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
|
|---|---|---|---|---|---|---|---|---|
2023-05-02T02:04:33Z | http://arxiv.org/pdf/2305.01155v1 | http://arxiv.org/abs/2305.01155v1 | Juan Zuluaga-Gomez, Iuliia Nigmatulina, Amrutha Prasad, Petr Motlicek, Driss Khalil, Srikanth Madikeri, Allan Tart, Igor Szoke, Vincent Lenders, Mickael Rigault, Khalid Choukri | 2023-05-02T02:04:33Z | eess.AS, cs.CL, cs.HC, cs.SD | Voice communication between air traffic controllers (ATCos) and pilots is
critical for ensuring safe and efficient air traffic control (ATC). This task
requires high levels of awareness from ATCos and can be tedious and
error-prone. Recent attempts have been made to integrate artificial
intelligence (AI) into ATC in order to reduce the workload of ATCos. However,
the development of data-driven AI systems for ATC demands large-scale annotated
datasets, which are currently lacking in the field. This paper explores the
lessons learned from the ATCO2 project, a project that aimed to develop a
unique platform to collect and preprocess large amounts of ATC data from
airspace in real time. Audio and surveillance data were collected from publicly
accessible radio frequency channels with VHF receivers owned by a community of
volunteers and later uploaded to Opensky Network servers, which can be
considered an "unlimited source" of data. In addition, this paper reviews
previous work from ATCO2 partners, including (i) robust automatic speech
recognition, (ii) natural language processing, (iii) English language
identification of ATC communications, and (iv) the integration of surveillance
data such as ADS-B. We believe that the pipeline developed during the ATCO2
project, along with the open-sourcing of its data, will encourage research in
the ATC field. A sample of the ATCO2 corpus is available on the following
website: https://www.atco2.org/data, while the full corpus can be purchased
through ELDA at http://catalog.elra.info/en-us/repository/browse/ELRA-S0484. We
demonstrated that ATCO2 is an appropriate dataset to develop ASR engines when
little or near to no ATC in-domain data is available. For instance, with the
CNN-TDNNf kaldi model, we reached the performance of as low as 17.9% and 24.9%
WER on public ATC datasets which is 6.6/7.6% better than "out-of-domain" but
supervised CNN-TDNNf model. | eess.AS | Lessons Learned in ATCO2: 5000 hours of Air Traffic Control Communications for Robust Automatic Speech Recognition and Understanding |
2023-05-02T01:33:02Z | http://arxiv.org/pdf/2305.01146v3 | http://arxiv.org/abs/2305.01146v3 | Dave Van Veen, Cara Van Uden, Maayane Attias, Anuj Pareek, Christian Bluethgen, Malgorzata Polacin, Wah Chiu, Jean-Benoit Delbrouck, Juan Manuel Zambrano Chaves, Curtis P. Langlotz, Akshay S. Chaudhari, John Pauly | 2023-07-20T13:10:07Z | cs.CL | We systematically investigate lightweight strategies to adapt large language
models (LLMs) for the task of radiology report summarization (RRS).
Specifically, we focus on domain adaptation via pretraining (on natural
language, biomedical text, or clinical text) and via discrete prompting or
parameter-efficient fine-tuning. Our results consistently achieve best
performance by maximally adapting to the task via pretraining on clinical text
and fine-tuning on RRS examples. Importantly, this method fine-tunes a mere
0.32% of parameters throughout the model, in contrast to end-to-end fine-tuning
(100% of parameters). Additionally, we study the effect of in-context examples
and out-of-distribution (OOD) training before concluding with a radiologist
reader study and qualitative analysis. Our findings highlight the importance of
domain adaptation in RRS and provide valuable insights toward developing
effective natural language processing solutions for clinical tasks. | cs.CL | RadAdapt: Radiology Report Summarization via Lightweight Domain Adaptation of Large Language Models |
2023-05-02T01:29:53Z | http://arxiv.org/pdf/2305.01145v1 | http://arxiv.org/abs/2305.01145v1 | Kristen M. Edwards, Binyang Song, Jaron Porciello, Mark Engelbert, Carolyn Huang, Faez Ahmed | 2023-05-02T01:29:53Z | cs.CL | When designing evidence-based policies and programs, decision-makers must
distill key information from a vast and rapidly growing literature base.
Identifying relevant literature from raw search results is time and resource
intensive, and is often done by manual screening. In this study, we develop an
AI agent based on a bidirectional encoder representations from transformers
(BERT) model and incorporate it into a human team designing an evidence
synthesis product for global development. We explore the effectiveness of the
human-AI hybrid team in accelerating the evidence synthesis process. To further
improve team efficiency, we enhance the human-AI hybrid team through active
learning (AL). Specifically, we explore different sampling strategies,
including random sampling, least confidence (LC) sampling, and highest priority
(HP) sampling, to study their influence on the collaborative screening process.
Results show that incorporating the BERT-based AI agent into the human team can
reduce the human screening effort by 68.5% compared to the case of no AI
assistance and by 16.8% compared to the case of using a support vector machine
(SVM)-based AI agent for identifying 80% of all relevant documents. When we
apply the HP sampling strategy for AL, the human screening effort can be
reduced even more: by 78.3% for identifying 80% of all relevant documents
compared to no AI assistance. We apply the AL-enhanced human-AI hybrid teaming
workflow in the design process of three evidence gap maps (EGMs) for USAID and
find it to be highly effective. These findings demonstrate how AI can
accelerate the development of evidence synthesis products and promote timely
evidence-based decision making in global development in a human-AI hybrid
teaming context. | cs.CL | ADVISE: AI-accelerated Design of Evidence Synthesis for Global Development |
2023-05-01T21:56:25Z | http://arxiv.org/pdf/2305.01099v1 | http://arxiv.org/abs/2305.01099v1 | Charlie Cowen-Breen, Creston Brooks, Johannes Haubold, Barbara Graziosi | 2023-05-01T21:56:25Z | cs.CL, cs.LG, I.2.7 | This paper presents machine-learning methods to address various problems in
Greek philology. After training a BERT model on the largest premodern Greek
dataset used for this purpose to date, we identify and correct previously
undetected errors made by scribes in the process of textual transmission, in
what is, to our knowledge, the first successful identification of such errors
via machine learning. Additionally, we demonstrate the model's capacity to fill
gaps caused by material deterioration of premodern manuscripts and compare the
model's performance to that of a domain expert. We find that best performance
is achieved when the domain expert is provided with model suggestions for
inspiration. With such human-computer collaborations in mind, we explore the
model's interpretability and find that certain attention heads appear to encode
select grammatical features of premodern Greek. | cs.CL | Logion: Machine Learning for Greek Philology |
2023-05-01T20:29:59Z | http://arxiv.org/pdf/2305.01082v2 | http://arxiv.org/abs/2305.01082v2 | Sanat Sharma, Josep Valls-Vargas, Tracy Holloway King, Francois Guerin, Chirag Arora | 2023-06-14T14:29:58Z | cs.CL, cs.IR, cs.LG | Spellchecking is one of the most fundamental and widely used search features.
Correcting incorrectly spelled user queries not only enhances the user
experience but is expected by the user. However, most widely available
spellchecking solutions are either lower accuracy than state-of-the-art
solutions or too slow to be used for search use cases where latency is a key
requirement. Furthermore, most innovative recent architectures focus on English
and are not trained in a multilingual fashion and are trained for spell
correction in longer text, which is a different paradigm from spell correction
for user queries, where context is sparse (most queries are 1-2 words long).
Finally, since most enterprises have unique vocabularies such as product names,
off-the-shelf spelling solutions fall short of users' needs. In this work, we
build a multilingual spellchecker that is extremely fast and scalable and that
adapts its vocabulary and hence speller output based on a specific product's
needs. Furthermore, our speller out-performs general purpose spellers by a wide
margin on in-domain datasets. Our multilingual speller is used in search in
Adobe products, powering autocomplete in various applications. | cs.CL | Contextual Multilingual Spellchecker for User Queries |
2023-05-01T19:30:32Z | http://arxiv.org/pdf/2305.01050v1 | http://arxiv.org/abs/2305.01050v1 | Sadat Shahriar, Thamar Solorio | 2023-05-01T19:30:32Z | cs.CL, cs.LG, cs.SI | Subjectivity and difference of opinion are key social phenomena, and it is
crucial to take these into account in the annotation and detection process of
derogatory textual content. In this paper, we use four datasets provided by
SemEval-2023 Task 11 and fine-tune a BERT model to capture the disagreement in
the annotation. We find individual annotator modeling and aggregation lowers
the Cross-Entropy score by an average of 0.21, compared to the direct training
on the soft labels. Our findings further demonstrate that annotator metadata
contributes to the average 0.029 reduction in the Cross-Entropy score. | cs.CL | SafeWebUH at SemEval-2023 Task 11: Learning Annotator Disagreement in Derogatory Text: Comparison of Direct Training vs Aggregation |
2023-05-01T18:36:06Z | http://arxiv.org/pdf/2305.01028v2 | http://arxiv.org/abs/2305.01028v2 | Maryan Rizinski, Andrej Jankov, Vignesh Sankaradas, Eugene Pinsky, Igor Miskovski, Dimitar Trajanov | 2023-10-26T20:19:37Z | cs.CL, cs.LG | In recent years, natural language processing (NLP) has become increasingly
important in a variety of business applications, including sentiment analysis,
text classification, and named entity recognition. In this paper, we propose an
approach for company classification using NLP and zero-shot learning. Our
method utilizes pre-trained transformer models to extract features from company
descriptions, and then applies zero-shot learning to classify companies into
relevant categories without the need for specific training data for each
category. We evaluate our approach on a dataset obtained through the Wharton
Research Data Services (WRDS), which comprises textual descriptions of publicly
traded companies. We demonstrate that the approach can streamline the process
of company classification, thereby reducing the time and resources required in
traditional approaches such as the Global Industry Classification Standard
(GICS). The results show that this method has potential for automation of
company classification, making it a promising avenue for future research in
this area. | cs.CL | Company classification using zero-shot learning |
2023-05-01T18:22:10Z | http://arxiv.org/pdf/2305.01020v1 | http://arxiv.org/abs/2305.01020v1 | Benjamin Lipkin, Lionel Wong, Gabriel Grand, Joshua B Tenenbaum | 2023-05-01T18:22:10Z | cs.CL, cs.AI | The relationship between communicated language and intended meaning is often
probabilistic and sensitive to context. Numerous strategies attempt to estimate
such a mapping, often leveraging recursive Bayesian models of communication. In
parallel, large language models (LLMs) have been increasingly applied to
semantic parsing applications, tasked with inferring logical representations
from natural language. While existing LLM explorations have been largely
restricted to literal language use, in this work, we evaluate the capacity of
LLMs to infer the meanings of pragmatic utterances. Specifically, we explore
the case of threshold estimation on the gradable adjective ``strong'',
contextually conditioned on a strength prior, then extended to composition with
qualification, negation, polarity inversion, and class comparison. We find that
LLMs can derive context-grounded, human-like distributions over the
interpretations of several complex pragmatic utterances, yet struggle composing
with negation. These results inform the inferential capacity of statistical
language models, and their use in pragmatic and semantic parsing applications.
All corresponding code is made publicly available
(https://github.com/benlipkin/probsem/tree/CogSci2023). | cs.CL | Evaluating statistical language models as pragmatic reasoners |
2023-05-01T18:00:54Z | http://arxiv.org/pdf/2305.01011v1 | http://arxiv.org/abs/2305.01011v1 | Sadat Shahriar, Arjun Mukherjee, Omprakash Gnawali | 2023-05-01T18:00:54Z | cs.CL, cs.LG | In this era of information explosion, deceivers use different domains or
mediums of information to exploit the users, such as News, Emails, and Tweets.
Although numerous research has been done to detect deception in all these
domains, information shortage in a new event necessitates these domains to
associate with each other to battle deception. To form this association, we
propose a feature augmentation method by harnessing the intermediate layer
representation of neural models. Our approaches provide an improvement over the
self-domain baseline models by up to 6.60%. We find Tweets to be the most
helpful information provider for Fake News and Phishing Email detection,
whereas News helps most in Tweet Rumor detection. Our analysis provides a
useful insight for domain knowledge transfer which can help build a stronger
deception detection system than the existing literature. | cs.CL | Deception Detection with Feature-Augmentation by soft Domain Transfer |
2023-05-01T17:56:32Z | http://arxiv.org/pdf/2305.00969v7 | http://arxiv.org/abs/2305.00969v7 | David Budaghyan, Charles C. Onu, Arsenii Gorin, Cem Subakan, Doina Precup | 2024-03-21T17:52:22Z | cs.SD, cs.AI, cs.CL, eess.AS | This paper describes the Ubenwa CryCeleb dataset - a labeled collection of
infant cries - and the accompanying CryCeleb 2023 task, which is a public
speaker verification challenge based on cry sounds. We released more than 6
hours of manually segmented cry sounds from 786 newborns for academic use,
aiming to encourage research in infant cry analysis. The inaugural public
competition attracted 59 participants, 11 of whom improved the baseline
performance. The top-performing system achieved a significant improvement
scoring 25.8% equal error rate, which is still far from the performance of
state-of-the-art adult speaker verification systems. Therefore, we believe
there is room for further research on this dataset, potentially extending
beyond the verification task. | cs.SD | CryCeleb: A Speaker Verification Dataset Based on Infant Cry Sounds |
2023-05-01T17:36:06Z | http://arxiv.org/pdf/2305.00955v2 | http://arxiv.org/abs/2305.00955v2 | Patrick Fernandes, Aman Madaan, Emmy Liu, António Farinhas, Pedro Henrique Martins, Amanda Bertsch, José G. C. de Souza, Shuyan Zhou, Tongshuang Wu, Graham Neubig, André F. T. Martins | 2023-06-01T01:24:53Z | cs.CL, cs.AI, cs.LG | Many recent advances in natural language generation have been fueled by
training large language models on internet-scale data. However, this paradigm
can lead to models that generate toxic, inaccurate, and unhelpful content, and
automatic evaluation metrics often fail to identify these behaviors. As models
become more capable, human feedback is an invaluable signal for evaluating and
improving models. This survey aims to provide an overview of the recent
research that has leveraged human feedback to improve natural language
generation. First, we introduce an encompassing formalization of feedback, and
identify and organize existing research into a taxonomy following this
formalization. Next, we discuss how feedback can be described by its format and
objective, and cover the two approaches proposed to use feedback (either for
training or decoding): directly using the feedback or training feedback models.
