publication_date
stringlengths 20
20
| doc_url
stringlengths 33
34
| id
stringlengths 33
34
| authors
stringlengths 5
20.1k
| update_date
stringlengths 20
20
| category_all
stringlengths 5
138
| abstract
stringlengths 123
3.09k
| category_primary
stringclasses 78
values | title
stringlengths 10
211
|
|---|---|---|---|---|---|---|---|---|
2024-02-22T13:26:56Z | http://arxiv.org/pdf/2402.14536v1 | http://arxiv.org/abs/2402.14536v1 | Siyin Wang, Jie Zhou, Qin Chen, Qi Zhang, Tao Gui, Xuanjing Huang | 2024-02-22T13:26:56Z | cs.CL | Domain adaption has been widely adapted for cross-domain sentiment analysis
to transfer knowledge from the source domain to the target domain. Whereas,
most methods are proposed under the assumption that the target (test) domain is
known, making them fail to generalize well on unknown test data that is not
always available in practice. In this paper, we focus on the problem of domain
generalization for cross-domain sentiment analysis. Specifically, we propose a
backdoor adjustment-based causal model to disentangle the domain-specific and
domain-invariant representations that play essential roles in tackling domain
shift. First, we rethink the cross-domain sentiment analysis task in a causal
view to model the causal-and-effect relationships among different variables.
Then, to learn an invariant feature representation, we remove the effect of
domain confounders (e.g., domain knowledge) using the backdoor adjustment. A
series of experiments over many homologous and diverse datasets show the great
performance and robustness of our model by comparing it with the
state-of-the-art domain generalization baselines. | cs.CL | Domain Generalization via Causal Adjustment for Cross-Domain Sentiment Analysis |
2024-02-22T13:25:17Z | http://arxiv.org/pdf/2402.14533v1 | http://arxiv.org/abs/2402.14533v1 | Ariel Rosenfeld, Teddy Lazebnik | 2024-02-22T13:25:17Z | cs.CL | Large Language Models (LLMs) are capable of generating text that is similar
to or surpasses human quality. However, it is unclear whether LLMs tend to
exhibit distinctive linguistic styles akin to how human authors do. Through a
comprehensive linguistic analysis, we compare the vocabulary, Part-Of-Speech
(POS) distribution, dependency distribution, and sentiment of texts generated
by three of the most popular LLMS today (GPT-3.5, GPT-4, and Bard) to diverse
inputs. The results point to significant linguistic variations which, in turn,
enable us to attribute a given text to its LLM origin with a favorable 88\%
accuracy using a simple off-the-shelf classification model. Theoretical and
practical implications of this intriguing finding are discussed. | cs.CL | Whose LLM is it Anyway? Linguistic Comparison and LLM Attribution for GPT-3.5, GPT-4 and Bard |
2024-02-22T13:24:10Z | http://arxiv.org/pdf/2402.14531v1 | http://arxiv.org/abs/2402.14531v1 | Ziqi Yin, Hao Wang, Kaito Horio, Daisuke Kawahara, Satoshi Sekine | 2024-02-22T13:24:10Z | cs.CL | We investigate the impact of politeness levels in prompts on the performance
of large language models (LLMs). Polite language in human communications often
garners more compliance and effectiveness, while rudeness can cause aversion,
impacting response quality. We consider that LLMs mirror human communication
traits, suggesting they align with human cultural norms. We assess the impact
of politeness in prompts on LLMs across English, Chinese, and Japanese tasks.
We observed that impolite prompts often result in poor performance, but overly
polite language does not guarantee better outcomes. The best politeness level
is different according to the language. This phenomenon suggests that LLMs not
only reflect human behavior but are also influenced by language, particularly
in different cultural contexts. Our findings highlight the need to factor in
politeness for cross-cultural natural language processing and LLM usage. | cs.CL | Should We Respect LLMs? A Cross-Lingual Study on the Influence of Prompt Politeness on LLM Performance |
2024-02-22T13:20:53Z | http://arxiv.org/pdf/2402.14526v2 | http://arxiv.org/abs/2402.14526v2 | Yunfan Shao, Linyang Li, Zhaoye Fei, Hang Yan, Dahua Lin, Xipeng Qiu | 2024-06-03T06:48:34Z | cs.CL, cs.AI | Data plays a fundamental role in the training of Large Language Models
(LLMs). While attention has been paid to the collection and composition of
datasets, determining the data sampling strategy in training remains an open
question. Most LLMs are trained with a simple strategy, random sampling.
However, this sampling strategy ignores the unbalanced nature of training data
distribution, which can be sub-optimal. In this paper, we propose ClusterClip
Sampling to balance the text distribution of training data for better model
training. Specifically, ClusterClip Sampling utilizes data clustering to
reflect the data distribution of the training set and balances the common
samples and rare samples during training based on the cluster results. A
repetition clip operation is introduced to mitigate the overfitting issue led
by samples from certain clusters. Extensive experiments validate the
effectiveness of ClusterClip Sampling, which outperforms random sampling and
other cluster-based sampling variants under various training datasets and large
language models. | cs.CL | Balanced Data Sampling for Language Model Training with Clustering |
2024-02-22T13:15:49Z | http://arxiv.org/pdf/2402.14523v2 | http://arxiv.org/abs/2402.14523v2 | Rendi Chevi, Alham Fikri Aji | 2024-06-27T15:14:58Z | cs.CL, cs.SD, eess.AS | We often verbally express emotions in a multifaceted manner, they may vary in
their intensities and may be expressed not just as a single but as a mixture of
emotions. This wide spectrum of emotions is well-studied in the structural
model of emotions, which represents variety of emotions as derivative products
of primary emotions with varying degrees of intensity. In this paper, we
propose an emotional text-to-speech design to simulate a wider spectrum of
emotions grounded on the structural model. Our proposed design, Daisy-TTS,
incorporates a prosody encoder to learn emotionally-separable prosody embedding
as a proxy for emotion. This emotion representation allows the model to
simulate: (1) Primary emotions, as learned from the training samples, (2)
Secondary emotions, as a mixture of primary emotions, (3) Intensity-level, by
scaling the emotion embedding, and (4) Emotions polarity, by negating the
emotion embedding. Through a series of perceptual evaluations, Daisy-TTS
demonstrated overall higher emotional speech naturalness and emotion
perceiveability compared to the baseline. | cs.CL | Daisy-TTS: Simulating Wider Spectrum of Emotions via Prosody Embedding Decomposition |
2024-02-22T13:13:31Z | http://arxiv.org/pdf/2402.14522v1 | http://arxiv.org/abs/2402.14522v1 | Xinyu Wang, Hainiu Xu, Lin Gui, Yulan He | 2024-02-22T13:13:31Z | cs.CL, cs.LG | Task embedding, a meta-learning technique that captures task-specific
information, has become prevalent, especially in areas such as multi-task
learning, model editing, and interpretability. However, it faces challenges
with the emergence of prompt-guided Large Language Models (LLMs) operating in a
gradientfree manner. Existing task embedding methods rely on fine-tuned,
task-specific language models, which hinders the adaptability of task
embeddings across diverse models, especially prompt-based LLMs. To unleash the
power of task embedding in the era of LLMs, we propose a framework for unified
task embeddings (FUTE), harmonizing task embeddings from various models,
including smaller language models and LLMs with varied prompts, within a single
vector space. Such uniformity enables the comparison and analysis of
similarities amongst different models, extending the scope and utility of
existing task embedding methods in addressing multi-model scenarios, whilst
maintaining their performance to be comparable to architecture-specific
methods. | cs.CL | Towards Unified Task Embeddings Across Multiple Models: Bridging the Gap for Prompt-Based Large Language Models and Beyond |
2024-02-22T13:12:05Z | http://arxiv.org/pdf/2402.14521v1 | http://arxiv.org/abs/2402.14521v1 | Mohan Raj Chanthran, Lay-Ki Soon, Huey Fang Ong, Bhawani Selvaretnam | 2024-02-22T13:12:05Z | cs.CL | Standard English and Malaysian English exhibit notable differences, posing
challenges for natural language processing (NLP) tasks on Malaysian English.
Unfortunately, most of the existing datasets are mainly based on standard
English and therefore inadequate for improving NLP tasks in Malaysian English.
An experiment using state-of-the-art Named Entity Recognition (NER) solutions
on Malaysian English news articles highlights that they cannot handle
morphosyntactic variations in Malaysian English. To the best of our knowledge,
there is no annotated dataset available to improvise the model. To address
these issues, we constructed a Malaysian English News (MEN) dataset, which
contains 200 news articles that are manually annotated with entities and
relations. We then fine-tuned the spaCy NER tool and validated that having a
dataset tailor-made for Malaysian English could improve the performance of NER
in Malaysian English significantly. This paper presents our effort in the data
acquisition, annotation methodology, and thorough analysis of the annotated
dataset. To validate the quality of the annotation, inter-annotator agreement
was used, followed by adjudication of disagreements by a subject matter expert.
Upon completion of these tasks, we managed to develop a dataset with 6,061
entities and 3,268 relation instances. Finally, we discuss on spaCy fine-tuning
setup and analysis on the NER performance. This unique dataset will contribute
significantly to the advancement of NLP research in Malaysian English, allowing
researchers to accelerate their progress, particularly in NER and relation
extraction. The dataset and annotation guideline has been published on Github. | cs.CL | Malaysian English News Decoded: A Linguistic Resource for Named Entity and Relation Extraction |
2024-02-22T12:47:33Z | http://arxiv.org/pdf/2402.14499v1 | http://arxiv.org/abs/2402.14499v1 | Xinpeng Wang, Bolei Ma, Chengzhi Hu, Leon Weber-Genzel, Paul Röttger, Frauke Kreuter, Dirk Hovy, Barbara Plank | 2024-02-22T12:47:33Z | cs.CL | The open-ended nature of language generation makes the evaluation of
autoregressive large language models (LLMs) challenging. One common evaluation
approach uses multiple-choice questions (MCQ) to limit the response space. The
model is then evaluated by ranking the candidate answers by the log probability
of the first token prediction. However, first-tokens may not consistently
reflect the final response output, due to model's diverse response styles such
as starting with "Sure" or refusing to answer. Consequently, MCQ evaluation is
not indicative of model behaviour when interacting with users. But by how much?
We evaluate how aligned first-token evaluation is with the text output along
several dimensions, namely final option choice, refusal rate, choice
distribution and robustness under prompt perturbation. Our results show that
the two approaches are severely misaligned on all dimensions, reaching mismatch
rates over 60%. Models heavily fine-tuned on conversational or safety data are
especially impacted. Crucially, models remain misaligned even when we
increasingly constrain prompts, i.e., force them to start with an option letter
or example template. Our findings i) underscore the importance of inspecting
the text output, too and ii) caution against relying solely on first-token
evaluation. | cs.CL | "My Answer is C": First-Token Probabilities Do Not Match Text Answers in Instruction-Tuned Language Models |
2024-02-22T12:39:50Z | http://arxiv.org/pdf/2402.14494v3 | http://arxiv.org/abs/2402.14494v3 | Jinxu Zhao, Guanting Dong, Yueyan Qiu, Tingfeng Hui, Xiaoshuai Song, Daichi Guo, Weiran Xu | 2024-03-06T07:17:03Z | cs.CL | In a realistic dialogue system, the input information from users is often
subject to various types of input perturbations, which affects the slot-filling
task. Although rule-based data augmentation methods have achieved satisfactory
results, they fail to exhibit the desired generalization when faced with
unknown noise disturbances. In this study, we address the challenges posed by
input perturbations in slot filling by proposing Noise-BERT, a unified
Perturbation-Robust Framework with Noise Alignment Pre-training. Our framework
incorporates two Noise Alignment Pre-training tasks: Slot Masked Prediction and
Sentence Noisiness Discrimination, aiming to guide the pre-trained language
model in capturing accurate slot information and noise distribution. During
fine-tuning, we employ a contrastive learning loss to enhance the semantic
representation of entities and labels. Additionally, we introduce an
adversarial attack training strategy to improve the model's robustness.
Experimental results demonstrate the superiority of our proposed approach over
state-of-the-art models, and further analysis confirms its effectiveness and
generalization ability. | cs.CL | Noise-BERT: A Unified Perturbation-Robust Framework with Noise Alignment Pre-training for Noisy Slot Filling Task |
2024-02-22T12:36:31Z | http://arxiv.org/pdf/2402.14891v5 | http://arxiv.org/abs/2402.14891v5 | Bin Zhu, Munan Ning, Peng Jin, Bin Lin, Jinfa Huang, Qi Song, Junwu Zhang, Zhenyu Tang, Mingjun Pan, Xing Zhou, Li Yuan | 2024-04-19T03:07:59Z | cs.CL, cs.AI | In the multi-modal domain, the dependence of various models on specific input
formats leads to user confusion and hinders progress. To address this
challenge, we introduce \textbf{LLMBind}, a novel framework designed to unify a
diverse array of multi-modal tasks. By harnessing a Mixture-of-Experts (MoE)
Large Language Model (LLM), LLMBind processes multi-modal inputs and generates
task-specific tokens, enabling the invocation of corresponding models to
accomplish tasks. This unique approach empowers LLMBind to interpret inputs and
generate outputs across various modalities, including image, text, video, and
audio. Furthermore, we have constructed an interaction dataset comprising 400k
instructions, which unlocks the ability of LLMBind for interactive visual
generation and editing tasks. Extensive experimentation demonstrates that
LLMBind achieves very superior performance across diverse tasks and outperforms
existing models in user evaluations conducted in real-world scenarios.
Moreover, the adaptability of LLMBind allows for seamless integration with the
latest models and extension to new modality tasks, highlighting its potential
to serve as a unified AI agent for modeling universal modalities. | cs.CL | LLMBind: A Unified Modality-Task Integration Framework |
2024-02-22T12:35:50Z | http://arxiv.org/pdf/2402.14492v2 | http://arxiv.org/abs/2402.14492v2 | Wei Han, Hui Chen, Soujanya Poria | 2024-06-14T13:38:14Z | cs.CL, cs.AI | Fine-tuning large language models (LLMs) on multi-task instruction-following
data has been proven to be a powerful learning paradigm for improving their
zero-shot capabilities on new tasks. Recent works about high-quality
instruction-following data generation and selection require amounts of human
labor to conceive model-understandable instructions for the given tasks and
carefully filter the LLM-generated data. In this work, we introduce an
automatic instruction augmentation method named INSTRAUG in multimodal tasks.
It starts from a handful of basic and straightforward meta instructions but can
expand an instruction-following dataset by 30 times. Results on two popular
multimodal instructionfollowing benchmarks MULTIINSTRUCT and InstructBLIP show
that INSTRAUG can significantly improve the alignment of multimodal large
language models (MLLMs) across 12 multimodal tasks, which is even equivalent to
the benefits of scaling up training data multiple times. | cs.CL | Towards Robust Instruction Tuning on Multimodal Large Language Models |
2024-02-22T12:26:07Z | http://arxiv.org/pdf/2402.14488v1 | http://arxiv.org/abs/2402.14488v1 | Xinshuo Hu, Baotian Hu, Dongfang Li, Xiaoguang Li, Lifeng Shang | 2024-02-22T12:26:07Z | cs.CL | The present study introduces the knowledge-augmented generator, which is
specifically designed to produce information that remains grounded in
contextual knowledge, regardless of alterations in the context. Previous
research has predominantly focused on examining hallucinations stemming from
static input, such as in the domains of summarization or machine translation.
However, our investigation delves into the faithfulness of generative question
answering in the presence of dynamic knowledge. Our objective is to explore the
existence of hallucinations arising from parametric memory when contextual
knowledge undergoes changes, while also analyzing the underlying causes for
their occurrence. In order to efficiently address this issue, we propose a
straightforward yet effective measure for detecting such hallucinations.
Intriguingly, our investigation uncovers that all models exhibit a tendency to
generate previous answers as hallucinations. To gain deeper insights into the
underlying causes of this phenomenon, we conduct a series of experiments that
verify the critical role played by context in hallucination, both during
training and testing, from various perspectives. | cs.CL | Does the Generator Mind its Contexts? An Analysis of Generative Model Faithfulness under Context Transfer |
2024-02-22T12:19:04Z | http://arxiv.org/pdf/2402.14484v2 | http://arxiv.org/abs/2402.14484v2 | Takehiro Takayanagi, Masahiro Suzuki, Ryotaro Kobayashi, Hiroki Sakaji, Kiyoshi Izumi | 2024-02-23T11:50:18Z | cs.CL | Causality is fundamental in human cognition and has drawn attention in
diverse research fields. With growing volumes of textual data, discerning
causalities within text data is crucial, and causal text mining plays a pivotal
role in extracting meaningful patterns. This study conducts comprehensive
evaluations of ChatGPT's causal text mining capabilities. Firstly, we introduce
a benchmark that extends beyond general English datasets, including
domain-specific and non-English datasets. We also provide an evaluation
framework to ensure fair comparisons between ChatGPT and previous approaches.
