publication_date
stringlengths 20
20
| doc_url
stringlengths 33
34
| id
stringlengths 33
34
| authors
stringlengths 5
20.1k
| update_date
stringlengths 20
20
| category_all
stringlengths 5
138
| abstract
stringlengths 123
3.09k
| category_primary
stringclasses 78
values | title
stringlengths 10
211
|
|---|---|---|---|---|---|---|---|---|
2024-04-30T03:31:03Z | http://arxiv.org/pdf/2404.19234v1 | http://arxiv.org/abs/2404.19234v1 | Abir Chakraborty | 2024-04-30T03:31:03Z | cs.AI, cs.CL, cs.DB | Knowledge graphs (KGs) are large datasets with specific structures
representing large knowledge bases (KB) where each node represents a key entity
and relations amongst them are typed edges. Natural language queries formed to
extract information from a KB entail starting from specific nodes and reasoning
over multiple edges of the corresponding KG to arrive at the correct set of
answer nodes. Traditional approaches of question answering on KG are based on
(a) semantic parsing (SP), where a logical form (e.g., S-expression, SPARQL
query, etc.) is generated using node and edge embeddings and then reasoning
over these representations or tuning language models to generate the final
answer directly, or (b) information-retrieval based that works by extracting
entities and relations sequentially. In this work, we evaluate the capability
of (LLMs) to answer questions over KG that involve multiple hops. We show that
depending upon the size and nature of the KG we need different approaches to
extract and feed the relevant information to an LLM since every LLM comes with
a fixed context window. We evaluate our approach on six KGs with and without
the availability of example-specific sub-graphs and show that both the IR and
SP-based methods can be adopted by LLMs resulting in an extremely competitive
performance. | cs.AI | Multi-hop Question Answering over Knowledge Graphs using Large Language Models |
2024-04-30T03:29:30Z | http://arxiv.org/pdf/2404.19232v4 | http://arxiv.org/abs/2404.19232v4 | Xinzhe Li, Ming Liu, Shang Gao | 2024-05-29T11:12:21Z | cs.CL, cs.AI | Retrieval-augmented Generation (RAG) systems have been actively studied and
deployed across various industries to query on domain-specific knowledge base.
However, evaluating these systems presents unique challenges due to the
scarcity of domain-specific queries and corresponding ground truths, as well as
a lack of systematic approaches to diagnosing the cause of failure cases --
whether they stem from knowledge deficits or issues related to system
robustness. To address these challenges, we introduce GRAMMAR (GRounded And
Modular Methodology for Assessment of RAG), an evaluation framework comprising
two key elements: 1) a data generation process that leverages relational
databases and LLMs to efficiently produce scalable query-answer pairs. This
method facilitates the separation of query logic from linguistic variations for
enhanced debugging capabilities; and 2) an evaluation framework that
differentiates knowledge gaps from robustness and enables the identification of
defective modules. Our empirical results underscore the limitations of current
reference-free evaluation approaches and the reliability of GRAMMAR to
accurately identify model vulnerabilities. | cs.CL | GRAMMAR: Grounded and Modular Methodology for Assessment of Domain-Specific Retrieval-Augmented Language Model |
2024-04-30T02:48:20Z | http://arxiv.org/pdf/2404.19221v1 | http://arxiv.org/abs/2404.19221v1 | Jiading Fang, Xiangshan Tan, Shengjie Lin, Igor Vasiljevic, Vitor Guizilini, Hongyuan Mei, Rares Ambrus, Gregory Shakhnarovich, Matthew R Walter | 2024-04-30T02:48:20Z | cs.CV, cs.CL | If robots are to work effectively alongside people, they must be able to
interpret natural language references to objects in their 3D environment.
