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
|
|---|---|---|---|---|---|---|---|---|
2023-09-12T19:32:45Z | http://arxiv.org/pdf/2309.06541v1 | http://arxiv.org/abs/2309.06541v1 | Joseph Gatto, Omar Sharif, Parker Seegmiller, Philip Bohlman, Sarah Masud Preum | 2023-09-12T19:32:45Z | cs.CL | Amidst the sharp rise in the evaluation of large language models (LLMs) on
various tasks, we find that semantic textual similarity (STS) has been
under-explored. In this study, we show that STS can be cast as a text
generation problem while maintaining strong performance on multiple STS
benchmarks. Additionally, we show generative LLMs significantly outperform
existing encoder-based STS models when characterizing the semantic similarity
between two texts with complex semantic relationships dependent on world
knowledge. We validate this claim by evaluating both generative LLMs and
existing encoder-based STS models on three newly collected STS challenge sets
which require world knowledge in the domains of Health, Politics, and Sports.
All newly collected data is sourced from social media content posted after May
2023 to ensure the performance of closed-source models like ChatGPT cannot be
credited to memorization. Our results show that, on average, generative LLMs
outperform the best encoder-only baselines by an average of 22.3% on STS tasks
requiring world knowledge. Our results suggest generative language models with
STS-specific prompting strategies achieve state-of-the-art performance in
complex, domain-specific STS tasks. | cs.CL | Text Encoders Lack Knowledge: Leveraging Generative LLMs for Domain-Specific Semantic Textual Similarity |
2023-09-12T19:23:40Z | http://arxiv.org/pdf/2309.15827v1 | http://arxiv.org/abs/2309.15827v1 | Angela Schöpke-Gonzalez, Siqi Wu, Sagar Kumar, Paul J. Resnick, Libby Hemphill | 2023-09-12T19:23:40Z | cs.CL, cs.CY, K.4.1 | Computational social science research has made advances in machine learning
and natural language processing that support content moderators in detecting
harmful content. These advances often rely on training datasets annotated by
crowdworkers for harmful content. In designing instructions for annotation
tasks to generate training data for these algorithms, researchers often treat
the harm concepts that we train algorithms to detect - 'hateful', 'offensive',
'toxic', 'racist', 'sexist', etc. - as interchangeable. In this work, we
studied whether the way that researchers define 'harm' affects annotation
outcomes. Using Venn diagrams, information gain comparisons, and content
analyses, we reveal that annotators do not use the concepts 'hateful',
'offensive', and 'toxic' interchangeably. We identify that features of harm
definitions and annotators' individual characteristics explain much of how
annotators use these terms differently. Our results offer empirical evidence
discouraging the common practice of using harm concepts interchangeably in
content moderation research. Instead, researchers should make specific choices
about which harm concepts to analyze based on their research goals. Recognizing
that researchers are often resource constrained, we also encourage researchers
to provide information to bound their findings when their concepts of interest
differ from concepts that off-the-shelf harmful content detection algorithms
identify. Finally, we encourage algorithm providers to ensure their instruments
can adapt to contextually-specific content detection goals (e.g., soliciting
instrument users' feedback). | cs.CL | How We Define Harm Impacts Data Annotations: Explaining How Annotators Distinguish Hateful, Offensive, and Toxic Comments |
2023-09-12T18:51:10Z | http://arxiv.org/pdf/2309.06520v2 | http://arxiv.org/abs/2309.06520v2 | Vyas Raina, Mark Gales | 2023-10-27T14:42:29Z | cs.CL, cs.AI | For sequence-to-sequence tasks it is challenging to combine individual system
outputs. Further, there is also often a mismatch between the decoding criterion
and the one used for assessment. Minimum Bayes' Risk (MBR) decoding can be used
to combine system outputs in a manner that encourages better alignment with the
final assessment criterion. This paper examines MBR decoding for Grammatical
Error Correction (GEC) systems, where performance is usually evaluated in terms
of edits and an associated F-score. Hence, we propose a novel MBR loss function
directly linked to this form of criterion. Furthermore, an approach to expand
the possible set of candidate sentences is described. This builds on a current
max-voting combination scheme, as well as individual edit-level selection.
Experiments on three popular GEC datasets and with state-of-the-art GEC systems
demonstrate the efficacy of the proposed MBR approach. Additionally, the paper
highlights how varying reward metrics within the MBR decoding framework can
provide control over precision, recall, and the F-score in combined GEC
systems. | cs.CL | Minimum Bayes' Risk Decoding for System Combination of Grammatical Error Correction Systems |
2023-09-12T18:47:29Z | http://arxiv.org/pdf/2309.06517v1 | http://arxiv.org/abs/2309.06517v1 | Shreyash Mishra, S Suryavardan, Megha Chakraborty, Parth Patwa, Anku Rani, Aman Chadha, Aishwarya Reganti, Amitava Das, Amit Sheth, Manoj Chinnakotla, Asif Ekbal, Srijan Kumar | 2023-09-12T18:47:29Z | cs.CL | Analyzing memes on the internet has emerged as a crucial endeavor due to the
impact this multi-modal form of content wields in shaping online discourse.
Memes have become a powerful tool for expressing emotions and sentiments,
possibly even spreading hate and misinformation, through humor and sarcasm. In
this paper, we present the overview of the Memotion 3 shared task, as part of
the DeFactify 2 workshop at AAAI-23. The task released an annotated dataset of
Hindi-English code-mixed memes based on their Sentiment (Task A), Emotion (Task
B), and Emotion intensity (Task C). Each of these is defined as an individual
task and the participants are ranked separately for each task. Over 50 teams
registered for the shared task and 5 made final submissions to the test set of
the Memotion 3 dataset. CLIP, BERT modifications, ViT etc. were the most
popular models among the participants along with approaches such as
Student-Teacher model, Fusion, and Ensembling. The best final F1 score for Task
A is 34.41, Task B is 79.77 and Task C is 59.82. | cs.CL | Overview of Memotion 3: Sentiment and Emotion Analysis of Codemixed Hinglish Memes |
2023-09-12T18:18:23Z | http://arxiv.org/pdf/2309.06503v1 | http://arxiv.org/abs/2309.06503v1 | Ramya Tekumalla, Juan M. Banda | 2023-09-12T18:18:23Z | cs.CL, cs.LG, cs.SI | The COVID-19 pandemic has presented significant challenges to the healthcare
industry and society as a whole. With the rapid development of COVID-19
vaccines, social media platforms have become a popular medium for discussions
on vaccine-related topics. Identifying vaccine-related tweets and analyzing
them can provide valuable insights for public health research-ers and
policymakers. However, manual annotation of a large number of tweets is
time-consuming and expensive. In this study, we evaluate the usage of Large
Language Models, in this case GPT-4 (March 23 version), and weak supervision,
to identify COVID-19 vaccine-related tweets, with the purpose of comparing
performance against human annotators. We leveraged a manu-ally curated
gold-standard dataset and used GPT-4 to provide labels without any additional
fine-tuning or instructing, in a single-shot mode (no additional prompting). | cs.CL | Leveraging Large Language Models and Weak Supervision for Social Media data annotation: an evaluation using COVID-19 self-reported vaccination tweets |
2023-09-12T18:03:55Z | http://arxiv.org/pdf/2309.06490v1 | http://arxiv.org/abs/2309.06490v1 | Sarah E. Finch, Ellie S. Paek, Jinho D. Choi | 2023-09-12T18:03:55Z | cs.CL | Developing high-performing dialogue systems benefits from the automatic
identification of undesirable behaviors in system responses. However, detecting
such behaviors remains challenging, as it draws on a breadth of general
knowledge and understanding of conversational practices. Although recent
research has focused on building specialized classifiers for detecting specific
dialogue behaviors, the behavior coverage is still incomplete and there is a
lack of testing on real-world human-bot interactions. This paper investigates
the ability of a state-of-the-art large language model (LLM), ChatGPT-3.5, to
perform dialogue behavior detection for nine categories in real human-bot
dialogues. We aim to assess whether ChatGPT can match specialized models and
approximate human performance, thereby reducing the cost of behavior detection
tasks. Our findings reveal that neither specialized models nor ChatGPT have yet
achieved satisfactory results for this task, falling short of human
performance. Nevertheless, ChatGPT shows promising potential and often
outperforms specialized detection models. We conclude with an in-depth
examination of the prevalent shortcomings of ChatGPT, offering guidance for
future research to enhance LLM capabilities. | cs.CL | Leveraging Large Language Models for Automated Dialogue Analysis |
2023-09-12T17:44:40Z | http://arxiv.org/pdf/2309.06460v1 | http://arxiv.org/abs/2309.06460v1 | Lydia Feng, Gregor Williamson, Han He, Jinho D. Choi | 2023-09-12T17:44:40Z | cs.CL | This paper presents a novel semantic representation, WISeR, that overcomes
challenges for Abstract Meaning Representation (AMR). Despite its strengths,
AMR is not easily applied to languages or domains without predefined semantic
frames, and its use of numbered arguments results in semantic role labels,
which are not directly interpretable and are semantically overloaded for
parsers. We examine the numbered arguments of predicates in AMR and convert
them to thematic roles that do not require reference to semantic frames. We
create a new corpus of 1K English dialogue sentences annotated in both WISeR
and AMR. WISeR shows stronger inter-annotator agreement for beginner and
experienced annotators, with beginners becoming proficient in WISeR annotation
more quickly. Finally, we train a state-of-the-art parser on the AMR 3.0 corpus
and a WISeR corpus converted from AMR 3.0. The parser is evaluated on these
corpora and our dialogue corpus. The WISeR model exhibits higher accuracy than
its AMR counterpart across the board, demonstrating that WISeR is easier for
parsers to learn. | cs.CL | Widely Interpretable Semantic Representation: Frameless Meaning Representation for Broader Applicability |
2023-09-12T17:01:02Z | http://arxiv.org/pdf/2309.07172v1 | http://arxiv.org/abs/2309.07172v1 | Yuan He, Jiaoyan Chen, Hang Dong, Ian Horrocks | 2023-09-12T17:01:02Z | cs.AI, cs.CL, cs.LG | This work investigates the applicability of recent generative Large Language
Models (LLMs), such as the GPT series and Flan-T5, to ontology alignment for
identifying concept equivalence mappings across ontologies. To test the
zero-shot performance of Flan-T5-XXL and GPT-3.5-turbo, we leverage challenging
subsets from two equivalence matching datasets of the OAEI Bio-ML track, taking
into account concept labels and structural contexts. Preliminary findings
suggest that LLMs have the potential to outperform existing ontology alignment
systems like BERTMap, given careful framework and prompt design. | cs.AI | Exploring Large Language Models for Ontology Alignment |
2023-09-12T16:39:41Z | http://arxiv.org/pdf/2309.08628v3 | http://arxiv.org/abs/2309.08628v3 | Arpita Vats, Zhe Liu, Peng Su, Debjyoti Paul, Yingyi Ma, Yutong Pang, Zeeshan Ahmed, Ozlem Kalinli | 2023-12-14T00:45:39Z | cs.CL, cs.CR, cs.LG | Model adaptation is crucial to handle the discrepancy between proxy training
data and actual users data received. To effectively perform adaptation, textual
data of users is typically stored on servers or their local devices, where
downstream natural language processing (NLP) models can be directly trained
using such in-domain data. However, this might raise privacy and security
concerns due to the extra risks of exposing user information to adversaries.
Replacing identifying information in textual data with a generic marker has
been recently explored. In this work, we leverage large language models (LLMs)
to suggest substitutes of masked tokens and have their effectiveness evaluated
on downstream language modeling tasks. Specifically, we propose multiple
pre-trained and fine-tuned LLM-based approaches and perform empirical studies
on various datasets for the comparison of these methods. Experimental results
show that models trained on the obfuscation corpora are able to achieve
comparable performance with the ones trained on the original data without
privacy-preserving token masking. | cs.CL | Recovering from Privacy-Preserving Masking with Large Language Models |
2023-09-12T16:28:36Z | http://arxiv.org/pdf/2309.06365v2 | http://arxiv.org/abs/2309.06365v2 | Xiangci Li, Yi-Hui Lee, Jessica Ouyang | 2024-02-20T23:31:22Z | cs.CL | An automatic citation generation system aims to concisely and accurately
describe the relationship between two scientific articles. To do so, such a
system must ground its outputs to the content of the cited paper to avoid
non-factual hallucinations. Due to the length of scientific documents, existing
abstractive approaches have conditioned only on cited paper abstracts. We
demonstrate empirically that the abstract is not always the most appropriate
input for citation generation and that models trained in this way learn to
hallucinate. We propose to condition instead on the cited text span (CTS) as an
alternative to the abstract. Because manual CTS annotation is extremely time-
and labor-intensive, we experiment with distant labeling of candidate CTS
sentences, achieving sufficiently strong performance to substitute for
expensive human annotations in model training, and we propose a
human-in-the-loop, keyword-based CTS retrieval approach that makes generating
citation texts grounded in the full text of cited papers both promising and
practical. | cs.CL | Cited Text Spans for Citation Text Generation |
2023-09-12T16:27:18Z | http://arxiv.org/pdf/2309.06363v1 | http://arxiv.org/abs/2309.06363v1 | Tianhui Zhang, Danushka Bollegala, Bei Peng | 2023-09-12T16:27:18Z | cs.CL | Prior work has shown that the ordering in which concepts are shown to a
commonsense generator plays an important role, affecting the quality of the
generated sentence. However, it remains a challenge to determine the optimal
ordering of a given set of concepts such that a natural sentence covering all
the concepts could be generated from a pretrained generator. To understand the
relationship between the ordering of the input concepts and the quality of the
generated sentences, we conduct a systematic study considering multiple
language models (LMs) and concept ordering strategies. We find that BART-large
model consistently outperforms all other LMs considered in this study when
fine-tuned using the ordering of concepts as they appear in CommonGen training
data as measured using multiple evaluation metrics. Moreover, the larger
GPT3-based large language models (LLMs) variants do not necessarily outperform
much smaller LMs on this task, even when fine-tuned on task-specific training
data. Interestingly, human annotators significantly reorder input concept sets
when manually writing sentences covering those concepts, and this ordering
provides the best sentence generations independently of the LM used for the
generation, outperforming a probabilistic concept ordering baseline | cs.CL | Learning to Predict Concept Ordering for Common Sense Generation |
2023-09-12T14:36:23Z | http://arxiv.org/pdf/2309.06275v2 | http://arxiv.org/abs/2309.06275v2 | Xiaohan Xu, Chongyang Tao, Tao Shen, Can Xu, Hongbo Xu, Guodong Long, Jian-guang Lou | 2024-02-29T06:28:55Z | cs.CL | To enhance the reasoning capabilities of off-the-shelf Large Language Models
(LLMs), we introduce a simple, yet general and effective prompting method, Re2,
i.e., \textbf{Re}-\textbf{Re}ading the question as input. Unlike most
thought-eliciting prompting methods, such as Chain-of-Thought (CoT), which aim
to elicit the reasoning process in the output, Re2 shifts the focus to the
input by processing questions twice, thereby enhancing the understanding
process. Consequently, Re2 demonstrates strong generality and compatibility
with most thought-eliciting prompting methods, including CoT. Crucially, Re2
facilitates a "bidirectional" encoding in unidirectional decoder-only LLMs
because the first pass could provide global information for the second pass. We
begin with a preliminary empirical study as the foundation of Re2, illustrating
its potential to enable "bidirectional" attention mechanisms. We then evaluate
Re2 on extensive reasoning benchmarks across 14 datasets, spanning 112
experiments, to validate its effectiveness and generality. Our findings
indicate that, with the exception of a few scenarios on vanilla ChatGPT, Re2
consistently enhances the reasoning performance of LLMs through a simple
re-reading strategy. Further analyses reveal Re2's adaptability, showing how it
can be effectively integrated with different LLMs, thought-eliciting prompting,
and ensemble strategies. Our code is available at
\url{https://github.com/Tebmer/Rereading-LLM-Reasoning/} | cs.CL | Re-Reading Improves Reasoning in Large Language Models |
2023-09-12T13:51:29Z | http://arxiv.org/pdf/2309.06236v1 | http://arxiv.org/abs/2309.06236v1 | Dimitris Spathis, Fahim Kawsar | 2023-09-12T13:51:29Z | cs.LG, cs.CL | Large Language Models (LLMs) have demonstrated remarkable generalization
across diverse tasks, leading individuals to increasingly use them as personal
assistants and universal computing engines. Nevertheless, a notable obstacle
emerges when feeding numerical/temporal data into these models, such as data
sourced from wearables or electronic health records. LLMs employ tokenizers in
their input that break down text into smaller units. However, tokenizers are
not designed to represent numerical values and might struggle to understand
repetitive patterns and context, treating consecutive values as separate tokens
and disregarding their temporal relationships. Here, we discuss recent works
that employ LLMs for human-centric tasks such as in mobile health sensing and