Finally, our analysis outlines the limitations and future challenges in
employing ChatGPT for causal text mining. Specifically, our analysis reveals
that ChatGPT serves as a good starting point for various datasets. However,
when equipped with a sufficient amount of training data, previous models still
surpass ChatGPT's performance. Additionally, ChatGPT suffers from the tendency
to falsely recognize non-causal sequences as causal sequences. These issues
become even more pronounced with advanced versions of the model, such as GPT-4.
In addition, we highlight the constraints of ChatGPT in handling complex
causality types, including both intra/inter-sentential and implicit causality.
The model also faces challenges with effectively leveraging in-context learning
and domain adaptation. We release our code to support further research and
development in this field. | cs.CL | Is ChatGPT the Future of Causal Text Mining? A Comprehensive Evaluation and Analysis |
2024-02-22T12:04:15Z | http://arxiv.org/pdf/2402.14474v1 | http://arxiv.org/abs/2402.14474v1 | Sebastian Bordt, Ben Lengerich, Harsha Nori, Rich Caruana | 2024-02-22T12:04:15Z | cs.LG, cs.CL | Recent years have seen important advances in the building of interpretable
models, machine learning models that are designed to be easily understood by
humans. In this work, we show that large language models (LLMs) are remarkably
good at working with interpretable models, too. In particular, we show that
LLMs can describe, interpret, and debug Generalized Additive Models (GAMs).
Combining the flexibility of LLMs with the breadth of statistical patterns
accurately described by GAMs enables dataset summarization, question answering,
and model critique. LLMs can also improve the interaction between domain
experts and interpretable models, and generate hypotheses about the underlying
phenomenon. We release \url{https://github.com/interpretml/TalkToEBM} as an
open-source LLM-GAM interface. | cs.LG | Data Science with LLMs and Interpretable Models |
2024-02-22T12:00:32Z | http://arxiv.org/pdf/2402.14890v2 | http://arxiv.org/abs/2402.14890v2 | Maxim K. Surkov, Ivan P. Yamshchikov | 2024-02-26T12:09:43Z | cs.CL, cs.AI, cs.LG, 68T01, 97P80, 97C30, 68Q32, H.1.1; I.2.4; I.2.6; F.2.0 | Evaluation plays a significant role in modern natural language processing.
Most modern NLP benchmarks consist of arbitrary sets of tasks that neither
guarantee any generalization potential for the model once applied outside the
test set nor try to minimize the resource consumption needed for model
evaluation. This paper presents a theoretical instrument and a practical
algorithm to calculate similarity between benchmark tasks, we call this
similarity measure "Vygotsky distance". The core idea of this similarity
measure is that it is based on relative performance of the "students" on a
given task, rather that on the properties of the task itself. If two tasks are
close to each other in terms of Vygotsky distance the models tend to have
similar relative performance on them. Thus knowing Vygotsky distance between
tasks one can significantly reduce the number of evaluation tasks while
maintaining a high validation quality. Experiments on various benchmarks,
including GLUE, SuperGLUE, CLUE, and RussianSuperGLUE, demonstrate that a vast
majority of NLP benchmarks could be at least 40% smaller in terms of the tasks
included. Most importantly, Vygotsky distance could also be used for the
validation of new tasks thus increasing the generalization potential of the
future NLP models. | cs.CL | Vygotsky Distance: Measure for Benchmark Task Similarity |
2024-02-22T11:31:50Z | http://arxiv.org/pdf/2402.14458v1 | http://arxiv.org/abs/2402.14458v1 | Ramon Ruiz-Dolz, Joaquin Taverner, John Lawrence, Chris Reed | 2024-02-22T11:31:50Z | cs.CL, cs.AI | Some of the major limitations identified in the areas of argument mining,
argument generation, and natural language argument analysis are related to the
complexity of annotating argumentatively rich data, the limited size of these
corpora, and the constraints that represent the different languages and domains
in which these data is annotated. To address these limitations, in this paper
we present the following contributions: (i) an effective methodology for the
automatic generation of natural language arguments in different topics and
languages, (ii) the largest publicly available corpus of natural language
argumentation schemes, and (iii) a set of solid baselines and fine-tuned models
for the automatic identification of argumentation schemes. | cs.CL | NLAS-multi: A Multilingual Corpus of Automatically Generated Natural Language Argumentation Schemes |
2024-02-22T11:24:45Z | http://arxiv.org/pdf/2402.14457v2 | http://arxiv.org/abs/2402.14457v2 | Pietro Giovanni Bizzaro, Elena Della Valentina, Maurizio Napolitano, Nadia Mana, Massimo Zancanaro | 2024-05-27T09:06:41Z | cs.CL | In this paper, we propose a new annotation scheme to classify different types
of clauses in Terms-and-Conditions contracts with the ultimate goal of
supporting legal experts to quickly identify and assess problematic issues in
this type of legal documents. To this end, we built a small corpus of
Terms-and-Conditions contracts and finalized an annotation scheme of 14
categories, eventually reaching an inter-annotator agreement of 0.92. Then, for
11 of them, we experimented with binary classification tasks using few-shot
prompting with a multilingual T5 and two fine-tuned versions of two BERT-based
LLMs for Italian. Our experiments showed the feasibility of automatic
classification of our categories by reaching accuracies ranging from .79 to .95
on validation tasks. | cs.CL | Annotation and Classification of Relevant Clauses in Terms-and-Conditions Contracts |
2024-02-22T11:16:23Z | http://arxiv.org/pdf/2402.14453v1 | http://arxiv.org/abs/2402.14453v1 | Seiji Gobara, Hidetaka Kamigaito, Taro Watanabe | 2024-02-22T11:16:23Z | cs.CL | Education that suits the individual learning level is necessary to improve
students' understanding. The first step in achieving this purpose by using
large language models (LLMs) is to adjust the textual difficulty of the
response to students. This work analyzes how LLMs can implicitly adjust text
difficulty between user input and its generated text. To conduct the
experiments, we created a new dataset from Stack-Overflow to explore the
performance of question-answering-based conversation. Experimental results on
the Stack-Overflow dataset and the TSCC dataset, including multi-turn
conversation show that LLMs can implicitly handle text difficulty between user
input and its generated response. We also observed that some LLMs can surpass
humans in handling text difficulty and the importance of instruction-tuning. | cs.CL | Do LLMs Implicitly Determine the Suitable Text Difficulty for Users? |
2024-02-22T10:46:11Z | http://arxiv.org/pdf/2402.14889v2 | http://arxiv.org/abs/2402.14889v2 | Priyanshul Govil, Hemang Jain, Vamshi Krishna Bonagiri, Aman Chadha, Ponnurangam Kumaraguru, Manas Gaur, Sanorita Dey | 2024-06-17T09:01:25Z | cs.CL, cs.AI | Large Language Models (LLMs) are trained on extensive web corpora, which
enable them to understand and generate human-like text. However, this training
process also results in inherent biases within the models. These biases arise
from web data's diverse and often uncurated nature, containing various
stereotypes and prejudices. Previous works on debiasing models rely on
benchmark datasets to measure their method's performance. However, these
datasets suffer from several pitfalls due to the highly subjective
understanding of bias, highlighting a critical need for contextual exploration.
We propose understanding the context of inputs by considering the diverse
situations in which they may arise. Our contribution is two-fold: (i) we
augment 2,291 stereotyped statements from two existing bias-benchmark datasets
with points for adding context; (ii) we develop the Context-Oriented Bias
Indicator and Assessment Score (COBIAS) to assess a statement's contextual
reliability in measuring bias. Our metric aligns with human judgment on
contextual reliability of statements (Spearman's $\rho = 0.65, p = 3.4 *
10^{-60}$) and can be used to create reliable datasets, which would assist bias
mitigation works. | cs.CL | COBIAS: Contextual Reliability in Bias Assessment |
2024-02-22T10:25:14Z | http://arxiv.org/pdf/2402.14433v1 | http://arxiv.org/abs/2402.14433v1 | Dimitri von Rütte, Sotiris Anagnostidis, Gregor Bachmann, Thomas Hofmann | 2024-02-22T10:25:14Z | cs.CL, cs.AI, I.2 | Concept guidance has emerged as a cheap and simple way to control the
behavior of language models by probing their hidden representations for concept
vectors and using them to perturb activations at inference time. While the
focus of previous work has largely been on truthfulness, in this paper we
extend this framework to a richer set of concepts such as appropriateness,
humor, creativity and quality, and explore to what degree current detection and
guidance strategies work in these challenging settings. To facilitate
evaluation, we develop a novel metric for concept guidance that takes into
account both the success of concept elicitation as well as the potential
degradation in fluency of the guided model. Our extensive experiments reveal
that while some concepts such as truthfulness more easily allow for guidance
with current techniques, novel concepts such as appropriateness or humor either
remain difficult to elicit, need extensive tuning to work, or even experience
confusion. Moreover, we find that probes with optimal detection accuracies do
not necessarily make for the optimal guides, contradicting previous
observations for truthfulness. Our work warrants a deeper investigation into
the interplay between detectability, guidability, and the nature of the
concept, and we hope that our rich experimental test-bed for guidance research
inspires stronger follow-up approaches. | cs.CL | A Language Model's Guide Through Latent Space |
2024-02-22T10:17:57Z | http://arxiv.org/pdf/2402.14428v2 | http://arxiv.org/abs/2402.14428v2 | Yumin Kim, Heejae Suh, Mingi Kim, Dongyeon Won, Hwanhee Lee | 2024-03-22T06:29:26Z | cs.CL, cs.AI | Sarcasm is a way of verbal irony where someone says the opposite of what they
mean, often to ridicule a person, situation, or idea. It is often difficult to
detect sarcasm in the dialogue since detecting sarcasm should reflect the
context (i.e., dialogue history). In this paper, we introduce a new dataset for
the Korean dialogue sarcasm detection task, KoCoSa (Korean Context-aware
Sarcasm Detection Dataset), which consists of 12.8K daily Korean dialogues and
the labels for this task on the last response. To build the dataset, we propose
an efficient sarcasm detection dataset generation pipeline: 1) generating new
sarcastic dialogues from source dialogues with large language models, 2)
automatic and manual filtering of abnormal and toxic dialogues, and 3) human
annotation for the sarcasm detection task. We also provide a simple but
effective baseline for the Korean sarcasm detection task trained on our
dataset. Experimental results on the dataset show that our baseline system
outperforms strong baselines like large language models, such as GPT-3.5, in
the Korean sarcasm detection task. We show that the sarcasm detection task
relies deeply on the existence of sufficient context. We will release the
dataset at https://github.com/Yu-billie/KoCoSa_sarcasm_detection. | cs.CL | KoCoSa: Korean Context-aware Sarcasm Detection Dataset |
2024-02-22T10:14:59Z | http://arxiv.org/pdf/2402.14427v1 | http://arxiv.org/abs/2402.14427v1 | Lala Shakti Swarup Ray, Bo Zhou, Sungho Suh, Lars Krupp, Vitor Fortes Rey, Paul Lukowicz | 2024-02-22T10:14:59Z | cs.LG, cs.CL, eess.SP | In human activity recognition (HAR), the availability of substantial ground
truth is necessary for training efficient models. However, acquiring ground
pressure data through physical sensors itself can be cost-prohibitive,
time-consuming. To address this critical need, we introduce Text-to-Pressure
(T2P), a framework designed to generate extensive ground pressure sequences
from textual descriptions of human activities using deep learning techniques.
We show that the combination of vector quantization of sensor data along with
simple text conditioned auto regressive strategy allows us to obtain
high-quality generated pressure sequences from textual descriptions with the
help of discrete latent correlation between text and pressure maps. We achieved
comparable performance on the consistency between text and generated motion
with an R squared value of 0.722, Masked R squared value of 0.892, and FID
score of 1.83. Additionally, we trained a HAR model with the the synthesized
data and evaluated it on pressure dynamics collected by a real pressure sensor
which is on par with a model trained on only real data. Combining both real and
synthesized training data increases the overall macro F1 score by 5.9 percent. | cs.LG | Text me the data: Generating Ground Pressure Sequence from Textual Descriptions for HAR |
2024-02-22T09:56:51Z | http://arxiv.org/pdf/2402.14411v1 | http://arxiv.org/abs/2402.14411v1 | Kosuke Matsuzaki, Masaya Taniguchi, Kentaro Inui, Keisuke Sakaguchi | 2024-02-22T09:56:51Z | cs.CL | We introduce a Japanese Morphology dataset, J-UniMorph, developed based on
the UniMorph feature schema. This dataset addresses the unique and rich verb
forms characteristic of the language's agglutinative nature. J-UniMorph
distinguishes itself from the existing Japanese subset of UniMorph, which is
automatically extracted from Wiktionary. On average, the Wiktionary Edition
features around 12 inflected forms for each word and is primarily dominated by
denominal verbs (i.e., [noun] +suru (do-PRS)). Morphologically, this form is
equivalent to the verb suru (do). In contrast, J-UniMorph explores a much
broader and more frequently used range of verb forms, offering 118 inflected
forms for each word on average. It includes honorifics, a range of politeness
levels, and other linguistic nuances, emphasizing the distinctive
characteristics of the Japanese language. This paper presents detailed
statistics and characteristics of J-UniMorph, comparing it with the Wiktionary
Edition. We release J-UniMorph and its interactive visualizer publicly
available, aiming to support cross-linguistic research and various
applications. | cs.CL | J-UniMorph: Japanese Morphological Annotation through the Universal Feature Schema |
2024-02-22T09:52:45Z | http://arxiv.org/pdf/2403.05572v1 | http://arxiv.org/abs/2403.05572v1 | Anuradha Welivita, Pearl Pu | 2024-02-22T09:52:45Z | cs.HC, cs.AI, cs.CL | This paper investigates the empathetic responding capabilities of ChatGPT,
particularly its latest iteration, GPT-4, in comparison to human-generated
responses to a wide range of emotional scenarios, both positive and negative.