We also discuss existing datasets for human-feedback data collection, and
concerns surrounding feedback collection. Finally, we provide an overview of
the nascent field of AI feedback, which exploits large language models to make
judgments based on a set of principles and minimize the need for human
intervention. | cs.CL | Bridging the Gap: A Survey on Integrating (Human) Feedback for Natural Language Generation |
2023-05-01T17:09:33Z | http://arxiv.org/pdf/2305.00948v2 | http://arxiv.org/abs/2305.00948v2 | Gašper Beguš, Maksymilian Dąbkowski, Ryan Rhodes | 2023-08-21T16:52:29Z | cs.CL, cs.AI | The performance of large language models (LLMs) has recently improved to the
point where the models can perform well on many language tasks. We show here
that for the first time, the models can also generate coherent and valid formal
analyses of linguistic data and illustrate the vast potential of large language
models for analyses of their metalinguistic abilities. LLMs are primarily
trained on language data in the form of text; analyzing and evaluating their
metalinguistic abilities improves our understanding of their general
capabilities and sheds new light on theoretical models in linguistics. In this
paper, we probe into GPT-4's metalinguistic capabilities by focusing on three
subfields of formal linguistics: syntax, phonology, and semantics. We outline a
research program for metalinguistic analyses of large language models, propose
experimental designs, provide general guidelines, discuss limitations, and
offer future directions for this line of research. This line of inquiry also
exemplifies behavioral interpretability of deep learning, where models'
representations are accessed by explicit prompting rather than internal
representations. | cs.CL | Large Linguistic Models: Analyzing theoretical linguistic abilities of LLMs |
2023-05-01T16:57:33Z | http://arxiv.org/pdf/2305.00944v1 | http://arxiv.org/abs/2305.00944v1 | Alexander Wan, Eric Wallace, Sheng Shen, Dan Klein | 2023-05-01T16:57:33Z | cs.CL, cs.CR, cs.LG | Instruction-tuned LMs such as ChatGPT, FLAN, and InstructGPT are finetuned on
datasets that contain user-submitted examples, e.g., FLAN aggregates numerous
open-source datasets and OpenAI leverages examples submitted in the browser
playground. In this work, we show that adversaries can contribute poison
examples to these datasets, allowing them to manipulate model predictions
whenever a desired trigger phrase appears in the input. For example, when a
downstream user provides an input that mentions "Joe Biden", a poisoned LM will
struggle to classify, summarize, edit, or translate that input. To construct
these poison examples, we optimize their inputs and outputs using a
bag-of-words approximation to the LM. We evaluate our method on open-source
instruction-tuned LMs. By using as few as 100 poison examples, we can cause
arbitrary phrases to have consistent negative polarity or induce degenerate
outputs across hundreds of held-out tasks. Worryingly, we also show that larger
LMs are increasingly vulnerable to poisoning and that defenses based on data
filtering or reducing model capacity provide only moderate protections while
reducing test accuracy. | cs.CL | Poisoning Language Models During Instruction Tuning |
2023-05-01T16:26:18Z | http://arxiv.org/pdf/2305.00926v1 | http://arxiv.org/abs/2305.00926v1 | Siddhant Arora, Hayato Futami, Emiru Tsunoo, Brian Yan, Shinji Watanabe | 2023-05-01T16:26:18Z | cs.CL, cs.SD, eess.AS | Most human interactions occur in the form of spoken conversations where the
semantic meaning of a given utterance depends on the context. Each utterance in
spoken conversation can be represented by many semantic and speaker attributes,
and there has been an interest in building Spoken Language Understanding (SLU)
systems for automatically predicting these attributes. Recent work has shown
that incorporating dialogue history can help advance SLU performance. However,
separate models are used for each SLU task, leading to an increase in inference
time and computation cost. Motivated by this, we aim to ask: can we jointly
model all the SLU tasks while incorporating context to facilitate low-latency
and lightweight inference? To answer this, we propose a novel model
architecture that learns dialog context to jointly predict the intent, dialog
act, speaker role, and emotion for the spoken utterance. Note that our joint
prediction is based on an autoregressive model and we need to decide the
prediction order of dialog attributes, which is not trivial. To mitigate the
issue, we also propose an order agnostic training method. Our experiments show
that our joint model achieves similar results to task-specific classifiers and
can effectively integrate dialog context to further improve the SLU
performance. | cs.CL | Joint Modelling of Spoken Language Understanding Tasks with Integrated Dialog History |
2023-05-01T14:16:37Z | http://arxiv.org/pdf/2305.00844v1 | http://arxiv.org/abs/2305.00844v1 | Eddie Guo, Mehul Gupta, Jiawen Deng, Ye-Jean Park, Mike Paget, Christopher Naugler | 2023-05-01T14:16:37Z | cs.CL, cs.AI | Objective: To assess the performance of the OpenAI GPT API in accurately and
efficiently identifying relevant titles and abstracts from real-world clinical
review datasets and compare its performance against ground truth labelling by
two independent human reviewers.
Methods: We introduce a novel workflow using the OpenAI GPT API for screening
titles and abstracts in clinical reviews. A Python script was created to make
calls to the GPT API with the screening criteria in natural language and a
corpus of title and abstract datasets that have been filtered by a minimum of
two human reviewers. We compared the performance of our model against
human-reviewed papers across six review papers, screening over 24,000 titles
and abstracts.
Results: Our results show an accuracy of 0.91, a sensitivity of excluded
papers of 0.91, and a sensitivity of included papers of 0.76. On a randomly
selected subset of papers, the GPT API demonstrated the ability to provide
reasoning for its decisions and corrected its initial decision upon being asked
to explain its reasoning for a subset of incorrect classifications.
Conclusion: The GPT API has the potential to streamline the clinical review
process, save valuable time and effort for researchers, and contribute to the
overall quality of clinical reviews. By prioritizing the workflow and acting as
an aid rather than a replacement for researchers and reviewers, the GPT API can
enhance efficiency and lead to more accurate and reliable conclusions in
medical research. | cs.CL | Automated Paper Screening for Clinical Reviews Using Large Language Models |
2023-05-01T14:02:48Z | http://arxiv.org/pdf/2305.00833v2 | http://arxiv.org/abs/2305.00833v2 | Jack Lanchantin, Shubham Toshniwal, Jason Weston, Arthur Szlam, Sainbayar Sukhbaatar | 2023-10-31T04:06:28Z | cs.LG, cs.AI, cs.CL | Large language models have been shown to struggle with multi-step reasoning,
and do not retain previous reasoning steps for future use. We propose a simple
method for solving both of these problems by allowing the model to take
Self-Notes. Unlike recent chain-of-thought or scratchpad approaches, the model
can deviate from the input context at any time to explicitly think and write
down its thoughts. This allows the model to perform reasoning on the fly as it
reads the context and even integrate previous reasoning steps, thus enhancing
its memory with useful information and enabling multi-step reasoning.
Experiments across a wide variety of tasks demonstrate that our method can
outperform chain-of-thought and scratchpad methods by taking Self-Notes that
interleave the input text. | cs.LG | Learning to Reason and Memorize with Self-Notes |
2023-05-01T02:37:59Z | http://arxiv.org/pdf/2305.00633v3 | http://arxiv.org/abs/2305.00633v3 | Yuxi Xie, Kenji Kawaguchi, Yiran Zhao, Xu Zhao, Min-Yen Kan, Junxian He, Qizhe Xie | 2023-10-26T01:43:17Z | cs.CL, cs.AI, cs.LG | Breaking down a problem into intermediate steps has demonstrated impressive
performance in Large Language Model (LLM) reasoning. However, the growth of the
reasoning chain introduces uncertainty and error accumulation, making it
challenging to elicit accurate final results. To tackle this challenge of
uncertainty in multi-step reasoning, we introduce a stepwise self-evaluation
mechanism to guide and calibrate the reasoning process of LLMs. We propose a
decoding algorithm integrating the self-evaluation guidance via stochastic beam
search. The self-evaluation guidance serves as a better-calibrated automatic
criterion, facilitating an efficient search in the reasoning space and
resulting in superior prediction quality. Stochastic beam search balances
exploitation and exploration of the search space with temperature-controlled
randomness. Our approach surpasses the corresponding Codex-backboned baselines
in few-shot accuracy by $6.34\%$, $9.56\%$, and $5.46\%$ on the GSM8K, AQuA,
and StrategyQA benchmarks, respectively. Experiment results with Llama-2 on
arithmetic reasoning demonstrate the efficiency of our method in outperforming
the baseline methods with comparable computational budgets. Further analysis in
multi-step reasoning finds our self-evaluation guidance pinpoints logic
failures and leads to higher consistency and robustness. Our code is publicly
available at https://guideddecoding.github.io/. | cs.CL | Self-Evaluation Guided Beam Search for Reasoning |
2023-05-01T00:04:19Z | http://arxiv.org/pdf/2305.00606v1 | http://arxiv.org/abs/2305.00606v1 | Derguene Mbaye, Moussa Diallo, Thierno Ibrahima Diop | 2023-05-01T00:04:19Z | cs.CL | Natural Language Processing (NLP) research has made great advancements in
recent years with major breakthroughs that have established new benchmarks.
However, these advances have mainly benefited a certain group of languages
commonly referred to as resource-rich such as English and French. Majority of
other languages with weaker resources are then left behind which is the case
for most African languages including Wolof. In this work, we present a parallel
Wolof/French corpus of 123,000 sentences on which we conducted experiments on
machine translation models based on Recurrent Neural Networks (RNN) in
different data configurations. We noted performance gains with the models
trained on subworded data as well as those trained on the French-English
language pair compared to those trained on the French-Wolof pair under the same
experimental conditions. | cs.CL | Low-Resourced Machine Translation for Senegalese Wolof Language |
2023-04-30T22:33:08Z | http://arxiv.org/pdf/2305.00593v1 | http://arxiv.org/abs/2305.00593v1 | Maohao Shen, Soumya Ghosh, Prasanna Sattigeri, Subhro Das, Yuheng Bu, Gregory Wornell | 2023-04-30T22:33:08Z | cs.LG, cs.CL | Due to privacy or commercial constraints, large pre-trained language models
(PLMs) are often offered as black-box APIs. Fine-tuning such models to
downstream tasks is challenging because one can neither access the model's
internal representations nor propagate gradients through it. This paper
addresses these challenges by developing techniques for adapting PLMs with only
API access. Building on recent work on soft prompt tuning, we develop methods
to tune the soft prompts without requiring gradient computation. Further, we
develop extensions that in addition to not requiring gradients also do not need
to access any internal representation of the PLM beyond the input embeddings.
Moreover, instead of learning a single prompt, our methods learn a distribution
over prompts allowing us to quantify predictive uncertainty. Ours is the first
work to consider uncertainty in prompts when only having API access to the PLM.
Finally, through extensive experiments, we carefully vet the proposed methods
and find them competitive with (and sometimes even improving on) gradient-based
approaches with full access to the PLM. | cs.LG | Reliable Gradient-free and Likelihood-free Prompt Tuning |
2023-04-30T21:44:21Z | http://arxiv.org/pdf/2305.00586v5 | http://arxiv.org/abs/2305.00586v5 | Michael Hanna, Ollie Liu, Alexandre Variengien | 2023-11-02T10:55:18Z | cs.CL, cs.AI, cs.LG | Pre-trained language models can be surprisingly adept at tasks they were not
explicitly trained on, but how they implement these capabilities is poorly
understood. In this paper, we investigate the basic mathematical abilities
often acquired by pre-trained language models. Concretely, we use mechanistic
interpretability techniques to explain the (limited) mathematical abilities of
GPT-2 small. As a case study, we examine its ability to take in sentences such
as "The war lasted from the year 1732 to the year 17", and predict valid
two-digit end years (years > 32). We first identify a circuit, a small subset
of GPT-2 small's computational graph that computes this task's output. Then, we
explain the role of each circuit component, showing that GPT-2 small's final
multi-layer perceptrons boost the probability of end years greater than the
start year. Finally, we find related tasks that activate our circuit. Our
results suggest that GPT-2 small computes greater-than using a complex but
general mechanism that activates across diverse contexts. | cs.CL | How does GPT-2 compute greater-than?: Interpreting mathematical abilities in a pre-trained language model |
2023-04-30T21:22:35Z | http://arxiv.org/pdf/2305.00581v1 | http://arxiv.org/abs/2305.00581v1 | Xuehai He, Xin Eric Wang | 2023-04-30T21:22:35Z | cs.CV, cs.AI, cs.CL | Despite the success of Transformer models in vision and language tasks, they
often learn knowledge from enormous data implicitly and cannot utilize
structured input data directly. On the other hand, structured learning
approaches such as graph neural networks (GNNs) that integrate prior
information can barely compete with Transformer models. In this work, we aim to
benefit from both worlds and propose a novel Multimodal Graph Transformer for
question answering tasks that requires performing reasoning across multiple
modalities. We introduce a graph-involved plug-and-play quasi-attention
mechanism to incorporate multimodal graph information, acquired from text and
visual data, to the vanilla self-attention as effective prior. In particular,
we construct the text graph, dense region graph, and semantic graph to generate
adjacency matrices, and then compose them with input vision and language
features to perform downstream reasoning. Such a way of regularizing
self-attention with graph information significantly improves the inferring
ability and helps align features from different modalities. We validate the
effectiveness of Multimodal Graph Transformer over its Transformer baselines on
GQA, VQAv2, and MultiModalQA datasets. | cs.CV | Multimodal Graph Transformer for Multimodal Question Answering |
2023-04-30T16:40:54Z | http://arxiv.org/pdf/2305.00522v1 | http://arxiv.org/abs/2305.00522v1 | Steven T. Piantadosi | 2023-04-30T16:40:54Z | cs.CL | I present a simple algorithm for enumerating the trees generated by a Context
Free Grammar (CFG). The algorithm uses a pairing function to form a bijection
between CFG derivations and natural numbers, so that trees can be uniquely
decoded from counting. This provides a general way to number expressions in
natural logical languages, and potentially can be extended to other
combinatorial problems. I also show how this algorithm may be generalized to
more general forms of derivation, including analogs of Lempel-Ziv coding on
trees. | cs.CL | How to enumerate trees from a context-free grammar |
2023-04-30T16:14:23Z | http://arxiv.org/pdf/2305.03169v2 | http://arxiv.org/abs/2305.03169v2 | Kai Zhang, Xiaoqian Jiang | 2023-05-22T00:07:33Z | cs.CR, cs.CL, cs.LG | In the era of big data, there is an increasing need for healthcare providers,
communities, and researchers to share data and collaborate to improve health
outcomes, generate valuable insights, and advance research. The Health
Insurance Portability and Accountability Act of 1996 (HIPAA) is a federal law
designed to protect sensitive health information by defining regulations for
protected health information (PHI). However, it does not provide efficient
tools for detecting or removing PHI before data sharing. One of the challenges
in this area of research is the heterogeneous nature of PHI fields in data
across different parties. This variability makes rule-based sensitive variable
identification systems that work on one database fail on another. To address
this issue, our paper explores the use of machine learning algorithms to
identify sensitive variables in structured data, thus facilitating the
de-identification process. We made a key observation that the distributions of
metadata of PHI fields and non-PHI fields are very different. Based on this
novel finding, we engineered over 30 features from the metadata of the original
features and used machine learning to build classification models to
automatically identify PHI fields in structured Electronic Health Record (EHR)
data. We trained the model on a variety of large EHR databases from different
data sources and found that our algorithm achieves 99% accuracy when detecting
PHI-related fields for unseen datasets. The implications of our study are
significant and can benefit industries that handle sensitive data. | cs.CR | Sensitive Data Detection with High-Throughput Machine Learning Models in Electrical Health Records |
2023-04-30T11:54:40Z | http://arxiv.org/pdf/2305.03731v4 | http://arxiv.org/abs/2305.03731v4 | Dongyu Gong, Xingchen Wan, Dingmin Wang | 2024-02-01T21:39:50Z | cs.AI, cs.CL, q-bio.NC | Working memory is a critical aspect of both human intelligence and artificial
intelligence, serving as a workspace for the temporary storage and manipulation
of information. In this paper, we systematically assess the working memory
capacity of ChatGPT, a large language model developed by OpenAI, by examining
its performance in verbal and spatial n-back tasks under various conditions.
Our experiments reveal that ChatGPT has a working memory capacity limit
strikingly similar to that of humans. Furthermore, we investigate the impact of
different instruction strategies on ChatGPT's performance and observe that the
fundamental patterns of a capacity limit persist. From our empirical findings,
we propose that n-back tasks may serve as tools for benchmarking the working
memory capacity of large language models and hold potential for informing
future efforts aimed at enhancing AI working memory. | cs.AI | Working Memory Capacity of ChatGPT: An Empirical Study |
2023-04-30T11:26:10Z | http://arxiv.org/pdf/2305.00450v2 | http://arxiv.org/abs/2305.00450v2 | Huachuan Qiu, Hongliang He, Shuai Zhang, Anqi Li, Zhenzhong Lan | 2024-02-22T10:21:56Z | cs.CL, cs.CY | Developing specialized dialogue systems for mental health support requires
multi-turn conversation data, which has recently garnered increasing attention.