present a case study showing that popular LLMs tokenize temporal data
incorrectly. To address that, we highlight potential solutions such as prompt
tuning with lightweight embedding layers as well as multimodal adapters, that
can help bridge this "modality gap". While the capability of language models to
generalize to other modalities with minimal or no finetuning is exciting, this
paper underscores the fact that their outputs cannot be meaningful if they
stumble over input nuances. | cs.LG | The first step is the hardest: Pitfalls of Representing and Tokenizing Temporal Data for Large Language Models |
2023-09-12T13:38:44Z | http://arxiv.org/pdf/2309.06219v3 | http://arxiv.org/abs/2309.06219v3 | Oana Ignat, Santiago Castro, Weiji Li, Rada Mihalcea | 2024-06-19T03:18:41Z | cs.CV, cs.CL, cs.CY, cs.IR | We introduce the task of automatic human action co-occurrence identification,
i.e., determine whether two human actions can co-occur in the same interval of
time. We create and make publicly available the ACE (Action Co-occurrencE)
dataset, consisting of a large graph of ~12k co-occurring pairs of visual
actions and their corresponding video clips. We describe graph link prediction
models that leverage visual and textual information to automatically infer if
two actions are co-occurring. We show that graphs are particularly well suited
to capture relations between human actions, and the learned graph
representations are effective for our task and capture novel and relevant
information across different data domains. The ACE dataset and the code
introduced in this paper are publicly available at
https://github.com/MichiganNLP/vlog_action_co-occurrence. | cs.CV | Human Action Co-occurrence in Lifestyle Vlogs using Graph Link Prediction |
2023-09-12T13:30:25Z | http://arxiv.org/pdf/2309.08627v1 | http://arxiv.org/abs/2309.08627v1 | Charu James, Mayank Nagda, Nooshin Haji Ghassemi, Marius Kloft, Sophie Fellenz | 2023-09-12T13:30:25Z | cs.CL, cs.IR, cs.LG | There is a lack of quantitative measures to evaluate the progression of
topics through time in dynamic topic models (DTMs). Filling this gap, we
propose a novel evaluation measure for DTMs that analyzes the changes in the
quality of each topic over time. Additionally, we propose an extension
combining topic quality with the model's temporal consistency. We demonstrate
the utility of the proposed measure by applying it to synthetic data and data
from existing DTMs. We also conducted a human evaluation, which indicates that
the proposed measure correlates well with human judgment. Our findings may help
in identifying changing topics, evaluating different DTMs, and guiding future
research in this area. | cs.CL | Evaluating Dynamic Topic Models |
2023-09-12T13:01:20Z | http://arxiv.org/pdf/2309.06192v2 | http://arxiv.org/abs/2309.06192v2 | Alessandra Polimeno, Myrthe Reuver, Sanne Vrijenhoek, Antske Fokkens | 2023-09-18T13:05:08Z | cs.CL, cs.CY, cs.IR | News recommender systems play an increasingly influential role in shaping
information access within democratic societies. However, tailoring
recommendations to users' specific interests can result in the divergence of
information streams. Fragmented access to information poses challenges to the
integrity of the public sphere, thereby influencing democracy and public
discourse. The Fragmentation metric quantifies the degree of fragmentation of
information streams in news recommendations. Accurate measurement of this
metric requires the application of Natural Language Processing (NLP) to
identify distinct news events, stories, or timelines. This paper presents an
extensive investigation of various approaches for quantifying Fragmentation in
news recommendations. These approaches are evaluated both intrinsically, by
measuring performance on news story clustering, and extrinsically, by assessing
the Fragmentation scores of different simulated news recommender scenarios. Our
findings demonstrate that agglomerative hierarchical clustering coupled with
SentenceBERT text representation is substantially better at detecting
Fragmentation than earlier implementations. Additionally, the analysis of
simulated scenarios yields valuable insights and recommendations for
stakeholders concerning the measurement and interpretation of Fragmentation. | cs.CL | Improving and Evaluating the Detection of Fragmentation in News Recommendations with the Clustering of News Story Chains |
2023-09-12T12:46:20Z | http://arxiv.org/pdf/2309.06179v1 | http://arxiv.org/abs/2309.06179v1 | Shoutao Guo, Shaolei Zhang, Yang Feng | 2023-09-12T12:46:20Z | cs.CL, cs.AI | Simultaneous machine translation (SiMT) outputs translation while reading the
source sentence. Unlike conventional sequence-to-sequence (seq2seq) training,
existing SiMT methods adopt the prefix-to-prefix (prefix2prefix) training,
where the model predicts target tokens based on partial source tokens. However,
the prefix2prefix training diminishes the ability of the model to capture
global information and introduces forced predictions due to the absence of
essential source information. Consequently, it is crucial to bridge the gap
between the prefix2prefix training and seq2seq training to enhance the
translation capability of the SiMT model. In this paper, we propose a novel
method that glances future in curriculum learning to achieve the transition
from the seq2seq training to prefix2prefix training. Specifically, we gradually
reduce the available source information from the whole sentence to the prefix
corresponding to that latency. Our method is applicable to a wide range of SiMT
methods and experiments demonstrate that our method outperforms strong
baselines. | cs.CL | Glancing Future for Simultaneous Machine Translation |
2023-09-12T12:37:37Z | http://arxiv.org/pdf/2309.06175v2 | http://arxiv.org/abs/2309.06175v2 | Di Lu, Zhongping Liang, Caixia Yuan, Xiaojie Wang | 2023-09-13T03:53:43Z | cs.CL, cs.IR | This paper presents a novel approach to address the Entity Recognition and
Linking Challenge at NLPCC 2015. The task involves extracting named entity
mentions from short search queries and linking them to entities within a
reference Chinese knowledge base. To tackle this problem, we first expand the
existing knowledge base and utilize external knowledge to identify candidate
entities, thereby improving the recall rate. Next, we extract features from the
candidate entities and utilize Support Vector Regression and Multiple Additive
Regression Tree as scoring functions to filter the results. Additionally, we
apply rules to further refine the results and enhance precision. Our method is
computationally efficient and achieves an F1 score of 0.535. | cs.CL | AKEM: Aligning Knowledge Base to Queries with Ensemble Model for Entity Recognition and Linking |
2023-09-12T12:18:18Z | http://arxiv.org/pdf/2309.06163v1 | http://arxiv.org/abs/2309.06163v1 | Luis Chiruzzo, Marvin Agüero-Torales, Gustavo Giménez-Lugo, Aldo Alvarez, Yliana Rodríguez, Santiago Góngora, Thamar Solorio | 2023-09-12T12:18:18Z | cs.CL | We present the first shared task for detecting and analyzing code-switching
in Guarani and Spanish, GUA-SPA at IberLEF 2023. The challenge consisted of
three tasks: identifying the language of a token, NER, and a novel task of
classifying the way a Spanish span is used in the code-switched context. We
annotated a corpus of 1500 texts extracted from news articles and tweets,
around 25 thousand tokens, with the information for the tasks. Three teams took
part in the evaluation phase, obtaining in general good results for Task 1, and
more mixed results for Tasks 2 and 3. | cs.CL | Overview of GUA-SPA at IberLEF 2023: Guarani-Spanish Code Switching Analysis |
2023-09-12T11:19:36Z | http://arxiv.org/pdf/2309.06135v2 | http://arxiv.org/abs/2309.06135v2 | Zhi-Yi Chin, Chieh-Ming Jiang, Ching-Chun Huang, Pin-Yu Chen, Wei-Chen Chiu | 2024-06-08T05:30:26Z | cs.CL, cs.CV | Text-to-image diffusion models, e.g. Stable Diffusion (SD), lately have shown
remarkable ability in high-quality content generation, and become one of the
representatives for the recent wave of transformative AI. Nevertheless, such
advance comes with an intensifying concern about the misuse of this generative
technology, especially for producing copyrighted or NSFW (i.e. not safe for
work) images. Although efforts have been made to filter inappropriate
images/prompts or remove undesirable concepts/styles via model fine-tuning, the
reliability of these safety mechanisms against diversified problematic prompts
remains largely unexplored. In this work, we propose Prompting4Debugging (P4D)
as a debugging and red-teaming tool that automatically finds problematic
prompts for diffusion models to test the reliability of a deployed safety
mechanism. We demonstrate the efficacy of our P4D tool in uncovering new
vulnerabilities of SD models with safety mechanisms. Particularly, our result
shows that around half of prompts in existing safe prompting benchmarks which
were originally considered "safe" can actually be manipulated to bypass many
deployed safety mechanisms, including concept removal, negative prompt, and
safety guidance. Our findings suggest that, without comprehensive testing, the
evaluations on limited safe prompting benchmarks can lead to a false sense of
safety for text-to-image models. | cs.CL | Prompting4Debugging: Red-Teaming Text-to-Image Diffusion Models by Finding Problematic Prompts |
2023-09-12T11:18:29Z | http://arxiv.org/pdf/2309.06132v2 | http://arxiv.org/abs/2309.06132v2 | Benjamin Icard, Vincent Claveau, Ghislain Atemezing, Paul Égré | 2023-10-23T18:16:19Z | cs.CL, cs.AI, 68T07, 68T50 | We present a hybrid approach to the automated measurement of vagueness and
subjectivity in texts. We first introduce the expert system VAGO, we illustrate
it on a small benchmark of fact vs. opinion sentences, and then test it on the
larger French press corpus FreSaDa to confirm the higher prevalence of
subjective markers in satirical vs. regular texts. We then build a neural clone
of VAGO, based on a BERT-like architecture, trained on the symbolic VAGO scores
obtained on FreSaDa. Using explainability tools (LIME), we show the interest of
this neural version for the enrichment of the lexicons of the symbolic version,
and for the production of versions in other languages. | cs.CL | Measuring vagueness and subjectivity in texts: from symbolic to neural VAGO |
2023-09-12T11:17:42Z | http://arxiv.org/pdf/2309.06131v1 | http://arxiv.org/abs/2309.06131v1 | Sophia Althammer, Guido Zuccon, Sebastian Hofstätter, Suzan Verberne, Allan Hanbury | 2023-09-12T11:17:42Z | cs.IR, cs.CL | Search methods based on Pretrained Language Models (PLM) have demonstrated
great effectiveness gains compared to statistical and early neural ranking
models. However, fine-tuning PLM-based rankers requires a great amount of
annotated training data. Annotating data involves a large manual effort and
thus is expensive, especially in domain specific tasks. In this paper we
investigate fine-tuning PLM-based rankers under limited training data and
budget. We investigate two scenarios: fine-tuning a ranker from scratch, and
domain adaptation starting with a ranker already fine-tuned on general data,
and continuing fine-tuning on a target dataset. We observe a great variability
in effectiveness when fine-tuning on different randomly selected subsets of
training data. This suggests that it is possible to achieve effectiveness gains
by actively selecting a subset of the training data that has the most positive
effect on the rankers. This way, it would be possible to fine-tune effective
PLM rankers at a reduced annotation budget. To investigate this, we adapt
existing Active Learning (AL) strategies to the task of fine-tuning PLM rankers
and investigate their effectiveness, also considering annotation and
computational costs. Our extensive analysis shows that AL strategies do not
significantly outperform random selection of training subsets in terms of
effectiveness. We further find that gains provided by AL strategies come at the
expense of more assessments (thus higher annotation costs) and AL strategies
underperform random selection when comparing effectiveness given a fixed
annotation cost. Our results highlight that ``optimal'' subsets of training
data that provide high effectiveness at low annotation cost do exist, but
current mainstream AL strategies applied to PLM rankers are not capable of
identifying them. | cs.IR | Annotating Data for Fine-Tuning a Neural Ranker? Current Active Learning Strategies are not Better than Random Selection |
2023-09-12T11:10:30Z | http://arxiv.org/pdf/2310.05960v1 | http://arxiv.org/abs/2310.05960v1 | Qiongkai Xu, Trevor Cohn, Olga Ohrimenko | 2023-09-12T11:10:30Z | cs.CR, cs.AI, cs.CL, cs.LG | Federated Learning allows collaborative training without data sharing in
settings where participants do not trust the central server and one another.
Privacy can be further improved by ensuring that communication between the
participants and the server is anonymized through a shuffle; decoupling the
participant identity from their data. This paper seeks to examine whether such
a defense is adequate to guarantee anonymity, by proposing a novel
fingerprinting attack over gradients sent by the participants to the server. We
show that clustering of gradients can easily break the anonymization in an
empirical study of learning federated language models on two language corpora.
We then show that training with differential privacy can provide a practical
defense against our fingerprint attack. | cs.CR | Fingerprint Attack: Client De-Anonymization in Federated Learning |
2023-09-12T11:02:27Z | http://arxiv.org/pdf/2309.06126v1 | http://arxiv.org/abs/2309.06126v1 | Tuan Dung Nguyen, Yuan-Sen Ting, Ioana Ciucă, Charlie O'Neill, Ze-Chang Sun, Maja Jabłońska, Sandor Kruk, Ernest Perkowski, Jack Miller, Jason Li, Josh Peek, Kartheik Iyer, Tomasz Różański, Pranav Khetarpal, Sharaf Zaman, David Brodrick, Sergio J. Rodríguez Méndez, Thang Bui, Alyssa Goodman, Alberto Accomazzi, Jill Naiman, Jesse Cranney, Kevin Schawinski, UniverseTBD | 2023-09-12T11:02:27Z | astro-ph.IM, astro-ph.CO, astro-ph.GA, astro-ph.HE, cs.CL, cs.LG | Large language models excel in many human-language tasks but often falter in
highly specialized domains like scholarly astronomy. To bridge this gap, we
introduce AstroLLaMA, a 7-billion-parameter model fine-tuned from LLaMA-2 using
over 300,000 astronomy abstracts from arXiv. Optimized for traditional causal
language modeling, AstroLLaMA achieves a 30% lower perplexity than Llama-2,
showing marked domain adaptation. Our model generates more insightful and
scientifically relevant text completions and embedding extraction than
state-of-the-arts foundation models despite having significantly fewer
parameters. AstroLLaMA serves as a robust, domain-specific model with broad
fine-tuning potential. Its public release aims to spur astronomy-focused
research, including automatic paper summarization and conversational agent
development. | astro-ph.IM | AstroLLaMA: Towards Specialized Foundation Models in Astronomy |
2023-09-12T10:27:39Z | http://arxiv.org/pdf/2309.06112v1 | http://arxiv.org/abs/2309.06112v1 | Sharath Srivatsa, Srinath Srinivasa | 2023-09-12T10:27:39Z | cs.CL, cs.IR | Media houses reporting on public figures, often come with their own biases
stemming from their respective worldviews. A characterization of these
underlying patterns helps us in better understanding and interpreting news
stories. For this, we need diverse or subjective summarizations, which may not
be amenable for classifying into predefined class labels. This work proposes a
zero-shot approach for non-extractive or generative characterizations of person
entities from a corpus using GPT-2. We use well-articulated articles from
several well-known news media houses as a corpus to build a sound argument for
this approach. First, we fine-tune a GPT-2 pre-trained language model with a
corpus where specific person entities are characterized. Second, we further
fine-tune this with demonstrations of person entity characterizations, created
from a corpus of programmatically constructed characterizations. This twice
fine-tuned model is primed with manual prompts consisting of entity names that
were not previously encountered in the second fine-tuning, to generate a simple
sentence about the entity. The results were encouraging, when compared against
actual characterizations from the corpus. | cs.CL | Characterizing Latent Perspectives of Media Houses Towards Public Figures |
2023-09-12T10:17:28Z | http://arxiv.org/pdf/2309.06105v1 | http://arxiv.org/abs/2309.06105v1 | Tinghui Zhu, Jingping Liu, Jiaqing Liang, Haiyun Jiang, Yanghua Xiao, Zongyu Wang, Rui Xie, Yunsen Xian | 2023-09-12T10:17:28Z | cs.CV, cs.CL | Taxonomy expansion task is essential in organizing the ever-increasing volume
of new concepts into existing taxonomies. Most existing methods focus
exclusively on using textual semantics, leading to an inability to generalize
to unseen terms and the "Prototypical Hypernym Problem." In this paper, we
propose Visual Taxonomy Expansion (VTE), introducing visual features into the
taxonomy expansion task. We propose a textual hypernymy learning task and a
visual prototype learning task to cluster textual and visual semantics. In
addition to the tasks on respective modalities, we introduce a hyper-proto
constraint that integrates textual and visual semantics to produce fine-grained
visual semantics. Our method is evaluated on two datasets, where we obtain
compelling results. Specifically, on the Chinese taxonomy dataset, our method
significantly improves accuracy by 8.75 %. Additionally, our approach performs
better than ChatGPT on the Chinese taxonomy dataset. | cs.CV | Towards Visual Taxonomy Expansion |
2023-09-12T09:37:08Z | http://arxiv.org/pdf/2309.06089v2 | http://arxiv.org/abs/2309.06089v2 | Boshko Koloski, Blaž Škrlj, Marko Robnik-Šikonja, Senja Pollak | 2024-04-15T12:53:46Z | cs.CL, cs.LG | The cross-lingual transfer is a promising technique to solve tasks in
less-resourced languages. In this empirical study, we compare two fine-tuning
approaches combined with zero-shot and full-shot learning approaches for large
language models in a cross-lingual setting. As fine-tuning strategies, we
compare parameter-efficient adapter methods with fine-tuning of all parameters.