We employ a rigorous evaluation methodology, involving a between-groups study
with 600 participants, to evaluate the level of empathy in responses generated
by humans and ChatGPT. ChatGPT is prompted in two distinct ways: a standard
approach and one explicitly detailing empathy's cognitive, affective, and
compassionate counterparts. Our findings indicate that the average empathy
rating of responses generated by ChatGPT exceeds those crafted by humans by
approximately 10%. Additionally, instructing ChatGPT to incorporate a clear
understanding of empathy in its responses makes the responses align
approximately 5 times more closely with the expectations of individuals
possessing a high degree of empathy, compared to human responses. The proposed
evaluation framework serves as a scalable and adaptable framework to assess the
empathetic capabilities of newer and updated versions of large language models,
eliminating the need to replicate the current study's results in future
research. | cs.HC | Is ChatGPT More Empathetic than Humans? |
2024-02-22T09:51:08Z | http://arxiv.org/pdf/2402.14409v1 | http://arxiv.org/abs/2402.14409v1 | Zhuoran Jin, Pengfei Cao, Yubo Chen, Kang Liu, Xiaojian Jiang, Jiexin Xu, Qiuxia Li, Jun Zhao | 2024-02-22T09:51:08Z | cs.CL, cs.AI, cs.IR | Retrieval-augmented language models (RALMs) have demonstrated significant
potential in refining and expanding their internal memory by retrieving
evidence from external sources. However, RALMs will inevitably encounter
knowledge conflicts when integrating their internal memory with external
sources. Knowledge conflicts can ensnare RALMs in a tug-of-war between
knowledge, limiting their practical applicability. In this paper, we focus on
exploring and resolving knowledge conflicts in RALMs. First, we present an
evaluation framework for assessing knowledge conflicts across various
dimensions. Then, we investigate the behavior and preference of RALMs from the
following two perspectives: (1) Conflicts between internal memory and external
sources: We find that stronger RALMs emerge with the Dunning-Kruger effect,
persistently favoring their faulty internal memory even when correct evidence
is provided. Besides, RALMs exhibit an availability bias towards common
knowledge; (2) Conflicts between truthful, irrelevant and misleading evidence:
We reveal that RALMs follow the principle of majority rule, leaning towards
placing trust in evidence that appears more frequently. Moreover, we find that
RALMs exhibit confirmation bias, and are more willing to choose evidence that
is consistent with their internal memory. To solve the challenge of knowledge
conflicts, we propose a method called Conflict-Disentangle Contrastive Decoding
(CD2) to better calibrate the model's confidence. Experimental results
demonstrate that our CD2 can effectively resolve knowledge conflicts in RALMs. | cs.CL | Tug-of-War Between Knowledge: Exploring and Resolving Knowledge Conflicts in Retrieval-Augmented Language Models |
2024-02-22T09:49:26Z | http://arxiv.org/pdf/2402.14408v1 | http://arxiv.org/abs/2402.14408v1 | Piotr Rybak | 2024-02-22T09:49:26Z | cs.CL | Pre-trained language models have revolutionized the natural language
understanding landscape, most notably BERT (Bidirectional Encoder
Representations from Transformers). However, a significant challenge remains
for low-resource languages, where limited data hinders the effective training
of such models. This work presents a novel approach to bridge this gap by
transferring BERT capabilities from high-resource to low-resource languages
using vocabulary matching. We conduct experiments on the Silesian and Kashubian
languages and demonstrate the effectiveness of our approach to improve the
performance of BERT models even when the target language has minimal training
data. Our results highlight the potential of the proposed technique to
effectively train BERT models for low-resource languages, thus democratizing
access to advanced language understanding models. | cs.CL | Transferring BERT Capabilities from High-Resource to Low-Resource Languages Using Vocabulary Matching |
2024-02-22T09:45:26Z | http://arxiv.org/pdf/2402.14404v2 | http://arxiv.org/abs/2402.14404v2 | Ningyu Xu, Qi Zhang, Menghan Zhang, Peng Qian, Xuanjing Huang | 2024-02-26T11:40:45Z | cs.CL, cs.AI | Probing and enhancing large language models' reasoning capacity remains a
crucial open question. Here we re-purpose the reverse dictionary task as a case
study to probe LLMs' capacity for conceptual inference. We use in-context
learning to guide the models to generate the term for an object concept implied
in a linguistic description. Models robustly achieve high accuracy in this
task, and their representation space encodes information about object
categories and fine-grained features. Further experiments suggest that the
conceptual inference ability as probed by the reverse-dictionary task predicts
model's general reasoning performance across multiple benchmarks, despite
similar syntactic generalization behaviors across models. Explorative analyses
suggest that prompting LLMs with description$\Rightarrow$word examples may
induce generalization beyond surface-level differences in task construals and
facilitate models on broader commonsense reasoning problems. | cs.CL | On the Tip of the Tongue: Analyzing Conceptual Representation in Large Language Models with Reverse-Dictionary Probe |
2024-02-22T09:43:53Z | http://arxiv.org/pdf/2402.14888v1 | http://arxiv.org/abs/2402.14888v1 | Roxana Petcu, Subhadeep Maji | 2024-02-22T09:43:53Z | cs.LG, cs.AI, cs.CL | Recent developments in natural language processing (NLP) have highlighted the
need for substantial amounts of data for models to capture textual information
accurately. This raises concerns regarding the computational resources and time
required for training such models. This paper introduces Semantics for data
SAliency in Model performance Estimation (SeSaME). It is an efficient data
sampling mechanism solely based on textual information without passing the data
through a compute-heavy model or other intensive pre-processing
transformations. The application of this approach is demonstrated in the use
case of low-resource automated speech recognition (ASR) models, which
excessively rely on text-to-speech (TTS) calls when using augmented data.
SeSaME learns to categorize new incoming data points into speech recognition
difficulty buckets by employing semantic similarity-based graph structures and
discrete ASR information from homophilous neighbourhoods through message
passing. The results indicate reliable projections of ASR performance, with a
93% accuracy increase when using the proposed method compared to random
predictions, bringing non-trivial information on the impact of textual
representations in speech models. Furthermore, a series of experiments show
both the benefits and challenges of using the ASR information on incoming data
to fine-tune the model. We report a 7% drop in validation loss compared to
random sampling, 7% WER drop with non-local aggregation when evaluating against
a highly difficult dataset, and 1.8% WER drop with local aggregation and high
semantic similarity between datasets. | cs.LG | Efficient data selection employing Semantic Similarity-based Graph Structures for model training |
2024-02-22T08:51:39Z | http://arxiv.org/pdf/2402.14382v2 | http://arxiv.org/abs/2402.14382v2 | Yuwei Xia, Ding Wang, Qiang Liu, Liang Wang, Shu Wu, Xiaoyu Zhang | 2024-06-07T08:15:18Z | cs.CL | Temporal Knowledge Graph (TKG) forecasting aims to predict future facts based
on given histories. Most recent graph-based models excel at capturing
structural information within TKGs but lack semantic comprehension abilities.
Nowadays, with the surge of LLMs, the LLM-based TKG prediction model has
emerged. However, the existing LLM-based model exhibits three shortcomings: (1)
It only focuses on the first-order history for prediction while ignoring
high-order historical information, resulting in the provided information for
LLMs being extremely limited. (2) LLMs struggle with optimal reasoning
performance under heavy historical information loads. (3) For TKG prediction,
the temporal reasoning capability of LLM alone is limited. To address the first
two challenges, we propose Chain-of-History (CoH) reasoning which explores
high-order histories step-by-step, achieving effective utilization of
high-order historical information for LLMs on TKG prediction. To address the
third issue, we design CoH as a plug-and-play module to enhance the performance
of graph-based models for TKG prediction. Extensive experiments on three
datasets and backbones demonstrate the effectiveness of CoH. | cs.CL | Chain-of-History Reasoning for Temporal Knowledge Graph Forecasting |
2024-02-22T08:48:21Z | http://arxiv.org/pdf/2402.14379v2 | http://arxiv.org/abs/2402.14379v2 | Mihailo Škorić | 2024-02-23T09:00:49Z | cs.CL | The paper will briefly present the development history of transformer-based
language models for the Serbian language. Several new models for text
generation and vectorization, trained on the resources of the Society for
Language Resources and Technologies, will also be presented. Ten selected
vectorization models for Serbian, including two new ones, will be compared on
four natural language processing tasks. Paper will analyze which models are the
best for each selected task, how does their size and the size of their training
sets affect the performance on those tasks, and what is the optimal setting to
train the best language models for the Serbian language. | cs.CL | Novi jezički modeli za srpski jezik |
2024-02-22T08:26:56Z | http://arxiv.org/pdf/2402.14373v1 | http://arxiv.org/abs/2402.14373v1 | Xuemei Tang, Jun Wang, Qi Su | 2024-02-22T08:26:56Z | cs.CL | Recently, large language models (LLMs) have been successful in relational
extraction (RE) tasks, especially in the few-shot learning. An important
problem in the field of RE is long-tailed data, while not much attention is
currently paid to this problem using LLM approaches. Therefore, in this paper,
we propose SLCoLM, a model collaboration framework, to mitigate the data
long-tail problem. In our framework, We use the
``\textit{Training-Guide-Predict}'' strategy to combine the strengths of
pre-trained language models (PLMs) and LLMs, where a task-specific PLM
framework acts as a tutor, transfers task knowledge to the LLM, and guides the
LLM in performing RE tasks. Our experiments on a RE dataset rich in relation
types show that the approach in this paper facilitates RE of long-tail relation
types. | cs.CL | Small Language Model Is a Good Guide for Large Language Model in Chinese Entity Relation Extraction |
2024-02-22T07:58:29Z | http://arxiv.org/pdf/2402.14359v1 | http://arxiv.org/abs/2402.14359v1 | Xiuying Chen, Tairan Wang, Qingqing Zhu, Taicheng Guo, Shen Gao, Zhiyong Lu, Xin Gao, Xiangliang Zhang | 2024-02-22T07:58:29Z | cs.CL | The summarization capabilities of pretrained and large language models (LLMs)
have been widely validated in general areas, but their use in scientific
corpus, which involves complex sentences and specialized knowledge, has been
less assessed. This paper presents conceptual and experimental analyses of
scientific summarization, highlighting the inadequacies of traditional
evaluation methods, such as $n$-gram, embedding comparison, and QA,
particularly in providing explanations, grasping scientific concepts, or
identifying key content. Subsequently, we introduce the Facet-aware Metric
(FM), employing LLMs for advanced semantic matching to evaluate summaries based
on different aspects. This facet-aware approach offers a thorough evaluation of
abstracts by decomposing the evaluation task into simpler subtasks.Recognizing
the absence of an evaluation benchmark in this domain, we curate a Facet-based
scientific summarization Dataset (FD) with facet-level annotations. Our
findings confirm that FM offers a more logical approach to evaluating
scientific summaries. In addition, fine-tuned smaller models can compete with
LLMs in scientific contexts, while LLMs have limitations in learning from
in-context information in scientific domains. This suggests an area for future
enhancement of LLMs. | cs.CL | Rethinking Scientific Summarization Evaluation: Grounding Explainable Metrics on Facet-aware Benchmark |
2024-02-22T07:55:26Z | http://arxiv.org/pdf/2402.14355v2 | http://arxiv.org/abs/2402.14355v2 | Ning Bian, Xianpei Han, Hongyu Lin, Yaojie Lu, Ben He, Le Sun | 2024-06-04T08:05:51Z | cs.CL | Building machines with commonsense has been a longstanding challenge in NLP
due to the reporting bias of commonsense rules and the exposure bias of
rule-based commonsense reasoning. In contrast, humans convey and pass down
commonsense implicitly through stories. This paper investigates the inherent
commonsense ability of large language models (LLMs) expressed through
storytelling. We systematically investigate and compare stories and rules for
retrieving and leveraging commonsense in LLMs. Experimental results on 28
commonsense QA datasets show that stories outperform rules as the expression
for retrieving commonsense from LLMs, exhibiting higher generation confidence
and commonsense accuracy. Moreover, stories are the more effective commonsense
expression for answering questions regarding daily events, while rules are more
effective for scientific questions. This aligns with the reporting bias of
commonsense in text corpora. We further show that the correctness and relevance
of commonsense stories can be further improved via iterative self-supervised
fine-tuning. These findings emphasize the importance of using appropriate
language to express, retrieve, and leverage commonsense for LLMs, highlighting
a promising direction for better exploiting their commonsense abilities. | cs.CL | Rule or Story, Which is a Better Commonsense Expression for Talking with Large Language Models? |
2024-02-22T07:12:34Z | http://arxiv.org/pdf/2402.14337v1 | http://arxiv.org/abs/2402.14337v1 | Hazel Kim | 2024-02-22T07:12:34Z | cs.CL, cs.LG | Rationales behind answers not only explain model decisions but boost language
models to reason well on complex reasoning tasks. However, obtaining impeccable
rationales is often impossible. Besides, it is non-trivial to estimate the
degree to which the rationales are faithful enough to encourage model
performance. Thus, such reasoning tasks often compel models to output correct
answers under undesirable rationales and are sub-optimal compared to what the
models are fully capable of. In this work, we propose how to deal with
imperfect rationales causing aleatoric uncertainty. We first define the
ambiguous rationales with entropy scores of given rationales, using model prior
beliefs as informativeness. We then guide models to select one of two different
reasoning models according to the ambiguity of rationales. We empirically argue
that our proposed method produces robust performance superiority against the
adversarial quality of rationales and low-resource settings. | cs.CL | AURA: Natural Language Reasoning for Aleatoric Uncertainty in Rationales |
2024-02-22T06:59:50Z | http://arxiv.org/pdf/2402.14334v1 | http://arxiv.org/abs/2402.14334v1 | Hanseok Oh, Hyunji Lee, Seonghyeon Ye, Haebin Shin, Hansol Jang, Changwook Jun, Minjoon Seo | 2024-02-22T06:59:50Z | cs.CL | Despite the critical need to align search targets with users' intention,
retrievers often only prioritize query information without delving into the
users' intended search context. Enhancing the capability of retrievers to
understand intentions and preferences of users, akin to language model
instructions, has the potential to yield more aligned search targets. Prior
studies restrict the application of instructions in information retrieval to a
task description format, neglecting the broader context of diverse and evolving
search scenarios. Furthermore, the prevailing benchmarks utilized for
evaluation lack explicit tailoring to assess instruction-following ability,
thereby hindering progress in this field. In response to these limitations, we
propose a novel benchmark,INSTRUCTIR, specifically designed to evaluate
instruction-following ability in information retrieval tasks. Our approach
focuses on user-aligned instructions tailored to each query instance,
reflecting the diverse characteristics inherent in real-world search scenarios.
Through experimental analysis, we observe that retrievers fine-tuned to follow
task-style instructions, such as INSTRUCTOR, can underperform compared to their
non-instruction-tuned counterparts. This underscores potential overfitting
issues inherent in constructing retrievers trained on existing
instruction-aware retrieval datasets. | cs.CL | INSTRUCTIR: A Benchmark for Instruction Following of Information Retrieval Models |
2024-02-22T06:47:56Z | http://arxiv.org/pdf/2402.14328v2 | http://arxiv.org/abs/2402.14328v2 | Zhaoyi Li, Gangwei Jiang, Hong Xie, Linqi Song, Defu Lian, Ying Wei | 2024-06-06T14:06:41Z | cs.CL | LLMs have marked a revolutonary shift, yet they falter when faced with
compositional reasoning tasks. Our research embarks on a quest to uncover the
root causes of compositional reasoning failures of LLMs, uncovering that most
of them stem from the improperly generated or leveraged implicit reasoning
results. Inspired by our empirical findings, we resort to Logit Lens and an
intervention experiment to dissect the inner hidden states of LLMs. This deep
dive reveals that implicit reasoning results indeed surface within middle
layers and play a causative role in shaping the final explicit reasoning
results. Our exploration further locates multi-head self-attention (MHSA)
modules within these layers, which emerge as the linchpins in accurate
generation and leveraing of implicit reasoning results. Grounded on the above
findings, we develop CREME, a lightweight method to patch errors in
compositional reasoning via editing the located MHSA modules. Our empirical
evidence stands testament to CREME's effectiveness, paving the way for
autonomously and continuously enhancing compositional reasoning capabilities in
language models. | cs.CL | Understanding and Patching Compositional Reasoning in LLMs |
2024-02-22T06:47:44Z | http://arxiv.org/pdf/2402.14327v2 | http://arxiv.org/abs/2402.14327v2 | Delong Chen, Samuel Cahyawijaya, Jianfeng Liu, Baoyuan Wang, Pascale Fung | 2024-04-23T13:41:47Z | cs.CV, cs.CL | Transformer-based vision models typically tokenize images into fixed-size
square patches as input units, which lacks the adaptability to image content
and overlooks the inherent pixel grouping structure. Inspired by the subword
tokenization widely adopted in language models, we propose an image tokenizer
at a subobject level, where the subobjects are represented by semantically
meaningful image segments obtained by segmentation models (e.g., segment
anything models). To implement a learning system based on subobject
tokenization, we first introduced a Direct Segment Anything Model (DirectSAM)
that efficiently produces comprehensive segmentation of subobjects, then embed
subobjects into compact latent vectors and fed them into a large language model
for vision language learning. Empirical results demonstrated that our
subobject-level tokenization significantly facilitates efficient learning of
translating images into object and attribute descriptions compared to the
traditional patch-level tokenization. Codes and models are open-sourced at
https://github.com/ChenDelong1999/subobjects. | cs.CV | Subobject-level Image Tokenization |
2024-02-22T06:23:37Z | http://arxiv.org/pdf/2402.14320v5 | http://arxiv.org/abs/2402.14320v5 | Chang Zong, Yuchen Yan, Weiming Lu, Jian Shao, Eliot Huang, Heng Chang, Yueting Zhuang | 2024-04-16T02:46:10Z | cs.CL, cs.AI, 68T50, I.2.7 | Recent progress with LLM-based agents has shown promising results across
various tasks. However, their use in answering questions from knowledge bases
remains largely unexplored. Implementing a KBQA system using traditional
methods is challenging due to the shortage of task-specific training data and
the complexity of creating task-focused model structures. In this paper, we
present Triad, a unified framework that utilizes an LLM-based agent with three
roles for KBQA tasks. The agent is assigned three roles to tackle different
KBQA subtasks: agent as a generalist for mastering various subtasks, as a
decision maker for the selection of candidates, and as an advisor for answering
questions with knowledge. Our KBQA framework is executed in four phases,
involving the collaboration of the agent's multiple roles. We evaluated the
performance of our framework using three benchmark datasets, and the results
show that our framework outperforms state-of-the-art systems on the LC-QuAD and
YAGO-QA benchmarks, yielding F1 scores of 11.8% and 20.7%, respectively. | cs.CL | Triad: A Framework Leveraging a Multi-Role LLM-based Agent to Solve Knowledge Base Question Answering |
2024-02-22T06:21:41Z | http://arxiv.org/pdf/2402.14318v1 | http://arxiv.org/abs/2402.14318v1 | Sławomir Dadas, Małgorzata Grębowiec | 2024-02-22T06:21:41Z | cs.CL | Retrieval-augmented generation (RAG) is becoming an increasingly popular
technique for integrating internal knowledge bases with large language models.