However, gathering and releasing large-scale and real-life multi-turn
conversations to facilitate advancements in mental health presents challenges
due to data privacy protection, as well as the time and cost involved. To
address the challenges related to data scarcity, we introduce SMILE, a
single-turn to multi-turn inclusive language expansion technique that prompts
ChatGPT to rewrite public single-turn dialogues into multi-turn ones. Our work
begins with the analysis of language transformation, validating the feasibility
of the proposed method when compared with other baseline methods. We then
conduct a study on dialogue diversity, including lexical features, semantic
features, and dialogue topics, demonstrating the effectiveness of our proposed
method. Furthermore, we implement an expert evaluation and the results
demonstrate that the dialogues generated with our proposed method are of higher
quality than those generated with other baseline methods. Thus, we employ our
method to generate a large-scale, diverse, and high-quality dialogue dataset
named SmileChat, comprising 55,165 dialogues in total with an average of 10.4
turns per dialogue. Finally, we utilize the collected corpus to develop a
mental health chatbot, MeChat. To better assess the overall quality of
SmileChat, we collect a real-life chat dataset comprising 82 counseling
dialogues for model evaluation. Both automatic and human evaluations
demonstrate that our trained dialogue system exhibits significant improvements,
showcasing that SmileChat is high-quality and practical. | cs.CL | SMILE: Single-turn to Multi-turn Inclusive Language Expansion via ChatGPT for Mental Health Support |
2023-04-30T10:41:43Z | http://arxiv.org/pdf/2305.00446v2 | http://arxiv.org/abs/2305.00446v2 | Hiuchung Hung, Andreas Maier, Thorsten Piske | 2024-01-07T17:17:00Z | cs.CL | This paper introduces a non-native speech corpus consisting of narratives
from fifty 5- to 6-year-old Chinese-English children. Transcripts totaling 6.5
hours of children taking a narrative comprehension test in English (L2) are
presented, along with human-rated scores and annotations of grammatical and
pronunciation errors. The children also completed the parallel MAIN tests in
Chinese (L1) for reference purposes. For all tests we recorded audio and video
with our innovative self-developed remote collection methods. The video
recordings serve to mitigate the challenge of low intelligibility in L2
narratives produced by young children during the transcription process. This
corpus offers valuable resources for second language teaching and has the
potential to enhance the overall performance of automatic speech recognition
(ASR). | cs.CL | Building a Non-native Speech Corpus Featuring Chinese-English Bilingual Children: Compilation and Rationale |
2023-04-30T07:15:29Z | http://arxiv.org/pdf/2305.01503v1 | http://arxiv.org/abs/2305.01503v1 | Sedrick Scott Keh, Zheyuan Ryan Shi, David J. Patterson, Nirmal Bhagabati, Karun Dewan, Areendran Gopala, Pablo Izquierdo, Debojyoti Mallick, Ambika Sharma, Pooja Shrestha, Fei Fang | 2023-04-30T07:15:29Z | cs.IR, cs.CL, cs.CY | Non-governmental organizations for environmental conservation have a
significant interest in monitoring conservation-related media and getting
timely updates about infrastructure construction projects as they may cause
massive impact to key conservation areas. Such monitoring, however, is
difficult and time-consuming. We introduce NewsPanda, a toolkit which
automatically detects and analyzes online articles related to environmental
conservation and infrastructure construction. We fine-tune a BERT-based model
using active learning methods and noise correction algorithms to identify
articles that are relevant to conservation and infrastructure construction. For
the identified articles, we perform further analysis, extracting keywords and
finding potentially related sources. NewsPanda has been successfully deployed
by the World Wide Fund for Nature teams in the UK, India, and Nepal since
February 2022. It currently monitors over 80,000 websites and 1,074
conservation sites across India and Nepal, saving more than 30 hours of human
efforts weekly. We have now scaled it up to cover 60,000 conservation sites
globally. | cs.IR | NewsPanda: Media Monitoring for Timely Conservation Action |
2023-04-30T04:36:05Z | http://arxiv.org/pdf/2305.01505v2 | http://arxiv.org/abs/2305.01505v2 | Guijin Son, Hanearl Jung, Moonjeong Hahm, Keonju Na, Sol Jin | 2023-06-25T18:06:25Z | cs.CL, cs.AI, cs.CY | Large Language Models (LLMs), consisting of 100 billion or more parameters,
have demonstrated remarkable ability in complex multi-step reasoning tasks.
However, the application of such generic advancements has been limited to a few
fields, such as clinical or legal, with the field of financial reasoning
remaining largely unexplored. To the best of our knowledge, the ability of LLMs
to solve financial reasoning problems has never been dealt with, and whether it
can be performed at any scale remains unknown. To address this knowledge gap,
this research presents a comprehensive investigation into the potential
application of LLMs in the financial domain. The investigation includes a
detailed exploration of a range of subjects, including task formulation,
synthetic data generation, prompting methods, and evaluation capability.
Furthermore, the study benchmarks various GPT variants with parameter scales
ranging from 2.8B to 13B, with and without instruction tuning, on diverse
dataset sizes. By analyzing the results, we reveal that the ability to generate
coherent financial reasoning first emerges at 6B parameters, and continues to
improve with better instruction-tuning or larger datasets. Additionally, the
study provides a publicly accessible dataset named sFIOG (Synthetic-Financial
Investment Opinion Generation), consisting of 11,802 synthetic investment
thesis samples, to support further research in the field of financial
reasoning. Overall, this research seeks to contribute to the understanding of
the efficacy of language models in the field of finance, with a particular
emphasis on their ability to engage in sophisticated reasoning and analysis
within the context of investment decision-making. | cs.CL | Beyond Classification: Financial Reasoning in State-of-the-Art Language Models |
2023-04-30T04:23:40Z | http://arxiv.org/pdf/2305.00382v2 | http://arxiv.org/abs/2305.00382v2 | Anders Mølmen Høst, Pierre Lison, Leon Moonen | 2023-05-15T07:36:11Z | cs.CR, cs.AI, cs.CL, cs.SE | Knowledge graphs have shown promise for several cybersecurity tasks, such as
vulnerability assessment and threat analysis. In this work, we present a new
method for constructing a vulnerability knowledge graph from information in the
National Vulnerability Database (NVD). Our approach combines named entity
recognition (NER), relation extraction (RE), and entity prediction using a
combination of neural models, heuristic rules, and knowledge graph embeddings.
We demonstrate how our method helps to fix missing entities in knowledge graphs
used for cybersecurity and evaluate the performance. | cs.CR | Constructing a Knowledge Graph from Textual Descriptions of Software Vulnerabilities in the National Vulnerability Database |
2023-04-30T02:07:22Z | http://arxiv.org/pdf/2305.00366v1 | http://arxiv.org/abs/2305.00366v1 | Yuze Lou, Bailey Kuehl, Erin Bransom, Sergey Feldman, Aakanksha Naik, Doug Downey | 2023-04-30T02:07:22Z | cs.CL, cs.IR, cs.LG | Entity linking (EL) is the task of linking a textual mention to its
corresponding entry in a knowledge base, and is critical for many
knowledge-intensive NLP applications. When applied to tables in scientific
papers, EL is a step toward large-scale scientific knowledge bases that could
enable advanced scientific question answering and analytics. We present the
first dataset for EL in scientific tables. EL for scientific tables is
especially challenging because scientific knowledge bases can be very
incomplete, and disambiguating table mentions typically requires understanding
the papers's tet in addition to the table. Our dataset, S2abEL, focuses on EL
in machine learning results tables and includes hand-labeled cell types,
attributed sources, and entity links from the PaperswithCode taxonomy for 8,429
cells from 732 tables. We introduce a neural baseline method designed for EL on
scientific tables containing many out-of-knowledge-base mentions, and show that
it significantly outperforms a state-of-the-art generic table EL method. The
best baselines fall below human performance, and our analysis highlights
avenues for improvement. | cs.CL | S2abEL: A Dataset for Entity Linking from Scientific Tables |
2023-04-29T22:05:22Z | http://arxiv.org/pdf/2305.00350v1 | http://arxiv.org/abs/2305.00350v1 | Korawat Tanwisuth, Shujian Zhang, Huangjie Zheng, Pengcheng He, Mingyuan Zhou | 2023-04-29T22:05:22Z | cs.LG, cs.AI, cs.CL, cs.CV, stat.ML | Through prompting, large-scale pre-trained models have become more expressive
and powerful, gaining significant attention in recent years. Though these big
models have zero-shot capabilities, in general, labeled data are still required
to adapt them to downstream tasks. To overcome this critical limitation, we
propose an unsupervised fine-tuning framework to directly fine-tune the model
or prompt on the unlabeled target data. We demonstrate how to apply our method
to both language-augmented vision and masked-language models by aligning the
discrete distributions extracted from the prompts and target data. To verify
our approach's applicability, we conduct extensive experiments on image
classification, sentiment analysis, and natural language inference tasks.
Across 13 image-related tasks and 15 language-related ones, the proposed
approach achieves consistent improvements over the baselines. | cs.LG | POUF: Prompt-oriented unsupervised fine-tuning for large pre-trained models |
2023-04-29T16:13:13Z | http://arxiv.org/pdf/2305.00296v1 | http://arxiv.org/abs/2305.00296v1 | Xerxes D. Arsiwalla | 2023-04-29T16:13:13Z | q-bio.NC, cs.CL | In this commentary article to 'The Puzzle of Ideography' by Morin, we put
forth a new cognitive account of the puzzle of ideography, that complements the
standardization account of Morin. Efficient standardization of spoken language
is phenomenologically attributed to a modality effect coupled with chunking of
cognitive representations, further aided by multi-sensory integration and the
serialized nature of attention. These cognitive mechanisms are crucial for
explaining why languages dominate graphic codes for general-purpose human
communication. | q-bio.NC | A Cognitive Account of the Puzzle of Ideography |
2023-04-29T13:53:48Z | http://arxiv.org/pdf/2305.00262v1 | http://arxiv.org/abs/2305.00262v1 | Xiao Liu, Jian Zhang, Heng Zhang, Fuzhao Xue, Yang You | 2023-04-29T13:53:48Z | cs.CL, cs.AI | Compared with standard text, understanding dialogue is more challenging for
machines as the dynamic and unexpected semantic changes in each turn. To model
such inconsistent semantics, we propose a simple but effective Hierarchical
Dialogue Understanding model, HiDialog. Specifically, we first insert multiple
special tokens into a dialogue and propose the turn-level attention to learn
turn embeddings hierarchically. Then, a heterogeneous graph module is leveraged
to polish the learned embeddings. We evaluate our model on various dialogue
understanding tasks including dialogue relation extraction, dialogue emotion
recognition, and dialogue act classification. Results show that our simple
approach achieves state-of-the-art performance on all three tasks above. All
our source code is publicly available at https://github.com/ShawX825/HiDialog. | cs.CL | Hierarchical Dialogue Understanding with Special Tokens and Turn-level Attention |
2023-04-29T09:58:32Z | http://arxiv.org/pdf/2305.00217v6 | http://arxiv.org/abs/2305.00217v6 | Alexander Koplenig | 2023-05-30T05:42:08Z | cs.CL, stat.AP | In a recent paper published in the Journal of Language Evolution, Kauhanen,
Einhaus & Walkden (https://doi.org/10.1093/jole/lzad005, KEW) challenge the
results presented in one of my papers (Koplenig, Royal Society Open Science, 6,
181274 (2019), https://doi.org/10.1098/rsos.181274), in which I tried to show
through a series of statistical analyses that large numbers of L2 (second
language) speakers do not seem to affect the (grammatical or statistical)
complexity of a language. To this end, I focus on the way in which the
Ethnologue assesses language status: a language is characterised as vehicular
if, in addition to being used by L1 (first language) speakers, it should also
have a significant number of L2 users. KEW criticise both the use of
vehicularity as a (binary) indicator of whether a language has a significant
number of L2 users and the idea of imputing a zero proportion of L2 speakers to
non-vehicular languages whenever a direct estimate of that proportion is
unavailable. While I recognise the importance of post-publication commentary on
published research, I show in this rejoinder that both points of criticism are
explicitly mentioned and analysed in my paper. In addition, I also comment on
other points raised by KEW and demonstrate that both alternative analyses
offered by KEW do not stand up to closer scrutiny. | cs.CL | Still no evidence for an effect of the proportion of non-native speakers on language complexity -- A response to Kauhanen, Einhaus & Walkden (2023) |
2023-04-29T09:14:55Z | http://arxiv.org/pdf/2305.06173v1 | http://arxiv.org/abs/2305.06173v1 | Elwin Huaman, David Lindemann, Valeria Caruso, Jorge Luis Huaman | 2023-04-29T09:14:55Z | cs.CL, cs.AI | Over the last decade, the Web has increasingly become a space of language and
knowledge representation. However, it is only true for well-spread languages
and well-established communities, while minority communities and their
resources received less attention. In this paper, we propose QICHWABASE to
support the harmonization process of the Quechua language and knowledge, and
its community. For doing it, we adopt methods and tools that could become a
game changer in favour of Quechua communities around the world. We conclude
that the methodology and tools adopted on building QICHWABASE, which is a
Wikibase instance, could enhance the presence of minorities on the Web. | cs.CL | QICHWABASE: A Quechua Language and Knowledge Base for Quechua Communities |
2023-04-29T06:26:03Z | http://arxiv.org/pdf/2305.00182v1 | http://arxiv.org/abs/2305.00182v1 | David Alonso del Barrio, Daniel Gatica-Perez | 2023-04-29T06:26:03Z | cs.CL | This paper examines how the European press dealt with the no-vax reactions
against the Covid-19 vaccine and the dis- and misinformation associated with
this movement. Using a curated dataset of 1786 articles from 19 European
newspapers on the anti-vaccine movement over a period of 22 months in
2020-2021, we used Natural Language Processing techniques including topic
modeling, sentiment analysis, semantic relationship with word embeddings,
political analysis, named entity recognition, and semantic networks, to
understand the specific role of the European traditional press in the
disinformation ecosystem. The results of this multi-angle analysis demonstrate
that the European well-established press actively opposed a variety of hoaxes
mainly spread on social media, and was critical of the anti-vax trend,
regardless of the political orientation of the newspaper. This confirms the
relevance of studying the role of high-quality press in the disinformation
ecosystem. | cs.CL | Examining European Press Coverage of the Covid-19 No-Vax Movement: An NLP Framework |
2023-04-28T22:35:03Z | http://arxiv.org/pdf/2305.00118v2 | http://arxiv.org/abs/2305.00118v2 | Kent K. Chang, Mackenzie Cramer, Sandeep Soni, David Bamman | 2023-10-20T21:23:21Z | cs.CL | In this work, we carry out a data archaeology to infer books that are known
to ChatGPT and GPT-4 using a name cloze membership inference query. We find
that OpenAI models have memorized a wide collection of copyrighted materials,
and that the degree of memorization is tied to the frequency with which
passages of those books appear on the web. The ability of these models to
memorize an unknown set of books complicates assessments of measurement
validity for cultural analytics by contaminating test data; we show that models
perform much better on memorized books than on non-memorized books for
downstream tasks. We argue that this supports a case for open models whose
training data is known. | cs.CL | Speak, Memory: An Archaeology of Books Known to ChatGPT/GPT-4 |
2023-04-28T21:02:58Z | http://arxiv.org/pdf/2305.00090v1 | http://arxiv.org/abs/2305.00090v1 | Mingyang Wang, Heike Adel, Lukas Lange, Jannik Strötgen, Hinrich Schütze | 2023-04-28T21:02:58Z | cs.CL | This paper describes our system developed for the SemEval-2023 Task 12
"Sentiment Analysis for Low-resource African Languages using Twitter Dataset".
Sentiment analysis is one of the most widely studied applications in natural
language processing. However, most prior work still focuses on a small number
of high-resource languages. Building reliable sentiment analysis systems for
low-resource languages remains challenging, due to the limited training data in
this task. In this work, we propose to leverage language-adaptive and
task-adaptive pretraining on African texts and study transfer learning with
source language selection on top of an African language-centric pretrained
language model. Our key findings are: (1) Adapting the pretrained model to the
target language and task using a small yet relevant corpus improves performance
remarkably by more than 10 F1 score points. (2) Selecting source languages with
positive transfer gains during training can avoid harmful interference from
dissimilar languages, leading to better results in multilingual and
cross-lingual settings. In the shared task, our system wins 8 out of 15 tracks
and, in particular, performs best in the multilingual evaluation. | cs.CL | NLNDE at SemEval-2023 Task 12: Adaptive Pretraining and Source Language Selection for Low-Resource Multilingual Sentiment Analysis |
2023-04-28T20:03:46Z | http://arxiv.org/pdf/2305.00076v1 | http://arxiv.org/abs/2305.00076v1 | Saminu Mohammad Aliyu, Idris Abdulmumin, Shamsuddeen Hassan Muhammad, Ibrahim Said Ahmad, Saheed Abdullahi Salahudeen, Aliyu Yusuf, Falalu Ibrahim Lawan | 2023-04-28T20:03:46Z | cs.CL | We present the findings of our participation in the SemEval-2023 Task 10:
Explainable Detection of Online Sexism (EDOS) task, a shared task on offensive
language (sexism) detection on English Gab and Reddit dataset. We investigated
the effects of transferring two language models: XLM-T (sentiment
classification) and HateBERT (same domain -- Reddit) for multi-level
classification into Sexist or not Sexist, and other subsequent
sub-classifications of the sexist data. We also use synthetic classification of
unlabelled dataset and intermediary class information to maximize the
performance of our models. We submitted a system in Task A, and it ranked 49th
with F1-score of 0.82. This result showed to be competitive as it only
under-performed the best system by 0.052% F1-score. | cs.CL | HausaNLP at SemEval-2023 Task 10: Transfer Learning, Synthetic Data and Side-Information for Multi-Level Sexism Classification |
2023-04-28T19:27:26Z | http://arxiv.org/pdf/2305.00061v1 | http://arxiv.org/abs/2305.00061v1 | Zhengzhong Liang, Zeyu Zhang, Steven Bethard, Mihai Surdeanu | 2023-04-28T19:27:26Z | cs.CL, cs.AI | Languages models have been successfully applied to a variety of reasoning
tasks in NLP, yet the language models still suffer from compositional
generalization. In this paper we present Explainable Verbal Reasoner Plus
(EVR+), a reasoning framework that enhances language models' compositional
reasoning ability by (1) allowing the model to explicitly generate and execute
symbolic operators, and (2) allowing the model to decompose a complex task into
several simpler ones in a flexible manner. Compared with its predecessor
Explainable Verbal Reasoner (EVR) and other previous approaches adopting
similar ideas, our framework supports more diverse types of reasoning such as
nested loops and different types of recursion. To evaluate our reasoning
framework, we build a synthetic dataset with five tasks that require
compositional reasoning. Results show that our reasoning framework can enhance
the language model's compositional generalization performance on the five
tasks, using a fine-tuned language model. We also discussed the possibility and
the challenges to combine our reasoning framework with a few-shot prompted
language model. | cs.CL | Explainable Verbal Reasoner Plus (EVR+): A Natural Language Reasoning Framework that Supports Diverse Compositional Reasoning |
2023-04-28T19:00:43Z | http://arxiv.org/pdf/2305.00050v2 | http://arxiv.org/abs/2305.00050v2 | Emre Kıcıman, Robert Ness, Amit Sharma, Chenhao Tan | 2023-05-08T17:54:45Z | cs.AI, cs.CL, cs.CY, cs.HC, cs.LG, stat.ME | The causal capabilities of large language models (LLMs) is a matter of
significant debate, with critical implications for the use of LLMs in
societally impactful domains such as medicine, science, law, and policy. We
further our understanding of LLMs and their causal implications, considering
the distinctions between different types of causal reasoning tasks, as well as
the entangled threats of construct and measurement validity. LLM-based methods
establish new state-of-the-art accuracies on multiple causal benchmarks.