As cross-lingual transfer strategies, we compare the intermediate-training
(\textit{IT}) that uses each language sequentially and cross-lingual validation
(\textit{CLV}) that uses a target language already in the validation phase of
fine-tuning. We assess the success of transfer and the extent of catastrophic
forgetting in a source language due to cross-lingual transfer, i.e., how much
previously acquired knowledge is lost when we learn new information in a
different language. The results on two different classification problems, hate
speech detection and product reviews, each containing datasets in several
languages, show that the \textit{IT} cross-lingual strategy outperforms
\textit{CLV} for the target language. Our findings indicate that, in the
majority of cases, the \textit{CLV} strategy demonstrates superior retention of
knowledge in the base language (English) compared to the \textit{IT} strategy,
when evaluating catastrophic forgetting in multiple cross-lingual transfers. | cs.CL | Measuring Catastrophic Forgetting in Cross-Lingual Transfer Paradigms: Exploring Tuning Strategies |
2023-09-12T09:31:25Z | http://arxiv.org/pdf/2309.06085v2 | http://arxiv.org/abs/2309.06085v2 | Wei Qi Leong, Jian Gang Ngui, Yosephine Susanto, Hamsawardhini Rengarajan, Kengatharaiyer Sarveswaran, William Chandra Tjhi | 2023-09-19T03:44:17Z | cs.CL | The rapid development of Large Language Models (LLMs) and the emergence of
novel abilities with scale have necessitated the construction of holistic,
diverse and challenging benchmarks such as HELM and BIG-bench. However, at the
moment, most of these benchmarks focus only on performance in English and
evaluations that include Southeast Asian (SEA) languages are few in number. We
therefore propose BHASA, a holistic linguistic and cultural evaluation suite
for LLMs in SEA languages. It comprises three components: (1) a NLP benchmark
covering eight tasks across Natural Language Understanding (NLU), Generation
(NLG) and Reasoning (NLR) tasks, (2) LINDSEA, a linguistic diagnostic toolkit
that spans the gamut of linguistic phenomena including syntax, semantics and
pragmatics, and (3) a cultural diagnostics dataset that probes for both
cultural representation and sensitivity. For this preliminary effort, we
implement the NLP benchmark only for Indonesian, Vietnamese, Thai and Tamil,
and we only include Indonesian and Tamil for LINDSEA and the cultural
diagnostics dataset. As GPT-4 is purportedly one of the best-performing
multilingual LLMs at the moment, we use it as a yardstick to gauge the
capabilities of LLMs in the context of SEA languages. Our initial experiments
on GPT-4 with BHASA find it lacking in various aspects of linguistic
capabilities, cultural representation and sensitivity in the targeted SEA
languages. BHASA is a work in progress and will continue to be improved and
expanded in the future. The repository for this paper can be found at:
https://github.com/aisingapore/BHASA | cs.CL | BHASA: A Holistic Southeast Asian Linguistic and Cultural Evaluation Suite for Large Language Models |
2023-09-12T08:52:56Z | http://arxiv.org/pdf/2309.06057v1 | http://arxiv.org/abs/2309.06057v1 | Weishi Wang, Yue Wang, Shafiq Joty, Steven C. H. Hoi | 2023-09-12T08:52:56Z | cs.SE, cs.CL | Automatic program repair (APR) is crucial to reduce manual debugging efforts
for developers and improve software reliability. While conventional
search-based techniques typically rely on heuristic rules or a redundancy
assumption to mine fix patterns, recent years have witnessed the surge of deep
learning (DL) based approaches to automate the program repair process in a
data-driven manner. However, their performance is often limited by a fixed set
of parameters to model the highly complex search space of APR. To ease such
burden on the parametric models, in this work, we propose a novel
Retrieval-Augmented Patch Generation framework (RAP-Gen) by explicitly
leveraging relevant fix patterns retrieved from a codebase of previous bug-fix
pairs. Specifically, we build a hybrid patch retriever to account for both
lexical and semantic matching based on the raw source code in a
language-agnostic manner, which does not rely on any code-specific features. In
addition, we adapt a code-aware language model CodeT5 as our foundation model
to facilitate both patch retrieval and generation tasks in a unified manner. We
adopt a stage-wise approach where the patch retriever first retrieves a
relevant external bug-fix pair to augment the buggy input for the CodeT5 patch
generator, which synthesizes a ranked list of repair patch candidates. Notably,
RAP-Gen is a generic APR framework that can flexibly integrate different patch
retrievers and generators to repair various types of bugs. We thoroughly
evaluate RAP-Gen on three benchmarks in two programming languages, including
the TFix benchmark in JavaScript, and Code Refinement and Defects4J benchmarks
in Java, where the bug localization information may or may not be provided.
Experimental results show that RAP-Gen significantly outperforms previous
state-of-the-art approaches on all benchmarks, e.g., repairing 15 more bugs on
818 Defects4J bugs. | cs.SE | RAP-Gen: Retrieval-Augmented Patch Generation with CodeT5 for Automatic Program Repair |
2023-09-12T08:45:25Z | http://arxiv.org/pdf/2309.06054v3 | http://arxiv.org/abs/2309.06054v3 | Jingwen Fu, Tao Yang, Yuwang Wang, Yan Lu, Nanning Zheng | 2024-06-06T15:09:52Z | cs.LG, cs.CL, cs.CV | In-context learning, i.e., learning from context examples, is an impressive
ability of Transformer. Training Transformers to possess this in-context
learning skill is computationally intensive due to the occurrence of learning
plateaus, which are periods within the training process where there is minimal
or no enhancement in the model's in-context learning capability. To study the
mechanism behind the learning plateaus, we conceptually seperate a component
within the model's internal representation that is exclusively affected by the
model's weights. We call this the "weights component", and the remainder is
identified as the "context component". By conducting meticulous and controlled
experiments on synthetic tasks, we note that the persistence of learning
plateaus correlates with compromised functionality of the weights component.
Recognizing the impaired performance of the weights component as a fundamental
behavior drives learning plateaus, we have developed three strategies to
expedite the learning of Transformers. The effectiveness of these strategies is
further confirmed in natural language processing tasks. In conclusion, our
research demonstrates the feasibility of cultivating a powerful in-context
learning ability within AI systems in an eco-friendly manner. | cs.LG | Breaking through the learning plateaus of in-context learning in Transformer |
2023-09-12T08:16:58Z | http://arxiv.org/pdf/2309.06453v2 | http://arxiv.org/abs/2309.06453v2 | Mingxin Li, Richong Zhang, Zhijie Nie, Yongyi Mao | 2023-12-19T12:13:25Z | cs.CL, cs.LG | Sentence Representation Learning (SRL) is a fundamental task in Natural
Language Processing (NLP), with the Contrastive Learning of Sentence Embeddings
(CSE) being the mainstream technique due to its superior performance. An
intriguing phenomenon in CSE is the significant performance gap between
supervised and unsupervised methods, with their only difference lying in the
training data. Previous works attribute this performance gap to differences in
two representation properties (alignment and uniformity). However, since
alignment and uniformity only measure the results, they fail to answer "What
aspects of the training data contribute to the performance gap?" and "How can
the performance gap be narrowed?", In this paper, we conduct empirical
experiments to answer these "What" and "How" questions. We first answer the
"What" question by thoroughly comparing the behavior of supervised and
unsupervised CSE during their respective training processes. From the
comparison, we identify the similarity pattern as a key factor to the
performance gap, and introduce a metric, called Relative Fitting Difficulty
(RFD), to measure the complexity of the similarity pattern. Then, based on the
insights gained from the "What" question, we tackle the "How" question by
increasing the pattern complexity of the training data. We achieve this by
leveraging the In-Context Learning (ICL) capability of the Large Language Model
(LLM) to generate data that simulates complex patterns. By utilizing the
hierarchical patterns in the LLM-generated data, we effectively narrow the gap
between supervised and unsupervised CSE. We release our codes and appendix at
https://github.com/BDBC-KG-NLP/NGCSE. | cs.CL | Narrowing the Gap between Supervised and Unsupervised Sentence Representation Learning with Large Language Model |
2023-09-12T07:08:22Z | http://arxiv.org/pdf/2309.06009v1 | http://arxiv.org/abs/2309.06009v1 | Shaoxiong Ji, Wei Sun, Pekka Marttinen | 2023-09-12T07:08:22Z | cs.CL | Many computational linguistic methods have been proposed to study the
information content of languages. We consider two interesting research
questions: 1) how is information distributed over long documents, and 2) how
does content reduction, such as token selection and text summarization, affect
the information density in long documents. We present four criteria for
information density estimation for long documents, including surprisal,
entropy, uniform information density, and lexical density. Among those
criteria, the first three adopt the measures from information theory. We
propose an attention-based word selection method for clinical notes and study
machine summarization for multiple-domain documents. Our findings reveal the
systematic difference in information density of long text in various domains.
Empirical results on automated medical coding from long clinical notes show the
effectiveness of the attention-based word selection method. | cs.CL | Content Reduction, Surprisal and Information Density Estimation for Long Documents |
2023-09-12T06:58:18Z | http://arxiv.org/pdf/2309.07927v3 | http://arxiv.org/abs/2309.07927v3 | Ahmed Adel Attia, Jing Liu, Wei Ai, Dorottya Demszky, Carol Espy-Wilson | 2024-05-15T07:05:32Z | eess.AS, cs.CL, cs.SD | Recent advancements in Automatic Speech Recognition (ASR) systems,
exemplified by Whisper, have demonstrated the potential of these systems to
approach human-level performance given sufficient data. However, this progress
doesn't readily extend to ASR for children due to the limited availability of
suitable child-specific databases and the distinct characteristics of
children's speech. A recent study investigated leveraging the My Science Tutor
(MyST) children's speech corpus to enhance Whisper's performance in recognizing
children's speech. They were able to demonstrate some improvement on a limited
testset. This paper builds on these findings by enhancing the utility of the
MyST dataset through more efficient data preprocessing. We reduce the Word
Error Rate (WER) on the MyST testset 13.93% to 9.11% with Whisper-Small and
from 13.23% to 8.61% with Whisper-Medium and show that this improvement can be
generalized to unseen datasets. We also highlight important challenges towards
improving children's ASR performance. The results showcase the viable and
efficient integration of Whisper for effective children's speech recognition. | eess.AS | Kid-Whisper: Towards Bridging the Performance Gap in Automatic Speech Recognition for Children VS. Adults |
2023-09-12T06:05:57Z | http://arxiv.org/pdf/2309.08626v1 | http://arxiv.org/abs/2309.08626v1 | Juntae Kim, Minkyu Lim, Seokjin Hong | 2023-09-12T06:05:57Z | cs.CL | Inverse text normalization (ITN) is crucial for converting spoken-form into
written-form, especially in the context of automatic speech recognition (ASR).
While most downstream tasks of ASR rely on written-form, ASR systems often
output spoken-form, highlighting the necessity for robust ITN in product-level
ASR-based applications. Although neural ITN methods have shown promise, they
still encounter performance challenges, particularly when dealing with
ASR-generated spoken text. These challenges arise from the out-of-domain
problem between training data and ASR-generated text. To address this, we
propose a direct training approach that utilizes ASR-generated written or
spoken text, with pairs augmented through ASR linguistic context emulation and
a semi-supervised learning method enhanced by a large language model,
respectively. Additionally, we introduce a post-aligning method to manage
unpredictable errors, thereby enhancing the reliability of ITN. Our experiments
show that our proposed methods remarkably improved ITN performance in various
ASR scenarios. | cs.CL | Improving Robustness of Neural Inverse Text Normalization via Data-Augmentation, Semi-Supervised Learning, and Post-Aligning Method |
2023-09-12T05:54:45Z | http://arxiv.org/pdf/2309.08625v1 | http://arxiv.org/abs/2309.08625v1 | Myriam Safrai, Amos Azaria | 2023-09-12T05:54:45Z | cs.CL | As Large Language Models (LLMs) are predictive models building their response
based on the words in the prompts, there is a risk that small talk and
irrelevant information may alter the response and the suggestion given.
Therefore, this study aims to investigate the impact of medical data mixed with
small talk on the accuracy of medical advice provided by ChatGPT. USMLE step 3
questions were used as a model for relevant medical data. We use both multiple
choice and open ended questions. We gathered small talk sentences from human
participants using the Mechanical Turk platform. Both sets of USLME questions
were arranged in a pattern where each sentence from the original questions was
followed by a small talk sentence. ChatGPT 3.5 and 4 were asked to answer both
sets of questions with and without the small talk sentences. A board-certified
physician analyzed the answers by ChatGPT and compared them to the formal
correct answer. The analysis results demonstrate that the ability of
ChatGPT-3.5 to answer correctly was impaired when small talk was added to
medical data for multiple-choice questions (72.1\% vs. 68.9\%) and open
questions (61.5\% vs. 44.3\%; p=0.01), respectively. In contrast, small talk
phrases did not impair ChatGPT-4 ability in both types of questions (83.6\% and
66.2\%, respectively). According to these results, ChatGPT-4 seems more
accurate than the earlier 3.5 version, and it appears that small talk does not
impair its capability to provide medical recommendations. Our results are an
important first step in understanding the potential and limitations of
utilizing ChatGPT and other LLMs for physician-patient interactions, which
include casual conversations. | cs.CL | Performance of ChatGPT-3.5 and GPT-4 on the United States Medical Licensing Examination With and Without Distractions |
2023-09-12T05:51:56Z | http://arxiv.org/pdf/2309.05973v2 | http://arxiv.org/abs/2309.05973v2 | Maximilian Li, Xander Davies, Max Nadeau | 2024-01-29T22:19:12Z | cs.CL, cs.LG | Language models often exhibit behaviors that improve performance on a
pre-training objective but harm performance on downstream tasks. We propose a
novel approach to removing undesirable behaviors by ablating a small number of
causal pathways between model components, with the intention of disabling the
computational circuit responsible for the bad behavior. Given a small dataset
of inputs where the model behaves poorly, we learn to ablate a small number of
important causal pathways. In the setting of reducing GPT-2 toxic language
generation, we find ablating just 12 of the 11.6K causal edges mitigates toxic
generation with minimal degradation of performance on other inputs. | cs.CL | Circuit Breaking: Removing Model Behaviors with Targeted Ablation |
2023-09-12T05:03:28Z | http://arxiv.org/pdf/2309.05961v5 | http://arxiv.org/abs/2309.05961v5 | Rima Hazra, Agnik Saha, Somnath Banerjee, Animesh Mukherjee | 2024-03-15T18:07:46Z | cs.SI, cs.CL, cs.IR, cs.LG | Community Question Answering (CQA) platforms steadily gain popularity as they
provide users with fast responses to their queries. The swiftness of these
responses is contingent on a mixture of query-specific and user-related
elements. This paper scrutinizes these contributing factors within the context
of six highly popular CQA platforms, identified through their standout
answering speed. Our investigation reveals a correlation between the time taken
to yield the first response to a question and several variables: the metadata,
the formulation of the questions, and the level of interaction among users.
Additionally, by employing conventional machine learning models to analyze
these metadata and patterns of user interaction, we endeavor to predict which
queries will receive their initial responses promptly. | cs.SI | Evaluating the Ebb and Flow: An In-depth Analysis of Question-Answering Trends across Diverse Platforms |
2023-09-12T04:49:39Z | http://arxiv.org/pdf/2309.05958v1 | http://arxiv.org/abs/2309.05958v1 | Kazuhiro Takemoto | 2023-09-12T04:49:39Z | cs.CL, cs.CY, cs.HC | As large language models (LLMs) become more deeply integrated into various
sectors, understanding how they make moral judgments has become crucial,
particularly in the realm of autonomous driving. This study utilized the Moral
Machine framework to investigate the ethical decision-making tendencies of
prominent LLMs, including GPT-3.5, GPT-4, PaLM 2, and Llama 2, comparing their
responses to human preferences. While LLMs' and humans' preferences such as
prioritizing humans over pets and favoring saving more lives are broadly
aligned, PaLM 2 and Llama 2, especially, evidence distinct deviations.