Understanding 3D referring expressions is challenging -- it requires the
ability to both parse the 3D structure of the scene and correctly ground
free-form language in the presence of distraction and clutter. We introduce
Transcrib3D, an approach that brings together 3D detection methods and the
emergent reasoning capabilities of large language models (LLMs). Transcrib3D
uses text as the unifying medium, which allows us to sidestep the need to learn
shared representations connecting multi-modal inputs, which would require
massive amounts of annotated 3D data. As a demonstration of its effectiveness,
Transcrib3D achieves state-of-the-art results on 3D reference resolution
benchmarks, with a great leap in performance from previous multi-modality
baselines. To improve upon zero-shot performance and facilitate local
deployment on edge computers and robots, we propose self-correction for
fine-tuning that trains smaller models, resulting in performance close to that
of large models. We show that our method enables a real robot to perform
pick-and-place tasks given queries that contain challenging referring
expressions. Project site is at https://ripl.github.io/Transcrib3D. | cs.CV | Transcrib3D: 3D Referring Expression Resolution through Large Language Models |
2024-04-30T01:41:03Z | http://arxiv.org/pdf/2404.19192v1 | http://arxiv.org/abs/2404.19192v1 | Xinwei Chen, Kun Li, Tianyou Song, Jiangjian Guo | 2024-04-30T01:41:03Z | cs.CL, cs.AI | Named Entity Recognition (NER) is an essential steppingstone in the field of
natural language processing. Although promising performance has been achieved
by various distantly supervised models, we argue that distant supervision
inevitably introduces incomplete and noisy annotations, which may mislead the
model training process. To address this issue, we propose a robust NER model
named BOND-MoE based on Mixture of Experts (MoE). Instead of relying on a
single model for NER prediction, multiple models are trained and ensembled
under the Expectation-Maximization (EM) framework, so that noisy supervision
can be dramatically alleviated. In addition, we introduce a fair assignment
module to balance the document-model assignment process. Extensive experiments
on real-world datasets show that the proposed method achieves state-of-the-art
performance compared with other distantly supervised NER. | cs.CL | Mix of Experts Language Model for Named Entity Recognition |
2024-04-30T01:19:18Z | http://arxiv.org/pdf/2405.00740v3 | http://arxiv.org/abs/2405.00740v3 | Samuel Lavoie, Polina Kirichenko, Mark Ibrahim, Mahmoud Assran, Andrew Gordon Wilson, Aaron Courville, Nicolas Ballas | 2024-05-14T12:48:45Z | cs.CV, cs.AI, cs.CL, cs.LG | There are a thousand ways to caption an image. Contrastive Language
Pretraining (CLIP) on the other hand, works by mapping an image and its caption
to a single vector -- limiting how well CLIP-like models can represent the
diverse ways to describe an image. In this work, we introduce Llip, Latent
Language Image Pretraining, which models the diversity of captions that could
match an image. Llip's vision encoder outputs a set of visual features that are
mixed into a final representation by conditioning on information derived from
the text. We show that Llip outperforms non-contextualized baselines like CLIP
and SigLIP on a variety of tasks even with large-scale encoders. Llip improves
zero-shot classification by an average of 2.9% zero-shot classification
benchmarks with a ViT-G/14 encoder. Specifically, Llip attains a zero-shot
top-1 accuracy of 83.5% on ImageNet outperforming a similarly sized CLIP by
1.4%. We also demonstrate improvement on zero-shot retrieval on MS-COCO by
6.0%. We provide a comprehensive analysis of the components introduced by the
method and demonstrate that Llip leads to richer visual representations. | cs.CV | Modeling Caption Diversity in Contrastive Vision-Language Pretraining |
2024-04-30T01:02:15Z | http://arxiv.org/pdf/2404.19178v1 | http://arxiv.org/abs/2404.19178v1 | James A. Michaelov, Catherine Arnett, Benjamin K. Bergen | 2024-04-30T01:02:15Z | cs.CL | Transformers have supplanted Recurrent Neural Networks as the dominant
architecture for both natural language processing tasks and, despite criticisms
of cognitive implausibility, for modelling the effect of predictability on
online human language comprehension. However, two recently developed recurrent
neural network architectures, RWKV and Mamba, appear to perform natural
language tasks comparably to or better than transformers of equivalent scale.