Algorithms based on GPT-3.5 and 4 outperform existing algorithms on a pairwise
causal discovery task (97%, 13 points gain), counterfactual reasoning task
(92%, 20 points gain), and actual causality (86% accuracy in determining
necessary and sufficient causes in vignettes). At the same time, LLMs exhibit
unpredictable failure modes and we provide some techniques to interpret their
robustness.
Crucially, LLMs perform these causal tasks while relying on sources of
knowledge and methods distinct from and complementary to non-LLM based
approaches. Specifically, LLMs bring capabilities so far understood to be
restricted to humans, such as using collected knowledge to generate causal
graphs or identifying background causal context from natural language. We
envision LLMs to be used alongside existing causal methods, as a proxy for
human domain knowledge and to reduce human effort in setting up a causal
analysis, one of the biggest impediments to the widespread adoption of causal
methods. We also see existing causal methods as promising tools for LLMs to
formalize, validate, and communicate their reasoning especially in high-stakes
scenarios.
In capturing common sense and domain knowledge about causal mechanisms and
supporting translation between natural language and formal methods, LLMs open
new frontiers for advancing the research, practice, and adoption of causality. | cs.AI | Causal Reasoning and Large Language Models: Opening a New Frontier for Causality |
2023-04-28T18:14:48Z | http://arxiv.org/pdf/2305.00034v1 | http://arxiv.org/abs/2305.00034v1 | Fantine Huot, Joshua Maynez, Shashi Narayan, Reinald Kim Amplayo, Kuzman Ganchev, Annie Louis, Anders Sandholm, Dipanjan Das, Mirella Lapata | 2023-04-28T18:14:48Z | cs.CL | While conditional generation models can now generate natural language well
enough to create fluent text, it is still difficult to control the generation
process, leading to irrelevant, repetitive, and hallucinated content. Recent
work shows that planning can be a useful intermediate step to render
conditional generation less opaque and more grounded. We present a web
browser-based demonstration for query-focused summarization that uses a
sequence of question-answer pairs, as a blueprint plan for guiding text
generation (i.e., what to say and in what order). We illustrate how users may
interact with the generated text and associated plan visualizations, e.g., by
editing and modifying the blueprint in order to improve or control the
generated output.
A short video demonstrating our system is available at
https://goo.gle/text-blueprint-demo. | cs.CL | Text-Blueprint: An Interactive Platform for Plan-based Conditional Generation |
2023-04-28T17:59:25Z | http://arxiv.org/pdf/2304.15010v1 | http://arxiv.org/abs/2304.15010v1 | Peng Gao, Jiaming Han, Renrui Zhang, Ziyi Lin, Shijie Geng, Aojun Zhou, Wei Zhang, Pan Lu, Conghui He, Xiangyu Yue, Hongsheng Li, Yu Qiao | 2023-04-28T17:59:25Z | cs.CV, cs.AI, cs.CL, cs.LG, cs.MM | How to efficiently transform large language models (LLMs) into instruction
followers is recently a popular research direction, while training LLM for
multi-modal reasoning remains less explored. Although the recent LLaMA-Adapter
demonstrates the potential to handle visual inputs with LLMs, it still cannot
generalize well to open-ended visual instructions and lags behind GPT-4. In
this paper, we present LLaMA-Adapter V2, a parameter-efficient visual
instruction model. Specifically, we first augment LLaMA-Adapter by unlocking
more learnable parameters (e.g., norm, bias and scale), which distribute the
instruction-following ability across the entire LLaMA model besides adapters.
Secondly, we propose an early fusion strategy to feed visual tokens only into
the early LLM layers, contributing to better visual knowledge incorporation.
Thirdly, a joint training paradigm of image-text pairs and
instruction-following data is introduced by optimizing disjoint groups of
learnable parameters. This strategy effectively alleviates the interference
between the two tasks of image-text alignment and instruction following and
achieves strong multi-modal reasoning with only a small-scale image-text and
instruction dataset. During inference, we incorporate additional expert models
(e.g. captioning/OCR systems) into LLaMA-Adapter to further enhance its image
understanding capability without incurring training costs. Compared to the
original LLaMA-Adapter, our LLaMA-Adapter V2 can perform open-ended multi-modal
instructions by merely introducing 14M parameters over LLaMA. The newly
designed framework also exhibits stronger language-only instruction-following
capabilities and even excels in chat interactions. Our code and models are
available at https://github.com/ZrrSkywalker/LLaMA-Adapter. | cs.CV | LLaMA-Adapter V2: Parameter-Efficient Visual Instruction Model |
2023-04-28T17:39:49Z | http://arxiv.org/pdf/2304.14999v1 | http://arxiv.org/abs/2304.14999v1 | George Pu, Anirudh Jain, Jihan Yin, Russell Kaplan | 2023-04-28T17:39:49Z | cs.CL, cs.AI | As foundation models continue to exponentially scale in size, efficient
methods of adaptation become increasingly critical. Parameter-efficient
fine-tuning (PEFT), a recent class of techniques that require only modifying a
small percentage of the model parameters, is currently the most popular method
for adapting large language models (LLMs). Several PEFT techniques have
recently been proposed with varying tradeoffs. We provide a comprehensive and
uniform benchmark of various PEFT techniques across a representative LLM, the
FLAN-T5 model, and evaluate model performance across different data scales of
classification and generation datasets. Based on this, we provide a framework
for choosing the optimal fine-tuning techniques given the task type and data
availability. Contrary to popular belief, we also empirically prove that PEFT
techniques converge slower than full tuning in low data scenarios, and posit
the amount of data required for PEFT methods to both perform well and converge
efficiently. Lastly, we further optimize these PEFT techniques by selectively
choosing which parts of the model to train, and find that these techniques can
be applied with significantly fewer parameters while maintaining and even
improving performance. | cs.CL | Empirical Analysis of the Strengths and Weaknesses of PEFT Techniques for LLMs |
2023-04-28T17:10:08Z | http://arxiv.org/pdf/2304.14986v2 | http://arxiv.org/abs/2304.14986v2 | Michele Cafagna, Lina M. Rojas-Barahona, Kees van Deemter, Albert Gatt | 2023-05-04T15:25:00Z | cs.CV, cs.AI, cs.CL | When applied to Image-to-text models, interpretability methods often provide
token-by-token explanations namely, they compute a visual explanation for each
token of the generated sequence. Those explanations are expensive to compute
and unable to comprehensively explain the model's output. Therefore, these
models often require some sort of approximation that eventually leads to
misleading explanations. We develop a framework based on SHAP, that allows for
generating comprehensive, meaningful explanations leveraging the meaning
representation of the output sequence as a whole. Moreover, by exploiting
semantic priors in the visual backbone, we extract an arbitrary number of
features that allows the efficient computation of Shapley values on large-scale
models, generating at the same time highly meaningful visual explanations. We
demonstrate that our method generates semantically more expressive explanations
than traditional methods at a lower compute cost and that it can be generalized
over other explainability methods. | cs.CV | Interpreting Vision and Language Generative Models with Semantic Visual Priors |
2023-04-28T16:12:18Z | http://arxiv.org/pdf/2304.14953v2 | http://arxiv.org/abs/2304.14953v2 | Michał Turski, Tomasz Stanisławek, Karol Kaczmarek, Paweł Dyda, Filip Graliński | 2023-06-06T07:35:17Z | cs.CL | In recent years, the field of document understanding has progressed a lot. A
significant part of this progress has been possible thanks to the use of
language models pretrained on large amounts of documents. However, pretraining
corpora used in the domain of document understanding are single domain,
monolingual, or nonpublic. Our goal in this paper is to propose an efficient
pipeline for creating a big-scale, diverse, multilingual corpus of PDF files
from all over the Internet using Common Crawl, as PDF files are the most
canonical types of documents as considered in document understanding. We
analysed extensively all of the steps of the pipeline and proposed a solution
which is a trade-off between data quality and processing time. We also share a
CCpdf corpus in a form or an index of PDF files along with a script for
downloading them, which produces a collection useful for language model
pretraining. The dataset and tools published with this paper offer researchers
the opportunity to develop even better multilingual language models. | cs.CL | CCpdf: Building a High Quality Corpus for Visually Rich Documents from Web Crawl Data |
2023-04-28T15:48:26Z | http://arxiv.org/pdf/2304.14936v1 | http://arxiv.org/abs/2304.14936v1 | Seif Laatiri, Pirashanth Ratnamogan, Joel Tang, Laurent Lam, William Vanhuffel, Fabien Caspani | 2023-04-28T15:48:26Z | cs.CL, cs.AI | Advances in the Visually-rich Document Understanding (VrDU) field and
particularly the Key-Information Extraction (KIE) task are marked with the
emergence of efficient Transformer-based approaches such as the LayoutLM
models. Despite the good performance of KIE models when fine-tuned on public
benchmarks, they still struggle to generalize on complex real-life use-cases
lacking sufficient document annotations. Our research highlighted that KIE
standard benchmarks such as SROIE and FUNSD contain significant similarity
between training and testing documents and can be adjusted to better evaluate
the generalization of models. In this work, we designed experiments to quantify
the information redundancy in public benchmarks, revealing a 75% template
replication in SROIE official test set and 16% in FUNSD. We also proposed
resampling strategies to provide benchmarks more representative of the
generalization ability of models. We showed that models not suited for document
analysis struggle on the adjusted splits dropping on average 10,5% F1 score on
SROIE and 3.5% on FUNSD compared to multi-modal models dropping only 7,5% F1 on
SROIE and 0.5% F1 on FUNSD. | cs.CL | Information Redundancy and Biases in Public Document Information Extraction Benchmarks |
2023-04-28T15:43:21Z | http://arxiv.org/pdf/2304.14933v2 | http://arxiv.org/abs/2304.14933v2 | Yi-Lin Sung, Linjie Li, Kevin Lin, Zhe Gan, Mohit Bansal, Lijuan Wang | 2023-10-11T15:08:51Z | cs.CV, cs.AI, cs.CL, cs.LG | Model merging (e.g., via interpolation or task arithmetic) fuses multiple
models trained on different tasks to generate a multi-task solution. The
technique has been proven successful in previous studies, where the models are
trained on similar tasks and with the same initialization. In this paper, we
expand on this concept to a multimodal setup by merging transformers trained on
different modalities. Furthermore, we conduct our study for a novel goal where
we can merge vision, language, and cross-modal transformers of a
modality-specific architecture to create a parameter-efficient
modality-agnostic architecture. Through comprehensive experiments, we
systematically investigate the key factors impacting model performance after
merging, including initialization, merging mechanisms, and model architectures.
We also propose two metrics that assess the distance between weights to be
merged and can serve as an indicator of the merging outcomes. Our analysis
leads to an effective training recipe for matching the performance of the
modality-agnostic baseline (i.e., pre-trained from scratch) via model merging.
Our method also outperforms naive merging significantly on various tasks, with
improvements of 3% on VQA, 7% on COCO retrieval, 25% on NLVR2, 14% on Flickr30k
and 3% on ADE20k. Our code is available at https://github.com/ylsung/vl-merging | cs.CV | An Empirical Study of Multimodal Model Merging |
2023-04-28T15:42:55Z | http://arxiv.org/pdf/2304.14931v2 | http://arxiv.org/abs/2304.14931v2 | Abdurahman Maarouf, Dominik Bär, Dominique Geissler, Stefan Feuerriegel | 2023-05-01T08:29:51Z | cs.CL | Online propaganda poses a severe threat to the integrity of societies.
However, existing datasets for detecting online propaganda have a key
limitation: they were annotated using weak labels that can be noisy and even
incorrect. To address this limitation, our work makes the following
contributions: (1) We present HQP: a novel dataset (N=30,000) for detecting
online propaganda with high-quality labels. To the best of our knowledge, HQP
is the first dataset for detecting online propaganda that was created through
human annotation. (2) We show empirically that state-of-the-art language models
fail in detecting online propaganda when trained with weak labels (AUC: 64.03).
In contrast, state-of-the-art language models can accurately detect online
propaganda when trained with our high-quality labels (AUC: 92.25), which is an
improvement of ~44%. (3) To address the cost of labeling, we extend our work to
few-shot learning. Specifically, we show that prompt-based learning using a
small sample of high-quality labels can still achieve a reasonable performance
(AUC: 80.27). Finally, we discuss implications for the NLP community to balance
the cost and quality of labeling. Crucially, our work highlights the importance
of high-quality labels for sensitive NLP tasks such as propaganda detection. | cs.CL | HQP: A Human-Annotated Dataset for Detecting Online Propaganda |
2023-04-28T13:14:36Z | http://arxiv.org/pdf/2304.14827v3 | http://arxiv.org/abs/2304.14827v3 | Chunkit Chan, Jiayang Cheng, Weiqi Wang, Yuxin Jiang, Tianqing Fang, Xin Liu, Yangqiu Song | 2024-01-26T10:33:08Z | cs.CL | This paper aims to quantitatively evaluate the performance of ChatGPT, an
interactive large language model, on inter-sentential relations such as
temporal relations, causal relations, and discourse relations. Given ChatGPT's
promising performance across various tasks, we proceed to carry out thorough
evaluations on the whole test sets of 11 datasets, including temporal and
causal relations, PDTB2.0-based, and dialogue-based discourse relations. To
ensure the reliability of our findings, we employ three tailored prompt
templates for each task, including the zero-shot prompt template, zero-shot
prompt engineering (PE) template, and in-context learning (ICL) prompt
template, to establish the initial baseline scores for all popular
sentence-pair relation classification tasks for the first time. Through our
study, we discover that ChatGPT exhibits exceptional proficiency in detecting
and reasoning about causal relations, albeit it may not possess the same level
of expertise in identifying the temporal order between two events. While it is
capable of identifying the majority of discourse relations with existing
explicit discourse connectives, the implicit discourse relation remains a
formidable challenge. Concurrently, ChatGPT demonstrates subpar performance in
the dialogue discourse parsing task that requires structural understanding in a
dialogue before being aware of the discourse relation. | cs.CL | ChatGPT Evaluation on Sentence Level Relations: A Focus on Temporal, Causal, and Discourse Relations |
2023-04-28T12:20:35Z | http://arxiv.org/pdf/2304.14803v1 | http://arxiv.org/abs/2304.14803v1 | Elisa Leonardelli, Alexandra Uma, Gavin Abercrombie, Dina Almanea, Valerio Basile, Tommaso Fornaciari, Barbara Plank, Verena Rieser, Massimo Poesio | 2023-04-28T12:20:35Z | cs.CL | NLP datasets annotated with human judgments are rife with disagreements
between the judges. This is especially true for tasks depending on subjective
judgments such as sentiment analysis or offensive language detection.