Additionally, despite the qualitative similarities between the LLM and human
preferences, there are significant quantitative disparities, suggesting that
LLMs might lean toward more uncompromising decisions, compared to the milder
inclinations of humans. These insights elucidate the ethical frameworks of LLMs
and their potential implications for autonomous driving. | cs.CL | The Moral Machine Experiment on Large Language Models |
2023-09-12T04:15:34Z | http://arxiv.org/pdf/2309.05951v1 | http://arxiv.org/abs/2309.05951v1 | Yan Jiang, Ruihong Qiu, Yi Zhang, Peng-Fei Zhang | 2023-09-12T04:15:34Z | cs.CL | As social media becomes increasingly popular, more and more public health
activities emerge, which is worth noting for pandemic monitoring and government
decision-making. Current techniques for public health analysis involve popular
models such as BERT and large language models (LLMs). Although recent progress
in LLMs has shown a strong ability to comprehend knowledge by being fine-tuned
on specific domain datasets, the costs of training an in-domain LLM for every
specific public health task are especially expensive. Furthermore, such kinds
of in-domain datasets from social media are generally highly imbalanced, which
will hinder the efficiency of LLMs tuning. To tackle these challenges, the data
imbalance issue can be overcome by sophisticated data augmentation methods for
social media datasets. In addition, the ability of the LLMs can be effectively
utilised by prompting the model properly. In light of the above discussion, in
this paper, a novel ALEX framework is proposed for social media analysis on
public health. Specifically, an augmentation pipeline is developed to resolve
the data imbalance issue. Furthermore, an LLMs explanation mechanism is
proposed by prompting an LLM with the predicted results from BERT models.
Extensive experiments conducted on three tasks at the Social Media Mining for
Health 2023 (SMM4H) competition with the first ranking in two tasks demonstrate
the superior performance of the proposed ALEX method. Our code has been
released in https://github.com/YanJiangJerry/ALEX. | cs.CL | Balanced and Explainable Social Media Analysis for Public Health with Large Language Models |
2023-09-12T04:03:41Z | http://arxiv.org/pdf/2309.05950v5 | http://arxiv.org/abs/2309.05950v5 | Shihong Liu, Zhiqiu Lin, Samuel Yu, Ryan Lee, Tiffany Ling, Deepak Pathak, Deva Ramanan | 2024-05-14T03:20:42Z | cs.CL, cs.CV, cs.LG, cs.MM | Vision-language models (VLMs) pre-trained on web-scale datasets have
demonstrated remarkable capabilities on downstream tasks when fine-tuned with
minimal data. However, many VLMs rely on proprietary data and are not
open-source, which restricts the use of white-box approaches for fine-tuning.
As such, we aim to develop a black-box approach to optimize VLMs through
natural language prompts, thereby avoiding the need to access model parameters,
feature embeddings, or even output logits. We propose employing chat-based LLMs
to search for the best text prompt for VLMs. Specifically, we adopt an
automatic hill-climbing procedure that converges to an effective prompt by
evaluating the performance of current prompts and asking LLMs to refine them
based on textual feedback, all within a conversational process without
human-in-the-loop. In a challenging 1-shot image classification setup, our
simple approach surpasses the white-box continuous prompting method (CoOp) by
an average of 1.5% across 11 datasets including ImageNet. Our approach also
outperforms both human-engineered and LLM-generated prompts. We highlight the
advantage of conversational feedback that incorporates both positive and
negative prompts, suggesting that LLMs can utilize the implicit gradient
direction in textual feedback for a more efficient search. In addition, we find
that the text prompts generated through our strategy are not only more
interpretable but also transfer well across different VLM architectures in a
black-box manner. Lastly, we apply our framework to optimize the
state-of-the-art black-box VLM (DALL-E 3) for text-to-image generation, prompt
inversion, and personalization. | cs.CL | Language Models as Black-Box Optimizers for Vision-Language Models |
2023-09-12T03:27:08Z | http://arxiv.org/pdf/2309.05938v2 | http://arxiv.org/abs/2309.05938v2 | Yufeng Zhang, Meng-xiang Wang, Jianxing Yu | 2023-10-06T12:35:20Z | cs.CL, cs.AI | This paper proposes a new task in the field of Answering Subjective Induction
Question on Products (SUBJPQA). The answer to this kind of question is
non-unique, but can be interpreted from many perspectives. For example, the
answer to 'whether the phone is heavy' has a variety of different viewpoints. A
satisfied answer should be able to summarize these subjective opinions from
multiple sources and provide objective knowledge, such as the weight of a
phone. That is quite different from the traditional QA task, in which the
answer to a factoid question is unique and can be found from a single data
source. To address this new task, we propose a three-steps method. We first
retrieve all answer-related clues from multiple knowledge sources on facts and
opinions. The implicit commonsense facts are also collected to supplement the
necessary but missing contexts. We then capture their relevance with the
questions by interactive attention. Next, we design a reinforcement-based
summarizer to aggregate all these knowledgeable clues. Based on a
template-controlled decoder, we can output a comprehensive and
multi-perspective answer. Due to the lack of a relevant evaluated benchmark set
for the new task, we construct a large-scale dataset, named SupQA, consisting
of 48,352 samples across 15 product domains. Evaluation results show the
effectiveness of our approach. | cs.CL | Answering Subjective Induction Questions on Products by Summarizing Multi-sources Multi-viewpoints Knowledge |
2023-09-12T03:20:50Z | http://arxiv.org/pdf/2309.05936v1 | http://arxiv.org/abs/2309.05936v1 | Weiqi Wu, Chengyue Jiang, Yong Jiang, Pengjun Xie, Kewei Tu | 2023-09-12T03:20:50Z | cs.CL | Ontological knowledge, which comprises classes and properties and their
relationships, is integral to world knowledge. It is significant to explore
whether Pretrained Language Models (PLMs) know and understand such knowledge.
However, existing PLM-probing studies focus mainly on factual knowledge,
lacking a systematic probing of ontological knowledge. In this paper, we focus
on probing whether PLMs store ontological knowledge and have a semantic
understanding of the knowledge rather than rote memorization of the surface
form. To probe whether PLMs know ontological knowledge, we investigate how well
PLMs memorize: (1) types of entities; (2) hierarchical relationships among
classes and properties, e.g., Person is a subclass of Animal and Member of
Sports Team is a subproperty of Member of ; (3) domain and range constraints of
properties, e.g., the subject of Member of Sports Team should be a Person and
the object should be a Sports Team. To further probe whether PLMs truly
understand ontological knowledge beyond memorization, we comprehensively study
whether they can reliably perform logical reasoning with given knowledge
according to ontological entailment rules. Our probing results show that PLMs
can memorize certain ontological knowledge and utilize implicit knowledge in
reasoning. However, both the memorizing and reasoning performances are less
than perfect, indicating incomplete knowledge and understanding. | cs.CL | Do PLMs Know and Understand Ontological Knowledge? |
2023-09-12T02:34:06Z | http://arxiv.org/pdf/2309.05922v1 | http://arxiv.org/abs/2309.05922v1 | Vipula Rawte, Amit Sheth, Amitava Das | 2023-09-12T02:34:06Z | cs.AI, cs.CL, cs.IR | Hallucination in a foundation model (FM) refers to the generation of content
that strays from factual reality or includes fabricated information. This
survey paper provides an extensive overview of recent efforts that aim to
identify, elucidate, and tackle the problem of hallucination, with a particular
focus on ``Large'' Foundation Models (LFMs). The paper classifies various types
of hallucination phenomena that are specific to LFMs and establishes evaluation
criteria for assessing the extent of hallucination. It also examines existing
strategies for mitigating hallucination in LFMs and discusses potential
directions for future research in this area. Essentially, the paper offers a
comprehensive examination of the challenges and solutions related to
hallucination in LFMs. | cs.AI | A Survey of Hallucination in Large Foundation Models |
2023-09-12T02:24:16Z | http://arxiv.org/pdf/2309.05920v1 | http://arxiv.org/abs/2309.05920v1 | Athanasios N. Nikolakopoulos, Swati Kaul, Siva Karthik Gade, Bella Dubrov, Umit Batur, Suleiman Ali Khan | 2023-09-12T02:24:16Z | cs.IR, cs.AI, cs.CL | We introduce SAGE; a Generative LLM for inferring attribute values for
products across world-wide e-Commerce catalogs. We introduce a novel
formulation of the attribute-value prediction problem as a Seq2Seq
summarization task, across languages, product types and target attributes. Our
novel modeling approach lifts the restriction of predicting attribute values
within a pre-specified set of choices, as well as, the requirement that the
sought attribute values need to be explicitly mentioned in the text. SAGE can
infer attribute values even when such values are mentioned implicitly using
periphrastic language, or not-at-all-as is the case for common-sense defaults.
Additionally, SAGE is capable of predicting whether an attribute is
inapplicable for the product at hand, or non-obtainable from the available
information. SAGE is the first method able to tackle all aspects of the
attribute-value-prediction task as they arise in practical settings in
e-Commerce catalogs. A comprehensive set of experiments demonstrates the
effectiveness of the proposed approach, as well as, its superiority against
state-of-the-art competing alternatives. Moreover, our experiments highlight
SAGE's ability to tackle the task of predicting attribute values in zero-shot
setting; thereby, opening up opportunities for significantly reducing the
overall number of labeled examples required for training. | cs.IR | SAGE: Structured Attribute Value Generation for Billion-Scale Product Catalogs |
2023-09-12T02:14:05Z | http://arxiv.org/pdf/2309.05918v3 | http://arxiv.org/abs/2309.05918v3 | Walid S. Saba | 2023-09-14T12:58:39Z | cs.CL, cs.AI | In our opinion the exuberance surrounding the relative success of data-driven
large language models (LLMs) is slightly misguided and for several reasons (i)
LLMs cannot be relied upon for factual information since for LLMs all ingested
text (factual or non-factual) was created equal; (ii) due to their subsymbolic
na-ture, whatever 'knowledge' these models acquire about language will always
be buried in billions of microfeatures (weights), none of which is meaningful
on its own; and (iii) LLMs will often fail to make the correct inferences in
several linguistic contexts (e.g., nominal compounds, copredication, quantifier
scope ambi-guities, intensional contexts. Since we believe the relative success
of data-driven large language models (LLMs) is not a reflection on the symbolic
vs. subsymbol-ic debate but a reflection on applying the successful strategy of
a bottom-up reverse engineering of language at scale, we suggest in this paper
applying the effective bottom-up strategy in a symbolic setting resulting in
symbolic, explainable, and ontologically grounded language models. | cs.CL | Stochastic LLMs do not Understand Language: Towards Symbolic, Explainable and Ontologically Based LLMs |
2023-09-11T22:24:39Z | http://arxiv.org/pdf/2309.08624v1 | http://arxiv.org/abs/2309.08624v1 | Vithya Yogarajan, Gillian Dobbie, Timothy Pistotti, Joshua Bensemann, Kobe Knowles | 2023-09-11T22:24:39Z | cs.CL, cs.AI | Recent advances in artificial intelligence, including the development of
highly sophisticated large language models (LLM), have proven beneficial in
many real-world applications. However, evidence of inherent bias encoded in
these LLMs has raised concerns about equity. In response, there has been an
increase in research dealing with bias, including studies focusing on
quantifying bias and developing debiasing techniques. Benchmark bias datasets
have also been developed for binary gender classification and ethical/racial
considerations, focusing predominantly on American demographics. However, there
is minimal research in understanding and quantifying bias related to
under-represented societies. Motivated by the lack of annotated datasets for
quantifying bias in under-represented societies, we endeavoured to create
benchmark datasets for the New Zealand (NZ) population. We faced many
challenges in this process, despite the availability of three annotators. This
research outlines the manual annotation process, provides an overview of the
challenges we encountered and lessons learnt, and presents recommendations for
future research. | cs.CL | Challenges in Annotating Datasets to Quantify Bias in Under-represented Society |
2023-09-11T22:11:46Z | http://arxiv.org/pdf/2309.07062v1 | http://arxiv.org/abs/2309.07062v1 | Chris Cummins, Volker Seeker, Dejan Grubisic, Mostafa Elhoushi, Youwei Liang, Baptiste Roziere, Jonas Gehring, Fabian Gloeckle, Kim Hazelwood, Gabriel Synnaeve, Hugh Leather | 2023-09-11T22:11:46Z | cs.PL, cs.AI, cs.CL, cs.LG | We explore the novel application of Large Language Models to code
optimization. We present a 7B-parameter transformer model trained from scratch
to optimize LLVM assembly for code size. The model takes as input unoptimized
assembly and outputs a list of compiler options to best optimize the program.
Crucially, during training, we ask the model to predict the instruction counts
before and after optimization, and the optimized code itself. These auxiliary
learning tasks significantly improve the optimization performance of the model
and improve the model's depth of understanding.
We evaluate on a large suite of test programs. Our approach achieves a 3.0%
improvement in reducing instruction counts over the compiler, outperforming two
state-of-the-art baselines that require thousands of compilations. Furthermore,
the model shows surprisingly strong code reasoning abilities, generating
compilable code 91% of the time and perfectly emulating the output of the
compiler 70% of the time. | cs.PL | Large Language Models for Compiler Optimization |
2023-09-11T21:24:00Z | http://arxiv.org/pdf/2309.05833v3 | http://arxiv.org/abs/2309.05833v3 | Dylan Zhang, Xuchao Zhang, Chetan Bansal, Pedro Las-Casas, Rodrigo Fonseca, Saravan Rajmohan | 2023-09-29T16:25:16Z | cs.CL, cs.AI, cs.LG, cs.SE | Major cloud providers have employed advanced AI-based solutions like large
language models to aid humans in identifying the root causes of cloud
incidents. Despite the growing prevalence of AI-driven assistants in the root
cause analysis process, their effectiveness in assisting on-call engineers is
constrained by low accuracy due to the intrinsic difficulty of the task, a
propensity for LLM-based approaches to hallucinate, and difficulties in
distinguishing these well-disguised hallucinations. To address this challenge,
we propose to perform confidence estimation for the predictions to help on-call
engineers make decisions on whether to adopt the model prediction. Considering
the black-box nature of many LLM-based root cause predictors, fine-tuning or
temperature-scaling-based approaches are inapplicable. We therefore design an
innovative confidence estimation framework based on prompting
retrieval-augmented large language models (LLMs) that demand a minimal amount
of information from the root cause predictor. This approach consists of two
scoring phases: the LLM-based confidence estimator first evaluates its
confidence in making judgments in the face of the current incident that
reflects its ``grounded-ness" level in reference data, then rates the root
cause prediction based on historical references. An optimization step combines
these two scores for a final confidence assignment. We show that our method is
able to produce calibrated confidence estimates for predicted root causes,
validate the usefulness of retrieved historical data and the prompting strategy
as well as the generalizability across different root cause prediction models.
Our study takes an important move towards reliably and effectively embedding
LLMs into cloud incident management systems. | cs.CL | PACE-LM: Prompting and Augmentation for Calibrated Confidence Estimation with GPT-4 in Cloud Incident Root Cause Analysis |
2023-09-11T20:16:38Z | http://arxiv.org/pdf/2309.05804v2 | http://arxiv.org/abs/2309.05804v2 | Abhisek Tiwari, Muhammed Sinan, Kaushik Roy, Amit Sheth, Sriparna Saha, Pushpak Bhattacharyya | 2024-05-29T18:17:12Z | cs.CL | Over the past two decades, dialogue modeling has made significant strides,
moving from simple rule-based responses to personalized and persuasive response
generation. However, despite these advancements, the objective functions and
evaluation metrics for dialogue generation have remained stagnant. These
lexical-based metrics, e.g., cross-entropy and BLEU, have two key limitations:
(a) word-to-word matching without semantic consideration: It assigns the same
credit for failure to generate "nice" and "rice" for "good", (b) missing
context attribute for evaluating the generated response: Even if a generated
response is relevant to the ongoing dialogue context, it may still be penalized
for not matching the gold utterance provided in the corpus. In this paper, we
first investigate these limitations comprehensively and propose a new loss
function called Semantic Infused Contextualized diaLogue (SemTextualLogue) loss
function. We also formulate an evaluation metric called Dialuation,
incorporating both context and semantic relevance. We experimented with both
non-pretrained and pre-trained models on two dialogue corpora, encompassing
task-oriented and open-domain scenarios. We found that the dialogue generation
models trained with SemTextualLogueloss attained superior performance compared
to the traditional cross-entropy loss function. The findings establish that the
effective training of a dialogue generation model hinges significantly on
incorporating semantics and context. This pattern is also mirrored in the
introduced Dialuation metric, where the consideration of both context and
semantics correlates more strongly with human evaluation compared to
traditional metrics. | cs.CL | Hi Model, generating 'nice' instead of 'good' is not as bad as generating 'rice'! Towards Context and Semantic Infused Dialogue Generation Loss Function and Evaluation Metric |
2023-09-11T18:05:12Z | http://arxiv.org/pdf/2309.13061v3 | http://arxiv.org/abs/2309.13061v3 | Armando D. Diaz Gonzalez, Kevin S. Hughes, Songhui Yue, Sean T. Hayes | 2024-04-22T20:26:25Z | cs.CL, cs.CY | Published biomedical information has and continues to rapidly increase. The
recent advancements in Natural Language Processing (NLP), have generated
considerable interest in automating the extraction, normalization, and
representation of biomedical knowledge about entities such as genes and
diseases. Our study analyzes germline abstracts in the construction of
knowledge graphs of the of the immense work that has been done in this area for
genes and diseases. This paper presents SimpleGermKG, an automatic knowledge
graph construction approach that connects germline genes and diseases. For the
extraction of genes and diseases, we employ BioBERT, a pre-trained BERT model
on biomedical corpora. We propose an ontology-based and rule-based algorithm to
standardize and disambiguate medical terms. For semantic relationships between
articles, genes, and diseases, we implemented a part-whole relation approach to
connect each entity with its data source and visualize them in a graph-based
knowledge representation. Lastly, we discuss the knowledge graph applications,
limitations, and challenges to inspire the future research of germline corpora.