In a typical RAG pipeline, three models are used, responsible for the
retrieval, reranking, and generation stages. In this article, we focus on the
reranking problem for the Polish language, examining the performance of
rerankers and comparing their results with available retrieval models. We
conduct a comprehensive evaluation of existing models and those trained by us,
utilizing a benchmark of 41 diverse information retrieval tasks for the Polish
language. The results of our experiments show that most models struggle with
out-of-domain generalization. However, a combination of effective optimization
method and a large training dataset allows for building rerankers that are both
compact in size and capable of generalization. The best of our models
establishes a new state-of-the-art for reranking in the Polish language,
outperforming existing models with up to 30 times more parameters. | cs.CL | Assessing generalization capability of text ranking models in Polish |
2024-02-22T05:58:03Z | http://arxiv.org/pdf/2402.14310v1 | http://arxiv.org/abs/2402.14310v1 | Jinlan Fu, Shenzhen Huangfu, Hang Yan, See-Kiong Ng, Xipeng Qiu | 2024-02-22T05:58:03Z | cs.CL | Large Language Models (LLMs) have recently showcased remarkable
generalizability in various domains. Despite their extensive knowledge, LLMs
still face challenges in efficiently utilizing encoded knowledge to develop
accurate and logical reasoning processes. To mitigate this problem, we
introduced Hint-before-Solving Prompting (HSP), which guides the model to
generate hints (e.g., specific knowledge or key ideas) for solving the problem
and then generate solutions containing intermediate reasoning steps. Since HSP
is orthogonal to prompting methods (e.g., Chain-of-Thought (CoT)), we applied
HSP to CoT, Least-to-Most, Plan-and-Solve, and Standard promptings. The results
of extensive experiments on 6 reasoning benchmarks and 4 open-source LLMs
demonstrate that HSP can effectively improve the accuracy of reasoning tasks:
(1) By applying high-quality hint-enhanced HSP to CoT prompting,
Llama2-70B-Chat shows an improvement of 9.7. (2) Beyond exploring training-free
LLM capabilities, we built the HSPMATH dataset based on HSP and fine-tuned
Llemma-7B, reaching 64.3 accuracy, surpassing GPT-3.5 and WizardMath-13B. We
make our code and dataset publicly available at
\url{https://github.com/jinlanfu/HSP}. | cs.CL | Hint-before-Solving Prompting: Guiding LLMs to Effectively Utilize Encoded Knowledge |
2024-02-22T05:48:03Z | http://arxiv.org/pdf/2402.15444v1 | http://arxiv.org/abs/2402.15444v1 | Yichi Zhang, Zhuo Chen, Lei Liang, Huajun Chen, Wen Zhang | 2024-02-22T05:48:03Z | cs.AI, cs.CL, cs.LG, cs.MM | Multi-modal knowledge graph completion (MMKGC) aims to predict the missing
triples in the multi-modal knowledge graphs by incorporating structural,
visual, and textual information of entities into the discriminant models. The
information from different modalities will work together to measure the triple
plausibility. Existing MMKGC methods overlook the imbalance problem of modality
information among entities, resulting in inadequate modal fusion and
inefficient utilization of the raw modality information. To address the
mentioned problems, we propose Adaptive Multi-modal Fusion and Modality
Adversarial Training (AdaMF-MAT) to unleash the power of imbalanced modality
information for MMKGC. AdaMF-MAT achieves multi-modal fusion with adaptive
modality weights and further generates adversarial samples by
modality-adversarial training to enhance the imbalanced modality information.
Our approach is a co-design of the MMKGC model and training strategy which can
outperform 19 recent MMKGC methods and achieve new state-of-the-art results on
three public MMKGC benchmarks. Our code and data have been released at
https://github.com/zjukg/AdaMF-MAT. | cs.AI | Unleashing the Power of Imbalanced Modality Information for Multi-modal Knowledge Graph Completion |
2024-02-22T05:24:19Z | http://arxiv.org/pdf/2402.14298v3 | http://arxiv.org/abs/2402.14298v3 | Bin Liang, Ang Li, Jingqian Zhao, Lin Gui, Min Yang, Yue Yu, Kam-Fai Wong, Ruifeng Xu | 2024-06-06T15:17:58Z | cs.CL | Stance detection is a challenging task that aims to identify public opinion
from social media platforms with respect to specific targets. Previous work on
stance detection largely focused on pure texts. In this paper, we study
multi-modal stance detection for tweets consisting of texts and images, which
are prevalent in today's fast-growing social media platforms where people often
post multi-modal messages. To this end, we create five new multi-modal stance
detection datasets of different domains based on Twitter, in which each example
consists of a text and an image. In addition, we propose a simple yet effective
Targeted Multi-modal Prompt Tuning framework (TMPT), where target information
is leveraged to learn multi-modal stance features from textual and visual
modalities. Experimental results on our five benchmark datasets show that the
proposed TMPT achieves state-of-the-art performance in multi-modal stance
detection. | cs.CL | Multi-modal Stance Detection: New Datasets and Model |
2024-02-22T05:17:49Z | http://arxiv.org/pdf/2402.14296v2 | http://arxiv.org/abs/2402.14296v2 | Ang Li, Jingqian Zhao, Bin Liang, Lin Gui, Hui Wang, Xi Zeng, Xingwei Liang, Kam-Fai Wong, Ruifeng Xu | 2024-06-16T12:04:24Z | cs.CL | Large language models (LLMs) have achieved remarkable progress in many
natural language processing tasks. However, our experiment reveals that, in
stance detection tasks, LLMs may generate biased stances due to
sentiment-stance spurious correlations and preference towards certain
individuals and topics, thus harming their performance. Therefore, in this
paper, we propose to Mitigate Biases of LLMs in stance detection with
Calibration (MB-Cal). To be specific, a novel calibration network is devised to
calibrate potential bias in the stance prediction of LLMs. Further, to address
the challenge of effectively learning bias representations and the difficulty
in the generalizability of debiasing, we construct counterfactual augmented
data. This approach enhances the calibration network, facilitating the
debiasing and out-of-domain generalization. Experimental results on in-target
and zero-shot stance detection tasks show that the proposed MB-Cal can
effectively mitigate biases of LLMs, achieving state-of-the-art results. | cs.CL | Mitigating Biases of Large Language Models in Stance Detection with Calibration |
2024-02-22T05:15:27Z | http://arxiv.org/pdf/2402.14293v1 | http://arxiv.org/abs/2402.14293v1 | Rui Yang, Boming Yang, Sixun Ouyang, Tianwei She, Aosong Feng, Yuang Jiang, Freddy Lecue, Jinghui Lu, Irene Li | 2024-02-22T05:15:27Z | cs.CL | In the domain of Natural Language Processing (NLP), Large Language Models
(LLMs) have demonstrated promise in text-generation tasks. However, their
educational applications, particularly for domain-specific queries, remain
underexplored. This study investigates LLMs' capabilities in educational
scenarios, focusing on concept graph recovery and question-answering (QA). We
assess LLMs' zero-shot performance in creating domain-specific concept graphs
and introduce TutorQA, a new expert-verified NLP-focused benchmark for
scientific graph reasoning and QA. TutorQA consists of five tasks with 500 QA
pairs. To tackle TutorQA queries, we present CGLLM, a pipeline integrating
concept graphs with LLMs for answering diverse questions. Our results indicate
that LLMs' zero-shot concept graph recovery is competitive with supervised
methods, showing an average 3% F1 score improvement. In TutorQA tasks, LLMs
achieve up to 26% F1 score enhancement. Moreover, human evaluation and analysis
show that CGLLM generates answers with more fine-grained concepts. | cs.CL | Leveraging Large Language Models for Concept Graph Recovery and Question Answering in NLP Education |
2024-02-22T05:07:31Z | http://arxiv.org/pdf/2402.14290v1 | http://arxiv.org/abs/2402.14290v1 | Samraj Moorjani, Adit Krishnan, Hari Sundaram | 2024-02-22T05:07:31Z | cs.CL, cs.LG | As large-scale language models become the standard for text generation, there
is a greater need to tailor the generations to be more or less concise,
targeted, and informative, depending on the audience/application. Existing
control approaches primarily adjust the semantic (e.g., emotion, topics),
structural (e.g., syntax tree, parts-of-speech), and lexical (e.g.,
keyword/phrase inclusion) properties of text, but are insufficient to
accomplish complex objectives such as pacing which control the complexity and
readability of the text. In this paper, we introduce CEV-LM - a lightweight,
semi-autoregressive language model that utilizes constrained edit vectors to
control three complementary metrics (speed, volume, and circuitousness) that
quantify the shape of text (e.g., pacing of content). We study an extensive set
of state-of-the-art CTG models and find that CEV-LM provides significantly more
targeted and precise control of these three metrics while preserving semantic
content, using less training data, and containing fewer parameters. | cs.CL | CEV-LM: Controlled Edit Vector Language Model for Shaping Natural Language Generations |
2024-02-22T05:05:30Z | http://arxiv.org/pdf/2402.14289v1 | http://arxiv.org/abs/2402.14289v1 | Baichuan Zhou, Ying Hu, Xi Weng, Junlong Jia, Jie Luo, Xien Liu, Ji Wu, Lei Huang | 2024-02-22T05:05:30Z | cs.LG, cs.CL | We present the TinyLLaVA framework that provides a unified perspective in
designing and analyzing the small-scale Large Multimodal Models (LMMs). We
empirically study the effects of different vision encoders, connection modules,
language models, training data and training recipes. Our extensive experiments
showed that better quality of data combined with better training recipes,
smaller LMMs can consistently achieve on-par performances compared to bigger
LMMs. Under our framework, we train a family of small-scale LMMs. Our best
model, TinyLLaVA-3.1B, achieves better overall performance against existing 7B
models such as LLaVA-1.5 and Qwen-VL. We hope our findings can serve as
baselines for future research in terms of data scaling, training setups and
model selections. Our model weights and codes will be made public. | cs.LG | TinyLLaVA: A Framework of Small-scale Large Multimodal Models |
2024-02-22T04:41:52Z | http://arxiv.org/pdf/2402.14279v1 | http://arxiv.org/abs/2402.14279v1 | Haeji Jung, Changdae Oh, Jooeon Kang, Jimin Sohn, Kyungwoo Song, Jinkyu Kim, David R. Mortensen | 2024-02-22T04:41:52Z | cs.CL, cs.AI | Approaches to improving multilingual language understanding often require
multiple languages during the training phase, rely on complicated training
techniques, and -- importantly -- struggle with significant performance gaps
between high-resource and low-resource languages. We hypothesize that the
performance gaps between languages are affected by linguistic gaps between
those languages and provide a novel solution for robust multilingual language
modeling by employing phonemic representations (specifically, using phonemes as
input tokens to LMs rather than subwords). We present quantitative evidence
from three cross-lingual tasks that demonstrate the effectiveness of phonemic
representation, which is further justified by a theoretical analysis of the
cross-lingual performance gap. | cs.CL | Mitigating the Linguistic Gap with Phonemic Representations for Robust Multilingual Language Understanding |
2024-02-22T04:36:14Z | http://arxiv.org/pdf/2402.14277v1 | http://arxiv.org/abs/2402.14277v1 | Spencer Rarrick, Ranjita Naik, Sundar Poudel, Vishal Chowdhary | 2024-02-22T04:36:14Z | cs.CL, cs.AI | Neural Machine Translation (NMT) continues to improve in quality and
adoption, yet the inadvertent perpetuation of gender bias remains a significant
concern. Despite numerous studies on gender bias in translations into English
from weakly gendered-languages, there are no benchmarks for evaluating this
phenomenon or for assessing mitigation strategies. To address this gap, we
introduce GATE X-E, an extension to the GATE (Rarrick et al., 2023) corpus,
that consists of human translations from Turkish, Hungarian, Finnish, and
Persian into English. Each translation is accompanied by feminine, masculine,
and neutral variants. The dataset, which contains between 1250 and 1850
instances for each of the four language pairs, features natural sentences with
a wide range of sentence lengths and domains, challenging translation rewriters
on various linguistic phenomena. Additionally, we present a translation gender
rewriting solution built with GPT-4 and use GATE X-E to evaluate it. We open
source our contributions to encourage further research on gender debiasing. | cs.CL | GATE X-E : A Challenge Set for Gender-Fair Translations from Weakly-Gendered Languages |
2024-02-22T04:20:14Z | http://arxiv.org/pdf/2402.14273v1 | http://arxiv.org/abs/2402.14273v1 | Qiyuan He, Yizhong Wang, Wenya Wang | 2024-02-22T04:20:14Z | cs.CL | Large language models (LLMs) have demonstrated remarkable proficiency in
understanding and generating responses to complex queries through large-scale
pre-training. However, the efficacy of these models in memorizing and reasoning
among large-scale structured knowledge, especially world knowledge that
explicitly covers abundant factual information remains questionable. Addressing
this gap, our research investigates whether LLMs can effectively store, recall,
and reason with knowledge on a large scale comparable to latest knowledge bases
(KBs) such as Wikidata. Specifically, we focus on three crucial aspects to
study the viability: (1) the efficiency of LLMs with different sizes in
memorizing the exact knowledge in the large-scale KB; (2) the flexibility of
recalling the memorized knowledge in response to natural language queries; (3)
the capability to infer new knowledge through reasoning. Our findings indicate
that while LLMs hold promise as large-scale KBs capable of retrieving and
responding with flexibility, enhancements in their reasoning capabilities are
necessary to fully realize their potential. | cs.CL | Can Language Models Act as Knowledge Bases at Scale? |
2024-02-22T04:14:10Z | http://arxiv.org/pdf/2402.14272v1 | http://arxiv.org/abs/2402.14272v1 | Yan Lei, Liang Pang, Yuanzhuo Wang, Huawei Shen, Xueqi Cheng | 2024-02-22T04:14:10Z | cs.CL | The questionnaire is a professional research methodology used for both
qualitative and quantitative analysis of human opinions, preferences,
attitudes, and behaviors. However, designing and evaluating questionnaires
demands significant effort due to their intricate and complex structure.
Questionnaires entail a series of questions that must conform to intricate
constraints involving the questions, options, and overall structure.
Specifically, the questions should be relevant and specific to the given
research topic and intent. The options should be tailored to the questions,
ensuring they are mutually exclusive, completed, and ordered sensibly.
Moreover, the sequence of questions should follow a logical order, grouping
similar topics together. As a result, automatically generating questionnaires
presents a significant challenge and this area has received limited attention
primarily due to the scarcity of high-quality datasets. To address these
issues, we present Qsnail, the first dataset specifically constructed for the
questionnaire generation task, which comprises 13,168 human-written
questionnaires gathered from online platforms. We further conduct experiments
on Qsnail, and the results reveal that retrieval models and traditional
generative models do not fully align with the given research topic and intents.
Large language models, while more closely related to the research topic and
intents, exhibit significant limitations in terms of diversity and specificity.