Particularly in these latter cases, the NLP community has come to realize that
the approach of 'reconciling' these different subjective interpretations is
inappropriate. Many NLP researchers have therefore concluded that rather than
eliminating disagreements from annotated corpora, we should preserve
them-indeed, some argue that corpora should aim to preserve all annotator
judgments. But this approach to corpus creation for NLP has not yet been widely
accepted. The objective of the LeWiDi series of shared tasks is to promote this
approach to developing NLP models by providing a unified framework for training
and evaluating with such datasets. We report on the second LeWiDi shared task,
which differs from the first edition in three crucial respects: (i) it focuses
entirely on NLP, instead of both NLP and computer vision tasks in its first
edition; (ii) it focuses on subjective tasks, instead of covering different
types of disagreements-as training with aggregated labels for subjective NLP
tasks is a particularly obvious misrepresentation of the data; and (iii) for
the evaluation, we concentrate on soft approaches to evaluation. This second
edition of LeWiDi attracted a wide array of participants resulting in 13 shared
task submission papers. | cs.CL | SemEval-2023 Task 11: Learning With Disagreements (LeWiDi) |
2023-04-28T12:19:47Z | http://arxiv.org/pdf/2304.14802v1 | http://arxiv.org/abs/2304.14802v1 | Shufang Xie, Huishuai Zhang, Junliang Guo, Xu Tan, Jiang Bian, Hany Hassan Awadalla, Arul Menezes, Tao Qin, Rui Yan | 2023-04-28T12:19:47Z | cs.CL, cs.AI, cs.LG, cs.NE | Transformer networks have become the preferred architecture for many tasks
due to their state-of-the-art performance. However, the optimal way to
implement residual connections in Transformer, which are essential for
effective training, is still debated. Two widely used variants are the
Post-Layer-Normalization (Post-LN) and Pre-Layer-Normalization (Pre-LN)
Transformers, which apply layer normalization after each residual block's
output or before each residual block's input, respectively. While both variants
enjoy their advantages, they also suffer from severe limitations: Post-LN
causes gradient vanishing issue that hinders training deep Transformers, and
Pre-LN causes representation collapse issue that limits model capacity. In this
paper, we propose ResiDual, a novel Transformer architecture with Pre-Post-LN
(PPLN), which fuses the connections in Post-LN and Pre-LN together and inherits
their advantages while avoids their limitations. We conduct both theoretical
analyses and empirical experiments to verify the effectiveness of ResiDual.
Theoretically, we prove that ResiDual has a lower bound on the gradient to
avoid the vanishing issue due to the residual connection from Pre-LN. Moreover,
ResiDual also has diverse model representations to avoid the collapse issue due
to the residual connection from Post-LN. Empirically, ResiDual outperforms both
Post-LN and Pre-LN on several machine translation benchmarks across different
network depths and data sizes. Thanks to the good theoretical and empirical
performance, ResiDual Transformer can serve as a foundation architecture for
different AI models (e.g., large language models). Our code is available at
https://github.com/microsoft/ResiDual. | cs.CL | ResiDual: Transformer with Dual Residual Connections |
2023-04-28T12:11:21Z | http://arxiv.org/pdf/2304.14796v1 | http://arxiv.org/abs/2304.14796v1 | Sonal Sannigrahi, Josef van Genabith, Cristina Espana-Bonet | 2023-04-28T12:11:21Z | cs.CL, cs.IR | Dense vector representations for textual data are crucial in modern NLP. Word
embeddings and sentence embeddings estimated from raw texts are key in
achieving state-of-the-art results in various tasks requiring semantic
understanding. However, obtaining embeddings at the document level is
challenging due to computational requirements and lack of appropriate data.
Instead, most approaches fall back on computing document embeddings based on
sentence representations. Although there exist architectures and models to
encode documents fully, they are in general limited to English and few other
high-resourced languages. In this work, we provide a systematic comparison of
methods to produce document-level representations from sentences based on
LASER, LaBSE, and Sentence BERT pre-trained multilingual models. We compare
input token number truncation, sentence averaging as well as some simple
windowing and in some cases new augmented and learnable approaches, on 3 multi-
and cross-lingual tasks in 8 languages belonging to 3 different language
families. Our task-based extrinsic evaluations show that, independently of the
language, a clever combination of sentence embeddings is usually better than
encoding the full document as a single unit, even when this is possible. We
demonstrate that while a simple sentence average results in a strong baseline
for classification tasks, more complex combinations are necessary for semantic
tasks. | cs.CL | Are the Best Multilingual Document Embeddings simply Based on Sentence Embeddings? |
2023-04-28T11:40:48Z | http://arxiv.org/pdf/2304.14780v1 | http://arxiv.org/abs/2304.14780v1 | Felix Stollenwerk | 2023-04-28T11:40:48Z | cs.CL, cs.AI | This paper provides a detailed discussion of the multilingual tokenizer used
for GPT-SW3. It was trained on the Nordic Pile using the SentencePiece library
and the BPE algorithm. We outline the tokenizer's most important features and
share details on its learned vocabulary. In addition, we systematically analyze
the properties and evaluate the performance of the tokenizer with regard to the
different languages present in the data. | cs.CL | Training and Evaluation of a Multilingual Tokenizer for GPT-SW3 |
2023-04-28T11:28:56Z | http://arxiv.org/pdf/2304.14770v2 | http://arxiv.org/abs/2304.14770v2 | Chengyuan Liu, Fubang Zhao, Yangyang Kang, Jingyuan Zhang, Xiang Zhou, Changlong Sun, Kun Kuang, Fei Wu | 2023-10-18T02:30:55Z | cs.CL | Universal Information Extraction (UIE) is an area of interest due to the
challenges posed by varying targets, heterogeneous structures, and
demand-specific schemas. However, previous works have only achieved limited
success by unifying a few tasks, such as Named Entity Recognition (NER) and
Relation Extraction (RE), which fall short of being authentic UIE models
particularly when extracting other general schemas such as quadruples and
quintuples. Additionally, these models used an implicit structural schema
instructor, which could lead to incorrect links between types, hindering the
model's generalization and performance in low-resource scenarios. In this
paper, we redefine the authentic UIE with a formal formulation that encompasses
almost all extraction schemas. To the best of our knowledge, we are the first
to introduce UIE for any kind of schemas. In addition, we propose RexUIE, which
is a Recursive Method with Explicit Schema Instructor for UIE. To avoid
interference between different types, we reset the position ids and attention
mask matrices. RexUIE shows strong performance under both full-shot and
few-shot settings and achieves State-of-the-Art results on the tasks of
extracting complex schemas. | cs.CL | RexUIE: A Recursive Method with Explicit Schema Instructor for Universal Information Extraction |
2023-04-28T11:26:17Z | http://arxiv.org/pdf/2304.14767v3 | http://arxiv.org/abs/2304.14767v3 | Mor Geva, Jasmijn Bastings, Katja Filippova, Amir Globerson | 2023-10-13T19:01:20Z | cs.CL | Transformer-based language models (LMs) are known to capture factual
knowledge in their parameters. While previous work looked into where factual
associations are stored, only little is known about how they are retrieved
internally during inference. We investigate this question through the lens of
information flow. Given a subject-relation query, we study how the model
aggregates information about the subject and relation to predict the correct
attribute. With interventions on attention edges, we first identify two
critical points where information propagates to the prediction: one from the
relation positions followed by another from the subject positions. Next, by
analyzing the information at these points, we unveil a three-step internal
mechanism for attribute extraction. First, the representation at the
last-subject position goes through an enrichment process, driven by the early
MLP sublayers, to encode many subject-related attributes. Second, information
from the relation propagates to the prediction. Third, the prediction
representation "queries" the enriched subject to extract the attribute. Perhaps
surprisingly, this extraction is typically done via attention heads, which
often encode subject-attribute mappings in their parameters. Overall, our
findings introduce a comprehensive view of how factual associations are stored
and extracted internally in LMs, facilitating future research on knowledge
localization and editing. | cs.CL | Dissecting Recall of Factual Associations in Auto-Regressive Language Models |
2023-04-28T10:40:34Z | http://arxiv.org/pdf/2304.14746v1 | http://arxiv.org/abs/2304.14746v1 | Liam Daly Manocchio, Siamak Layeghy, Wai Weng Lo, Gayan K. Kulatilleke, Mohanad Sarhan, Marius Portmann | 2023-04-28T10:40:34Z | cs.CR, cs.AI, cs.CL, cs.NE, cs.NI | This paper presents the FlowTransformer framework, a novel approach for
implementing transformer-based Network Intrusion Detection Systems (NIDSs).
FlowTransformer leverages the strengths of transformer models in identifying
the long-term behaviour and characteristics of networks, which are often
overlooked by most existing NIDSs. By capturing these complex patterns in
network traffic, FlowTransformer offers a flexible and efficient tool for
researchers and practitioners in the cybersecurity community who are seeking to
implement NIDSs using transformer-based models. FlowTransformer allows the
direct substitution of various transformer components, including the input
encoding, transformer, classification head, and the evaluation of these across
any flow-based network dataset. To demonstrate the effectiveness and efficiency
of the FlowTransformer framework, we utilise it to provide an extensive
evaluation of various common transformer architectures, such as GPT 2.0 and
BERT, on three commonly used public NIDS benchmark datasets. We provide results
for accuracy, model size and speed. A key finding of our evaluation is that the
choice of classification head has the most significant impact on the model
performance. Surprisingly, Global Average Pooling, which is commonly used in
text classification, performs very poorly in the context of NIDS. In addition,
we show that model size can be reduced by over 50\%, and inference and training
times improved, with no loss of accuracy, by making specific choices of input
encoding and classification head instead of other commonly used alternatives. | cs.CR | FlowTransformer: A Transformer Framework for Flow-based Network Intrusion Detection Systems |
2023-04-28T10:39:44Z | http://arxiv.org/pdf/2304.14745v1 | http://arxiv.org/abs/2304.14745v1 | Annerose Eichel, Helena Schlipf, Sabine Schulte im Walde | 2023-04-28T10:39:44Z | cs.CL, cs.AI, cs.IR | We propose a novel approach to learn domain-specific plausible materials for
components in the vehicle repair domain by probing Pretrained Language Models
(PLMs) in a cloze task style setting to overcome the lack of annotated
datasets. We devise a new method to aggregate salient predictions from a set of
cloze query templates and show that domain-adaptation using either a small,
high-quality or a customized Wikipedia corpus boosts performance. When
exploring resource-lean alternatives, we find a distilled PLM clearly
outperforming a classic pattern-based algorithm. Further, given that 98% of our
domain-specific components are multiword expressions, we successfully exploit
the compositionality assumption as a way to address data sparsity. | cs.CL | Made of Steel? Learning Plausible Materials for Components in the Vehicle Repair Domain |
2023-04-28T10:31:12Z | http://arxiv.org/pdf/2304.14738v1 | http://arxiv.org/abs/2304.14738v1 | Harsh Rangwani, Shrinivas Ramasubramanian, Sho Takemori, Kato Takashi, Yuhei Umeda, Venkatesh Babu Radhakrishnan | 2023-04-28T10:31:12Z | cs.LG, cs.CL, cs.CV | Self-training based semi-supervised learning algorithms have enabled the
learning of highly accurate deep neural networks, using only a fraction of
labeled data. However, the majority of work on self-training has focused on the
objective of improving accuracy, whereas practical machine learning systems can
have complex goals (e.g. maximizing the minimum of recall across classes, etc.)
that are non-decomposable in nature. In this work, we introduce the
Cost-Sensitive Self-Training (CSST) framework which generalizes the
self-training-based methods for optimizing non-decomposable metrics. We prove
that our framework can better optimize the desired non-decomposable metric
utilizing unlabeled data, under similar data distribution assumptions made for
the analysis of self-training. Using the proposed CSST framework, we obtain
practical self-training methods (for both vision and NLP tasks) for optimizing
different non-decomposable metrics using deep neural networks. Our results
demonstrate that CSST achieves an improvement over the state-of-the-art in
majority of the cases across datasets and objectives. | cs.LG | Cost-Sensitive Self-Training for Optimizing Non-Decomposable Metrics |
2023-04-28T10:15:25Z | http://arxiv.org/pdf/2304.14732v7 | http://arxiv.org/abs/2304.14732v7 | Shicheng Xu, Liang Pang, Huawei Shen, Xueqi Cheng, Tat-Seng Chua | 2024-02-24T16:54:29Z | cs.CL | Making the content generated by Large Language Model (LLM), accurate,
credible and traceable is crucial, especially in complex knowledge-intensive
tasks that require multi-step reasoning and each step needs knowledge to solve.
Retrieval-augmented generation is good potential to solve this problem.
However, where and how to introduce Information Retrieval (IR) to LLM is a big
challenge. Previous work has the problems that wrong knowledge retrieved by IR
misleads the LLM and interaction between IR and LLM breaks the reasoning chain
of LLM. This paper proposes a novel framework named
\textbf{Search-in-the-Chain} (SearChain) for the interaction between LLM and IR
to solve the challenges. First, LLM generates the reasoning chain named
Chain-of-Query (CoQ) where each node consists of an IR-oriented query-answer
pair. Second, IR verifies the answer of each node of CoQ. It corrects the
answer that is not consistent with the retrieved information when IR gives high
confidence, which improves the credibility. Third, LLM can indicate its missing
knowledge in CoQ and rely on IR to provide this knowledge to LLM. These
operations improve the accuracy in terms of reasoning and knowledge. Finally,
SearChain generates the reasoning process and marks references to supporting
documents for each reasoning step, which improves traceability. Interaction
with IR in SearChain forms a novel reasoning path based on a tree, which
enables LLM to dynamically modify the direction of reasoning. Experiments show
that SearChain outperforms state-of-the-art baselines on complex
knowledge-intensive tasks including multi-hop Q\&A, slot filling, fact
checking, and long-form Q\&A. | cs.CL | Search-in-the-Chain: Interactively Enhancing Large Language Models with Search for Knowledge-intensive Tasks |
2023-04-28T09:42:18Z | http://arxiv.org/pdf/2304.14721v4 | http://arxiv.org/abs/2304.14721v4 | Yuchen Xia, Manthan Shenoy, Nasser Jazdi, Michael Weyrich | 2023-07-24T09:49:55Z | cs.RO, cs.CL, cs.SE, cs.SY, eess.SY | In this paper, we present a novel framework that combines large language
models (LLMs), digital twins and industrial automation system to enable
intelligent planning and control of production processes. We retrofit the
automation system for a modular production facility and create executable
control interfaces of fine-granular functionalities and coarse-granular skills.
Low-level functionalities are executed by automation components, and high-level
skills are performed by automation modules. Subsequently, a digital twin system
is developed, registering these interfaces and containing additional
descriptive information about the production system. Based on the retrofitted
automation system and the created digital twins, LLM-agents are designed to
interpret descriptive information in the digital twins and control the physical
system through service interfaces. These LLM-agents serve as intelligent agents
on different levels within an automation system, enabling autonomous planning
and control of flexible production. Given a task instruction as input, the
LLM-agents orchestrate a sequence of atomic functionalities and skills to
accomplish the task. We demonstrate how our implemented prototype can handle
un-predefined tasks, plan a production process, and execute the operations.
This research highlights the potential of integrating LLMs into industrial
automation systems in the context of smart factory for more agile, flexible,
and adaptive production processes, while it also underscores the critical
insights and limitations for future work. Demos at:
https://github.com/YuchenXia/GPT4IndustrialAutomation | cs.RO | Towards autonomous system: flexible modular production system enhanced with large language model agents |
2023-04-28T07:32:00Z | http://arxiv.org/pdf/2304.14662v1 | http://arxiv.org/abs/2304.14662v1 | Tong Zhu, Guoliang Zhang, Zechang Li, Zijian Yu, Junfei Ren, Mengsong Wu, Zhefeng Wang, Baoxing Huai, Pingfu Chao, Wenliang Chen | 2023-04-28T07:32:00Z | cs.CL | Sentence-by-sentence information extraction from long documents is an
exhausting and error-prone task. As the indicator of document skeleton,
catalogs naturally chunk documents into segments and provide informative
cascade semantics, which can help to reduce the search space. Despite their
usefulness, catalogs are hard to be extracted without the assist from external
knowledge. For documents that adhere to a specific template, regular
expressions are practical to extract catalogs. However, handcrafted heuristics
are not applicable when processing documents from different sources with
diverse formats. To address this problem, we build a large manually annotated
corpus, which is the first dataset for the Catalog Extraction from Documents
(CED) task. Based on this corpus, we propose a transition-based framework for
parsing documents into catalog trees. The experimental results demonstrate that
our proposed method outperforms baseline systems and shows a good ability to
transfer. We believe the CED task could fill the gap between raw text segments
and information extraction tasks on extremely long documents. Data and code are
available at \url{https://github.com/Spico197/CatalogExtraction} | cs.CL | CED: Catalog Extraction from Documents |
2023-04-28T02:52:38Z | http://arxiv.org/pdf/2304.14599v1 | http://arxiv.org/abs/2304.14599v1 | Gunther Jikeli, Sameer Karali, Daniel Miehling, Katharina Soemer | 2023-04-28T02:52:38Z | cs.CL, cs.CY | One of the major challenges in automatic hate speech detection is the lack of
datasets that cover a wide range of biased and unbiased messages and that are
consistently labeled. We propose a labeling procedure that addresses some of
the common weaknesses of labeled datasets. We focus on antisemitic speech on
Twitter and create a labeled dataset of 6,941 tweets that cover a wide range of
topics common in conversations about Jews, Israel, and antisemitism between
January 2019 and December 2021 by drawing from representative samples with
relevant keywords. Our annotation process aims to strictly apply a commonly
used definition of antisemitism by forcing annotators to specify which part of
the definition applies, and by giving them the option to personally disagree
with the definition on a case-by-case basis. Labeling tweets that call out
antisemitism, report antisemitism, or are otherwise related to antisemitism
(such as the Holocaust) but are not actually antisemitic can help reduce false
positives in automated detection. The dataset includes 1,250 tweets (18%) that
are antisemitic according to the International Holocaust Remembrance Alliance
(IHRA) definition of antisemitism. It is important to note, however, that the
dataset is not comprehensive. Many topics are still not covered, and it only
includes tweets collected from Twitter between January 2019 and December 2021.