In this paper, we show that contemporary recurrent models are now also able to
match - and in some cases, exceed - performance of comparably sized
transformers at modeling online human language comprehension. This suggests
that transformer language models are not uniquely suited to this task, and
opens up new directions for debates about the extent to which architectural
features of language models make them better or worse models of human language
comprehension. | cs.CL | Revenge of the Fallen? Recurrent Models Match Transformers at Predicting Human Language Comprehension Metrics |
2024-04-30T00:39:26Z | http://arxiv.org/pdf/2404.19175v1 | http://arxiv.org/abs/2404.19175v1 | Zhihao Zhang, Feiqi Cao, Yingbin Mo, Yiran Zhang, Josiah Poon, Caren Han | 2024-04-30T00:39:26Z | cs.CL | The dynamic nature of esports makes the situation relatively complicated for
average viewers. Esports broadcasting involves game expert casters, but the
caster-dependent game commentary is not enough to fully understand the game
situation. It will be richer by including diverse multimodal esports
information, including audiences' talks/emotions, game audio, and game match
event information. This paper introduces GAME-MUG, a new multimodal game
situation understanding and audience-engaged commentary generation dataset and
its strong baseline. Our dataset is collected from 2020-2022 LOL game live
streams from YouTube and Twitch, and includes multimodal esports game
information, including text, audio, and time-series event logs, for detecting
the game situation. In addition, we also propose a new audience conversation
augmented commentary dataset by covering the game situation and audience
conversation understanding, and introducing a robust joint multimodal dual
learning model as a baseline. We examine the model's game situation/event
understanding ability and commentary generation capability to show the
effectiveness of the multimodal aspects coverage and the joint integration
learning approach. | cs.CL | Game-MUG: Multimodal Oriented Game Situation Understanding and Commentary Generation Dataset |
2024-04-29T23:49:19Z | http://arxiv.org/pdf/2404.19159v1 | http://arxiv.org/abs/2404.19159v1 | Sina Bagheri Nezhad, Ameeta Agrawal | 2024-04-29T23:49:19Z | cs.CL, I.2.7 | This study investigates the factors influencing the performance of
multilingual large language models (MLLMs) across diverse languages. We study 6
MLLMs, including masked language models, autoregressive models, and
instruction-tuned LLMs, on the SIB-200 dataset, a topic classification dataset
encompassing 204 languages. Our analysis considers three scenarios: ALL
languages, SEEN languages (present in the model's pretraining data), and UNSEEN
languages (not present or documented in the model's pretraining data in any
meaningful way). We examine the impact of factors such as pretraining data
size, general resource availability, language family, and script type on model
performance. Decision tree analysis reveals that pretraining data size is the
most influential factor for SEEN languages. However, interestingly, script type
and language family are crucial for UNSEEN languages, highlighting the
importance of cross-lingual transfer learning. Notably, model size and
architecture do not significantly alter the most important features identified.