Our knowledge graph contains 297 genes, 130 diseases, and 46,747 triples.
Graph-based visualizations are used to show the results. | cs.CL | Applying BioBERT to Extract Germline Gene-Disease Associations for Building a Knowledge Graph from the Biomedical Literature |
2023-09-11T17:56:57Z | http://arxiv.org/pdf/2309.05660v2 | http://arxiv.org/abs/2309.05660v2 | Ruocheng Wang, Eric Zelikman, Gabriel Poesia, Yewen Pu, Nick Haber, Noah D. Goodman | 2024-05-30T23:10:00Z | cs.LG, cs.AI, cs.CL | Inductive reasoning is a core problem-solving capacity: humans can identify
underlying principles from a few examples, which robustly generalize to novel
scenarios. Recent work evaluates large language models (LLMs) on inductive
reasoning tasks by directly prompting them yielding "in context learning." This
works well for straightforward inductive tasks but performs poorly on complex
tasks such as the Abstraction and Reasoning Corpus (ARC). In this work, we
propose to improve the inductive reasoning ability of LLMs by generating
explicit hypotheses at multiple levels of abstraction: we prompt the LLM to
propose multiple abstract hypotheses about the problem, in natural language,
then implement the natural language hypotheses as concrete Python programs.
These programs can be verified by running on observed examples and generalized
to novel inputs. To reduce the hypothesis search space, we explore steps to
filter the set of hypotheses to implement: we either ask the LLM to summarize
them into a smaller set of hypotheses or ask human annotators to select a
subset. We verify our pipeline's effectiveness on the ARC visual inductive
reasoning benchmark, its variant 1D-ARC, string transformation dataset SyGuS,
and list transformation dataset List Functions. On a random 100-problem subset
of ARC, our automated pipeline using LLM summaries achieves 30% accuracy,
outperforming the direct prompting baseline (accuracy of 17%). With the minimal
human input of selecting from LLM-generated candidates, performance is boosted
to 33%. Our ablations show that both abstract hypothesis generation and
concrete program representations benefit LLMs on inductive reasoning tasks. | cs.LG | Hypothesis Search: Inductive Reasoning with Language Models |
2023-09-11T17:49:27Z | http://arxiv.org/pdf/2309.05689v1 | http://arxiv.org/abs/2309.05689v1 | Qingxiu Dong, Li Dong, Ke Xu, Guangyan Zhou, Yaru Hao, Zhifang Sui, Furu Wei | 2023-09-11T17:49:27Z | cs.CL, cs.AI | In this work, we use large language models (LLMs) to augment and accelerate
research on the P versus NP problem, one of the most important open problems in
theoretical computer science and mathematics. Specifically, we propose Socratic
reasoning, a general framework that promotes in-depth thinking with LLMs for
complex problem-solving. Socratic reasoning encourages LLMs to recursively
discover, solve, and integrate problems while facilitating self-evaluation and
refinement. Our pilot study on the P vs. NP problem shows that GPT-4
successfully produces a proof schema and engages in rigorous reasoning
throughout 97 dialogue turns, concluding "P $\neq$ NP", which is in alignment
with (Xu and Zhou, 2023). The investigation uncovers novel insights within the
extensive solution space of LLMs, shedding light on LLM for Science. | cs.CL | Large Language Model for Science: A Study on P vs. NP |
2023-09-11T17:47:22Z | http://arxiv.org/pdf/2309.05653v3 | http://arxiv.org/abs/2309.05653v3 | Xiang Yue, Xingwei Qu, Ge Zhang, Yao Fu, Wenhao Huang, Huan Sun, Yu Su, Wenhu Chen | 2023-10-03T02:48:42Z | cs.CL | We introduce MAmmoTH, a series of open-source large language models (LLMs)
specifically tailored for general math problem-solving. The MAmmoTH models are
trained on MathInstruct, our meticulously curated instruction tuning dataset.
MathInstruct is compiled from 13 math datasets with intermediate rationales,
six of which have rationales newly curated by us. It presents a unique hybrid
of chain-of-thought (CoT) and program-of-thought (PoT) rationales, and also
ensures extensive coverage of diverse fields in math. The hybrid of CoT and PoT
not only unleashes the potential of tool use but also allows different thought
processes for different math problems. As a result, the MAmmoTH series
substantially outperform existing open-source models on nine mathematical
reasoning datasets across all scales with an average accuracy gain between 16%
and 32%. Remarkably, our MAmmoTH-7B model reaches 33% on MATH (a
competition-level dataset), which exceeds the best open-source 7B model
(WizardMath) by 23%, and the MAmmoTH-34B model achieves 44% accuracy on MATH,
even surpassing GPT-4's CoT result. Our work underscores the importance of
diverse problem coverage and the use of hybrid rationales in developing
superior math generalist models. | cs.CL | MAmmoTH: Building Math Generalist Models through Hybrid Instruction Tuning |
2023-09-11T17:07:01Z | http://arxiv.org/pdf/2309.05619v2 | http://arxiv.org/abs/2309.05619v2 | Wei Du, Laksh Advani, Yashmeet Gambhir, Daniel J Perry, Prashant Shiralkar, Zhengzheng Xing, Aaron Colak | 2023-11-20T01:24:40Z | cs.CL | Large language models (LLMs) have demonstrated significant capability to
generalize across a large number of NLP tasks. For industry applications, it is
imperative to assess the performance of the LLM on unlabeled production data
from time to time to validate for a real-world setting. Human labeling to
assess model error requires considerable expense and time delay. Here we
demonstrate that ensemble disagreement scores work well as a proxy for human
labeling for language models in zero-shot, few-shot, and fine-tuned settings,
per our evaluation on keyphrase extraction (KPE) task. We measure fidelity of
the results by comparing to true error measured from human labeled ground
truth. We contrast with the alternative of using another LLM as a source of
machine labels, or silver labels. Results across various languages and domains
show disagreement scores provide a better estimation of model performance with
mean average error (MAE) as low as 0.4% and on average 13.8% better than using
silver labels. | cs.CL | Effective Proxy for Human Labeling: Ensemble Disagreement Scores in Large Language Models for Industrial NLP |
2023-09-11T16:47:01Z | http://arxiv.org/pdf/2309.05608v1 | http://arxiv.org/abs/2309.05608v1 | Ruibo Chen, Zhiyuan Zhang, Yi Liu, Ruihan Bao, Keiko Harimoto, Xu Sun | 2023-09-11T16:47:01Z | cs.CL, cs.CE | Multimodal stock trading volume movement prediction with stock-related news
is one of the fundamental problems in the financial area. Existing multimodal
works that train models from scratch face the problem of lacking universal
knowledge when modeling financial news. In addition, the models ability may be
limited by the lack of domain-related knowledge due to insufficient data in the
datasets. To handle this issue, we propose the Prompt-based MUltimodal Stock
volumE prediction model (ProMUSE) to process text and time series modalities.
We use pre-trained language models for better comprehension of financial news
and adopt prompt learning methods to leverage their capability in universal
knowledge to model textual information. Besides, simply fusing two modalities
can cause harm to the unimodal representations. Thus, we propose a novel
cross-modality contrastive alignment while reserving the unimodal heads beside
the fusion head to mitigate this problem. Extensive experiments demonstrate
that our proposed ProMUSE outperforms existing baselines. Comprehensive
analyses further validate the effectiveness of our architecture compared to
potential variants and learning mechanisms. | cs.CL | Incorporating Pre-trained Model Prompting in Multimodal Stock Volume Movement Prediction |
2023-09-11T16:39:30Z | http://arxiv.org/pdf/2309.05605v3 | http://arxiv.org/abs/2309.05605v3 | Mansi Sakarvadia, Aswathy Ajith, Arham Khan, Daniel Grzenda, Nathaniel Hudson, André Bauer, Kyle Chard, Ian Foster | 2024-02-28T21:00:13Z | cs.CL, cs.AI, cs.LG | Answering multi-hop reasoning questions requires retrieving and synthesizing
information from diverse sources. Large Language Models (LLMs) struggle to
perform such reasoning consistently. Here we propose an approach to pinpoint
and rectify multi-hop reasoning failures through targeted memory injections on
LLM attention heads. First, we analyze the per-layer activations of GPT-2
models in response to single and multi-hop prompts. We then propose a mechanism
that allows users to inject pertinent prompt-specific information, which we
refer to as "memories," at critical LLM locations during inference. By thus
enabling the LLM to incorporate additional relevant information during
inference, we enhance the quality of multi-hop prompt completions. We show
empirically that a simple, efficient, and targeted memory injection into a key
attention layer can often increase the probability of the desired next token in
multi-hop tasks, by up to 424%. | cs.CL | Memory Injections: Correcting Multi-Hop Reasoning Failures during Inference in Transformer-Based Language Models |
2023-09-11T15:54:30Z | http://arxiv.org/pdf/2309.05569v1 | http://arxiv.org/abs/2309.05569v1 | Cheng Zhang, Xuanbai Chen, Siqi Chai, Chen Henry Wu, Dmitry Lagun, Thabo Beeler, Fernando De la Torre | 2023-09-11T15:54:30Z | cs.CV, cs.AI, cs.CL, cs.LG | Text-to-image generative models often reflect the biases of the training
data, leading to unequal representations of underrepresented groups. This study
investigates inclusive text-to-image generative models that generate images
based on human-written prompts and ensure the resulting images are uniformly
distributed across attributes of interest. Unfortunately, directly expressing
the desired attributes in the prompt often leads to sub-optimal results due to
linguistic ambiguity or model misrepresentation. Hence, this paper proposes a
drastically different approach that adheres to the maxim that "a picture is
worth a thousand words". We show that, for some attributes, images can
represent concepts more expressively than text. For instance, categories of
skin tones are typically hard to specify by text but can be easily represented
by example images. Building upon these insights, we propose a novel approach,
ITI-GEN, that leverages readily available reference images for Inclusive
Text-to-Image GENeration. The key idea is learning a set of prompt embeddings
to generate images that can effectively represent all desired attribute
categories. More importantly, ITI-GEN requires no model fine-tuning, making it
computationally efficient to augment existing text-to-image models. Extensive
experiments demonstrate that ITI-GEN largely improves over state-of-the-art
models to generate inclusive images from a prompt. Project page:
https://czhang0528.github.io/iti-gen. | cs.CV | ITI-GEN: Inclusive Text-to-Image Generation |
2023-09-11T15:45:40Z | http://arxiv.org/pdf/2309.05557v3 | http://arxiv.org/abs/2309.05557v3 | Yukai Miao, Yu Bai, Li Chen, Dan Li, Haifeng Sun, Xizheng Wang, Ziqiu Luo, Yanyu Ren, Dapeng Sun, Xiuting Xu, Qi Zhang, Chao Xiang, Xinchi Li | 2023-09-19T16:04:25Z | cs.CL, cs.AI, cs.NI | Nowadays, the versatile capabilities of Pre-trained Large Language Models
(LLMs) have attracted much attention from the industry. However, some vertical
domains are more interested in the in-domain capabilities of LLMs. For the
Networks domain, we present NetEval, an evaluation set for measuring the
comprehensive capabilities of LLMs in Network Operations (NetOps). NetEval is
designed for evaluating the commonsense knowledge and inference ability in
NetOps in a multi-lingual context. NetEval consists of 5,732 questions about
NetOps, covering five different sub-domains of NetOps. With NetEval, we
systematically evaluate the NetOps capability of 26 publicly available LLMs.
The results show that only GPT-4 can achieve a performance competitive to
humans. However, some open models like LLaMA 2 demonstrate significant
potential. | cs.CL | An Empirical Study of NetOps Capability of Pre-Trained Large Language Models |
2023-09-11T15:27:59Z | http://arxiv.org/pdf/2309.05542v1 | http://arxiv.org/abs/2309.05542v1 | Andrew Zhu, Liam Dugan, Alyssa Hwang, Chris Callison-Burch | 2023-09-11T15:27:59Z | cs.SE, cs.AI, cs.CL, I.2.7 | Language model applications are becoming increasingly popular and complex,
often including features like tool usage and retrieval augmentation. However,
existing frameworks for such applications are often opinionated, deciding for
developers how their prompts ought to be formatted and imposing limitations on
customizability and reproducibility. To solve this we present Kani: a
lightweight, flexible, and model-agnostic open-source framework for building
language model applications. Kani helps developers implement a variety of
complex features by supporting the core building blocks of chat interaction:
model interfacing, chat management, and robust function calling. All Kani core
functions are easily overridable and well documented to empower developers to
customize functionality for their own needs. Kani thus serves as a useful tool
for researchers, hobbyists, and industry professionals alike to accelerate
their development while retaining interoperability and fine-grained control. | cs.SE | Kani: A Lightweight and Highly Hackable Framework for Building Language Model Applications |
2023-09-11T15:18:28Z | http://arxiv.org/pdf/2309.05534v1 | http://arxiv.org/abs/2309.05534v1 | Chengyu Wang, Zhongjie Duan, Bingyan Liu, Xinyi Zou, Cen Chen, Kui Jia, Jun Huang | 2023-09-11T15:18:28Z | cs.CL, cs.AI, cs.CV | Text-to-image synthesis for the Chinese language poses unique challenges due
to its large vocabulary size, and intricate character relationships. While
existing diffusion models have shown promise in generating images from textual
descriptions, they often neglect domain-specific contexts and lack robustness
in handling the Chinese language. This paper introduces PAI-Diffusion, a
comprehensive framework that addresses these limitations. PAI-Diffusion
incorporates both general and domain-specific Chinese diffusion models,
enabling the generation of contextually relevant images. It explores the
potential of using LoRA and ControlNet for fine-grained image style transfer
and image editing, empowering users with enhanced control over image
generation. Moreover, PAI-Diffusion seamlessly integrates with Alibaba Cloud's
Machine Learning Platform for AI, providing accessible and scalable solutions.
All the Chinese diffusion model checkpoints, LoRAs, and ControlNets, including
domain-specific ones, are publicly available. A user-friendly Chinese WebUI and
the diffusers-api elastic inference toolkit, also open-sourced, further
facilitate the easy deployment of PAI-Diffusion models in various environments,
making it a valuable resource for Chinese text-to-image synthesis. | cs.CL | PAI-Diffusion: Constructing and Serving a Family of Open Chinese Diffusion Models for Text-to-image Synthesis on the Cloud |
2023-09-11T15:02:25Z | http://arxiv.org/pdf/2309.05519v3 | http://arxiv.org/abs/2309.05519v3 | Shengqiong Wu, Hao Fei, Leigang Qu, Wei Ji, Tat-Seng Chua | 2024-06-25T05:01:09Z | cs.AI, cs.CL, cs.LG | While recently Multimodal Large Language Models (MM-LLMs) have made exciting
strides, they mostly fall prey to the limitation of only input-side multimodal
understanding, without the ability to produce content in multiple modalities.
As we humans always perceive the world and communicate with people through
various modalities, developing any-to-any MM-LLMs capable of accepting and
delivering content in any modality becomes essential to human-level AI. To fill
the gap, we present an end-to-end general-purpose any-to-any MM-LLM system,
NExT-GPT. We connect an LLM with multimodal adaptors and different diffusion
decoders, enabling NExT-GPT to perceive inputs and generate outputs in
arbitrary combinations of text, images, videos, and audio. By leveraging the
existing well-trained highly-performing encoders and decoders, NExT-GPT is
tuned with only a small amount of parameter (1%) of certain projection layers,
which not only benefits low-cost training and also facilitates convenient
expansion to more potential modalities. Moreover, we introduce a
modality-switching instruction tuning (MosIT) and manually curate a
high-quality dataset for MosIT, based on which NExT-GPT is empowered with
complex cross-modal semantic understanding and content generation. Overall, our
research showcases the promising possibility of building an AI agent capable of
modeling universal modalities, paving the way for more human-like AI research
in the community. Project page: https://next-gpt.github.io/ | cs.AI | NExT-GPT: Any-to-Any Multimodal LLM |
2023-09-11T14:58:23Z | http://arxiv.org/pdf/2309.05516v4 | http://arxiv.org/abs/2309.05516v4 | Wenhua Cheng, Weiwei Zhang, Haihao Shen, Yiyang Cai, Xin He, Kaokao Lv, Yi Liu | 2024-05-23T10:43:09Z | cs.CL, cs.AI, cs.LG | Large Language Models (LLMs) have demonstrated exceptional proficiency in
language-related tasks, but their deployment poses significant challenges due
to substantial memory and storage requirements. Weight-only quantization has
emerged as a promising solution to address these challenges. Previous research
suggests that fine-tuning through up and down rounding can enhance performance.