Additionally, the dataset only includes tweets that were written in English.
Despite these limitations, we hope that this is a meaningful contribution to
improving the automated detection of antisemitic speech. | cs.CL | Antisemitic Messages? A Guide to High-Quality Annotation and a Labeled Dataset of Tweets |
2023-04-28T01:53:54Z | http://arxiv.org/pdf/2304.14590v2 | http://arxiv.org/abs/2304.14590v2 | Sean Deyo, Veit Elser | 2023-06-06T00:46:49Z | cs.CL | We introduce the logical grammar emdebbing (LGE), a model inspired by
pregroup grammars and categorial grammars to enable unsupervised inference of
lexical categories and syntactic rules from a corpus of text. LGE produces
comprehensible output summarizing its inferences, has a completely transparent
process for producing novel sentences, and can learn from as few as a hundred
sentences. | cs.CL | A logical word embedding for learning grammar |
2023-04-28T01:22:47Z | http://arxiv.org/pdf/2304.14585v1 | http://arxiv.org/abs/2304.14585v1 | Feng Xie, Xiang Zeng, Bin Zhou, Yusong Tan | 2023-04-28T01:22:47Z | cs.CL, cs.SI | Entity alignment (EA) which links equivalent entities across different
knowledge graphs (KGs) plays a crucial role in knowledge fusion. In recent
years, graph neural networks (GNNs) have been successfully applied in many
embedding-based EA methods. However, existing GNN-based methods either suffer
from the structural heterogeneity issue that especially appears in the real KG
distributions or ignore the heterogeneous representation learning for unseen
(unlabeled) entities, which would lead the model to overfit on few alignment
seeds (i.e., training data) and thus cause unsatisfactory alignment
performance. To enhance the EA ability, we propose GAEA, a novel EA approach
based on graph augmentation. In this model, we design a simple Entity-Relation
(ER) Encoder to generate latent representations for entities via jointly
modeling comprehensive structural information and rich relation semantics.
Moreover, we use graph augmentation to create two graph views for margin-based
alignment learning and contrastive entity representation learning, thus
mitigating structural heterogeneity and further improving the model's alignment
performance. Extensive experiments conducted on benchmark datasets demonstrate
the effectiveness of our method. | cs.CL | Improving Knowledge Graph Entity Alignment with Graph Augmentation |
2023-04-27T22:21:52Z | http://arxiv.org/pdf/2304.14553v1 | http://arxiv.org/abs/2304.14553v1 | Hendrik Kempt, Alon Lavie, Saskia K. Nagel | 2023-04-27T22:21:52Z | cs.AI, cs.CL, cs.HC | The strive to make AI applications "safe" has led to the development of
safety-measures as the main or even sole normative requirement of their
permissible use. Similar can be attested to the latest version of chatbots,
such as chatGPT. In this view, if they are "safe", they are supposed to be
permissible to deploy. This approach, which we call "safety-normativity", is
rather limited in solving the emerging issues that chatGPT and other chatbots
have caused thus far. In answering this limitation, in this paper we argue for
limiting chatbots in the range of topics they can chat about according to the
normative concept of appropriateness. We argue that rather than looking for
"safety" in a chatbot's utterances to determine what they may and may not say,
we ought to assess those utterances according to three forms of
appropriateness: technical-discursive, social, and moral. We then spell out
what requirements for chatbots follow from these forms of appropriateness to
avoid the limits of previous accounts: positionality, acceptability, and value
alignment (PAVA). With these in mind, we may be able to determine what a
chatbot may and may not say. Lastly, one initial suggestion is to use challenge
sets, specifically designed for appropriateness, as a validation method. | cs.AI | Appropriateness is all you need! |
2023-04-27T21:51:42Z | http://arxiv.org/pdf/2304.14543v1 | http://arxiv.org/abs/2304.14543v1 | S. M. Seals, Valerie L. Shalin | 2023-04-27T21:51:42Z | cs.CL | The summarization of conversation, that is, discourse over discourse,
elevates pragmatic considerations as a pervasive limitation of both
summarization and other applications of contemporary conversational AI.
Building on impressive progress in both semantics and syntax, pragmatics
concerns meaning in the practical sense. In this paper, we discuss several
challenges in both summarization of conversations and other conversational AI
applications, drawing on relevant theoretical work. We illustrate the
importance of pragmatics with so-called star sentences, syntactically
acceptable propositions that are pragmatically inappropriate in conversation or
its summary. Because the baseline for quality of AI is indistinguishability
from human behavior, we draw heavily on the psycho-linguistics literature, and
label our complaints as "Turing Test Triggers" (TTTs). We discuss implications
for the design and evaluation of conversation summarization methods and
conversational AI applications like voice assistants and chatbots | cs.CL | Discourse over Discourse: The Need for an Expanded Pragmatic Focus in Conversational AI |
2023-04-27T21:08:05Z | http://arxiv.org/pdf/2304.14535v2 | http://arxiv.org/abs/2304.14535v2 | Hamza Kheddar, Yassine Himeur, Somaya Al-Maadeed, Abbes Amira, Faycal Bensaali | 2023-07-31T11:58:18Z | cs.SD, cs.CL, cs.LG, eess.AS | Automatic speech recognition (ASR) has recently become an important challenge
when using deep learning (DL). It requires large-scale training datasets and
high computational and storage resources. Moreover, DL techniques and machine
learning (ML) approaches in general, hypothesize that training and testing data
come from the same domain, with the same input feature space and data
distribution characteristics. This assumption, however, is not applicable in
some real-world artificial intelligence (AI) applications. Moreover, there are
situations where gathering real data is challenging, expensive, or rarely
occurring, which can not meet the data requirements of DL models. deep transfer
learning (DTL) has been introduced to overcome these issues, which helps
develop high-performing models using real datasets that are small or slightly
different but related to the training data. This paper presents a comprehensive
survey of DTL-based ASR frameworks to shed light on the latest developments and
helps academics and professionals understand current challenges. Specifically,
after presenting the DTL background, a well-designed taxonomy is adopted to
inform the state-of-the-art. A critical analysis is then conducted to identify
the limitations and advantages of each framework. Moving on, a comparative
study is introduced to highlight the current challenges before deriving
opportunities for future research. | cs.SD | Deep Transfer Learning for Automatic Speech Recognition: Towards Better Generalization |
2023-04-27T20:30:46Z | http://arxiv.org/pdf/2304.14522v1 | http://arxiv.org/abs/2304.14522v1 | Hamed Zamani, Michael Bendersky | 2023-04-27T20:30:46Z | cs.IR, cs.CL, cs.LG | Dense retrieval models use bi-encoder network architectures for learning
query and document representations. These representations are often in the form
of a vector representation and their similarities are often computed using the
dot product function. In this paper, we propose a new representation learning
framework for dense retrieval. Instead of learning a vector for each query and
document, our framework learns a multivariate distribution and uses negative
multivariate KL divergence to compute the similarity between distributions. For
simplicity and efficiency reasons, we assume that the distributions are
multivariate normals and then train large language models to produce mean and
variance vectors for these distributions. We provide a theoretical foundation
for the proposed framework and show that it can be seamlessly integrated into
the existing approximate nearest neighbor algorithms to perform retrieval
efficiently. We conduct an extensive suite of experiments on a wide range of
datasets, and demonstrate significant improvements compared to competitive
dense retrieval models. | cs.IR | Multivariate Representation Learning for Information Retrieval |
2023-04-27T20:05:36Z | http://arxiv.org/pdf/2304.14514v1 | http://arxiv.org/abs/2304.14514v1 | Gary Wang, Kyle Kastner, Ankur Bapna, Zhehuai Chen, Andrew Rosenberg, Bhuvana Ramabhadran, Yu Zhang | 2023-04-27T20:05:36Z | cs.CL, cs.LG, cs.SD, eess.AS | Recently, a number of approaches to train speech models by incorpo-rating
text into end-to-end models have been developed, with Mae-stro advancing
state-of-the-art automatic speech recognition (ASR)and Speech Translation (ST)
performance. In this paper, we expandour understanding of the resulting shared
speech-text representationswith two types of analyses. First we examine the
limits of speech-free domain adaptation, finding that a corpus-specific
duration modelfor speech-text alignment is the most important component for
learn-ing a shared speech-text representation. Second, we inspect the
sim-ilarities between activations of unimodal (speech or text) encodersas
compared to the activations of a shared encoder. We find that theshared encoder
learns a more compact and overlapping speech-textrepresentation than the
uni-modal encoders. We hypothesize that thispartially explains the
effectiveness of the Maestro shared speech-textrepresentations. | cs.CL | Understanding Shared Speech-Text Representations |
2023-04-27T20:00:51Z | http://arxiv.org/pdf/2304.14511v1 | http://arxiv.org/abs/2304.14511v1 | Kevin Denamganaï, Sondess Missaoui, James Alfred Walker | 2023-04-27T20:00:51Z | cs.CL | Natural languages are powerful tools wielded by human beings to communicate
information. Among their desirable properties, compositionality has been the
main focus in the context of referential games and variants, as it promises to
enable greater systematicity to the agents which would wield it. The concept of
disentanglement has been shown to be of paramount importance to learned
representations that generalise well in deep learning, and is thought to be a
necessary condition to enable systematicity. Thus, this paper investigates how
do compositionality at the level of the emerging languages, disentanglement at
the level of the learned representations, and systematicity relate to each
other in the context of visual referential games. Firstly, we find that visual
referential games that are based on the Obverter architecture outperforms
state-of-the-art unsupervised learning approach in terms of many major
disentanglement metrics. Secondly, we expand the previously proposed Positional
Disentanglement (PosDis) metric for compositionality to (re-)incorporate some
concerns pertaining to informativeness and completeness features found in the
Mutual Information Gap (MIG) disentanglement metric it stems from. This
extension allows for further discrimination between the different kind of
compositional languages that emerge in the context of Obverter-based
referential games, in a way that neither the referential game accuracy nor
previous metrics were able to capture. Finally we investigate whether the
resulting (emergent) systematicity, as measured by zero-shot compositional
learning tests, correlates with any of the disentanglement and compositionality
metrics proposed so far. Throughout the training process, statically
significant correlation coefficients can be found both positive and negative
depending on the moment of the measure. | cs.CL | Visual Referential Games Further the Emergence of Disentangled Representations |
2023-04-27T18:30:18Z | http://arxiv.org/pdf/2304.14456v1 | http://arxiv.org/abs/2304.14456v1 | David Alonso del Barrio, Daniel Gatica-Perez | 2023-04-27T18:30:18Z | cs.CL, cs.HC | Identifying the frames of news is important to understand the articles'
vision, intention, message to be conveyed, and which aspects of the news are
emphasized. Framing is a widely studied concept in journalism, and has emerged
as a new topic in computing, with the potential to automate processes and
facilitate the work of journalism professionals. In this paper, we study this
issue with articles related to the Covid-19 anti-vaccine movement. First, to
understand the perspectives used to treat this theme, we developed a protocol
for human labeling of frames for 1786 headlines of No-Vax movement articles of
European newspapers from 5 countries. Headlines are key units in the written
press, and worth of analysis as many people only read headlines (or use them to
guide their decision for further reading.) Second, considering advances in
Natural Language Processing (NLP) with large language models, we investigated
two approaches for frame inference of news headlines: first with a GPT-3.5
fine-tuning approach, and second with GPT-3.5 prompt-engineering. Our work
contributes to the study and analysis of the performance that these models have
to facilitate journalistic tasks like classification of frames, while
understanding whether the models are able to replicate human perception in the
identification of these frames. | cs.CL | Framing the News:From Human Perception to Large Language Model Inferences |
2023-04-27T18:29:05Z | http://arxiv.org/pdf/2304.14454v3 | http://arxiv.org/abs/2304.14454v3 | Chaoyi Wu, Weixiong Lin, Xiaoman Zhang, Ya Zhang, Yanfeng Wang, Weidi Xie | 2023-08-25T14:08:38Z | cs.CL | Recently, Large Language Models (LLMs) have showcased remarkable capabilities
in natural language understanding. While demonstrating proficiency in everyday
conversations and question-answering situations, these models frequently
struggle in domains that require precision, such as medical applications, due
to their lack of domain-specific knowledge. In this paper, we describe the
procedure for building a powerful, open-source language model specifically
designed for medicine applications, termed as PMC-LLaMA. Our contributions are
threefold: (i) we systematically investigate the process of adapting a
general-purpose foundation language model towards medical domain, this involves
data-centric knowledge injection through the integration of 4.8M biomedical
academic papers and 30K medical textbooks, as well as comprehensive fine-tuning
for alignment with domain-specific instructions; (ii) we contribute a
large-scale, comprehensive dataset for instruction tuning. This dataset
encompasses medical question-answering (QA), rationale for reasoning, and
conversational dialogues, comprising a total of 202M tokens; (iii) we conduct
thorough ablation studies to demonstrate the effectiveness of each proposed
component. While evaluating on various public medical question-answering
benchmarks, our lightweight PMCLLaMA, which consists of only 13 billion
parameters, exhibits superior performance, even surpassing ChatGPT. All models,
codes, datasets can be found in https://github.com/chaoyi-wu/PMC-LLaMA. | cs.CL | PMC-LLaMA: Towards Building Open-source Language Models for Medicine |
2023-04-27T18:19:24Z | http://arxiv.org/pdf/2304.14447v1 | http://arxiv.org/abs/2304.14447v1 | Nguyen Anh Tu, Hoang Thi Thu Uyen, Tu Minh Phuong, Ngo Xuan Bach | 2023-04-27T18:19:24Z | cs.CL | In this paper, we propose using deep neural networks to extract important
information from Vietnamese legal questions, a fundamental task towards
building a question answering system in the legal domain. Given a legal
question in natural language, the goal is to extract all the segments that
contain the needed information to answer the question. We introduce a deep
model that solves the task in three stages. First, our model leverages recent
advanced autoencoding language models to produce contextual word embeddings,
which are then combined with character-level and POS-tag information to form
word representations. Next, bidirectional long short-term memory networks are
employed to capture the relations among words and generate sentence-level
representations. At the third stage, borrowing ideas from graph-based
dependency parsing methods which provide a global view on the input sentence,
we use biaffine classifiers to estimate the probability of each pair of
start-end words to be an important segment. Experimental results on a public
Vietnamese legal dataset show that our model outperforms the previous work by a
large margin, achieving 94.79% in the F1 score. The results also prove the
effectiveness of using contextual features extracted from pre-trained language
models combined with other types of features such as character-level and
POS-tag features when training on a limited dataset. | cs.CL | Analyzing Vietnamese Legal Questions Using Deep Neural Networks with Biaffine Classifiers |
2023-04-27T17:59:53Z | http://arxiv.org/pdf/2304.14405v1 | http://arxiv.org/abs/2304.14405v1 | Ta Duc Huy, Nguyen Anh Tu, Tran Hoang Vu, Nguyen Phuc Minh, Nguyen Phan, Trung H. Bui, Steven Q. H. Truong | 2023-04-27T17:59:53Z | cs.CL | Existing medical text datasets usually take the form of question and answer
pairs that support the task of natural language generation, but lacking the
composite annotations of the medical terms. In this study, we publish a
Vietnamese dataset of medical questions from patients with sentence-level and
entity-level annotations for the Intent Classification and Named Entity
Recognition tasks. The tag sets for two tasks are in medical domain and can
facilitate the development of task-oriented healthcare chatbots with better
comprehension of queries from patients. We train baseline models for the two
tasks and propose a simple self-supervised training strategy with span-noise
modelling that substantially improves the performance. Dataset and code will be
published at https://github.com/tadeephuy/ViMQ | cs.CL | ViMQ: A Vietnamese Medical Question Dataset for Healthcare Dialogue System Development |
2023-04-27T17:58:49Z | http://arxiv.org/pdf/2304.14402v3 | http://arxiv.org/abs/2304.14402v3 | Minghao Wu, Abdul Waheed, Chiyu Zhang, Muhammad Abdul-Mageed, Alham Fikri Aji | 2024-01-29T02:58:23Z | cs.CL | Large language models (LLMs) with instruction fine-tuning demonstrate
superior generative capabilities. However, these models are resource-intensive.