Despite enhancements through the chain-of-thought prompt and finetuning,
questionnaires generated by language models still fall short of human-written
questionnaires. Therefore, questionnaire generation is challenging and needs to
be further explored. The dataset is available at:
https://github.com/LeiyanGithub/qsnail. | cs.CL | Qsnail: A Questionnaire Dataset for Sequential Question Generation |
2024-02-22T04:07:00Z | http://arxiv.org/pdf/2402.14268v1 | http://arxiv.org/abs/2402.14268v1 | Yupeng Cao, Aishwarya Muralidharan Nair, Elyon Eyimife, Nastaran Jamalipour Soofi, K. P. Subbalakshmi, John R. Wullert II, Chumki Basu, David Shallcross | 2024-02-22T04:07:00Z | cs.CL, cs.AI, cs.SI | Scientific facts are often spun in the popular press with the intent to
influence public opinion and action, as was evidenced during the COVID-19
pandemic. Automatic detection of misinformation in the scientific domain is
challenging because of the distinct styles of writing in these two media types
and is still in its nascence. Most research on the validity of scientific
reporting treats this problem as a claim verification challenge. In doing so,
significant expert human effort is required to generate appropriate claims. Our
solution bypasses this step and addresses a more real-world scenario where such
explicit, labeled claims may not be available. The central research question of
this paper is whether it is possible to use large language models (LLMs) to
detect misinformation in scientific reporting. To this end, we first present a
new labeled dataset SciNews, containing 2.4k scientific news stories drawn from
trusted and untrustworthy sources, paired with related abstracts from the
CORD-19 database. Our dataset includes both human-written and LLM-generated
news articles, making it more comprehensive in terms of capturing the growing
trend of using LLMs to generate popular press articles. Then, we identify
dimensions of scientific validity in science news articles and explore how this
can be integrated into the automated detection of scientific misinformation. We
propose several baseline architectures using LLMs to automatically detect false
representations of scientific findings in the popular press. For each of these
architectures, we use several prompt engineering strategies including
zero-shot, few-shot, and chain-of-thought prompting. We also test these
architectures and prompting strategies on GPT-3.5, GPT-4, and Llama2-7B,
Llama2-13B. | cs.CL | Can Large Language Models Detect Misinformation in Scientific News Reporting? |
2024-02-22T03:46:08Z | http://arxiv.org/pdf/2402.14259v1 | http://arxiv.org/abs/2402.14259v1 | Zhiyuan Wang, Jinhao Duan, Chenxi Yuan, Qingyu Chen, Tianlong Chen, Huaxiu Yao, Yue Zhang, Ren Wang, Kaidi Xu, Xiaoshuang Shi | 2024-02-22T03:46:08Z | cs.CL, cs.AI, cs.LG | Uncertainty estimation plays a pivotal role in ensuring the reliability of
safety-critical human-AI interaction systems, particularly in the medical
domain. However, a general method for quantifying the uncertainty of free-form
answers has yet to be established in open-ended medical question-answering (QA)
tasks, where irrelevant words and sequences with limited semantic information
can be the primary source of uncertainty due to the presence of generative
inequality. In this paper, we propose the Word-Sequence Entropy (WSE), which
calibrates the uncertainty proportion at both the word and sequence levels
according to the semantic relevance, with greater emphasis placed on keywords
and more relevant sequences when performing uncertainty quantification. We
compare WSE with 6 baseline methods on 5 free-form medical QA datasets,
utilizing 7 "off-the-shelf" large language models (LLMs), and show that WSE
exhibits superior performance on accurate uncertainty measurement under two
standard criteria for correctness evaluation (e.g., WSE outperforms existing
state-of-the-art method by 3.23% AUROC on the MedQA dataset). Additionally, in
terms of the potential for real-world medical QA applications, we achieve a
significant enhancement in the performance of LLMs when employing sequences
with lower uncertainty, identified by WSE, as final answers (e.g., +6.36%
accuracy improvement on the COVID-QA dataset), without requiring any additional
task-specific fine-tuning or architectural modifications. | cs.CL | Word-Sequence Entropy: Towards Uncertainty Estimation in Free-Form Medical Question Answering Applications and Beyond |
2024-02-22T03:46:02Z | http://arxiv.org/pdf/2402.14258v1 | http://arxiv.org/abs/2402.14258v1 | Masahiro Kaneko, Danushka Bollegala, Timothy Baldwin | 2024-02-22T03:46:02Z | cs.CL | Recent studies have demonstrated that large language models (LLMs) have
ethical-related problems such as social biases, lack of moral reasoning, and
generation of offensive content. The existing evaluation metrics and methods to
address these ethical challenges use datasets intentionally created by
instructing humans to create instances including ethical problems. Therefore,
the data does not reflect prompts that users actually provide when utilizing
LLM services in everyday contexts. This may not lead to the development of safe
LLMs that can address ethical challenges arising in real-world applications. In
this paper, we create Eagle datasets extracted from real interactions between
ChatGPT and users that exhibit social biases, toxicity, and immoral problems.
Our experiments show that Eagle captures complementary aspects, not covered by
existing datasets proposed for evaluation and mitigation of such ethical
challenges. Our code is publicly available at
https://huggingface.co/datasets/MasahiroKaneko/eagle. | cs.CL | Eagle: Ethical Dataset Given from Real Interactions |
2024-02-22T02:07:21Z | http://arxiv.org/pdf/2402.14224v2 | http://arxiv.org/abs/2402.14224v2 | Alexandria Leto, Elliot Pickens, Coen D. Needell, David Rothschild, Maria Leonor Pacheco | 2024-02-23T02:19:22Z | cs.CL | The mainstream media has much leeway in what it chooses to cover and how it
covers it. These choices have real-world consequences on what people know and
their subsequent behaviors. However, the lack of objective measures to evaluate
editorial choices makes research in this area particularly difficult. In this
paper, we argue that there are newsworthy topics where objective measures exist
in the form of supporting data and propose a computational framework to analyze
editorial choices in this setup. We focus on the economy because the reporting
of economic indicators presents us with a relatively easy way to determine both
the selection and framing of various publications. Their values provide a
ground truth of how the economy is doing relative to how the publications
choose to cover it. To do this, we define frame prediction as a set of
interdependent tasks. At the article level, we learn to identify the reported
stance towards the general state of the economy. Then, for every numerical
quantity reported in the article, we learn to identify whether it corresponds
to an economic indicator and whether it is being reported in a positive or
negative way. To perform our analysis, we track six American publishers and
each article that appeared in the top 10 slots of their landing page between
2015 and 2023. | cs.CL | Framing in the Presence of Supporting Data: A Case Study in U.S. Economic News |
2024-02-22T01:20:51Z | http://arxiv.org/pdf/2402.14208v3 | http://arxiv.org/abs/2402.14208v3 | Wenlong Deng, Blair Chen, Beidi Zhao, Chiyu Zhang, Xiaoxiao Li, Christos Thrampoulidis | 2024-06-24T04:49:16Z | cs.CL, cs.AI, cs.CY, cs.LG | Mitigating biases in machine learning models has become an increasing concern
in Natural Language Processing (NLP), particularly in developing fair text
embeddings, which are crucial yet challenging for real-world applications like
search engines. In response, this paper proposes a novel method for learning
fair text embeddings. First, we define a novel content-conditional equal
distance (CCED) fairness for text embeddings, ensuring content-conditional
independence between sensitive attributes and text embeddings. Building on
CCED, we introduce a content-conditional debiasing (CCD) loss to ensure that
embeddings of texts with different sensitive attributes but identical content
maintain the same distance from the embedding of their corresponding neutral
text. Additionally, we tackle the issue of insufficient training data by using
Large Language Models (LLMs) with instructions to fairly augment texts into
different sensitive groups. Our extensive evaluations show that our approach
effectively enhances fairness while maintaining the utility of embeddings.
Furthermore, our augmented dataset, combined with the CCED metric, serves as an
new benchmark for evaluating fairness. | cs.CL | LLM-Assisted Content Conditional Debiasing for Fair Text Embedding |
2024-02-22T01:20:17Z | http://arxiv.org/pdf/2402.14207v2 | http://arxiv.org/abs/2402.14207v2 | Yijia Shao, Yucheng Jiang, Theodore A. Kanell, Peter Xu, Omar Khattab, Monica S. Lam | 2024-04-08T05:38:50Z | cs.CL, cs.AI | We study how to apply large language models to write grounded and organized
long-form articles from scratch, with comparable breadth and depth to Wikipedia
pages. This underexplored problem poses new challenges at the pre-writing
stage, including how to research the topic and prepare an outline prior to
writing. We propose STORM, a writing system for the Synthesis of Topic Outlines
through Retrieval and Multi-perspective Question Asking. STORM models the
pre-writing stage by (1) discovering diverse perspectives in researching the
given topic, (2) simulating conversations where writers carrying different
perspectives pose questions to a topic expert grounded on trusted Internet
sources, (3) curating the collected information to create an outline.
For evaluation, we curate FreshWiki, a dataset of recent high-quality
Wikipedia articles, and formulate outline assessments to evaluate the
pre-writing stage. We further gather feedback from experienced Wikipedia
editors. Compared to articles generated by an outline-driven
retrieval-augmented baseline, more of STORM's articles are deemed to be
organized (by a 25% absolute increase) and broad in coverage (by 10%). The
expert feedback also helps identify new challenges for generating grounded long
articles, such as source bias transfer and over-association of unrelated facts. | cs.CL | Assisting in Writing Wikipedia-like Articles From Scratch with Large Language Models |
2024-02-22T01:02:37Z | http://arxiv.org/pdf/2402.14200v1 | http://arxiv.org/abs/2402.14200v1 | Younghun Lee, Dan Goldwasser, Laura Schwab Reese | 2024-02-22T01:02:37Z | cs.CL | Understanding the dynamics of counseling conversations is an important task,
yet it is a challenging NLP problem regardless of the recent advance of
Transformer-based pre-trained language models. This paper proposes a systematic
approach to examine the efficacy of domain knowledge and large language models
(LLMs) in better representing conversations between a crisis counselor and a
help seeker. We empirically show that state-of-the-art language models such as
Transformer-based models and GPT models fail to predict the conversation
outcome. To provide richer context to conversations, we incorporate
human-annotated domain knowledge and LLM-generated features; simple integration
of domain knowledge and LLM features improves the model performance by
approximately 15%. We argue that both domain knowledge and LLM-generated
features can be exploited to better characterize counseling conversations when
they are used as an additional context to conversations. | cs.CL | Towards Understanding Counseling Conversations: Domain Knowledge and Large Language Models |
2024-02-22T00:41:23Z | http://arxiv.org/pdf/2402.14195v1 | http://arxiv.org/abs/2402.14195v1 | Younghun Lee, Sungchul Kim, Tong Yu, Ryan A. Rossi, Xiang Chen | 2024-02-22T00:41:23Z | cs.CL | Large Language Models (LLMs) have been widely used as general-purpose AI
agents showing comparable performance on many downstream tasks. However,
existing work shows that it is challenging for LLMs to integrate structured
data (e.g. KG, tables, DBs) into their prompts; LLMs need to either understand
long text data or select the most relevant evidence prior to inference, and
both approaches are not trivial.
In this paper, we propose a framework, Learning to Reduce, that fine-tunes a
language model to generate a reduced version of an input context, given a task
description and context input. The model learns to reduce the input context
using On-Policy Reinforcement Learning and aims to improve the reasoning
performance of a fixed LLM. Experimental results illustrate that our model not
only achieves comparable accuracies in selecting the relevant evidence from an
input context, but also shows generalizability on different datasets. We
further show that our model helps improve the LLM's performance on downstream
tasks especially when the context is long. | cs.CL | Learning to Reduce: Optimal Representations of Structured Data in Prompting Large Language Models |
2024-02-22T00:24:44Z | http://arxiv.org/pdf/2402.14187v1 | http://arxiv.org/abs/2402.14187v1 | Yuhang Zhou, Xuan Lu, Wei Ai | 2024-02-22T00:24:44Z | cs.CY, cs.AI, cs.CL, cs.HC | In the rapidly evolving landscape of social media, the introduction of new
emojis in Unicode release versions presents a structured opportunity to explore
digital language evolution. Analyzing a large dataset of sampled English
tweets, we examine how newly released emojis gain traction and evolve in
meaning. We find that community size of early adopters and emoji semantics are
crucial in determining their popularity. Certain emojis experienced notable
shifts in the meanings and sentiment associations during the diffusion process.
Additionally, we propose a novel framework utilizing language models to extract
words and pre-existing emojis with semantically similar contexts, which
enhances interpretation of new emojis. The framework demonstrates its
effectiveness in improving sentiment classification performance by substituting
unknown new emojis with familiar ones. This study offers a new perspective in
understanding how new language units are adopted, adapted, and integrated into
the fabric of online communication. | cs.CY | From Adoption to Adaption: Tracing the Diffusion of New Emojis on Twitter |
2024-02-22T00:04:21Z | http://arxiv.org/pdf/2402.14184v1 | http://arxiv.org/abs/2402.14184v1 | Polina Proskura, Alexey Zaytsev | 2024-02-22T00:04:21Z | cs.LG, cs.CL | Ensembles are important tools for improving the performance of machine
learning models. In cases related to natural language processing, ensembles
boost the performance of a method due to multiple large models available in
open source. However, existing approaches mostly rely on simple averaging of
predictions by ensembles with equal weights for each model, ignoring
differences in the quality and conformity of models. We propose to estimate
weights for ensembles of NLP models using not only knowledge of their
individual performance but also their similarity to each other. By adopting
distance measures based on Topological Data Analysis (TDA), we improve our
ensemble. The quality improves for both text classification accuracy and
relevant uncertainty estimation. | cs.LG | Diversity-Aware Ensembling of Language Models Based on Topological Data Analysis |
2024-02-21T23:51:06Z | http://arxiv.org/pdf/2402.14881v1 | http://arxiv.org/abs/2402.14881v1 | Shanker Ram, Chen Qian | 2024-02-21T23:51:06Z | cs.CL, cs.AI, cs.CY, I.2.7 | ChatGPT is a chatbot that can answer text prompts fairly accurately, even
performing very well on postgraduate-level questions. Many educators have found
that their take-home or remote tests and exams are vulnerable to ChatGPT-based
cheating because students may directly use answers provided by tools like
ChatGPT. In this paper, we try to provide an answer to an important question:
how well ChatGPT can answer test questions and how we can detect whether the
questions of a test can be answered correctly by ChatGPT. We generated
ChatGPT's responses to the MedMCQA dataset, which contains over 10,000 medical
school entrance exam questions. We analyzed the responses and uncovered certain
types of questions ChatGPT answers more inaccurately than others. In addition,
we have created a basic natural language processing model to single out the
most vulnerable questions to ChatGPT in a collection of questions or a sample
exam. Our tool can be used by test-makers to avoid ChatGPT-vulnerable test
questions. | cs.CL | A Study on the Vulnerability of Test Questions against ChatGPT-based Cheating |
2024-02-21T23:50:37Z | http://arxiv.org/pdf/2402.16611v2 | http://arxiv.org/abs/2402.16611v2 | Crystal Qian, Emily Reif, Minsuk Kahng | 2024-04-01T19:58:43Z | cs.CL, cs.AI, cs.HC | As large language models (LLMs) become more advanced and impactful, it is
increasingly important to scrutinize the data that they rely upon and produce.
What is it to be a dataset practitioner doing this work? We approach this in
two parts: first, we define the role of "dataset practitioners" by performing a
retrospective analysis on the responsibilities of teams contributing to LLM
development at a technology company, Google. Then, we conduct semi-structured
interviews with a cross-section of these practitioners (N=10). We find that
although data quality is a top priority, there is little consensus around what
data quality is and how to evaluate it. Consequently, practitioners either rely
on their own intuition or write custom code to evaluate their data. We discuss
potential reasons for this phenomenon and opportunities for alignment. | cs.CL | Understanding the Dataset Practitioners Behind Large Language Model Development |
2024-02-21T23:43:04Z | http://arxiv.org/pdf/2402.14179v1 | http://arxiv.org/abs/2402.14179v1 | MD Ashraful Goni, Fahad Mostafa, Kerk F. Kee | 2024-02-21T23:43:04Z | cs.CL, cs.AI | Ethnic media, which caters to diaspora communities in host nations, serves as
a vital platform for these communities to both produce content and access
information. Rather than utilizing the language of the host nation, ethnic
media delivers news in the language of the immigrant community. For instance,
in the USA, Bangla ethnic media presents news in Bangla rather than English.
This research delves into the prospective integration of large language models
(LLM) and multi-lingual machine translations (MMT) within the ethnic media
industry. It centers on the transformative potential of using LLM in MMT in
various facets of news translation, searching, and categorization. The paper
outlines a theoretical framework elucidating the integration of LLM and MMT
into the news searching and translation processes for ethnic media.