In this study, we introduce SignRound, a method that utilizes signed gradient
descent (SignSGD) to optimize rounding values and weight clipping within just
200 steps. SignRound integrates the advantages of Quantization-Aware Training
(QAT) and Post-Training Quantization (PTQ), achieving exceptional results
across 2 to 4 bits while maintaining low tuning costs and avoiding additional
inference overhead. For example, SignRound achieves absolute average accuracy
improvements ranging from 6.91\% to 33.22\% at 2 bits. It also demonstrates
robust generalization to recent models and achieves near-lossless quantization
in most scenarios at 4 bits. The source code is publicly available at
\url{https://github.com/intel/auto-round}. | cs.CL | Optimize Weight Rounding via Signed Gradient Descent for the Quantization of LLMs |
2023-09-11T14:45:24Z | http://arxiv.org/pdf/2309.05503v1 | http://arxiv.org/abs/2309.05503v1 | Thibault Douzon, Stefan Duffner, Christophe Garcia, Jérémy Espinas | 2023-09-11T14:45:24Z | cs.CL, 68T01, I.2.7 | Since their release, Transformers have revolutionized many fields from
Natural Language Understanding to Computer Vision. Document Understanding (DU)
was not left behind with first Transformer based models for DU dating from late
2019. However, the computational complexity of the self-attention operation
limits their capabilities to small sequences. In this paper we explore multiple
strategies to apply Transformer based models to long multi-page documents. We
introduce 2 new multi-modal (text + layout) long-range models for DU. They are
based on efficient implementations of Transformers for long sequences.
Long-range models can process whole documents at once effectively and are less
impaired by the document's length. We compare them to LayoutLM, a classical
Transformer adapted for DU and pre-trained on millions of documents. We further
propose 2D relative attention bias to guide self-attention towards relevant
tokens without harming model efficiency. We observe improvements on multi-page
business documents on Information Retrieval for a small performance cost on
smaller sequences. Relative 2D attention revealed to be effective on dense text
for both normal and long-range models. | cs.CL | Long-Range Transformer Architectures for Document Understanding |
2023-09-11T14:43:54Z | http://arxiv.org/pdf/2309.05501v1 | http://arxiv.org/abs/2309.05501v1 | Ha-Thanh Nguyen, Randy Goebel, Francesca Toni, Kostas Stathis, Ken Satoh | 2023-09-11T14:43:54Z | cs.CL, cs.AI | The evolution of Generative Pre-trained Transformer (GPT) models has led to
significant advancements in various natural language processing applications,
particularly in legal textual entailment. We present an analysis of GPT-3.5
(ChatGPT) and GPT-4 performances on COLIEE Task 4 dataset, a prominent
benchmark in this domain. The study encompasses data from Heisei 18 (2006) to
Reiwa 3 (2021), exploring the models' abilities to discern entailment
relationships within Japanese statute law across different periods. Our
preliminary experimental results unveil intriguing insights into the models'
strengths and weaknesses in handling legal textual entailment tasks, as well as
the patterns observed in model performance. In the context of proprietary
models with undisclosed architectures and weights, black-box analysis becomes
crucial for evaluating their capabilities. We discuss the influence of training
data distribution and the implications on the models' generalizability. This
analysis serves as a foundation for future research, aiming to optimize
GPT-based models and enable their successful adoption in legal information
extraction and entailment applications. | cs.CL | Black-Box Analysis: GPTs Across Time in Legal Textual Entailment Task |
2023-09-11T14:43:45Z | http://arxiv.org/pdf/2309.05500v1 | http://arxiv.org/abs/2309.05500v1 | Hai-Long Nguyen, Dieu-Quynh Nguyen, Hoang-Trung Nguyen, Thu-Trang Pham, Huu-Dong Nguyen, Thach-Anh Nguyen, Thi-Hai-Yen Vuong, Ha-Thanh Nguyen | 2023-09-11T14:43:45Z | cs.CL, cs.AI | In recent years, natural language processing has gained significant
popularity in various sectors, including the legal domain. This paper presents
NeCo Team's solutions to the Vietnamese text processing tasks provided in the
Automated Legal Question Answering Competition 2023 (ALQAC 2023), focusing on
legal domain knowledge acquisition for low-resource languages through data
enrichment. Our methods for the legal document retrieval task employ a
combination of similarity ranking and deep learning models, while for the
second task, which requires extracting an answer from a relevant legal article
in response to a question, we propose a range of adaptive techniques to handle
different question types. Our approaches achieve outstanding results on both
tasks of the competition, demonstrating the potential benefits and
effectiveness of question answering systems in the legal field, particularly
for low-resource languages. | cs.CL | NeCo@ALQAC 2023: Legal Domain Knowledge Acquisition for Low-Resource Languages through Data Enrichment |
2023-09-11T14:39:04Z | http://arxiv.org/pdf/2309.05497v1 | http://arxiv.org/abs/2309.05497v1 | Abhilash Datta, Souvic Chakraborty, Animesh Mukherjee | 2023-09-11T14:39:04Z | cs.CL, cs.CY | Personality types are important in various fields as they hold relevant
information about the characteristics of a human being in an explainable
format. They are often good predictors of a person's behaviors in a particular
environment and have applications ranging from candidate selection to marketing
and mental health. Recently automatic detection of personality traits from
texts has gained significant attention in computational linguistics. Most
personality detection and analysis methods have focused on small datasets
making their experimental observations often limited. To bridge this gap, we
focus on collecting and releasing the largest automatically curated dataset for
the research community which has 152 million tweets and 56 thousand data points
for the Myers-Briggs personality type (MBTI) prediction task. We perform a
series of extensive qualitative and quantitative studies on our dataset to
analyze the data patterns in a better way and infer conclusions. We show how
our intriguing analysis results often follow natural intuition. We also perform
a series of ablation studies to show how the baselines perform for our dataset. | cs.CL | Personality Detection and Analysis using Twitter Data |
2023-09-11T14:36:16Z | http://arxiv.org/pdf/2309.05494v3 | http://arxiv.org/abs/2309.05494v3 | Rabindra Lamsal, Maria Rodriguez Read, Shanika Karunasekera | 2024-04-11T05:25:17Z | cs.CL | Social media platforms play an essential role in crisis communication, but
analyzing crisis-related social media texts is challenging due to their
informal nature. Transformer-based pre-trained models like BERT and RoBERTa
have shown success in various NLP tasks, but they are not tailored for
crisis-related texts. Furthermore, general-purpose sentence encoders are used
to generate sentence embeddings, regardless of the textual complexities in
crisis-related texts. Advances in applications like text classification,
semantic search, and clustering contribute to the effective processing of
crisis-related texts, which is essential for emergency responders to gain a
comprehensive view of a crisis event, whether historical or real-time. To
address these gaps in crisis informatics literature, this study introduces
CrisisTransformers, an ensemble of pre-trained language models and sentence
encoders trained on an extensive corpus of over 15 billion word tokens from
tweets associated with more than 30 crisis events, including disease outbreaks,
natural disasters, conflicts, and other critical incidents. We evaluate
existing models and CrisisTransformers on 18 crisis-specific public datasets.
Our pre-trained models outperform strong baselines across all datasets in
classification tasks, and our best-performing sentence encoder improves the
state-of-the-art by 17.43% in sentence encoding tasks. Additionally, we
investigate the impact of model initialization on convergence and evaluate the
significance of domain-specific models in generating semantically meaningful
sentence embeddings. The models are publicly available at:
https://huggingface.co/crisistransformers | cs.CL | CrisisTransformers: Pre-trained language models and sentence encoders for crisis-related social media texts |
2023-09-11T14:16:27Z | http://arxiv.org/pdf/2309.05475v2 | http://arxiv.org/abs/2309.05475v2 | Neel Bhate, Ansh Mittal, Zhe He, Xiao Luo | 2023-09-13T20:30:22Z | cs.CL | Demographics, Social determinants of health, and family history documented in
the unstructured text within the electronic health records are increasingly
being studied to understand how this information can be utilized with the
structured data to improve healthcare outcomes. After the GPT models were
released, many studies have applied GPT models to extract this information from
the narrative clinical notes. Different from the existing work, our research
focuses on investigating the zero-shot learning on extracting this information
together by providing minimum information to the GPT model. We utilize
de-identified real-world clinical notes annotated for demographics, various
social determinants, and family history information. Given that the GPT model
might provide text different from the text in the original data, we explore two
sets of evaluation metrics, including the traditional NER evaluation metrics
and semantic similarity evaluation metrics, to completely understand the
performance. Our results show that the GPT-3.5 method achieved an average of
0.975 F1 on demographics extraction, 0.615 F1 on social determinants
extraction, and 0.722 F1 on family history extraction. We believe these results
can be further improved through model fine-tuning or few-shots learning.
Through the case studies, we also identified the limitations of the GPT models,
which need to be addressed in future research. | cs.CL | Zero-shot Learning with Minimum Instruction to Extract Social Determinants and Family History from Clinical Notes using GPT Model |
2023-09-11T14:13:09Z | http://arxiv.org/pdf/2309.05472v2 | http://arxiv.org/abs/2309.05472v2 | Titouan Parcollet, Ha Nguyen, Solene Evain, Marcely Zanon Boito, Adrien Pupier, Salima Mdhaffar, Hang Le, Sina Alisamir, Natalia Tomashenko, Marco Dinarelli, Shucong Zhang, Alexandre Allauzen, Maximin Coavoux, Yannick Esteve, Mickael Rouvier, Jerome Goulian, Benjamin Lecouteux, Francois Portet, Solange Rossato, Fabien Ringeval, Didier Schwab, Laurent Besacier | 2024-03-18T10:54:15Z | cs.CL, cs.AI, cs.SD, eess.AS | Self-supervised learning (SSL) is at the origin of unprecedented improvements
in many different domains including computer vision and natural language
processing. Speech processing drastically benefitted from SSL as most of the
current domain-related tasks are now being approached with pre-trained models.
This work introduces LeBenchmark 2.0 an open-source framework for assessing and
building SSL-equipped French speech technologies. It includes documented,
large-scale and heterogeneous corpora with up to 14,000 hours of heterogeneous
speech, ten pre-trained SSL wav2vec 2.0 models containing from 26 million to
one billion learnable parameters shared with the community, and an evaluation
protocol made of six downstream tasks to complement existing benchmarks.
LeBenchmark 2.0 also presents unique perspectives on pre-trained SSL models for
speech with the investigation of frozen versus fine-tuned downstream models,
task-agnostic versus task-specific pre-trained models as well as a discussion
on the carbon footprint of large-scale model training. Overall, the newly
introduced models trained on 14,000 hours of French speech outperform
multilingual and previous LeBenchmark SSL models across the benchmark but also
required up to four times more energy for pre-training. | cs.CL | LeBenchmark 2.0: a Standardized, Replicable and Enhanced Framework for Self-supervised Representations of French Speech |
2023-09-11T14:01:45Z | http://arxiv.org/pdf/2309.05463v1 | http://arxiv.org/abs/2309.05463v1 | Yuanzhi Li, Sébastien Bubeck, Ronen Eldan, Allie Del Giorno, Suriya Gunasekar, Yin Tat Lee | 2023-09-11T14:01:45Z | cs.CL, cs.AI | We continue the investigation into the power of smaller Transformer-based
language models as initiated by \textbf{TinyStories} -- a 10 million parameter
model that can produce coherent English -- and the follow-up work on
\textbf{phi-1}, a 1.3 billion parameter model with Python coding performance
close to the state-of-the-art. The latter work proposed to use existing Large
Language Models (LLMs) to generate ``textbook quality" data as a way to enhance
the learning process compared to traditional web data. We follow the
``Textbooks Are All You Need" approach, focusing this time on common sense
reasoning in natural language, and create a new 1.3 billion parameter model
named \textbf{phi-1.5}, with performance on natural language tasks comparable
to models 5x larger, and surpassing most non-frontier LLMs on more complex
reasoning tasks such as grade-school mathematics and basic coding. More
generally, \textbf{phi-1.5} exhibits many of the traits of much larger LLMs,
both good -- such as the ability to ``think step by step" or perform some
rudimentary in-context learning -- and bad, including hallucinations and the
potential for toxic and biased generations -- encouragingly though, we are
seeing improvement on that front thanks to the absence of web data. We
open-source \textbf{phi-1.5} to promote further research on these urgent
topics. | cs.CL | Textbooks Are All You Need II: phi-1.5 technical report |
2023-09-11T13:50:38Z | http://arxiv.org/pdf/2309.05454v2 | http://arxiv.org/abs/2309.05454v2 | Joseph Marvin Imperial, Harish Tayyar Madabushi | 2023-11-03T21:23:06Z | cs.CL | Readability metrics and standards such as Flesch Kincaid Grade Level (FKGL)
and the Common European Framework of Reference for Languages (CEFR) exist to
guide teachers and educators to properly assess the complexity of educational
materials before administering them for classroom use. In this study, we select
a diverse set of open and closed-source instruction-tuned language models and
investigate their performances in writing story completions and simplifying
narratives--tasks that teachers perform--using standard-guided prompts
controlling text readability. Our extensive findings provide empirical proof of
how globally recognized models like ChatGPT may be considered less effective
and may require more refined prompts for these generative tasks compared to
other open-sourced models such as BLOOMZ and FlanT5--which have shown promising
results. | cs.CL | Flesch or Fumble? Evaluating Readability Standard Alignment of Instruction-Tuned Language Models |
2023-09-11T13:47:07Z | http://arxiv.org/pdf/2309.05452v2 | http://arxiv.org/abs/2309.05452v2 | Spencer M. Seals, Valerie L. Shalin | 2024-04-15T13:01:30Z | cs.CL | The development of highly fluent large language models (LLMs) has prompted
increased interest in assessing their reasoning and problem-solving
capabilities. We investigate whether several LLMs can solve a classic type of
deductive reasoning problem from the cognitive science literature. The tested
LLMs have limited abilities to solve these problems in their conventional form.