To alleviate this issue, we explore distilling knowledge from instruction-tuned
LLMs into much smaller ones. To this end, we carefully develop a large set of
2.58M instructions based on both existing and newly-generated instructions. In
addition to being sizable, we design our instructions to cover a broad set of
topics to ensure diversity. Extensive analysis of our instruction dataset
confirms its diversity, and we generate responses for these instructions using
gpt-3.5-turbo. Leveraging these instructions, we fine-tune a diverse herd of
models, collectively referred to as LaMini-LM, which includes models from both
the encoder-decoder and decoder-only families, with varying sizes. We evaluate
the performance of our models using automatic metrics on 15 different natural
language processing (NLP) benchmarks, as well as through human assessment. The
results demonstrate that our proposed LaMini-LM models are comparable to
competitive baselines, while being much smaller in size. | cs.CL | LaMini-LM: A Diverse Herd of Distilled Models from Large-Scale Instructions |
2023-04-27T17:57:58Z | http://arxiv.org/pdf/2304.14399v2 | http://arxiv.org/abs/2304.14399v2 | Alisa Liu, Zhaofeng Wu, Julian Michael, Alane Suhr, Peter West, Alexander Koller, Swabha Swayamdipta, Noah A. Smith, Yejin Choi | 2023-10-20T05:46:14Z | cs.CL | Ambiguity is an intrinsic feature of natural language. Managing ambiguity is
a key part of human language understanding, allowing us to anticipate
misunderstanding as communicators and revise our interpretations as listeners.
As language models (LMs) are increasingly employed as dialogue interfaces and
writing aids, handling ambiguous language is critical to their success. We
characterize ambiguity in a sentence by its effect on entailment relations with
another sentence, and collect AmbiEnt, a linguist-annotated benchmark of 1,645
examples with diverse kinds of ambiguity. We design a suite of tests based on
AmbiEnt, presenting the first evaluation of pretrained LMs to recognize
ambiguity and disentangle possible meanings. We find that the task remains
extremely challenging, including for GPT-4, whose generated disambiguations are
considered correct only 32% of the time in human evaluation, compared to 90%
for disambiguations in our dataset. Finally, to illustrate the value of
ambiguity-sensitive tools, we show that a multilabel NLI model can flag
political claims in the wild that are misleading due to ambiguity. We encourage
the field to rediscover the importance of ambiguity for NLP. | cs.CL | We're Afraid Language Models Aren't Modeling Ambiguity |
2023-04-27T17:57:19Z | http://arxiv.org/pdf/2304.14395v1 | http://arxiv.org/abs/2304.14395v1 | Mirac Suzgun, Stuart M. Shieber, Dan Jurafsky | 2023-04-27T17:57:19Z | cs.CL, cs.DL | We introduce string2string, an open-source library that offers a
comprehensive suite of efficient algorithms for a broad range of
string-to-string problems. It includes traditional algorithmic solutions as
well as recent advanced neural approaches to tackle various problems in string
alignment, distance measurement, lexical and semantic search, and similarity
analysis -- along with several helpful visualization tools and metrics to
facilitate the interpretation and analysis of these methods. Notable algorithms
featured in the library include the Smith-Waterman algorithm for pairwise local
alignment, the Hirschberg algorithm for global alignment, the Wagner-Fisher
algorithm for edit distance, BARTScore and BERTScore for similarity analysis,
the Knuth-Morris-Pratt algorithm for lexical search, and Faiss for semantic
search. Besides, it wraps existing efficient and widely-used implementations of
certain frameworks and metrics, such as sacreBLEU and ROUGE, whenever it is
appropriate and suitable. Overall, the library aims to provide extensive
coverage and increased flexibility in comparison to existing libraries for
strings. It can be used for many downstream applications, tasks, and problems
in natural-language processing, bioinformatics, and computational social
sciences. It is implemented in Python, easily installable via pip, and
accessible through a simple API. Source code, documentation, and tutorials are
all available on our GitHub page: https://github.com/stanfordnlp/string2string. | cs.CL | string2string: A Modern Python Library for String-to-String Algorithms |
2023-04-27T17:55:13Z | http://arxiv.org/pdf/2304.14391v4 | http://arxiv.org/abs/2304.14391v4 | Nikolaos Gkanatsios, Ayush Jain, Zhou Xian, Yunchu Zhang, Christopher Atkeson, Katerina Fragkiadaki | 2024-01-23T15:52:28Z | cs.RO, cs.AI, cs.CL, cs.CV, cs.LG | Language is compositional; an instruction can express multiple relation
constraints to hold among objects in a scene that a robot is tasked to
rearrange. Our focus in this work is an instructable scene-rearranging
framework that generalizes to longer instructions and to spatial concept
compositions never seen at training time. We propose to represent
language-instructed spatial concepts with energy functions over relative object
arrangements. A language parser maps instructions to corresponding energy
functions and an open-vocabulary visual-language model grounds their arguments
to relevant objects in the scene. We generate goal scene configurations by
gradient descent on the sum of energy functions, one per language predicate in
the instruction. Local vision-based policies then re-locate objects to the
inferred goal locations. We test our model on established instruction-guided
manipulation benchmarks, as well as benchmarks of compositional instructions we
introduce. We show our model can execute highly compositional instructions
zero-shot in simulation and in the real world. It outperforms
language-to-action reactive policies and Large Language Model planners by a
large margin, especially for long instructions that involve compositions of
multiple spatial concepts. Simulation and real-world robot execution videos, as
well as our code and datasets are publicly available on our website:
https://ebmplanner.github.io. | cs.RO | Energy-based Models are Zero-Shot Planners for Compositional Scene Rearrangement |
2023-04-27T17:39:11Z | http://arxiv.org/pdf/2304.14364v2 | http://arxiv.org/abs/2304.14364v2 | Albert Yu Sun, Varun Nair, Elliot Schumacher, Anitha Kannan | 2024-04-03T23:57:12Z | cs.CL, cs.AI, cs.LG | A wave of new task-based virtual assistants has been fueled by increasingly
powerful large language models (LLMs), such as GPT-4 (OpenAI, 2023). A major
challenge in deploying LLM-based virtual conversational assistants in real
world settings is ensuring they operate within what is admissible for the task.
To overcome this challenge, the designers of these virtual assistants rely on
an independent guardrail system that verifies the virtual assistant's output
aligns with the constraints required for the task. However, relying on commonly
used, prompt-based guardrails can be difficult to engineer correctly and
comprehensively. To address these challenges, we propose CONSCENDI. We use
CONSCENDI to exhaustively generate training data with two key LLM-powered
components: scenario-augmented generation and contrastive training examples.
When generating conversational data, we generate a set of rule-breaking
scenarios, which enumerate a diverse set of high-level ways a rule can be
violated. This scenario-guided approach produces a diverse training set and
provides chatbot designers greater control. To generate contrastive examples,
we prompt the LLM to alter conversations with violations into acceptable
conversations to enable fine-grained distinctions. We then use this data,
generated by CONSCENDI, to train a smaller model. We find that CONSCENDI
results in guardrail models that improve over baselines in multiple dialogue
domains. | cs.CL | CONSCENDI: A Contrastive and Scenario-Guided Distillation Approach to Guardrail Models for Virtual Assistants |
2023-04-27T17:33:49Z | http://arxiv.org/pdf/2304.14354v1 | http://arxiv.org/abs/2304.14354v1 | Oluwatosin Ogundare, Srinath Madasu, Nathanial Wiggins | 2023-04-27T17:33:49Z | cs.CL | Large Language Models (LLMs) have shown great potential in solving complex
problems in various fields, including oil and gas engineering and other
industrial engineering disciplines like factory automation, PLC programming
etc. However, automatic identification of strong and weak solutions to
fundamental physics equations governing several industrial processes remain a
challenging task. This paper identifies the limitation of current LLM
approaches, particularly ChatGPT in selected practical problems native to oil
and gas engineering but not exclusively. The performance of ChatGPT in solving
complex problems in oil and gas engineering is discussed and the areas where
LLMs are most effective are presented. | cs.CL | Industrial Engineering with Large Language Models: A case study of ChatGPT's performance on Oil & Gas problems |
2023-04-27T17:06:20Z | http://arxiv.org/pdf/2304.14333v1 | http://arxiv.org/abs/2304.14333v1 | Filip Klubička, Vasudevan Nedumpozhimana, John D. Kelleher | 2023-04-27T17:06:20Z | cs.CL, cs.AI, cs.LG, 68T50 | The goal of this paper is to learn more about how idiomatic information is
structurally encoded in embeddings, using a structural probing method. We
repurpose an existing English verbal multi-word expression (MWE) dataset to
suit the probing framework and perform a comparative probing study of static
(GloVe) and contextual (BERT) embeddings. Our experiments indicate that both
encode some idiomatic information to varying degrees, but yield conflicting
evidence as to whether idiomaticity is encoded in the vector norm, leaving this
an open question. We also identify some limitations of the used dataset and
highlight important directions for future work in improving its suitability for
a probing analysis. | cs.CL | Idioms, Probing and Dangerous Things: Towards Structural Probing for Idiomaticity in Vector Space |
2023-04-27T16:39:15Z | http://arxiv.org/pdf/2304.14318v2 | http://arxiv.org/abs/2304.14318v2 | Yonatan Bitton, Shlomi Cohen-Ganor, Ido Hakimi, Yoad Lewenberg, Roee Aharoni, Enav Weinreb | 2023-12-26T16:00:48Z | cs.CL | One of the exciting capabilities of recent language models for dialog is
their ability to independently search for relevant information to ground a
given dialog response. However, obtaining training data to teach models how to
issue search queries is time and resource consuming. In this work, we propose
q2d: an automatic data generation pipeline that generates information-seeking
dialogs from questions. We prompt a large language model (PaLM) to create
conversational versions of question answering datasets, and use it to improve
query generation models that communicate with external search APIs to ground
dialog responses. Unlike previous approaches which relied on human written
dialogs with search queries, our method allows to automatically generate
query-based grounded dialogs with better control and scale. Our experiments
demonstrate that: (1) For query generation on the QReCC dataset, models trained
on our synthetically-generated data achieve 90%--97% of the performance of
models trained on the human-generated data; (2) We can successfully generate
data for training dialog models in new domains without any existing dialog data
as demonstrated on the multi-hop MuSiQue and Bamboogle QA datasets. (3) We
perform a thorough analysis of the generated dialogs showing that humans find
them of high quality and struggle to distinguish them from human-written
dialogs. | cs.CL | q2d: Turning Questions into Dialogs to Teach Models How to Search |
2023-04-27T16:38:17Z | http://arxiv.org/pdf/2304.14317v2 | http://arxiv.org/abs/2304.14317v2 | Terry Yue Zhuo | 2024-01-22T17:06:50Z | cs.AI, cs.CL, cs.SE | Recent advancements in the field of natural language generation have
facilitated the use of large language models to assess the quality of generated
text. Although these models have shown promising results in tasks such as
machine translation and summarization, their applicability in code intelligence
tasks remains limited without human involvement. The complexity of programming
concepts required for such tasks makes it difficult to develop evaluation
metrics that align with human judgment. Token-matching-based metrics, such as
BLEU, have demonstrated weak correlations with human practitioners in code
intelligence tasks. Moreover, utilizing human-written test suites to evaluate
functional correctness can be challenging in domains with low resources. To
overcome these obstacles, we propose \texttt{ICE-Score}, a new evaluation
metric via instructing large language models (LLMs) for code assessments. Our
metric addresses the limitations of existing approaches by achieving superior
correlations with functional correctness and human preferences, without the
need for test oracles or references. We evaluate the efficacy of our metric on
two different aspects (\textit{human preference} and \textit{execution
success}) and four programming languages. Our results demonstrate that our
metric surpasses state-of-the-art metrics for code generation, delivering high
levels of accuracy and consistency across various programming languages and
tasks. We also make our evaluation metric and datasets available to the
public\footnote{\url{https://github.com/terryyz/ice-score}}, encouraging
further research in evaluating code intelligence tasks. | cs.AI | ICE-Score: Instructing Large Language Models to Evaluate Code |
2023-04-27T15:56:34Z | http://arxiv.org/pdf/2304.14293v2 | http://arxiv.org/abs/2304.14293v2 | Wangchunshu Zhou, Yuchen Eleanor Jiang, Ethan Wilcox, Ryan Cotterell, Mrinmaya Sachan | 2023-06-08T06:33:23Z | cs.CL, cs.AI, cs.LG | Large language models generate fluent texts and can follow natural language
instructions to solve a wide range of tasks without task-specific training.