Additionally, it briefly addresses the potential ethical challenges associated
with the incorporation of LLM and MMT in news translation procedures. | cs.CL | Bangla AI: A Framework for Machine Translation Utilizing Large Language Models for Ethnic Media |
2024-02-21T22:41:38Z | http://arxiv.org/pdf/2402.14158v2 | http://arxiv.org/abs/2402.14158v2 | Dheeraj Mekala, Jason Weston, Jack Lanchantin, Roberta Raileanu, Maria Lomeli, Jingbo Shang, Jane Dwivedi-Yu | 2024-03-13T16:38:42Z | cs.CL | Teaching language models to use tools is an important milestone towards
building general assistants, but remains an open problem. While there has been
significant progress on learning to use specific tools via fine-tuning,
language models still struggle with learning how to robustly use new tools from
only a few demonstrations. In this work we introduce a self-verification method
which distinguishes between close candidates by self-asking contrastive
questions during (1) tool selection; and (2) parameter generation. We construct
synthetic, high-quality, self-generated data for this goal using Llama-2 70B,
which we intend to release publicly. Extensive experiments on 4 tasks from the
ToolBench benchmark, consisting of 17 unseen tools, demonstrate an average
improvement of 22% over few-shot baselines, even in scenarios where the
distinctions between candidate tools are finely nuanced. | cs.CL | TOOLVERIFIER: Generalization to New Tools via Self-Verification |
2024-02-21T22:30:57Z | http://arxiv.org/pdf/2402.14155v1 | http://arxiv.org/abs/2402.14155v1 | Amogh Mannekote, Xiaoyi Tian, Kristy Elizabeth Boyer, Bonnie J. Dorr | 2024-02-21T22:30:57Z | cs.CL, cs.AI | Task-oriented dialogue systems are expected to handle a constantly expanding
set of intents and domains even after they have been deployed to support more
and more functionalities. To live up to this expectation, it becomes critical
to mitigate the catastrophic forgetting problem (CF) that occurs in continual
learning (CL) settings for a task such as intent recognition. While existing
dialogue systems research has explored replay-based and regularization-based
methods to this end, the effect of domain ordering on the CL performance of
intent recognition models remains unexplored. If understood well, domain
ordering has the potential to be an orthogonal technique that can be leveraged
alongside existing techniques such as experience replay. Our work fills this
gap by comparing the impact of three domain-ordering strategies (min-sum path,
max-sum path, random) on the CL performance of a generative intent recognition
model. Our findings reveal that the min-sum path strategy outperforms the
others in reducing catastrophic forgetting when training on the 220M T5-Base
model. However, this advantage diminishes with the larger 770M T5-Large model.
These results underscores the potential of domain ordering as a complementary
strategy for mitigating catastrophic forgetting in continually learning intent
recognition models, particularly in resource-constrained scenarios. | cs.CL | Can Similarity-Based Domain-Ordering Reduce Catastrophic Forgetting for Intent Recognition? |
2024-02-21T22:29:16Z | http://arxiv.org/pdf/2402.14880v1 | http://arxiv.org/abs/2402.14880v1 | Emily Reif, Crystal Qian, James Wexler, Minsuk Kahng | 2024-02-21T22:29:16Z | cs.CL, cs.AI, cs.HC | Making sense of unstructured text datasets is perennially difficult, yet
increasingly relevant with Large Language Models. Data workers often rely on
dataset summaries, especially distributions of various derived features. Some
features, like toxicity or topics, are relevant to many datasets, but many
interesting features are domain specific: instruments and genres for a music
dataset, or diseases and symptoms for a medical dataset. Accordingly, data
workers often run custom analyses for each dataset, which is cumbersome and
difficult. We present AutoHistograms, a visualization tool leveragingLLMs.
AutoHistograms automatically identifies relevant features, visualizes them with
histograms, and allows the user to interactively query the dataset for
categories of entities and create new histograms. In a user study with 10 data
workers (n=10), we observe that participants can quickly identify insights and
explore the data using AutoHistograms, and conceptualize a broad range of
applicable use cases. Together, this tool and user study contributeto the
growing field of LLM-assisted sensemaking tools. | cs.CL | Automatic Histograms: Leveraging Language Models for Text Dataset Exploration |
2024-02-21T22:27:40Z | http://arxiv.org/pdf/2402.14154v1 | http://arxiv.org/abs/2402.14154v1 | Yiqiao Jin, Minje Choi, Gaurav Verma, Jindong Wang, Srijan Kumar | 2024-02-21T22:27:40Z | cs.CL, cs.CV, cs.CY | Social media platforms are hubs for multimodal information exchange,
encompassing text, images, and videos, making it challenging for machines to
comprehend the information or emotions associated with interactions in online
spaces. Multimodal Large Language Models (MLLMs) have emerged as a promising
solution to address these challenges, yet struggle with accurately interpreting
human emotions and complex contents like misinformation. This paper introduces
MM-Soc, a comprehensive benchmark designed to evaluate MLLMs' understanding of
multimodal social media content. MM-Soc compiles prominent multimodal datasets
and incorporates a novel large-scale YouTube tagging dataset, targeting a range
of tasks from misinformation detection, hate speech detection, and social
context generation. Through our exhaustive evaluation on ten size-variants of
four open-source MLLMs, we have identified significant performance disparities,
highlighting the need for advancements in models' social understanding
capabilities. Our analysis reveals that, in a zero-shot setting, various types
of MLLMs generally exhibit difficulties in handling social media tasks.
However, MLLMs demonstrate performance improvements post fine-tuning,
suggesting potential pathways for improvement. | cs.CL | MM-Soc: Benchmarking Multimodal Large Language Models in Social Media Platforms |
2024-02-21T22:22:30Z | http://arxiv.org/pdf/2402.14151v2 | http://arxiv.org/abs/2402.14151v2 | Xiaoyue Wang, Jianyou Wang, Weili Cao, Kaicheng Wang, Ramamohan Paturi, Leon Bergen | 2024-04-03T20:11:02Z | cs.IR, cs.AI, cs.CL, cs.LG | We present the Benchmark of Information Retrieval (IR) tasks with Complex
Objectives (BIRCO). BIRCO evaluates the ability of IR systems to retrieve
documents given multi-faceted user objectives. The benchmark's complexity and
compact size make it suitable for evaluating large language model (LLM)-based
information retrieval systems. We present a modular framework for investigating
factors that may influence LLM performance on retrieval tasks, and identify a
simple baseline model which matches or outperforms existing approaches and more
complex alternatives. No approach achieves satisfactory performance on all
benchmark tasks, suggesting that stronger models and new retrieval protocols
are necessary to address complex user needs. | cs.IR | BIRCO: A Benchmark of Information Retrieval Tasks with Complex Objectives |
2024-02-21T22:02:37Z | http://arxiv.org/pdf/2402.14146v1 | http://arxiv.org/abs/2402.14146v1 | Karin de Langis, Ryan Koo, Dongyeop Kang | 2024-02-21T22:02:37Z | cs.CL | Style is an integral component of text that expresses a diverse set of
information, including interpersonal dynamics (e.g. formality) and the author's
emotions or attitudes (e.g. disgust). Humans often employ multiple styles
simultaneously. An open question is how large language models can be explicitly
controlled so that they weave together target styles when generating text: for
example, to produce text that is both negative and non-toxic. Previous work
investigates the controlled generation of a single style, or else controlled
generation of a style and other attributes. In this paper, we expand this into
controlling multiple styles simultaneously. Specifically, we investigate
various formulations of multiple style rewards for a reinforcement learning
(RL) approach to controlled multi-style generation. These reward formulations
include calibrated outputs from discriminators and dynamic weighting by
discriminator gradient magnitudes. We find that dynamic weighting generally
outperforms static weighting approaches, and we explore its effectiveness in 2-
and 3-style control, even compared to strong baselines like plug-and-play
model. All code and data for RL pipelines with multiple style attributes will
be publicly available. | cs.CL | Reinforcement Learning with Dynamic Multi-Reward Weighting for Multi-Style Controllable Generation |
2024-02-21T21:29:57Z | http://arxiv.org/pdf/2402.14879v1 | http://arxiv.org/abs/2402.14879v1 | Lawrence J. Klinkert, Stephanie Buongiorno, Corey Clark | 2024-02-21T21:29:57Z | cs.CL, cs.AI | This research explores the potential of Large Language Models (LLMs) to
utilize psychometric values, specifically personality information, within the
context of video game character development. Affective Computing (AC) systems
quantify a Non-Player character's (NPC) psyche, and an LLM can take advantage
of the system's information by using the values for prompt generation. The
research shows an LLM can consistently represent a given personality profile,
thereby enhancing the human-like characteristics of game characters.
Repurposing a human examination, the International Personality Item Pool (IPIP)
questionnaire, to evaluate an LLM shows that the model can accurately generate
content concerning the personality provided. Results show that the improvement
of LLM, such as the latest GPT-4 model, can consistently utilize and interpret
a personality to represent behavior. | cs.CL | Driving Generative Agents With Their Personality |
2024-02-21T20:53:29Z | http://arxiv.org/pdf/2402.14129v1 | http://arxiv.org/abs/2402.14129v1 | Zhi Hong, Kyle Chard, Ian Foster | 2024-02-21T20:53:29Z | cs.IR, cs.CL | Relation extraction is an efficient way of mining the extraordinary wealth of
human knowledge on the Web. Existing methods rely on domain-specific training
data or produce noisy outputs. We focus here on extracting targeted relations
from semi-structured web pages given only a short description of the relation.
We present GraphScholarBERT, an open-domain information extraction method based
on a joint graph and language model structure. GraphScholarBERT can generalize
to previously unseen domains without additional data or training and produces
only clean extraction results matched to the search keyword. Experiments show
that GraphScholarBERT can improve extraction F1 scores by as much as 34.8\%
compared to previous work in a zero-shot domain and zero-shot website setting. | cs.IR | Combining Language and Graph Models for Semi-structured Information Extraction on the Web |
2024-02-21T20:30:45Z | http://arxiv.org/pdf/2402.14116v2 | http://arxiv.org/abs/2402.14116v2 | Andrew Zhu, Alyssa Hwang, Liam Dugan, Chris Callison-Burch | 2024-06-06T16:41:21Z | cs.CL, cs.AI | One type of question that is commonly found in day-to-day scenarios is
``fan-out'' questions, complex multi-hop, multi-document reasoning questions
that require finding information about a large number of entities. However,
there exist few resources to evaluate this type of question-answering
capability among large language models. To evaluate complex reasoning in LLMs
more fully, we present FanOutQA, a high-quality dataset of fan-out
question-answer pairs and human-annotated decompositions with English Wikipedia
as the knowledge base. We formulate three benchmark settings across our dataset
and benchmark 7 LLMs, including GPT-4, LLaMA 2, Claude-2.1, and Mixtral-8x7B,
finding that contemporary models still have room to improve reasoning over
inter-document dependencies in a long context. We provide our dataset and
open-source tools to run models to encourage evaluation at https://fanoutqa.com | cs.CL | FanOutQA: A Multi-Hop, Multi-Document Question Answering Benchmark for Large Language Models |
2024-02-21T19:53:36Z | http://arxiv.org/pdf/2402.14101v1 | http://arxiv.org/abs/2402.14101v1 | Preni Golazizian, Ali Omrani, Alireza S. Ziabari, Morteza Dehghani | 2024-02-21T19:53:36Z | cs.CL | In subjective NLP tasks, where a single ground truth does not exist, the
inclusion of diverse annotators becomes crucial as their unique perspectives
significantly influence the annotations. In realistic scenarios, the annotation
budget often becomes the main determinant of the number of perspectives (i.e.,
annotators) included in the data and subsequent modeling. We introduce a novel
framework for annotation collection and modeling in subjective tasks that aims
to minimize the annotation budget while maximizing the predictive performance
for each annotator. Our framework has a two-stage design: first, we rely on a
small set of annotators to build a multitask model, and second, we augment the
model for a new perspective by strategically annotating a few samples per
annotator. To test our framework at scale, we introduce and release a unique
dataset, Moral Foundations Subjective Corpus, of 2000 Reddit posts annotated by
24 annotators for moral sentiment. We demonstrate that our framework surpasses
the previous SOTA in capturing the annotators' individual perspectives with as
little as 25% of the original annotation budget on two datasets. Furthermore,
our framework results in more equitable models, reducing the performance
disparity among annotators. | cs.CL | Cost-Efficient Subjective Task Annotation and Modeling through Few-Shot Annotator Adaptation |
2024-02-21T19:20:06Z | http://arxiv.org/pdf/2402.14086v1 | http://arxiv.org/abs/2402.14086v1 | Zheng-Xin Yong, Cristina Menghini, Stephen H. Bach | 2024-02-21T19:20:06Z | cs.CL, cs.AI, cs.LG | Data scarcity in low-resource languages can be addressed with word-to-word
translations from labeled task data in high-resource languages using bilingual
lexicons. However, bilingual lexicons often have limited lexical overlap with
task data, which results in poor translation coverage and lexicon utilization.
We propose lexicon-conditioned data generation (LexC-Gen), a method that
generates low-resource-language classification task data at scale.
Specifically, LexC-Gen first uses high-resource-language words from bilingual
lexicons to generate lexicon-compatible task data, and then it translates them
into low-resource languages with bilingual lexicons via word translation.
Across 17 extremely low-resource languages, LexC-Gen generated data is
competitive with expert-translated gold data, and yields on average 5.6 and 8.9
points improvement over existing lexicon-based word translation methods on
sentiment analysis and topic classification tasks respectively. We show that
conditioning on bilingual lexicons is the key component of LexC-Gen. LexC-Gen
is also practical -- it only needs a single GPU to generate data at scale. It
works well with open-access LLMs, and its cost is one-fifth of the cost of
GPT4-based multilingual data generation. | cs.CL | LexC-Gen: Generating Data for Extremely Low-Resource Languages with Large Language Models and Bilingual Lexicons |
2024-02-21T19:01:03Z | http://arxiv.org/pdf/2402.14073v1 | http://arxiv.org/abs/2402.14073v1 | Tianyu Gao, Zirui Wang, Adithya Bhaskar, Danqi Chen | 2024-02-21T19:01:03Z | cs.CL, cs.CV, cs.LG | An emerging family of language models (LMs), capable of processing both text
and images within a single visual view, has the promise to unlock complex tasks
such as chart understanding and UI navigation. We refer to these models as
screenshot language models. Despite their appeal, existing screenshot LMs
substantially lag behind text-only models on language understanding tasks. To
close this gap, we adopt a simplified setting where the model inputs are
plain-text-rendered screenshots, and we focus on improving the text ability of
screenshot LMs. We propose a novel Patch-and-Text Prediction (PTP) objective,
which masks and recovers both image patches of screenshots and text within
screenshots. We also conduct extensive ablation studies on masking rates and
patch sizes, as well as designs for improving training stability. Our
pre-trained model, while solely taking visual inputs, achieves comparable
performance with BERT on 6 out of 8 GLUE tasks (within 2%) and improves up to
8% over prior work. Additionally, we extend PTP to train autoregressive
screenshot LMs and demonstrate its effectiveness--our models can significantly
reduce perplexity by utilizing the screenshot context. Together, we hope our
findings can inspire future research on developing powerful screenshot LMs and
extending their reach to broader applications. | cs.CL | Improving Language Understanding from Screenshots |
2024-02-21T18:59:13Z | http://arxiv.org/pdf/2402.14020v1 | http://arxiv.org/abs/2402.14020v1 | Jonas Geiping, Alex Stein, Manli Shu, Khalid Saifullah, Yuxin Wen, Tom Goldstein | 2024-02-21T18:59:13Z | cs.LG, cs.CL, cs.CR | It has recently been shown that adversarial attacks on large language models
(LLMs) can "jailbreak" the model into making harmful statements. In this work,
we argue that the spectrum of adversarial attacks on LLMs is much larger than
merely jailbreaking. We provide a broad overview of possible attack surfaces
and attack goals. Based on a series of concrete examples, we discuss,
categorize and systematize attacks that coerce varied unintended behaviors,
such as misdirection, model control, denial-of-service, or data extraction.
