their architectural
constraints. Hence, we propose Statement-Tuning, a technique that models
discriminative tasks as a set of finite statements and trains an Encoder model
to discriminate between the potential statements to determine the label. We do
Statement-Tuning on multiple tasks to enable cross-task generalization.
Experimental results demonstrate that Statement Tuning achieves competitive
performance compared to state-of-the-art LLMs with significantly fewer
parameters. Moreover, the study investigates the impact of several design
choices on few-shot and zero-shot generalization, revealing that Statement
Tuning can achieve sufficient performance with modest training data and
benefits from task and statement diversity for unseen task generalizability. | cs.CL | Enabling Natural Zero-Shot Prompting on Encoder Models via Statement-Tuning |
2024-04-19T13:27:38Z | http://arxiv.org/pdf/2404.12879v1 | http://arxiv.org/abs/2404.12879v1 | Guanhua Chen, Wenhan Yu, Lei Sha | 2024-04-19T13:27:38Z | cs.CL | While Retrieval-Augmented Generation (RAG) plays a crucial role in the
application of Large Language Models (LLMs), existing retrieval methods in
knowledge-dense domains like law and medicine still suffer from a lack of
multi-perspective views, which are essential for improving interpretability and
reliability. Previous research on multi-view retrieval often focused solely on
different semantic forms of queries, neglecting the expression of specific
domain knowledge perspectives. This paper introduces a novel multi-view RAG
framework, MVRAG, tailored for knowledge-dense domains that utilizes
intention-aware query rewriting from multiple domain viewpoints to enhance
retrieval precision, thereby improving the effectiveness of the final
inference. Experiments conducted on legal and medical case retrieval
demonstrate significant improvements in recall and precision rates with our
framework. Our multi-perspective retrieval approach unleashes the potential of
multi-view information enhancing RAG tasks, accelerating the further
application of LLMs in knowledge-intensive fields. | cs.CL | Unlocking Multi-View Insights in Knowledge-Dense Retrieval-Augmented Generation |
2024-04-19T13:17:07Z | http://arxiv.org/pdf/2404.12872v1 | http://arxiv.org/abs/2404.12872v1 | Zhaodonghui Li, Haitao Yuan, Huiming Wang, Gao Cong, Lidong Bing | 2024-04-19T13:17:07Z | cs.DB, cs.CL | Query rewrite, which aims to generate more efficient queries by altering a
SQL query's structure without changing the query result, has been an important
research problem. In order to maintain equivalence between the rewritten query
and the original one during rewriting, traditional query rewrite methods always
rewrite the queries following certain rewrite rules. However, some problems
still remain. Firstly, existing methods of finding the optimal choice or
sequence of rewrite rules are still limited and the process always costs a lot
of resources. Methods involving discovering new rewrite rules typically require
complicated proofs of structural logic or extensive user interactions.
Secondly, current query rewrite methods usually rely highly on DBMS cost
estimators which are often not accurate. In this paper, we address these
problems by proposing a novel method of query rewrite named LLM-R2, adopting a
large language model (LLM) to propose possible rewrite rules for a database
rewrite system. To further improve the inference ability of LLM in recommending
rewrite rules, we train a contrastive model by curriculum to learn query
representations and select effective query demonstrations for the LLM.
Experimental results have shown that our method can significantly improve the
query execution efficiency and outperform the baseline methods. In addition,
our method enjoys high robustness across different datasets. | cs.DB | LLM-R2: A Large Language Model Enhanced Rule-based Rewrite System for Boosting Query Efficiency |
2024-04-19T13:05:37Z | http://arxiv.org/pdf/2404.12866v1 | http://arxiv.org/abs/2404.12866v1 | Yang Luo, Zangwei Zheng, Zirui Zhu, Yang You | 2024-04-19T13:05:37Z | cs.CL, cs.AI, cs.CV | The increase in parameter size of multimodal large language models (MLLMs)
introduces significant capabilities, particularly in-context learning, where
MLLMs enhance task performance without updating pre-trained parameters. This
effectiveness, however, hinges on the appropriate selection of in-context
examples, a process that is currently biased towards visual data, overlooking
textual information. Furthermore, the area of supervised retrievers for MLLMs,
crucial for optimal in-context example selection, continues to be
uninvestigated. Our study offers an in-depth evaluation of the impact of
textual information on the unsupervised selection of in-context examples in
multimodal contexts, uncovering a notable sensitivity of retriever performance
to the employed modalities. Responding to this, we introduce a novel supervised
MLLM-retriever MSIER that employs a neural network to select examples that
enhance multimodal in-context learning efficiency. This approach is validated
through extensive testing across three distinct tasks, demonstrating the
method's effectiveness. Additionally, we investigate the influence of
modalities on our supervised retrieval method's training and pinpoint factors
contributing to our model's success. This exploration paves the way for future
advancements, highlighting the potential for refined in-context learning in
MLLMs through the strategic use of multimodal data. | cs.CL | How Does the Textual Information Affect the Retrieval of Multimodal In-Context Learning? |
2024-04-19T12:47:56Z | http://arxiv.org/pdf/2404.13104v1 | http://arxiv.org/abs/2404.13104v1 | Muhammad Osama Nusrat, Waseem Shahzad, Saad Ahmed Jamal | 2024-04-19T12:47:56Z | cs.CL, cs.AI | Depression is a significant issue nowadays. As per the World Health
Organization (WHO), in 2023, over 280 million individuals are grappling with
depression. This is a huge number; if not taken seriously, these numbers will
increase rapidly. About 4.89 billion individuals are social media users. People
express their feelings and emotions on platforms like Twitter, Facebook,
Reddit, Instagram, etc. These platforms contain valuable information which can
be used for research purposes. Considerable research has been conducted across
various social media platforms. However, certain limitations persist in these
endeavors. Particularly, previous studies were only focused on detecting
depression and the intensity of depression in tweets. Also, there existed
inaccuracies in dataset labeling. In this research work, five types of
depression (Bipolar, major, psychotic, atypical, and postpartum) were predicted
using tweets from the Twitter database based on lexicon labeling. Explainable
AI was used to provide reasoning by highlighting the parts of tweets that
represent type of depression. Bidirectional Encoder Representations from
Transformers (BERT) was used for feature extraction and training. Machine
learning and deep learning methodologies were used to train the model. The BERT
model presented the most promising results, achieving an overall accuracy of
0.96. | cs.CL | Multi Class Depression Detection Through Tweets using Artificial Intelligence |
2024-04-19T12:26:28Z | http://arxiv.org/pdf/2404.12845v1 | http://arxiv.org/abs/2404.12845v1 | Aleksei Dorkin, Kairit Sirts | 2024-04-19T12:26:28Z | cs.CL | We present our submission to the unconstrained subtask of the SIGTYP 2024
Shared Task on Word Embedding Evaluation for Ancient and Historical Languages
for morphological annotation, POS-tagging, lemmatization, character- and
word-level gap-filling. We developed a simple, uniform, and computationally
lightweight approach based on the adapters framework using parameter-efficient
fine-tuning. We applied the same adapter-based approach uniformly to all tasks
and 16 languages by fine-tuning stacked language- and task-specific adapters.