Nevertheless, it is notoriously difficult to control their generation to
satisfy the various constraints required by different applications. In this
work, we present InstructCTG, a controlled text generation framework that
incorporates different constraints by conditioning on natural language
descriptions and demonstrations of the constraints. In particular, we first
extract the underlying constraints of natural texts through a combination of
off-the-shelf NLP tools and simple heuristics. We then verbalize the
constraints into natural language instructions to form weakly supervised
training data. By prepending natural language descriptions of the constraints
and a few demonstrations, we fine-tune a pre-trained language model to
incorporate various types of constraints. Compared to existing search-based or
score-based methods, InstructCTG is more flexible to different constraint types
and has a much smaller impact on the generation quality and speed because it
does not modify the decoding procedure. Additionally, InstructCTG allows the
model to adapt to new constraints without re-training through the use of
few-shot task generalization and in-context learning abilities of
instruction-tuned language models. | cs.CL | Controlled Text Generation with Natural Language Instructions |
2023-04-27T15:46:09Z | http://arxiv.org/pdf/2304.14286v1 | http://arxiv.org/abs/2304.14286v1 | Kosuke Yamada, Ryohei Sasano, Koichi Takeda | 2023-04-27T15:46:09Z | cs.CL | Recent studies have demonstrated the usefulness of contextualized word
embeddings in unsupervised semantic frame induction. However, they have also
revealed that generic contextualized embeddings are not always consistent with
human intuitions about semantic frames, which causes unsatisfactory performance
for frame induction based on contextualized embeddings. In this paper, we
address supervised semantic frame induction, which assumes the existence of
frame-annotated data for a subset of predicates in a corpus and aims to build a
frame induction model that leverages the annotated data. We propose a model
that uses deep metric learning to fine-tune a contextualized embedding model,
and we apply the fine-tuned contextualized embeddings to perform semantic frame
induction. Our experiments on FrameNet show that fine-tuning with deep metric
learning considerably improves the clustering evaluation scores, namely, the
B-cubed F-score and Purity F-score, by about 8 points or more. We also
demonstrate that our approach is effective even when the number of training
instances is small. | cs.CL | Semantic Frame Induction with Deep Metric Learning |
2023-04-27T15:01:20Z | http://arxiv.org/pdf/2304.14241v1 | http://arxiv.org/abs/2304.14241v1 | Egil Rønningstad, Erik Velldal, Lilja Øvrelid | 2023-04-27T15:01:20Z | cs.CL | This paper explores the task of identifying the overall sentiment expressed
towards volitional entities (persons and organizations) in a document -- what
we refer to as Entity-Level Sentiment Analysis (ELSA). While identifying
sentiment conveyed towards an entity is well researched for shorter texts like
tweets, we find little to no research on this specific task for longer texts
with multiple mentions and opinions towards the same entity. This lack of
research would be understandable if ELSA can be derived from existing tasks and
models. To assess this, we annotate a set of professional reviews for their
overall sentiment towards each volitional entity in the text. We sample from
data already annotated for document-level, sentence-level, and target-level
sentiment in a multi-domain review corpus, and our results indicate that there
is no single proxy task that provides this overall sentiment we seek for the
entities at a satisfactory level of performance. We present a suite of
experiments aiming to assess the contribution towards ELSA provided by
document-, sentence-, and target-level sentiment analysis, and provide a
discussion of their shortcomings. We show that sentiment in our dataset is
expressed not only with an entity mention as target, but also towards targets
with a sentiment-relevant relation to a volitional entity. In our data, these
relations extend beyond anaphoric coreference resolution, and our findings call
for further research of the topic. Finally, we also present a survey of
previous relevant work. | cs.CL | Entity-Level Sentiment Analysis (ELSA): An exploratory task survey |
2023-04-27T14:55:23Z | http://arxiv.org/pdf/2304.14238v2 | http://arxiv.org/abs/2304.14238v2 | Michael Schlichtkrull, Nedjma Ousidhoum, Andreas Vlachos | 2023-11-08T12:10:49Z | cs.CL | Automated fact-checking is often presented as an epistemic tool that
fact-checkers, social media consumers, and other stakeholders can use to fight
misinformation. Nevertheless, few papers thoroughly discuss how. We document
this by analysing 100 highly-cited papers, and annotating epistemic elements
related to intended use, i.e., means, ends, and stakeholders. We find that
narratives leaving out some of these aspects are common, that many papers
propose inconsistent means and ends, and that the feasibility of suggested
strategies rarely has empirical backing. We argue that this vagueness actively
hinders the technology from reaching its goals, as it encourages overclaiming,
limits criticism, and prevents stakeholder feedback. Accordingly, we provide
several recommendations for thinking and writing about the use of fact-checking
artefacts. | cs.CL | The Intended Uses of Automated Fact-Checking Artefacts: Why, How and Who |
2023-04-27T14:45:55Z | http://arxiv.org/pdf/2304.14233v2 | http://arxiv.org/abs/2304.14233v2 | Tao Shen, Guodong Long, Xiubo Geng, Chongyang Tao, Tianyi Zhou, Daxin Jiang | 2023-08-02T02:06:28Z | cs.CL, cs.IR | In this work, we propose a simple method that applies a large language model
(LLM) to large-scale retrieval in zero-shot scenarios. Our method, the Language
language model as Retriever (LameR), is built upon no other neural models but
an LLM, while breaking brute-force combinations of retrievers with LLMs and
lifting the performance of zero-shot retrieval to be very competitive on
benchmark datasets. Essentially, we propose to augment a query with its
potential answers by prompting LLMs with a composition of the query and the
query's in-domain candidates. The candidates, regardless of correct or wrong,
are obtained by a vanilla retrieval procedure on the target collection. As a
part of the prompts, they are likely to help LLM generate more precise answers
by pattern imitation or candidate summarization. Even if all the candidates are
wrong, the prompts at least make LLM aware of in-collection patterns and
genres. Moreover, due to the low performance of a self-supervised retriever,
the LLM-based query augmentation becomes less effective as the retriever
bottlenecks the whole pipeline. Therefore, we propose to leverage a
non-parametric lexicon-based method (e.g., BM25) as the retrieval module to
capture query-document overlap in a literal fashion. As such, LameR makes the
retrieval procedure transparent to the LLM, thus circumventing the performance
bottleneck. | cs.CL | Large Language Models are Strong Zero-Shot Retriever |
2023-04-27T14:32:23Z | http://arxiv.org/pdf/2304.14221v1 | http://arxiv.org/abs/2304.14221v1 | Nayla Escribano, German Rigau, Rodrigo Agerri | 2023-04-27T14:32:23Z | cs.CL | Detecting and normalizing temporal expressions is an essential step for many
NLP tasks. While a variety of methods have been proposed for detection, best
normalization approaches rely on hand-crafted rules. Furthermore, most of them
have been designed only for English. In this paper we present a modular
multilingual temporal processing system combining a fine-tuned Masked Language
Model for detection, and a grammar-based normalizer. We experiment in Spanish
and English and compare with HeidelTime, the state-of-the-art in multilingual
temporal processing. We obtain best results in gold timex normalization, timex
detection and type recognition, and competitive performance in the combined
TempEval-3 relaxed value metric. A detailed error analysis shows that detecting
only those timexes for which it is feasible to provide a normalization is
highly beneficial in this last metric. This raises the question of which is the
best strategy for timex processing, namely, leaving undetected those timexes
for which is not easy to provide normalization rules or aiming for high
coverage. | cs.CL | A Modular Approach for Multilingual Timex Detection and Normalization using Deep Learning and Grammar-based methods |
2023-04-27T13:51:18Z | http://arxiv.org/pdf/2304.14189v1 | http://arxiv.org/abs/2304.14189v1 | Egil Rønningstad | 2023-04-27T13:51:18Z | cs.CL | Our contribution to the 2023 AfriSenti-SemEval shared task 12: Sentiment
Analysis for African Languages, provides insight into how a multilingual large
language model can be a resource for sentiment analysis in languages not seen
during pretraining. The shared task provides datasets of a variety of African
languages from different language families. The languages are to various
degrees related to languages used during pretraining, and the language data
contain various degrees of code-switching. We experiment with both monolingual
and multilingual datasets for the final fine-tuning, and find that with the
provided datasets that contain samples in the thousands, monolingual
fine-tuning yields the best results. | cs.CL | UIO at SemEval-2023 Task 12: Multilingual fine-tuning for sentiment classification in low-resource languages |
2023-04-27T13:33:08Z | http://arxiv.org/pdf/2304.14179v1 | http://arxiv.org/abs/2304.14179v1 | Neele Falk, Annerose Eichel, Prisca Piccirilli | 2023-04-27T13:33:08Z | cs.CL | Persuasion techniques detection in news in a multi-lingual setup is
non-trivial and comes with challenges, including little training data. Our
system successfully leverages (back-)translation as data augmentation
strategies with multi-lingual transformer models for the task of detecting
persuasion techniques. The automatic and human evaluation of our augmented data
allows us to explore whether (back-)translation aid or hinder performance. Our
in-depth analyses indicate that both data augmentation strategies boost
performance; however, balancing human-produced and machine-generated data seems
to be crucial. | cs.CL | NAP at SemEval-2023 Task 3: Is Less Really More? (Back-)Translation as Data Augmentation Strategies for Detecting Persuasion Techniques |
2023-04-27T13:27:01Z | http://arxiv.org/pdf/2304.14178v3 | http://arxiv.org/abs/2304.14178v3 | Qinghao Ye, Haiyang Xu, Guohai Xu, Jiabo Ye, Ming Yan, Yiyang Zhou, Junyang Wang, Anwen Hu, Pengcheng Shi, Yaya Shi, Chenliang Li, Yuanhong Xu, Hehong Chen, Junfeng Tian, Qi Qian, Ji Zhang, Fei Huang, Jingren Zhou | 2024-03-29T08:13:38Z | cs.CL, cs.CV, cs.LG | Large language models (LLMs) have demonstrated impressive zero-shot abilities
on a variety of open-ended tasks, while recent research has also explored the
use of LLMs for multi-modal generation. In this study, we introduce mPLUG-Owl,
a novel training paradigm that equips LLMs with multi-modal abilities through
modularized learning of foundation LLM, a visual knowledge module, and a visual
abstractor module. This approach can support multiple modalities and facilitate
diverse unimodal and multimodal abilities through modality collaboration. The
training paradigm of mPLUG-Owl involves a two-stage method for aligning image
and text, which learns visual knowledge with the assistance of LLM while
maintaining and even improving the generation abilities of LLM. In the first
stage, the visual knowledge module and abstractor module are trained with a
frozen LLM module to align the image and text. In the second stage,
language-only and multi-modal supervised datasets are used to jointly fine-tune
a low-rank adaption (LoRA) module on LLM and the abstractor module by freezing
the visual knowledge module. We carefully build a visually-related instruction
evaluation set OwlEval. Experimental results show that our model outperforms
existing multi-modal models, demonstrating mPLUG-Owl's impressive instruction
and visual understanding ability, multi-turn conversation ability, and
knowledge reasoning ability. Besides, we observe some unexpected and exciting
abilities such as multi-image correlation and scene text understanding, which
makes it possible to leverage it for harder real scenarios, such as vision-only
document comprehension. Our code, pre-trained model, instruction-tuned models,
and evaluation set are available at https://github.com/X-PLUG/mPLUG-Owl. The
online demo is available at https://www.modelscope.cn/studios/damo/mPLUG-Owl. | cs.CL | mPLUG-Owl: Modularization Empowers Large Language Models with Multimodality |
2023-04-27T13:25:43Z | http://arxiv.org/pdf/2304.14177v2 | http://arxiv.org/abs/2304.14177v2 | Roberto Martínez-Cruz, Alvaro J. López-López, José Portela | 2023-06-29T13:40:42Z | cs.CL, cs.AI | Transformer-based language models, including ChatGPT, have demonstrated
exceptional performance in various natural language generation tasks. However,
there has been limited research evaluating ChatGPT's keyphrase generation
ability, which involves identifying informative phrases that accurately reflect
a document's content. This study seeks to address this gap by comparing
ChatGPT's keyphrase generation performance with state-of-the-art models, while
also testing its potential as a solution for two significant challenges in the
field: domain adaptation and keyphrase generation from long documents. We
conducted experiments on six publicly available datasets from scientific
articles and news domains, analyzing performance on both short and long
documents. Our results show that ChatGPT outperforms current state-of-the-art
models in all tested datasets and environments, generating high-quality
keyphrases that adapt well to diverse domains and document lengths. | cs.CL | ChatGPT vs State-of-the-Art Models: A Benchmarking Study in Keyphrase Generation Task |
2023-04-27T11:37:18Z | http://arxiv.org/pdf/2304.14108v5 | http://arxiv.org/abs/2304.14108v5 | Samir Yitzhak Gadre, Gabriel Ilharco, Alex Fang, Jonathan Hayase, Georgios Smyrnis, Thao Nguyen, Ryan Marten, Mitchell Wortsman, Dhruba Ghosh, Jieyu Zhang, Eyal Orgad, Rahim Entezari, Giannis Daras, Sarah Pratt, Vivek Ramanujan, Yonatan Bitton, Kalyani Marathe, Stephen Mussmann, Richard Vencu, Mehdi Cherti, Ranjay Krishna, Pang Wei Koh, Olga Saukh, Alexander Ratner, Shuran Song, Hannaneh Hajishirzi, Ali Farhadi, Romain Beaumont, Sewoong Oh, Alex Dimakis, Jenia Jitsev, Yair Carmon, Vaishaal Shankar, Ludwig Schmidt | 2023-10-20T17:01:44Z | cs.CV, cs.CL, cs.LG | Multimodal datasets are a critical component in recent breakthroughs such as
Stable Diffusion and GPT-4, yet their design does not receive the same research
attention as model architectures or training algorithms. To address this
shortcoming in the ML ecosystem, we introduce DataComp, a testbed for dataset
experiments centered around a new candidate pool of 12.8 billion image-text
pairs from Common Crawl. Participants in our benchmark design new filtering
techniques or curate new data sources and then evaluate their new dataset by
running our standardized CLIP training code and testing the resulting model on
38 downstream test sets. Our benchmark consists of multiple compute scales
spanning four orders of magnitude, which enables the study of scaling trends
and makes the benchmark accessible to researchers with varying resources. Our
baseline experiments show that the DataComp workflow leads to better training
sets. In particular, our best baseline, DataComp-1B, enables training a CLIP
ViT-L/14 from scratch to 79.2% zero-shot accuracy on ImageNet, outperforming
OpenAI's CLIP ViT-L/14 by 3.7 percentage points while using the same training
procedure and compute. We release DataComp and all accompanying code at
www.datacomp.ai. | cs.CV | DataComp: In search of the next generation of multimodal datasets |
2023-04-27T11:33:48Z | http://arxiv.org/pdf/2304.14106v2 | http://arxiv.org/abs/2304.14106v2 | Shangqing Tu, Chunyang Li, Jifan Yu, Xiaozhi Wang, Lei Hou, Juanzi Li | 2024-06-18T00:33:25Z | cs.CL, cs.AI | ChatGPT has achieved great success and can be considered to have acquired an
infrastructural status. There are abundant works for evaluating ChatGPT on
benchmarks. However, existing benchmarks encounter two challenges: (1)
Disregard for periodical evaluation and (2) Lack of fine-grained features. In
this paper, we construct ChatLog, an ever-updating dataset with large-scale
records of diverse long-form ChatGPT responses for 21 NLP benchmarks from
March, 2023 to now. We conduct a comprehensive performance evaluation to find
that most capabilities of ChatGPT improve over time except for some abilities,
and there exists a step-wise evolving pattern of ChatGPT. We further analyze
the inherent characteristics of ChatGPT by extracting the knowledge and
linguistic features. We find some stable features that stay unchanged and apply
them on the detection of ChatGPT-generated texts to improve the robustness of
cross-version detection. We will continuously maintain our project at
\url{https://github.com/THU-KEG/ChatLog/}. | cs.CL | ChatLog: Carefully Evaluating the Evolution of ChatGPT Across Time |
2023-04-27T11:32:48Z | http://arxiv.org/pdf/2304.14104v4 | http://arxiv.org/abs/2304.14104v4 | Morris Alper, Hadar Averbuch-Elor | 2023-09-18T17:50:31Z | cs.CV, cs.CL, cs.LG | Interactions between humans are diverse and context-dependent, but previous
works have treated them as categorical, disregarding the heavy tail of possible
interactions. We propose a new paradigm of learning human-human interactions as
free text from a single still image, allowing for flexibility in modeling the
unlimited space of situations and relationships between people. To overcome the
absence of data labelled specifically for this task, we use knowledge
distillation applied to synthetic caption data produced by a large language
model without explicit supervision. We show that the pseudo-labels produced by
this procedure can be used to train a captioning model to effectively
understand human-human interactions in images, as measured by a variety of
metrics that measure textual and semantic faithfulness and factual groundedness
of our predictions. We further show that our approach outperforms SOTA image
captioning and situation recognition models on this task. We will release our
code and pseudo-labels along with Waldo and Wenda, a manually-curated test set
for still image human-human interaction understanding. | cs.CV | Learning Human-Human Interactions in Images from Weak Textual Supervision |
2023-04-27T10:05:57Z | http://arxiv.org/pdf/2304.14072v1 | http://arxiv.org/abs/2304.14072v1 | Linyang Li, Pengyu Wang, Ke Ren, Tianxiang Sun, Xipeng Qiu | 2023-04-27T10:05:57Z | cs.CL | The extraordinary performance of large language models (LLMs) heightens the
importance of detecting whether the context is generated by an AI system. More
importantly, while more and more companies and institutions release their LLMs,
the origin can be hard to trace. Since LLMs are heading towards the time of
AGI, similar to the origin tracing in anthropology, it is of great importance
to trace the origin of LLMs. In this paper, we first raise the concern of the
origin tracing of LLMs and propose an effective method to trace and detect
AI-generated contexts. We introduce a novel algorithm that leverages the
contrastive features between LLMs and extracts model-wise features to trace the
text origins. Our proposed method works under both white-box and black-box
settings therefore can be widely generalized to detect various LLMs.(e.g. can
be generalized to detect GPT-3 models without the GPT-3 models). Also, our
proposed method requires only limited data compared with the supervised
learning methods and can be extended to trace new-coming model origins. We
construct extensive experiments to examine whether we can trace the origins of
given texts. We provide valuable observations based on the experimental
results, such as the difficulty level of AI origin tracing, and the AI origin
similarities, and call for ethical concerns of LLM providers. We are releasing
all codes and data as a toolkit and benchmark for future AI origin tracing and
detecting studies. \footnote{We are releasing all available resource at
\url{https://github.com/OpenLMLab/}.} | cs.CL | Origin Tracing and Detecting of LLMs |
2023-04-27T07:32:37Z | http://arxiv.org/pdf/2304.13994v3 | http://arxiv.org/abs/2304.13994v3 | Dmytro Kalpakchi, Johan Boye | 2023-06-22T13:02:23Z | cs.CL | We present SweCTRL-Mini, a large Swedish language model that can be used for
inference and fine-tuning on a single consumer-grade GPU. The model is based on
the CTRL architecture by Keskar, McCann, Varshney, Xiong, and Socher (2019),
which means that users of the SweCTRL-Mini model can control the genre of the
generated text by inserting special tokens in the generation prompts.
SweCTRL-Mini is trained on a subset of the Swedish part of the mC4 corpus and a
set of Swedish novels. In this article, we provide (1) a detailed account of
the utilized training data and text pre-processing steps, to the extent that it
is possible to check whether a specific phrase/source was a part of the
training data, and (2) an evaluation of the model on both discriminative tasks,
using automatic evaluation methods, and generative tasks, using human referees.
We also compare the generative capabilities of the model with those of GPT-3.
SweCTRL-Mini is fully open and available for download. | cs.CL | SweCTRL-Mini: a data-transparent Transformer-based large language model for controllable text generation in Swedish |
2023-04-27T07:24:29Z | http://arxiv.org/pdf/2304.13989v1 | http://arxiv.org/abs/2304.13989v1 | Kia Kirstein Hansen, Rob van der Goot | 2023-04-27T07:24:29Z | cs.CL | The Wall Street Journal section of the Penn Treebank has been the de-facto
standard for evaluating POS taggers for a long time, and accuracies over 97\%
have been reported. However, less is known about out-of-domain tagger
performance, especially with fine-grained label sets. Using data from Elder
Scrolls Fandom, a wiki about the \textit{Elder Scrolls} video game universe, we
create a modest dataset for qualitatively evaluating the cross-domain
performance of two POS taggers: the Stanford tagger (Toutanova et al. 2003) and
Bilty (Plank et al. 2016), both trained on WSJ. Our analyses show that
performance on tokens seen during training is almost as good as in-domain
performance, but accuracy on unknown tokens decreases from 90.37% to 78.37%
(Stanford) and 87.84\% to 80.41\% (Bilty) across domains. Both taggers struggle
with proper nouns and inconsistent capitalization. | cs.CL | Cross-Domain Evaluation of POS Taggers: From Wall Street Journal to Fandom Wiki |
Subsets and Splits
No saved queries yet
Save your SQL queries to embed, download, and access them later. Queries will appear here once saved.