We analyze these attacks in controlled experiments, and find that many of
them stem from the practice of pre-training LLMs with coding capabilities, as
well as the continued existence of strange "glitch" tokens in common LLM
vocabularies that should be removed for security reasons. | cs.LG | Coercing LLMs to do and reveal (almost) anything |
2024-02-21T18:57:54Z | http://arxiv.org/pdf/2402.14052v1 | http://arxiv.org/abs/2402.14052v1 | Di Wu, Wasi Uddin Ahmad, Kai-Wei Chang | 2024-02-21T18:57:54Z | cs.CL | This study addresses the application of encoder-only Pre-trained Language
Models (PLMs) in keyphrase generation (KPG) amidst the broader availability of
domain-tailored encoder-only models compared to encoder-decoder models. We
investigate three core inquiries: (1) the efficacy of encoder-only PLMs in KPG,
(2) optimal architectural decisions for employing encoder-only PLMs in KPG, and
(3) a performance comparison between in-domain encoder-only and encoder-decoder
PLMs across varied resource settings. Our findings, derived from extensive
experimentation in two domains reveal that with encoder-only PLMs, although KPE
with Conditional Random Fields slightly excels in identifying present
keyphrases, the KPG formulation renders a broader spectrum of keyphrase
predictions. Additionally, prefix-LM fine-tuning of encoder-only PLMs emerges
as a strong and data-efficient strategy for KPG, outperforming general-domain
seq2seq PLMs. We also identify a favorable parameter allocation towards model
depth rather than width when employing encoder-decoder architectures
initialized with encoder-only PLMs. The study sheds light on the potential of
utilizing encoder-only PLMs for advancing KPG systems and provides a groundwork
for future KPG methods. Our code and pre-trained checkpoints are released at
https://github.com/uclanlp/DeepKPG. | cs.CL | On Leveraging Encoder-only Pre-trained Language Models for Effective Keyphrase Generation |
2024-02-21T18:55:20Z | http://arxiv.org/pdf/2402.14016v1 | http://arxiv.org/abs/2402.14016v1 | Vyas Raina, Adian Liusie, Mark Gales | 2024-02-21T18:55:20Z | cs.CL | Large Language Models (LLMs) are powerful zero-shot assessors and are
increasingly used in real-world situations such as for written exams or
benchmarking systems. Despite this, no existing work has analyzed the
vulnerability of judge-LLMs against adversaries attempting to manipulate
outputs. This work presents the first study on the adversarial robustness of
assessment LLMs, where we search for short universal phrases that when appended
to texts can deceive LLMs to provide high assessment scores. Experiments on
SummEval and TopicalChat demonstrate that both LLM-scoring and pairwise
LLM-comparative assessment are vulnerable to simple concatenation attacks,
where in particular LLM-scoring is very susceptible and can yield maximum
assessment scores irrespective of the input text quality. Interestingly, such
attacks are transferable and phrases learned on smaller open-source LLMs can be
applied to larger closed-source models, such as GPT3.5. This highlights the
pervasive nature of the adversarial vulnerabilities across different judge-LLM
sizes, families and methods. Our findings raise significant concerns on the
reliability of LLMs-as-a-judge methods, and underscore the importance of
addressing vulnerabilities in LLM assessment methods before deployment in
high-stakes real-world scenarios. | cs.CL | Is LLM-as-a-Judge Robust? Investigating Universal Adversarial Attacks on Zero-shot LLM Assessment |
2024-02-21T18:49:26Z | http://arxiv.org/pdf/2402.14008v2 | http://arxiv.org/abs/2402.14008v2 | Chaoqun He, Renjie Luo, Yuzhuo Bai, Shengding Hu, Zhen Leng Thai, Junhao Shen, Jinyi Hu, Xu Han, Yujie Huang, Yuxiang Zhang, Jie Liu, Lei Qi, Zhiyuan Liu, Maosong Sun | 2024-06-06T13:19:44Z | cs.CL | Recent advancements have seen Large Language Models (LLMs) and Large
Multimodal Models (LMMs) surpassing general human capabilities in various
tasks, approaching the proficiency level of human experts across multiple
domains. With traditional benchmarks becoming less challenging for these
models, new rigorous challenges are essential to gauge their advanced
abilities. In this work, we present OlympiadBench, an Olympiad-level bilingual
multimodal scientific benchmark, featuring 8,476 problems from Olympiad-level
mathematics and physics competitions, including the Chinese college entrance
exam. Each problem is detailed with expert-level annotations for step-by-step
reasoning. Evaluating top-tier models on OlympiadBench, we implement a
comprehensive assessment methodology to accurately evaluate model responses.
Notably, the best-performing model, GPT-4V, attains an average score of 17.97%
on OlympiadBench, with a mere 10.74% in physics, highlighting the benchmark
rigor and the intricacy of physical reasoning. Our analysis orienting GPT-4V
points out prevalent issues with hallucinations, knowledge omissions, and
logical fallacies. We hope that our challenging benchmark can serve as a
valuable resource for helping future AGI research endeavors. The data and
evaluation code are available at \url{https://github.com/OpenBMB/OlympiadBench} | cs.CL | OlympiadBench: A Challenging Benchmark for Promoting AGI with Olympiad-Level Bilingual Multimodal Scientific Problems |
2024-02-21T18:48:38Z | http://arxiv.org/pdf/2402.14007v2 | http://arxiv.org/abs/2402.14007v2 | Zhiwei He, Binglin Zhou, Hongkun Hao, Aiwei Liu, Xing Wang, Zhaopeng Tu, Zhuosheng Zhang, Rui Wang | 2024-06-04T14:24:15Z | cs.CL, cs.AI | Text watermarking technology aims to tag and identify content produced by
large language models (LLMs) to prevent misuse. In this study, we introduce the
concept of cross-lingual consistency in text watermarking, which assesses the
ability of text watermarks to maintain their effectiveness after being
translated into other languages. Preliminary empirical results from two LLMs
and three watermarking methods reveal that current text watermarking
technologies lack consistency when texts are translated into various languages.
Based on this observation, we propose a Cross-lingual Watermark Removal Attack
(CWRA) to bypass watermarking by first obtaining a response from an LLM in a
pivot language, which is then translated into the target language. CWRA can
effectively remove watermarks, decreasing the AUCs to a random-guessing level
without performance loss. Furthermore, we analyze two key factors that
contribute to the cross-lingual consistency in text watermarking and propose
X-SIR as a defense method against CWRA. Code: https://github.com/zwhe99/X-SIR. | cs.CL | Can Watermarks Survive Translation? On the Cross-lingual Consistency of Text Watermark for Large Language Models |
2024-02-21T18:40:24Z | http://arxiv.org/pdf/2402.14002v1 | http://arxiv.org/abs/2402.14002v1 | Aline Ioste | 2024-02-21T18:40:24Z | cs.CL | Large Language Models with transformer architecture have revolutionized the
domain of text generation, setting unprecedented benchmarks. Despite their
impressive capabilities, LLMs have been criticized for generating outcomes that
deviate from factual accuracy or display logical inconsistencies, phenomena
commonly referred to as hallucinations. This term, however, has often been
misapplied to any results deviating from the instructor's expectations, which
this paper defines as attention misdirection rather than true hallucinations.
Understanding the distinction between hallucinations and attention misdirection
becomes increasingly relevant in business contexts, where the ramifications of
such errors can significantly impact the value extraction from these inherently
pre-trained models. This paper highlights the best practices of the PGI,
Persona, Grouping, and Intelligence, method, a strategic framework that
achieved a remarkable error rate of only 3,15 percent across 4,000 responses
generated by GPT in response to a real business challenge. It emphasizes that
by equipping experimentation with knowledge, businesses can unlock
opportunities for innovation through the use of these natively pre-trained
models. This reinforces the notion that strategic application grounded in a
skilled team can maximize the benefits of emergent technologies such as the
LLMs. | cs.CL | Hallucinations or Attention Misdirection? The Path to Strategic Value Extraction in Business Using Large Language Models |
2024-02-21T18:25:25Z | http://arxiv.org/pdf/2402.14875v2 | http://arxiv.org/abs/2402.14875v2 | Amit Haim, Alejandro Salinas, Julian Nyarko | 2024-02-29T19:39:35Z | cs.CL, cs.AI, cs.CY, cs.LG | We employ an audit design to investigate biases in state-of-the-art large
language models, including GPT-4. In our study, we prompt the models for advice
involving a named individual across a variety of scenarios, such as during car
purchase negotiations or election outcome predictions. We find that the advice
systematically disadvantages names that are commonly associated with racial
minorities and women. Names associated with Black women receive the least
advantageous outcomes. The biases are consistent across 42 prompt templates and
several models, indicating a systemic issue rather than isolated incidents.
While providing numerical, decision-relevant anchors in the prompt can
successfully counteract the biases, qualitative details have inconsistent
effects and may even increase disparities. Our findings underscore the
importance of conducting audits at the point of LLM deployment and
implementation to mitigate their potential for harm against marginalized
communities. | cs.CL | What's in a Name? Auditing Large Language Models for Race and Gender Bias |
2024-02-21T18:23:16Z | http://arxiv.org/pdf/2402.13991v1 | http://arxiv.org/abs/2402.13991v1 | Yu Zhao, Yuanbin Qu, Konrad Staniszewski, Szymon Tworkowski, Wei Liu, Piotr Miłoś, Yuxiang Wu, Pasquale Minervini | 2024-02-21T18:23:16Z | cs.CL | Most language model pre-training frameworks concatenate multiple documents
into fixed-length sequences and use causal masking to compute the likelihood of
each token given its context; this strategy is widely adopted due to its
simplicity and efficiency. However, to this day, the influence of the
pre-training sequence composition strategy on the generalisation properties of
the model remains under-explored. In this work, we find that applying causal
masking can lead to the inclusion of distracting information from previous
documents during pre-training, which negatively impacts the performance of the
models on language modelling and downstream tasks. In intra-document causal
masking, the likelihood of each token is only conditioned on the previous
tokens in the same document, eliminating potential distracting information from
previous documents and significantly improving performance. Furthermore, we
find that concatenating related documents can reduce some potential
distractions during pre-training, and our proposed efficient retrieval-based
sequence construction method, BM25Chunk, can improve in-context learning
(+11.6\%), knowledge memorisation (+9.8\%), and context utilisation (+7.2\%)
abilities of language models without sacrificing efficiency. | cs.CL | Analysing The Impact of Sequence Composition on Language Model Pre-Training |
2024-02-21T17:47:20Z | http://arxiv.org/pdf/2402.13963v4 | http://arxiv.org/abs/2402.13963v4 | Pengcheng Qiu, Chaoyi Wu, Xiaoman Zhang, Weixiong Lin, Haicheng Wang, Ya Zhang, Yanfeng Wang, Weidi Xie | 2024-06-02T10:02:00Z | cs.CL | The development of open-source, multilingual medical language models can
benefit a wide, linguistically diverse audience from different regions. To
promote this domain, we present contributions from the following: First, we
construct a multilingual medical corpus, containing approximately 25.5B tokens
encompassing 6 main languages, termed as MMedC, enabling auto-regressive domain
adaptation for general LLMs; Second, to monitor the development of multilingual
medical LLMs, we propose a multilingual medical multi-choice question-answering
benchmark with rationale, termed as MMedBench; Third, we have assessed a number
of open-source large language models (LLMs) on our benchmark, along with those
further auto-regressive trained on MMedC. Our final model, MMed-Llama 3, with
only 8B parameters, achieves superior performance compared to all other
open-source models on both MMedBench and English benchmarks, even rivaling
GPT-4. In conclusion, in this work, we present a large-scale corpus, a
benchmark and a series of models to support the development of multilingual
medical LLMs. | cs.CL | Towards Building Multilingual Language Model for Medicine |
2024-02-21T17:36:07Z | http://arxiv.org/pdf/2402.13956v2 | http://arxiv.org/abs/2402.13956v2 | William Merrill, Zhaofeng Wu, Norihito Naka, Yoon Kim, Tal Linzen | 2024-02-29T22:18:03Z | cs.CL | Do LMs infer the semantics of text from co-occurrence patterns in their
training data? Merrill et al. (2022) argue that, in theory, probabilities
predicted by an optimal LM encode semantic information about entailment
relations, but it is unclear whether neural LMs trained on corpora learn
entailment in this way because of strong idealizing assumptions made by Merrill
et al. In this work, we investigate whether their theory can be used to decode
entailment judgments from neural LMs. We find that a test similar to theirs can
decode entailment relations between natural sentences, well above random
chance, though not perfectly, across many datasets and LMs. This suggests LMs
implicitly model aspects of semantics to predict semantic effects on sentence
co-occurrence patterns. However, we find the test that predicts entailment in
practice works in the opposite direction to the theoretical test. We thus
revisit the assumptions underlying the original test, finding its derivation
did not adequately account for redundancy in human-written text. We argue that
correctly accounting for redundancy related to explanations might derive the
observed flipped test and, more generally, improve linguistic theories of human
speakers. | cs.CL | Can You Learn Semantics Through Next-Word Prediction? The Case of Entailment |
2024-02-21T17:33:13Z | http://arxiv.org/pdf/2402.13954v1 | http://arxiv.org/abs/2402.13954v1 | Rahul Zalkikar, Kanchan Chandra | 2024-02-21T17:33:13Z | cs.CL | Social and political scientists often aim to discover and measure distinct
biases from text data representations (embeddings). Innovative
transformer-based language models produce contextually-aware token embeddings
and have achieved state-of-the-art performance for a variety of natural
language tasks, but have been shown to encode unwanted biases for downstream
applications. In this paper, we evaluate the social biases encoded by
transformers trained with the masked language modeling objective using proposed
proxy functions within an iterative masking experiment to measure the quality
of transformer models' predictions, and assess the preference of MLMs towards
disadvantaged and advantaged groups. We compare bias estimations with those
produced by other evaluation methods using two benchmark datasets, finding
relatively high religious and disability biases across considered MLMs and low
gender bias in one dataset relative to the other. Our measures outperform
others in their agreement with human annotators. We extend on previous work by
evaluating social biases introduced after re-training an MLM under the masked
language modeling objective (w.r.t. the model's pre-trained base), and find
that proposed measures produce more accurate estimations of relative preference
for biased sentences between transformers than others based on our methods. | cs.CL | Measuring Social Biases in Masked Language Models by Proxy of Prediction Quality |
2024-02-21T17:23:59Z | http://arxiv.org/pdf/2402.13950v2 | http://arxiv.org/abs/2402.13950v2 | Debjit Paul, Robert West, Antoine Bosselut, Boi Faltings | 2024-02-23T18:01:48Z | cs.CL | Large language models (LLMs) have been shown to perform better when asked to
reason step-by-step before answering a question. However, it is unclear to what
degree the model's final answer is faithful to the stated reasoning steps. In
this paper, we perform a causal mediation analysis on twelve LLMs to examine
how intermediate reasoning steps generated by the LLM influence the final
outcome and find that LLMs do not reliably use their intermediate reasoning
steps when generating an answer. To address this issue, we introduce FRODO, a
framework to tailor small-sized LMs to generate correct reasoning steps and
robustly reason over these steps. FRODO consists of an inference module that
learns to generate correct reasoning steps using an implicit causal reward
function and a reasoning module that learns to faithfully reason over these
intermediate inferences using a counterfactual and causal preference objective.