Our submission obtained an overall second place out of three submissions, with
the first place in word-level gap-filling. Our results show the feasibility of
adapting language models pre-trained on modern languages to historical and
ancient languages via adapter training. | cs.CL | TartuNLP @ SIGTYP 2024 Shared Task: Adapting XLM-RoBERTa for Ancient and Historical Languages |
2024-04-19T12:23:57Z | http://arxiv.org/pdf/2404.12843v1 | http://arxiv.org/abs/2404.12843v1 | Diego Calanzone, Stefano Teso, Antonio Vergari | 2024-04-19T12:23:57Z | cs.LG, cs.CL | Large language models (LLMs) are a promising venue for natural language
understanding and generation tasks. However, current LLMs are far from
reliable: they are prone to generate non-factual information and, more
crucially, to contradict themselves when prompted to reason about beliefs of
the world. These problems are currently addressed with large scale fine-tuning
or by delegating consistent reasoning to external tools. In this work, we
strive for a middle ground and introduce a training objective based on
principled probabilistic reasoning that teaches a LLM to be consistent with
external knowledge in the form of a set of facts and rules. Fine-tuning with
our loss on a limited set of facts enables our LLMs to be more logically
consistent than previous baselines and allows them to extrapolate to unseen but
semantically similar factual knowledge more systematically. | cs.LG | Towards Logically Consistent Language Models via Probabilistic Reasoning |
2024-04-19T12:06:28Z | http://arxiv.org/pdf/2404.12829v1 | http://arxiv.org/abs/2404.12829v1 | Alessandra Bassani, Beatrice Del Bo, Alfio Ferrara, Marta Mangini, Sergio Picascia, Ambra Stefanello | 2024-04-19T12:06:28Z | cs.CL | The Latin language has received attention from the computational linguistics
research community, which has built, over the years, several valuable
resources, ranging from detailed annotated corpora to sophisticated tools for
linguistic analysis. With the recent advent of large language models,
researchers have also started developing models capable of generating vector
representations of Latin texts. The performances of such models remain behind
the ones for modern languages, given the disparity in available data. In this
paper, we present the LiMe dataset, a corpus of 325 documents extracted from a
series of medieval manuscripts called Libri sententiarum potestatis Mediolani,
and thoroughly annotated by experts, in order to be employed for masked
language model, as well as supervised natural language processing tasks. | cs.CL | LiMe: a Latin Corpus of Late Medieval Criminal Sentences |
2024-04-19T12:04:32Z | http://arxiv.org/pdf/2404.12827v1 | http://arxiv.org/abs/2404.12827v1 | Anthony Yazdani, Alban Bornet, Boya Zhang, Philipp Khlebnikov, Poorya Amini, Douglas Teodoro | 2024-04-19T12:04:32Z | cs.CL | Adverse drug events (ADEs) significantly impact clinical research and public
health, contributing to failures in clinical trials and leading to increased
healthcare costs. The accurate prediction and management of ADEs are crucial
for improving the development of safer, more effective medications, and
enhancing patient outcomes. To support this effort, we introduce CT-ADE, a
novel dataset compiled to enhance the predictive modeling of ADEs. Encompassing
over 12,000 instances extracted from clinical trial results, the CT-ADE dataset
integrates drug, patient population, and contextual information for multilabel
ADE classification tasks in monopharmacy treatments, providing a comprehensive
resource for developing advanced predictive models. To mirror the complex
nature of ADEs, annotations are standardized at the system organ class level of
the Medical Dictionary for Regulatory Activities (MedDRA) ontology. Preliminary
analyses using baseline models have demonstrated promising results, achieving
73.33% F1 score and 81.54% balanced accuracy, highlighting CT-ADE's potential
to advance ADE prediction. CT-ADE provides an essential tool for researchers
aiming to leverage the power of artificial intelligence and machine learning to
enhance patient safety and minimize the impact of ADEs on pharmaceutical
research and development. Researchers interested in using the CT-ADE dataset
can find all necessary resources at https://github.com/xxxx/xxxx. | cs.CL | CT-ADE: An Evaluation Benchmark for Adverse Drug Event Prediction from Clinical Trial Results |
Subsets and Splits
No saved queries yet
Save your SQL queries to embed, download, and access them later. Queries will appear here once saved.