We performed follow up experiments to investigate if changes to the
presentation format and content improve model performance. We do find
performance differences between conditions; however, they do not improve
overall performance. Moreover, we find that performance interacts with
presentation format and content in unexpected ways that differ from human
performance. Overall, our results suggest that LLMs have unique reasoning
biases that are only partially predicted from human reasoning performance and
the human-generated language corpora that informs them. | cs.CL | Evaluating the Deductive Competence of Large Language Models |
2023-09-11T13:41:27Z | http://arxiv.org/pdf/2309.05448v2 | http://arxiv.org/abs/2309.05448v2 | Haoran Chen, Kenneth Blomqvist, Francesco Milano, Roland Siegwart | 2024-01-18T08:33:35Z | cs.CV, cs.AI, cs.CL | Recently, methods have been proposed for 3D open-vocabulary semantic
segmentation. Such methods are able to segment scenes into arbitrary classes
based on text descriptions provided during runtime. In this paper, we propose
to the best of our knowledge the first algorithm for open-vocabulary panoptic
segmentation in 3D scenes. Our algorithm, Panoptic Vision-Language Feature
Fields (PVLFF), learns a semantic feature field of the scene by distilling
vision-language features from a pretrained 2D model, and jointly fits an
instance feature field through contrastive learning using 2D instance segments
on input frames. Despite not being trained on the target classes, our method
achieves panoptic segmentation performance similar to the state-of-the-art
closed-set 3D systems on the HyperSim, ScanNet and Replica dataset and
additionally outperforms current 3D open-vocabulary systems in terms of
semantic segmentation. We ablate the components of our method to demonstrate
the effectiveness of our model architecture. Our code will be available at
https://github.com/ethz-asl/pvlff. | cs.CV | Panoptic Vision-Language Feature Fields |
2023-09-11T13:41:18Z | http://arxiv.org/pdf/2309.05447v2 | http://arxiv.org/abs/2309.05447v2 | Yongrui Chen, Haiyun Jiang, Xinting Huang, Shuming Shi, Guilin Qi | 2024-05-25T10:05:52Z | cs.CL | The improvement of LLMs' instruction-following capabilities relies heavily on
the availability of high-quality instruction-response pairs. Unfortunately, the
current methods used to collect the pairs suffer from either unaffordable labor
costs or severe hallucinations in the self-generation of LLM. To tackle these
challenges, this paper proposes a scalable solution. It involves training LLMs
to generate instruction-response pairs based on human-written documents, rather
than relying solely on self-generation without context. Our proposed method not
only exploits the advantages of human-written documents in reducing
hallucinations but also utilizes an LLM to wrap the expression of documents,
which enables us to bridge the gap between various document styles and the
standard AI response. Experiments demonstrate that our method outperforms
existing typical methods on multiple benchmarks. In particular, compared to the
best-performing baseline, the LLM trained using our generated dataset exhibits
a 10\% relative improvement in performance on AlpacaEval, despite utilizing
only 1/5 of its training data. Furthermore, a comprehensive manual evaluation
validates the quality of the data we generated. Our trained wrapper is publicly
available at https://github.com/Bahuia/Dog-Instruct. | cs.CL | DoG-Instruct: Towards Premium Instruction-Tuning Data via Text-Grounded Instruction Wrapping |
2023-09-11T13:31:00Z | http://arxiv.org/pdf/2309.05444v1 | http://arxiv.org/abs/2309.05444v1 | Ted Zadouri, Ahmet Üstün, Arash Ahmadian, Beyza Ermiş, Acyr Locatelli, Sara Hooker | 2023-09-11T13:31:00Z | cs.CL, cs.LG | The Mixture of Experts (MoE) is a widely known neural architecture where an
ensemble of specialized sub-models optimizes overall performance with a
constant computational cost. However, conventional MoEs pose challenges at
scale due to the need to store all experts in memory. In this paper, we push
MoE to the limit. We propose extremely parameter-efficient MoE by uniquely
combining MoE architecture with lightweight experts.Our MoE architecture
outperforms standard parameter-efficient fine-tuning (PEFT) methods and is on
par with full fine-tuning by only updating the lightweight experts -- less than
1% of an 11B parameters model. Furthermore, our method generalizes to unseen
tasks as it does not depend on any prior task knowledge. Our research
underscores the versatility of the mixture of experts architecture, showcasing
its ability to deliver robust performance even when subjected to rigorous
parameter constraints. Our code used in all the experiments is publicly
available here: https://github.com/for-ai/parameter-efficient-moe. | cs.CL | Pushing Mixture of Experts to the Limit: Extremely Parameter Efficient MoE for Instruction Tuning |
2023-09-11T13:05:23Z | http://arxiv.org/pdf/2309.05429v1 | http://arxiv.org/abs/2309.05429v1 | Thibault Douzon, Stefan Duffner, Christophe Garcia, Jérémy Espinas | 2023-09-11T13:05:23Z | cs.CL, cs.AI, 68T01, I.2.7 | Transformer-based Language Models are widely used in Natural Language
Processing related tasks. Thanks to their pre-training, they have been
successfully adapted to Information Extraction in business documents. However,
most pre-training tasks proposed in the literature for business documents are
too generic and not sufficient to learn more complex structures. In this paper,
we use LayoutLM, a language model pre-trained on a collection of business
documents, and introduce two new pre-training tasks that further improve its
capacity to extract relevant information. The first is aimed at better
understanding the complex layout of documents, and the second focuses on
numeric values and their order of magnitude. These tasks force the model to
learn better-contextualized representations of the scanned documents. We
further introduce a new post-processing algorithm to decode BIESO tags in
Information Extraction that performs better with complex entities. Our method
significantly improves extraction performance on both public (from 93.88 to
95.50 F1 score) and private (from 84.35 to 84.84 F1 score) datasets composed of
expense receipts, invoices, and purchase orders. | cs.CL | Improving Information Extraction on Business Documents with Specific Pre-Training Tasks |
2023-09-11T12:50:28Z | http://arxiv.org/pdf/2309.05423v2 | http://arxiv.org/abs/2309.05423v2 | Jinzuomu Zhong, Yang Li, Hui Huang, Korin Richmond, Jie Liu, Zhiba Su, Jing Guo, Benlai Tang, Fengjie Zhu | 2024-06-11T16:43:11Z | eess.AS, cs.AI, cs.CL, cs.SD | In expressive and controllable Text-to-Speech (TTS), explicit prosodic
features significantly improve the naturalness and controllability of
synthesised speech. However, manual prosody annotation is labor-intensive and
inconsistent. To address this issue, a two-stage automatic annotation pipeline
is novelly proposed in this paper. In the first stage, we use contrastive
pretraining of Speech-Silence and Word-Punctuation (SSWP) pairs to enhance
prosodic information in latent representations. In the second stage, we build a
multi-modal prosody annotator, comprising pretrained encoders, a text-speech
fusing scheme, and a sequence classifier. Experiments on English prosodic
boundaries demonstrate that our method achieves state-of-the-art (SOTA)
performance with 0.72 and 0.93 f1 score for Prosodic Word and Prosodic Phrase
boundary respectively, while bearing remarkable robustness to data scarcity. | eess.AS | Multi-Modal Automatic Prosody Annotation with Contrastive Pretraining of SSWP |
2023-09-11T08:57:02Z | http://arxiv.org/pdf/2309.05312v1 | http://arxiv.org/abs/2309.05312v1 | Olga Kellert, Mahmud Uz Zaman, Nicholas Hill Matlis, Carlos Gómez-Rodríguez | 2023-09-11T08:57:02Z | cs.CL, 68T50, I.2.7 | This paper summarizes the results of experimenting with Universal
Dependencies (UD) adaptation of an Unsupervised, Compositional and Recursive
(UCR) rule-based approach for Sentiment Analysis (SA) submitted to the Shared
Task at Rest-Mex 2023 (Team Olga/LyS-SALSA) (within the IberLEF 2023
conference). By using basic syntactic rules such as rules of modification and
negation applied on words from sentiment dictionaries, our approach exploits
some advantages of an unsupervised method for SA: (1) interpretability and
explainability of SA, (2) robustness across datasets, languages and domains and
(3) usability by non-experts in NLP. We compare our approach with other
unsupervised approaches of SA that in contrast to our UCR rule-based approach
use simple heuristic rules to deal with negation and modification. Our results
show a considerable improvement over these approaches. We discuss future
improvements of our results by using modality features as another shifting rule
of polarity and word disambiguation techniques to identify the right sentiment
words. | cs.CL | Experimenting with UD Adaptation of an Unsupervised Rule-based Approach for Sentiment Analysis of Mexican Tourist Texts |
2023-09-11T08:56:47Z | http://arxiv.org/pdf/2309.05311v1 | http://arxiv.org/abs/2309.05311v1 | Michael Beukman, Manuel Fokam | 2023-09-11T08:56:47Z | cs.CL | Transfer learning has led to large gains in performance for nearly all NLP
tasks while making downstream models easier and faster to train. This has also
been extended to low-resourced languages, with some success. We investigate the
properties of cross-lingual transfer learning between ten low-resourced
languages, from the perspective of a named entity recognition task. We
specifically investigate how much adaptive fine-tuning and the choice of
transfer language affect zero-shot transfer performance. We find that models
that perform well on a single language often do so at the expense of
generalising to others, while models with the best generalisation to other
languages suffer in individual language performance. Furthermore, the amount of
data overlap between the source and target datasets is a better predictor of
transfer performance than either the geographical or genetic distance between
the languages. | cs.CL | Analysing Cross-Lingual Transfer in Low-Resourced African Named Entity Recognition |
2023-09-11T07:10:47Z | http://arxiv.org/pdf/2309.05272v1 | http://arxiv.org/abs/2309.05272v1 | František Kmječ, Ondřej Bojar | 2023-09-11T07:10:47Z | cs.CL | Many meetings require creating a meeting summary to keep everyone up to date.
Creating minutes of sufficient quality is however very cognitively demanding.
Although we currently possess capable models for both audio speech recognition
(ASR) and summarization, their fully automatic use is still problematic. ASR
models frequently commit errors when transcribing named entities while the
summarization models tend to hallucinate and misinterpret the transcript. We
propose a novel tool -- Minuteman -- to enable efficient semi-automatic meeting
minuting. The tool provides a live transcript and a live meeting summary to the
users, who can edit them in a collaborative manner, enabling correction of ASR
errors and imperfect summary points in real time. The resulting application
eases the cognitive load of the notetakers and allows them to easily catch up
if they missed a part of the meeting due to absence or a lack of focus. We
conduct several tests of the application in varied settings, exploring the
worthiness of the concept and the possible user strategies. | cs.CL | Minuteman: Machine and Human Joining Forces in Meeting Summarization |
2023-09-11T07:02:13Z | http://arxiv.org/pdf/2309.05270v2 | http://arxiv.org/abs/2309.05270v2 | Mohsin Ali, Kandukuri Sai Teja, Neeharika Gupta, Parth Patwa, Anubhab Chatterjee, Vinija Jain, Aman Chadha, Amitava Das | 2023-10-18T23:48:40Z | cs.CL, cs.LG | The mixing of two or more languages is called Code-Mixing (CM). CM is a
social norm in multilingual societies. Neural Language Models (NLMs) like
transformers have been effective on many NLP tasks. However, NLM for CM is an
under-explored area. Though transformers are capable and powerful, they cannot
always encode positional information since they are non-recurrent. Therefore,
to enrich word information and incorporate positional information, positional
encoding is defined. We hypothesize that Switching Points (SPs), i.e.,
junctions in the text where the language switches (L1 -> L2 or L2 -> L1), pose
a challenge for CM Language Models (LMs), and hence give special emphasis to
SPs in the modeling process. We experiment with several positional encoding
mechanisms and show that rotatory positional encodings along with switching
point information yield the best results.
We introduce CONFLATOR: a neural language modeling approach for code-mixed
languages. CONFLATOR tries to learn to emphasize switching points using smarter
positional encoding, both at unigram and bigram levels. CONFLATOR outperforms
the state-of-the-art on two tasks based on code-mixed Hindi and English
(Hinglish): (i) sentiment analysis and (ii) machine translation. | cs.CL | CONFLATOR: Incorporating Switching Point based Rotatory Positional Encodings for Code-Mixed Language Modeling |
2023-09-11T06:51:09Z | http://arxiv.org/pdf/2309.06557v1 | http://arxiv.org/abs/2309.06557v1 | Adam M. Lehavi, William McCormack, Noah Kornfeld, Solomon Glazer | 2023-09-11T06:51:09Z | cs.CL, cs.AI, cs.LG, I.2.7; H.2.8; J.4 | This paper presents a pipeline with minimal human influence for scraping and
detecting bias on college newspaper archives. This paper introduces a framework
for scraping complex archive sites that automated tools fail to grab data from,
and subsequently generates a dataset of 14 student papers with 23,154 entries.
This data can also then be queried by keyword to calculate bias by comparing
the sentiment of a large language model summary to the original article. The
advantages of this approach are that it is less comparative than reconstruction
bias and requires less labelled data than generating keyword sentiment. Results
are calculated on politically charged words as well as control words to show
how conclusions can be drawn. The complete method facilitates the extraction of
nuanced insights with minimal assumptions and categorizations, paving the way
for a more objective understanding of bias within student newspaper sources. | cs.CL | Unsupervised Bias Detection in College Student Newspapers |
2023-09-11T04:20:36Z | http://arxiv.org/pdf/2309.05227v1 | http://arxiv.org/abs/2309.05227v1 | Md Abdul Aowal, Maliha T Islam, Priyanka Mary Mammen, Sandesh Shetty | 2023-09-11T04:20:36Z | cs.CL, cs.AI | In this project, we want to explore the newly emerging field of prompt
engineering and apply it to the downstream task of detecting LM biases. More
concretely, we explore how to design prompts that can indicate 4 different
types of biases: (1) gender, (2) race, (3) sexual orientation, and (4)
religion-based. Within our project, we experiment with different manually
crafted prompts that can draw out the subtle biases that may be present in the
language model. We apply these prompts to multiple variations of popular and
well-recognized models: BERT, RoBERTa, and T5 to evaluate their biases. We
provide a comparative analysis of these models and assess them using a two-fold
method: use human judgment to decide whether model predictions are biased and
utilize model-level judgment (through further prompts) to understand if a model
can self-diagnose the biases of its own prediction. | cs.CL | Detecting Natural Language Biases with Prompt-based Learning |
2023-09-11T03:54:38Z | http://arxiv.org/pdf/2309.05221v1 | http://arxiv.org/abs/2309.05221v1 | Fuqian Zhang, Zhenhua Wang | 2023-09-11T03:54:38Z | cs.CL | The renowned proverb, Numbers do not lie, underscores the reliability and
insight that lie beneath numbers, a concept of undisputed importance,
especially in economics and finance etc. Despite the prosperity of Benford's
Law in the first digit analysis, its scope fails to remain comprehensiveness
when it comes to deciphering the laws of number. This paper delves into number
laws by taking the financial statements of China real estate as a
representative, quantitatively study not only the first digit, but also depict
the other two dimensions of numbers: frequency and length. The research
outcomes transcend mere reservations about data manipulation and open the door
to discussions surrounding number diversity and the delineation of the usage
insights. This study wields both economic significance and the capacity to
foster a deeper comprehension of numerical phenomena. | cs.CL | Exploring the Law of Numbers: Evidence from China's Real Estate |
2023-09-11T03:35:00Z | http://arxiv.org/pdf/2309.05217v1 | http://arxiv.org/abs/2309.05217v1 | Li Du, Yequan Wang, Xingrun Xing, Yiqun Ya, Xiang Li, Xin Jiang, Xuezhi Fang | 2023-09-11T03:35:00Z | cs.AI, cs.CL | Although demonstrating superb performance on various NLP tasks, large
language models (LLMs) still suffer from the hallucination problem, which
threatens the reliability of LLMs. To measure the level of hallucination of
LLMs, previous works first categorize the hallucination according to the
phenomenon similarity, then quantify the proportion that model outputs contain
hallucinatory contents. However, such hallucination rates could easily be
distorted by confounders. Moreover, such hallucination rates could not reflect
the reasons for the hallucination, as similar hallucinatory phenomena may
originate from different sources. To address these issues, we propose to
combine the hallucination level quantification and hallucination reason
investigation through an association analysis, which builds the relationship
between the hallucination rate of LLMs with a set of risk factors. In this way,
we are able to observe the hallucination level under each value of each risk
factor, examining the contribution and statistical significance of each risk
factor, meanwhile excluding the confounding effect of other factors.
Additionally, by recognizing the risk factors according to a taxonomy of model
capability, we reveal a set of potential deficiencies in commonsense
memorization, relational reasoning, and instruction following, which may
further provide guidance for the pretraining and supervised fine-tuning process
of LLMs to mitigate the hallucination. | cs.AI | Quantifying and Attributing the Hallucination of Large Language Models via Association Analysis |
2023-09-11T02:58:32Z | http://arxiv.org/pdf/2309.05210v3 | http://arxiv.org/abs/2309.05210v3 | Somnath Roy | 2023-09-17T05:38:46Z | cs.CL | Large language models (LLMs) are rapidly increasing in size, with the number
of parameters becoming a key factor in the success of many commercial models,
such as ChatGPT, Claude, and Bard. Even the recently released publicly
accessible models for commercial usage, such as Falcon and Llama2, come
equipped with billions of parameters. This significant increase in the number
of parameters makes deployment and operation very costly. The remarkable
progress in the field of quantization for large neural networks in general and
LLMs in particular, has made these models more accessible by enabling them to
be deployed on consumer-grade GPUs. Quantized models generally demonstrate
comparable performance levels to their unquantized base counterparts.
Nonetheless, there exists a notable gap in our comprehensive understanding of
how these quantized models respond to hyperparameters, such as temperature, max
new tokens, and topk, particularly for next word prediction. The present
analysis reveals that nf4 and fp4 are equally proficient 4-bit quantization
techniques, characterized by similar attributes such as inference speed, memory
consumption, and the quality of generated content. the study identifies nf4 as
displaying greater resilience to temperature variations in the case of the
llama2 series of models at lower temperature, while fp4 and fp4-dq proves to be
a more suitable choice for falcon series of models. It is noteworthy that, in
general, 4-bit quantized models of varying sizes exhibit higher sensitivity to
temperature in the range of 0.5 to 0.8, unlike their unquantized counterparts.
Additionally, int8 quantization is associated with significantly slower
inference speeds, whereas unquantized bfloat16 models consistently yield the
fastest inference speeds across models of all sizes. | cs.CL | Understanding the Impact of Post-Training Quantization on Large Language Models |
2023-09-11T02:35:36Z | http://arxiv.org/pdf/2309.05203v3 | http://arxiv.org/abs/2309.05203v3 | Yuhan Chen, Nuwa Xi, Yanrui Du, Haochun Wang, Jianyu Chen, Sendong Zhao, Bing Qin | 2024-03-05T10:51:23Z | cs.CL | Molecule discovery serves as a cornerstone in numerous scientific domains,
fueling the development of new materials and innovative drug designs. Recent
developments of in-silico molecule discovery have highlighted the promising
results of cross-modal techniques, which bridge molecular structures with their
descriptive annotations. However, these cross-modal methods frequently
encounter the issue of data scarcity, hampering their performance and
application. In this paper, we address the low-resource challenge by utilizing
artificially-real data generated by Large Language Models (LLMs). We first
introduce a retrieval-based prompting strategy to construct high-quality pseudo
data, then explore the optimal method to effectively leverage this pseudo data.