Our experiments show that FRODO significantly outperforms four competitive
baselines. Furthermore, FRODO improves the robustness and generalization
ability of the reasoning LM, yielding higher performance on out-of-distribution
test sets. Finally, we find that FRODO's rationales are more faithful to its
final answer predictions than standard supervised fine-tuning. | cs.CL | Making Reasoning Matter: Measuring and Improving Faithfulness of Chain-of-Thought Reasoning |
2024-02-21T17:20:38Z | http://arxiv.org/pdf/2402.14874v1 | http://arxiv.org/abs/2402.14874v1 | Phuc Phan, Hieu Tran, Long Phan | 2024-02-21T17:20:38Z | cs.CL, cs.AI, cs.LG | We propose a straightforward approach called Distillation Contrastive
Decoding (DCD) to enhance the reasoning capabilities of Large Language Models
(LLMs) during inference. In contrast to previous approaches that relied on
smaller amateur models or analysis of hidden state differences, DCD employs
Contrastive Chain-of-thought Prompting and advanced distillation techniques,
including Dropout and Quantization. This approach effectively addresses the
limitations of Contrastive Decoding (CD), which typically requires both an
expert and an amateur model, thus increasing computational resource demands. By
integrating contrastive prompts with distillation, DCD obviates the need for an
amateur model and reduces memory usage. Our evaluations demonstrate that DCD
significantly enhances LLM performance across a range of reasoning benchmarks,
surpassing both CD and existing methods in the GSM8K and StrategyQA datasets. | cs.CL | Distillation Contrastive Decoding: Improving LLMs Reasoning with Contrastive Decoding and Distillation |
2024-02-21T17:13:41Z | http://arxiv.org/pdf/2402.14873v2 | http://arxiv.org/abs/2402.14873v2 | Bradley Emi, Max Spero | 2024-02-26T05:28:41Z | cs.CL, cs.AI, 68T50, I.2.7 | We present the CheckforAI text classifier, a transformer-based neural network
trained to distinguish text written by large language models from text written
by humans. CheckforAI outperforms zero-shot methods such as DetectGPT as well
as leading commercial AI detection tools with over 9 times lower error rates on
a comprehensive benchmark comprised of ten text domains (student writing,
creative writing, scientific writing, books, encyclopedias, news, email,
scientific papers, short-form Q&A) and 8 open- and closed-source large language
models. We propose a training algorithm, hard negative mining with synthetic
mirrors, that enables our classifier to achieve orders of magnitude lower false
positive rates on high-data domains such as reviews. Finally, we show that
CheckforAI is not biased against nonnative English speakers and generalizes to
domains and models unseen during training. | cs.CL | Technical Report on the Checkfor.ai AI-Generated Text Classifier |
2024-02-21T17:05:06Z | http://arxiv.org/pdf/2402.13936v1 | http://arxiv.org/abs/2402.13936v1 | Antoine Chaffin, Ewa Kijak, Vincent Claveau | 2024-02-21T17:05:06Z | cs.CL, cs.CV | Training image captioning models using teacher forcing results in very
generic samples, whereas more distinctive captions can be very useful in
retrieval applications or to produce alternative texts describing images for
accessibility. Reinforcement Learning (RL) allows to use cross-modal retrieval
similarity score between the generated caption and the input image as reward to
guide the training, leading to more distinctive captions. Recent studies show
that pre-trained cross-modal retrieval models can be used to provide this
reward, completely eliminating the need for reference captions. However, we
argue in this paper that Ground Truth (GT) captions can still be useful in this
RL framework. We propose a new image captioning model training strategy that
makes use of GT captions in different ways. Firstly, they can be used to train
a simple MLP discriminator that serves as a regularization to prevent reward
hacking and ensures the fluency of generated captions, resulting in a textual
GAN setup extended for multimodal inputs. Secondly, they can serve as
additional trajectories in the RL strategy, resulting in a teacher forcing loss
weighted by the similarity of the GT to the image. This objective acts as an
additional learning signal grounded to the distribution of the GT captions.
Thirdly, they can serve as strong baselines when added to the pool of captions
used to compute the proposed contrastive reward to reduce the variance of
gradient estimate. Experiments on MS-COCO demonstrate the interest of the
proposed training strategy to produce highly distinctive captions while
maintaining high writing quality. | cs.CL | Distinctive Image Captioning: Leveraging Ground Truth Captions in CLIP Guided Reinforcement Learning |
2024-02-21T17:00:56Z | http://arxiv.org/pdf/2402.13934v1 | http://arxiv.org/abs/2402.13934v1 | Kai Yang, Jan Ackermann, Zhenyu He, Guhao Feng, Bohang Zhang, Yunzhen Feng, Qiwei Ye, Di He, Liwei Wang | 2024-02-21T17:00:56Z | cs.LG, cs.AI, cs.CL, stat.ML | As transformer-based language models are trained on increasingly large
datasets and with vast numbers of parameters, finding more efficient
alternatives to the standard Transformer has become very valuable. While many
efficient Transformers and Transformer alternatives have been proposed, none
provide theoretical guarantees that they are a suitable replacement for the
standard Transformer. This makes it challenging to identify when to use a
specific model and what directions to prioritize for further investigation. In
this paper, we aim to understand the capabilities and limitations of efficient
Transformers, specifically the Sparse Transformer and the Linear Transformer.
We focus on their reasoning capability as exhibited by Chain-of-Thought (CoT)
prompts and follow previous works to model them as Dynamic Programming (DP)
problems. Our results show that while these models are expressive enough to
solve general DP tasks, contrary to expectations, they require a model size
that scales with the problem size. Nonetheless, we identify a class of DP
problems for which these models can be more efficient than the standard
Transformer. We confirm our theoretical results through experiments on
representative DP tasks, adding to the understanding of efficient Transformers'
practical strengths and weaknesses. | cs.LG | Do Efficient Transformers Really Save Computation? |
2024-02-21T16:46:36Z | http://arxiv.org/pdf/2402.13926v1 | http://arxiv.org/abs/2402.13926v1 | Federico Bianchi, James Zou | 2024-02-21T16:46:36Z | cs.CL, cs.AI | The risks derived from large language models (LLMs) generating deceptive and
damaging content have been the subject of considerable research, but even safe
generations can lead to problematic downstream impacts. In our study, we shift
the focus to how even safe text coming from LLMs can be easily turned into
potentially dangerous content through Bait-and-Switch attacks. In such attacks,
the user first prompts LLMs with safe questions and then employs a simple
find-and-replace post-hoc technique to manipulate the outputs into harmful
narratives. The alarming efficacy of this approach in generating toxic content
highlights a significant challenge in developing reliable safety guardrails for
LLMs. In particular, we stress that focusing on the safety of the verbatim LLM
outputs is insufficient and that we also need to consider post-hoc
transformations. | cs.CL | Large Language Models are Vulnerable to Bait-and-Switch Attacks for Generating Harmful Content |
2024-02-21T16:44:35Z | http://arxiv.org/pdf/2403.14643v2 | http://arxiv.org/abs/2403.14643v2 | Md. Asraful Haque, Shuai Li | 2024-03-25T05:35:12Z | cs.CY, cs.AI, cs.CL, 68Txx | Artificial intelligence has been around for a while, but suddenly it has
received more attention than ever before. Thanks to innovations from companies
like Google, Microsoft, Meta, and other major brands in technology. OpenAI,
though, has triggered the button with its ground-breaking invention ChatGPT.
ChatGPT is a Large Language Model (LLM) based on Transformer architecture that
has the ability to generate human-like responses in a conversational context.
It uses deep learning algorithms to generate natural language responses to
input text. Its large number of parameters, contextual generation, and
open-domain training make it a versatile and effective tool for a wide range of
applications, from chatbots to customer service to language translation. It has
the potential to revolutionize various industries and transform the way we
interact with technology. However, the use of ChatGPT has also raised several
concerns, including ethical, social, and employment challenges, which must be
carefully considered to ensure the responsible use of this technology. The
article provides an overview of ChatGPT, delving into its architecture and
training process. It highlights the potential impacts of ChatGPT on the
society. In this paper, we suggest some approaches involving technology,
regulation, education, and ethics in an effort to maximize ChatGPT's benefits
while minimizing its negative impacts. This study is expected to contribute to
a greater understanding of ChatGPT and aid in predicting the potential changes
it may bring about. | cs.CY | Exploring ChatGPT and its Impact on Society |
2024-02-21T16:33:22Z | http://arxiv.org/pdf/2402.13919v3 | http://arxiv.org/abs/2402.13919v3 | Prakamya Mishra, Zonghai Yao, Parth Vashisht, Feiyun Ouyang, Beining Wang, Vidhi Dhaval Mody, Hong Yu | 2024-04-18T16:50:43Z | cs.CL, cs.AI | Large Language Models (LLMs) such as GPT & Llama have demonstrated
significant achievements in summarization tasks but struggle with factual
inaccuracies, a critical issue in clinical NLP applications where errors could
lead to serious consequences. To counter the high costs and limited
availability of expert-annotated data for factual alignment, this study
introduces an innovative pipeline that utilizes >100B parameter GPT variants
like GPT-3.5 & GPT-4 to act as synthetic experts to generate high-quality
synthetics feedback aimed at enhancing factual consistency in clinical note
summarization. Our research primarily focuses on edit feedback generated by
these synthetic feedback experts without additional human annotations,
mirroring and optimizing the practical scenario in which medical professionals
refine AI system outputs. Although such 100B+ parameter GPT variants have
proven to demonstrate expertise in various clinical NLP tasks, such as the
Medical Licensing Examination, there is scant research on their capacity to act
as synthetic feedback experts and deliver expert-level edit feedback for
improving the generation quality of weaker (<10B parameter) LLMs like GPT-2
(1.5B) & Llama 2 (7B) in clinical domain. So in this work, we leverage 100B+
GPT variants to act as synthetic feedback experts offering expert-level edit
feedback, that is used to reduce hallucinations and align weaker (<10B
parameter) LLMs with medical facts using two distinct alignment algorithms (DPO
& SALT), endeavoring to narrow the divide between AI-generated content and
factual accuracy. This highlights the substantial potential of LLM-based
synthetic edits in enhancing the alignment of clinical factuality. | cs.CL | SYNFAC-EDIT: Synthetic Imitation Edit Feedback for Factual Alignment in Clinical Summarization |
2024-02-21T16:32:38Z | http://arxiv.org/pdf/2402.13917v2 | http://arxiv.org/abs/2402.13917v2 | Ryandito Diandaru, Lucky Susanto, Zilu Tang, Ayu Purwarianti, Derry Wijaya | 2024-04-05T05:26:05Z | cs.CL, cs.AI | Large Language Models (LLMs) demonstrate strong machine translation
capabilities on languages they are trained on. However, the impact of factors
beyond training data size on translation performance remains a topic of debate,
especially concerning languages not directly encountered during training. Our
study delves into Llama2's translation capabilities. By modeling a linear
relationship between linguistic feature distances and machine translation
scores, we ask ourselves if there are potentially better central languages for
LLMs other than English. Our experiments show that the 7B Llama2 model yields
above 10 BLEU when translating into all languages it has seen, which rarely
happens for languages it has not seen. Most translation improvements into
unseen languages come from scaling up the model size rather than instruction
tuning or increasing shot count. Furthermore, our correlation analysis reveals
that syntactic similarity is not the only linguistic factor that strongly
correlates with machine translation scores. Interestingly, we discovered that
under specific circumstances, some languages (e.g. Swedish, Catalan), despite
having significantly less training data, exhibit comparable correlation levels
to English. These insights challenge the prevailing landscape of LLMs,
suggesting that models centered around languages other than English could
provide a more efficient foundation for multilingual applications. | cs.CL | Could We Have Had Better Multilingual LLMs If English Was Not the Central Language? |
2024-02-21T16:22:21Z | http://arxiv.org/pdf/2402.13906v2 | http://arxiv.org/abs/2402.13906v2 | Gili Lior, Yoav Goldberg, Gabriel Stanovsky | 2024-06-20T14:43:51Z | cs.CL | Document collections of various domains, e.g., legal, medical, or financial,
often share some underlying collection-wide structure, which captures
information that can aid both human users and structure-aware models. We
propose to identify the typical structure of document within a collection,
which requires to capture recurring topics across the collection, while
abstracting over arbitrary header paraphrases, and ground each topic to
respective document locations. These requirements pose several challenges:
headers that mark recurring topics frequently differ in phrasing, certain
section headers are unique to individual documents and do not reflect the
typical structure, and the order of topics can vary between documents.
Subsequently, we develop an unsupervised graph-based method which leverages
both inter- and intra-document similarities, to extract the underlying
collection-wide structure. Our evaluations on three diverse domains in both
English and Hebrew indicate that our method extracts meaningful collection-wide
structure, and we hope that future work will leverage our method for
multi-document applications and structure-aware models. | cs.CL | Leveraging Collection-Wide Similarities for Unsupervised Document Structure Extraction |
2024-02-21T16:15:20Z | http://arxiv.org/pdf/2402.13904v1 | http://arxiv.org/abs/2402.13904v1 | Qing Lyu, Kumar Shridhar, Chaitanya Malaviya, Li Zhang, Yanai Elazar, Niket Tandon, Marianna Apidianaki, Mrinmaya Sachan, Chris Callison-Burch | 2024-02-21T16:15:20Z | cs.CL | Accurately gauging the confidence level of Large Language Models' (LLMs)
predictions is pivotal for their reliable application. However, LLMs are often
uncalibrated inherently and elude conventional calibration techniques due to
their proprietary nature and massive scale. In this work, we explore the
potential of deriving confidence from the distribution of multiple randomly
sampled model generations, via three measures of consistency. We perform an
extensive evaluation across various open and closed-source models on nine
reasoning datasets. Results show that consistency-based calibration methods
outperform existing post-hoc approaches. Meanwhile, we find that factors such
as intermediate explanations, model scaling, and larger sample sizes enhance
calibration, while instruction-tuning makes calibration more difficult.
Moreover, confidence scores obtained from consistency have the potential to
enhance model performance. Finally, we offer practical guidance on choosing
suitable consistency metrics for calibration, tailored to the characteristics
of various LMs. | cs.CL | Calibrating Large Language Models with Sample Consistency |
2024-02-21T16:09:25Z | http://arxiv.org/pdf/2402.13897v2 | http://arxiv.org/abs/2402.13897v2 | Loïc Rakotoson, Sylvain Massip, Fréjus A. A. Laleye | 2024-03-14T00:21:09Z | cs.IR, cs.AI, cs.CL, cs.LG, H.3.1; H.3.3; I.7; K.4 | Information retrieval is a rapidly evolving field. However it still faces
significant limitations in the scientific and industrial vast amounts of
information, such as semantic divergence and vocabulary gaps in sparse
retrieval, low precision and lack of interpretability in semantic search, or
hallucination and outdated information in generative models. In this paper, we
introduce a two-block approach to tackle these hurdles for long documents. The
first block enhances language understanding in sparse retrieval by query
expansion to retrieve relevant documents. The second block deepens the result
by providing comprehensive and informative answers to the complex question
using only the information spread in the long document, enabling bidirectional
engagement. At various stages of the pipeline, intermediate results are
presented to users to facilitate understanding of the system's reasoning. We
believe this bidirectional approach brings significant advancements in terms of
transparency, logical thinking, and comprehensive understanding in the field of
scientific information retrieval. | cs.IR | Science Checker Reloaded: A Bidirectional Paradigm for Transparency and Logical Reasoning |
2024-02-21T15:58:37Z | http://arxiv.org/pdf/2402.13887v1 | http://arxiv.org/abs/2402.13887v1 | Chenyang Lyu, Minghao Wu, Alham Fikri Aji | 2024-02-21T15:58:37Z | cs.CL | Large Language Models (LLMs) have demonstrated remarkable capabilities across
various applications, fundamentally reshaping the landscape of natural language
processing (NLP) research. However, recent evaluation frameworks often rely on
the output probabilities of LLMs for predictions, primarily due to
computational constraints, diverging from real-world LLM usage scenarios. While
widely employed, the efficacy of these probability-based evaluation strategies
remains an open research question. This study aims to scrutinize the validity
of such probability-based evaluation methods within the context of using LLMs
for Multiple Choice Questions (MCQs), highlighting their inherent limitations.
Our empirical investigation reveals that the prevalent probability-based
evaluation method inadequately aligns with generation-based prediction.
Furthermore, current evaluation frameworks typically assess LLMs through
predictive tasks based on output probabilities rather than directly generating
responses, owing to computational limitations. We illustrate that these
probability-based approaches do not effectively correspond with generative
predictions. The outcomes of our study can enhance the understanding of LLM
evaluation methodologies and provide insights for future research in this
domain. | cs.CL | Beyond Probabilities: Unveiling the Misalignment in Evaluating Large Language Models |
2024-02-21T15:35:04Z | http://arxiv.org/pdf/2402.13874v3 | http://arxiv.org/abs/2402.13874v3 | Haoyu Liu, Jianfeng Liu, Shaohan Huang, Yuefeng Zhan, Hao Sun, Weiwei Deng, Furu Wei, Qi Zhang | 2024-06-06T11:12:57Z | cs.CL | The remarkable capability of large language models (LLMs) for in-context
learning (ICL) needs to be activated by demonstration examples. Prior work has
extensively explored the selection of examples for ICL, predominantly following
the "select then organize" paradigm, such approaches often neglect the internal
relationships between examples and exist an inconsistency between the training
and inference. In this paper, we formulate the problem as a $Se$quential
$Se$lection problem and introduce $Se^2$, a sequential-aware method that
leverages the LLM's feedback on varying context, aiding in capturing
inter-relationships and sequential information among examples, significantly
enriching the contextuality and relevance of ICL prompts. Meanwhile, we utilize
beam search to seek and construct example sequences, enhancing both quality and
diversity. Extensive experiments across 23 NLP tasks from 8 distinct categories
illustrate that $Se^2$ markedly surpasses competitive baselines and achieves
42\% relative improvement over random selection. Further in-depth analysis
shows the effectiveness of proposed strategies, highlighting $Se^2$'s
exceptional stability and adaptability across various scenarios. Code available
at https://github.com/microsoft/LMOps. | cs.CL | $Se^2$: Sequential Example Selection for In-Context Learning |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.