Experiments show that using pseudo data for domain adaptation outperforms all
existing methods, while also requiring a smaller model scale, reduced data size
and lower training cost, highlighting its efficiency. Furthermore, our method
shows a sustained improvement as the volume of pseudo data increases, revealing
the great potential of pseudo data in advancing low-resource cross-modal
molecule discovery. Our code and data are available at
https://github.com/SCIR-HI/ArtificiallyR2R. | cs.CL | From Artificially Real to Real: Leveraging Pseudo Data from Large Language Models for Low-Resource Molecule Discovery |
2023-09-11T02:31:41Z | http://arxiv.org/pdf/2309.05201v1 | http://arxiv.org/abs/2309.05201v1 | Minhao Zhang, Yongliang Ma, Yanzeng Li, Ruoyu Zhang, Lei Zou, Ming Zhou | 2023-09-11T02:31:41Z | cs.CL | Incorporating multiple knowledge sources is proven to be beneficial for
answering complex factoid questions. To utilize multiple knowledge bases (KB),
previous works merge all KBs into a single graph via entity alignment and
reduce the problem to question-answering (QA) over the fused KB. In reality,
various link relations between KBs might be adopted in QA over multi-KBs. In
addition to the identity between the alignable entities (i.e. full link),
unalignable entities expressing the different aspects or types of an abstract
concept may also be treated identical in a question (i.e. partial link). Hence,
the KB fusion in prior works fails to represent all types of links, restricting
their ability to comprehend multi-KBs for QA. In this work, we formulate the
novel Multi-KB-QA task that leverages the full and partial links among multiple
KBs to derive correct answers, a benchmark with diversified link and query
types is also constructed to efficiently evaluate Multi-KB-QA performance.
Finally, we propose a method for Multi-KB-QA that encodes all link relations in
the KB embedding to score and rank candidate answers. Experiments show that our
method markedly surpasses conventional KB-QA systems in Multi-KB-QA, justifying
the necessity of devising this task. | cs.CL | Two is Better Than One: Answering Complex Questions by Multiple Knowledge Sources with Generalized Links |
2023-09-11T02:16:47Z | http://arxiv.org/pdf/2309.05196v3 | http://arxiv.org/abs/2309.05196v3 | Vishakh Padmakumar, He He | 2024-07-01T16:36:30Z | cs.CL, cs.CY, cs.HC, cs.LG | Large language models (LLMs) have led to a surge in collaborative writing
with model assistance. As different users incorporate suggestions from the same
model, there is a risk of decreased diversity in the produced content,
potentially limiting diverse perspectives in public discourse. In this work, we
measure the impact of co-writing on diversity via a controlled experiment,
where users write argumentative essays in three setups -- using a base LLM
(GPT3), a feedback-tuned LLM (InstructGPT), and writing without model help. We
develop a set of diversity metrics and find that writing with InstructGPT (but
not the GPT3) results in a statistically significant reduction in diversity.
Specifically, it increases the similarity between the writings of different
authors and reduces the overall lexical and content diversity. We additionally
find that this effect is mainly attributable to InstructGPT contributing less
diverse text to co-written essays. In contrast, the user-contributed text
remains unaffected by model collaboration. This suggests that the recent
improvement in generation quality from adapting models to human feedback might
come at the cost of more homogeneous and less diverse content. | cs.CL | Does Writing with Language Models Reduce Content Diversity? |
2023-09-11T00:02:05Z | http://arxiv.org/pdf/2309.05173v5 | http://arxiv.org/abs/2309.05173v5 | Zhengxiang Shi, Aldo Lipani | 2024-02-18T10:02:23Z | cs.CL, cs.AI, cs.CV, cs.LG | Prompt tuning (PT), where a small amount of trainable soft (continuous)
prompt vectors is affixed to the input of language models (LM), has shown
promising results across various tasks and models for parameter-efficient
fine-tuning (PEFT). PT stands out from other PEFT approaches because it
maintains competitive performance with fewer trainable parameters and does not
drastically scale up its parameters as the model size expands. However, PT
introduces additional soft prompt tokens, leading to longer input sequences,
which significantly impacts training and inference time and memory usage due to
the Transformer's quadratic complexity. Particularly concerning for Large
Language Models (LLMs) that face heavy daily querying. To address this issue,
we propose Decomposed Prompt Tuning (DePT), which decomposes the soft prompt
into a shorter soft prompt and a pair of low-rank matrices that are then
optimised with two different learning rates. This allows DePT to achieve better
performance while saving substantial memory and time costs compared to vanilla
PT and its variants, without changing trainable parameter sizes. Through
extensive experiments on 23 natural language processing (NLP) and
vision-language (VL) tasks, we demonstrate that DePT outperforms
state-of-the-art PEFT approaches, including the full fine-tuning baseline, in
some scenarios. Additionally, we empirically show that DEPT grows more
efficient as the model size increases. Our further study reveals that DePT
integrates seamlessly with parameter-efficient transfer learning in the
few-shot learning setting and highlights its adaptability to various model
architectures and sizes. | cs.CL | DePT: Decomposed Prompt Tuning for Parameter-Efficient Fine-tuning |
2023-09-10T23:00:35Z | http://arxiv.org/pdf/2309.05162v1 | http://arxiv.org/abs/2309.05162v1 | Bram Willemsen, Dmytro Kalpakchi, Gabriel Skantze | 2023-09-10T23:00:35Z | cs.CL, cs.AI, cs.CV | An idealized, though simplistic, view of the referring expression production
and grounding process in (situated) dialogue assumes that a speaker must merely
appropriately specify their expression so that the target referent may be
successfully identified by the addressee. However, referring in conversation is
a collaborative process that cannot be aptly characterized as an exchange of
minimally-specified referring expressions. Concerns have been raised regarding
assumptions made by prior work on visually-grounded dialogue that reveal an
oversimplified view of conversation and the referential process. We address
these concerns by introducing a collaborative image ranking task, a grounded
agreement game we call "A Game Of Sorts". In our game, players are tasked with
reaching agreement on how to rank a set of images given some sorting criterion
through a largely unrestricted, role-symmetric dialogue. By putting emphasis on
the argumentation in this mixed-initiative interaction, we collect discussions
that involve the collaborative referential process. We describe results of a
small-scale data collection experiment with the proposed task. All discussed
materials, which includes the collected data, the codebase, and a containerized
version of the application, are publicly available. | cs.CL | Collecting Visually-Grounded Dialogue with A Game Of Sorts |
2023-09-10T21:23:09Z | http://arxiv.org/pdf/2309.05142v1 | http://arxiv.org/abs/2309.05142v1 | Michalis Vlachos, Mircea Lungu, Yash Raj Shrestha, Johannes-Rudolf David | 2023-09-10T21:23:09Z | cs.CL, cs.AI | We use large language models to aid learners enhance proficiency in a foreign
language. This is accomplished by identifying content on topics that the user
is interested in, and that closely align with the learner's proficiency level
in that foreign language. Our work centers on French content, but our approach
is readily transferable to other languages. Our solution offers several
distinctive characteristics that differentiate it from existing
language-learning solutions, such as, a) the discovery of content across topics
that the learner cares about, thus increasing motivation, b) a more precise
estimation of the linguistic difficulty of the content than traditional
readability measures, and c) the availability of both textual and video-based
content. The linguistic complexity of video content is derived from the video
captions. It is our aspiration that such technology will enable learners to
remain engaged in the language-learning process by continuously adapting the
topics and the difficulty of the content to align with the learners' evolving
interests and learning objectives. | cs.CL | Large Language Models for Difficulty Estimation of Foreign Language Content with Application to Language Learning |
2023-09-10T18:13:11Z | http://arxiv.org/pdf/2309.05103v1 | http://arxiv.org/abs/2309.05103v1 | Son Quoc Tran, Gia-Huy Do, Phong Nguyen-Thuan Do, Matt Kretchmar, Xinya Du | 2023-09-10T18:13:11Z | cs.CL, cs.AI | The development of large high-quality datasets and high-performing models
have led to significant advancements in the domain of Extractive Question
Answering (EQA). This progress has sparked considerable interest in exploring
unanswerable questions within the EQA domain. Training EQA models with
unanswerable questions helps them avoid extracting misleading or incorrect
answers for queries that lack valid responses. However, manually annotating
unanswerable questions is labor-intensive. To address this, we propose AGent, a
novel pipeline that automatically creates new unanswerable questions by
re-matching a question with a context that lacks the necessary information for
a correct answer. In this paper, we demonstrate the usefulness of this AGent
pipeline by creating two sets of unanswerable questions from answerable
questions in SQuAD and HotpotQA. These created question sets exhibit low error
rates. Additionally, models fine-tuned on these questions show comparable
performance with those fine-tuned on the SQuAD 2.0 dataset on multiple EQA
benchmarks. | cs.CL | AGent: A Novel Pipeline for Automatically Creating Unanswerable Questions |
2023-09-10T17:13:25Z | http://arxiv.org/pdf/2309.05086v2 | http://arxiv.org/abs/2309.05086v2 | Zhijun Chen, Hailong Sun, Wanhao Zhang, Chunyi Xu, Qianren Mao, Pengpeng Chen | 2023-09-28T19:44:09Z | cs.CL, cs.AI | We propose a neuralized undirected graphical model called Neural-Hidden-CRF
to solve the weakly-supervised sequence labeling problem. Under the umbrella of
probabilistic undirected graph theory, the proposed Neural-Hidden-CRF embedded
with a hidden CRF layer models the variables of word sequence, latent ground
truth sequence, and weak label sequence with the global perspective that
undirected graphical models particularly enjoy. In Neural-Hidden-CRF, we can
capitalize on the powerful language model BERT or other deep models to provide
rich contextual semantic knowledge to the latent ground truth sequence, and use
the hidden CRF layer to capture the internal label dependencies.
Neural-Hidden-CRF is conceptually simple and empirically powerful. It obtains
new state-of-the-art results on one crowdsourcing benchmark and three
weak-supervision benchmarks, including outperforming the recent advanced model
CHMM by 2.80 F1 points and 2.23 F1 points in average generalization and
inference performance, respectively. | cs.CL | Neural-Hidden-CRF: A Robust Weakly-Supervised Sequence Labeler |
2023-09-10T16:55:49Z | http://arxiv.org/pdf/2309.05076v1 | http://arxiv.org/abs/2309.05076v1 | Maximilian Croissant, Madeleine Frister, Guy Schofield, Cade McCall | 2023-09-10T16:55:49Z | cs.AI, cs.CL, cs.HC | The development of believable, natural, and interactive digital artificial
agents is a field of growing interest. Theoretical uncertainties and technical
barriers present considerable challenges to the field, particularly with
regards to developing agents that effectively simulate human emotions. Large
language models (LLMs) might address these issues by tapping common patterns in
situational appraisal. In three empirical experiments, this study tests the
capabilities of LLMs to solve emotional intelligence tasks and to simulate
emotions. It presents and evaluates a new chain-of-emotion architecture for
emotion simulation within video games, based on psychological appraisal
research. Results show that it outperforms standard LLM architectures on a
range of user experience and content analysis metrics. This study therefore
provides early evidence of how to construct and test affective agents based on
cognitive processes represented in language models. | cs.AI | An Appraisal-Based Chain-Of-Emotion Architecture for Affective Language Model Game Agents |
2023-09-10T14:46:31Z | http://arxiv.org/pdf/2309.05044v1 | http://arxiv.org/abs/2309.05044v1 | Mohamed Anwar | 2023-09-10T14:46:31Z | cs.CL | One of the things that need to change when it comes to machine translation is
the models' ability to translate code-switching content, especially with the
rise of social media and user-generated content. In this paper, we are
proposing a way of training a single machine translation model that is able to
translate monolingual sentences from one language to another, along with
translating code-switched sentences to either language. This model can be
considered a bilingual model in the human sense. For better use of parallel
data, we generated synthetic code-switched (CSW) data along with an alignment
loss on the encoder to align representations across languages. Using the WMT14
English-French (En-Fr) dataset, the trained model strongly outperforms
bidirectional baselines on code-switched translation while maintaining quality
for non-code-switched (monolingual) data. | cs.CL | The Effect of Alignment Objectives on Code-Switching Translation |
2023-09-10T13:06:45Z | http://arxiv.org/pdf/2309.05021v1 | http://arxiv.org/abs/2309.05021v1 | Yaonai Wei, Tuo Zhang, Han Zhang, Tianyang Zhong, Lin Zhao, Zhengliang Liu, Chong Ma, Songyao Zhang, Muheng Shang, Lei Du, Xiao Li, Tianming Liu, Junwei Han | 2023-09-10T13:06:45Z | cs.CL | Over decades, neuroscience has accumulated a wealth of research results in
the text modality that can be used to explore cognitive processes.
Meta-analysis is a typical method that successfully establishes a link from
text queries to brain activation maps using these research results, but it
still relies on an ideal query environment. In practical applications, text
queries used for meta-analyses may encounter issues such as semantic redundancy
and ambiguity, resulting in an inaccurate mapping to brain images. On the other
hand, large language models (LLMs) like ChatGPT have shown great potential in
tasks such as context understanding and reasoning, displaying a high degree of
consistency with human natural language. Hence, LLMs could improve the
connection between text modality and neuroscience, resolving existing
challenges of meta-analyses. In this study, we propose a method called
Chat2Brain that combines LLMs to basic text-2-image model, known as Text2Brain,
to map open-ended semantic queries to brain activation maps in data-scarce and
complex query environments. By utilizing the understanding and reasoning
capabilities of LLMs, the performance of the mapping model is optimized by
transferring text queries to semantic queries. We demonstrate that Chat2Brain
can synthesize anatomically plausible neural activation patterns for more
complex tasks of text queries. | cs.CL | Chat2Brain: A Method for Mapping Open-Ended Semantic Queries to Brain Activation Maps |
2023-09-10T11:58:29Z | http://arxiv.org/pdf/2309.05007v2 | http://arxiv.org/abs/2309.05007v2 | Yan Meng, Liangming Pan, Yixin Cao, Min-Yen Kan | 2023-09-19T07:51:03Z | cs.CL, cs.AI | Humans ask follow-up questions driven by curiosity, which reflects a creative
human cognitive process. We introduce the task of real-world
information-seeking follow-up question generation (FQG), which aims to generate
follow-up questions seeking a more in-depth understanding of an initial
question and answer. We construct FOLLOWUPQG, a dataset of over 3K real-world
(initial question, answer, follow-up question) tuples collected from a Reddit
forum providing layman-friendly explanations for open-ended questions. In
contrast to existing datasets, questions in FOLLOWUPQG use more diverse
pragmatic strategies to seek information, and they also show higher-order
cognitive skills (such as applying and relating). We evaluate current question
generation models on their efficacy for generating follow-up questions,
exploring how to generate specific types of follow-up questions based on
step-by-step demonstrations. Our results validate FOLLOWUPQG as a challenging
benchmark, as model-generated questions are adequate but far from human-raised
questions in terms of informativeness and complexity. | cs.CL | FOLLOWUPQG: Towards Information-Seeking Follow-up Question Generation |
2023-09-10T11:22:59Z | http://arxiv.org/pdf/2309.06527v1 | http://arxiv.org/abs/2309.06527v1 | Pavel Burnyshev, Elizaveta Kostenok, Alexey Zaytsev | 2023-09-10T11:22:59Z | cs.CL, cs.CR, cs.LG | Adversarial attacks expose vulnerabilities of deep learning models by
introducing minor perturbations to the input, which lead to substantial
alterations in the output. Our research focuses on the impact of such
adversarial attacks on sequence-to-sequence (seq2seq) models, specifically
machine translation models. We introduce algorithms that incorporate basic text
perturbation heuristics and more advanced strategies, such as the
gradient-based attack, which utilizes a differentiable approximation of the
inherently non-differentiable translation metric. Through our investigation, we
provide evidence that machine translation models display robustness displayed
robustness against best performed known adversarial attacks, as the degree of
perturbation in the output is directly proportional to the perturbation in the
input. However, among underdogs, our attacks outperform alternatives, providing
the best relative performance. Another strong candidate is an attack based on
mixing of individual characters. | cs.CL | Machine Translation Models Stand Strong in the Face of Adversarial Attacks |
2023-09-10T10:57:41Z | http://arxiv.org/pdf/2309.04992v1 | http://arxiv.org/abs/2309.04992v1 | Adian Liusie, Potsawee Manakul, Mark J. F. Gales | 2023-09-10T10:57:41Z | cs.CL | Prompt-based classifiers are an attractive approach for zero-shot
classification. However, the precise choice of the prompt template and label
words can largely influence performance, with semantically equivalent settings
often showing notable performance difference. This discrepancy can be partly
attributed to word biases, where the classifier may be biased towards classes.
To address this problem, it is possible to optimise classification thresholds
on a labelled data set, however, this mitigates some of the advantages of
prompt-based classifiers. This paper instead approaches this problem by
examining the expected marginal probabilities of the classes. Here,
probabilities are reweighted to have a uniform prior over classes, in an
unsupervised fashion. Further, we draw a theoretical connection between the
class priors and the language models' word prior, and offer the ability to set
a threshold in a zero-resource fashion. We show that matching class priors
correlates strongly with the oracle upper bound performance and demonstrate
large consistent performance gains for prompt settings over a range of NLP
tasks. | cs.CL | Mitigating Word Bias in Zero-shot Prompt-based Classifiers |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.