Our findings provide valuable insights into the strengths and limitations of
current MLLMs and hope to guide the development of more effective and equitable
multilingual NLP systems. | cs.CL | What Drives Performance in Multilingual Language Models? |
2024-04-29T23:36:38Z | http://arxiv.org/pdf/2404.19154v2 | http://arxiv.org/abs/2404.19154v2 | Ning An, Lei Hei, Yong Jiang, Weiping Meng, Jingjing Hu, Boran Huang, Feiliang Ren | 2024-06-13T16:26:15Z | cs.CL | Relational triple extraction is crucial work for the automatic construction
of knowledge graphs. Existing methods only construct shallow representations
from a token or token pair-level. However, previous works ignore local spatial
dependencies of relational triples, resulting in a weakness of entity pair
boundary detection. To tackle this problem, we propose a novel Region-based
Table Filling method (RTF). We devise a novel region-based tagging scheme and
bi-directional decoding strategy, which regard each relational triple as a
region on the relation-specific table, and identifies triples by determining
two endpoints of each region. We also introduce convolution to construct
region-level table representations from a spatial perspective which makes
triples easier to be captured. In addition, we share partial tagging scores
among different relations to improve learning efficiency of relation
classifier. Experimental results show that our method achieves state-of-the-art
with better generalization capability on three variants of two widely used
benchmark datasets. | cs.CL | RTF: Region-based Table Filling Method for Relational Triple Extraction |
2024-04-29T22:06:17Z | http://arxiv.org/pdf/2404.19128v1 | http://arxiv.org/abs/2404.19128v1 | Navid Rajabi, Jana Kosecka | 2024-04-29T22:06:17Z | cs.CV, cs.CL, cs.LG | Vision and Language Models (VLMs) continue to demonstrate remarkable
zero-shot (ZS) performance across various tasks. However, many probing studies
have revealed that even the best-performing VLMs struggle to capture aspects of
compositional scene understanding, lacking the ability to properly ground and
localize linguistic phrases in images. Recent VLM advancements include scaling
up both model and dataset sizes, additional training objectives and levels of
supervision, and variations in the model architectures. To characterize the
grounding ability of VLMs, such as phrase grounding, referring expressions
comprehension, and relationship understanding, Pointing Game has been used as
an evaluation metric for datasets with bounding box annotations. In this paper,
we introduce a novel suite of quantitative metrics that utilize GradCAM
activations to rigorously evaluate the grounding capabilities of pre-trained
VLMs like CLIP, BLIP, and ALBEF. These metrics offer an explainable and
quantifiable approach for a more detailed comparison of the zero-shot
capabilities of VLMs and enable measuring models' grounding uncertainty. This
characterization reveals interesting tradeoffs between the size of the model,
the dataset size, and their performance. | cs.CV | Q-GroundCAM: Quantifying Grounding in Vision Language Models via GradCAM |
2024-04-29T21:59:07Z | http://arxiv.org/pdf/2404.19124v2 | http://arxiv.org/abs/2404.19124v2 | Davis Wertheimer, Joshua Rosenkranz, Thomas Parnell, Sahil Suneja, Pavithra Ranganathan, Raghu Ganti, Mudhakar Srivatsa | 2024-06-06T18:38:34Z | cs.CL | This technical report describes the design and training of novel speculative
decoding draft models, for accelerating the inference speeds of large language
models in a production environment. By conditioning draft predictions on both
context vectors and sampled tokens, we can train our speculators to efficiently
predict high-quality n-grams, which the base model then accepts or rejects.
This allows us to effectively predict multiple tokens per inference forward
pass, accelerating wall-clock inference speeds of highly optimized base model
implementations by a factor of 2-3x. We explore these initial results and
describe next steps for further improvements. | cs.CL | Accelerating Production LLMs with Combined Token/Embedding Speculators |
2024-04-29T21:36:16Z | http://arxiv.org/pdf/2404.19119v1 | http://arxiv.org/abs/2404.19119v1 | Arif Ahmed, Gondy Leroy, Stephen A. Rains, Philip Harber, David Kauchak, Prosanta Barai | 2024-04-29T21:36:16Z | cs.CL | Health literacy is crucial to supporting good health and is a major national
goal. Audio delivery of information is becoming more popular for informing
oneself. In this study, we evaluate the effect of audio enhancements in the
form of information emphasis and pauses with health texts of varying difficulty
and we measure health information comprehension and retention. We produced
audio snippets from difficult and easy text and conducted the study on Amazon
Mechanical Turk (AMT). Our findings suggest that emphasis matters for both
information comprehension and retention. When there is no added pause,
emphasizing significant information can lower the perceived difficulty for
difficult and easy texts. Comprehension is higher (54%) with correctly placed
emphasis for the difficult texts compared to not adding emphasis (50%). Adding
a pause lowers perceived difficulty and can improve retention but adversely
affects information comprehension. | cs.CL | Effects of Added Emphasis and Pause in Audio Delivery of Health Information |
2024-04-29T21:00:33Z | http://arxiv.org/pdf/2405.01592v1 | http://arxiv.org/abs/2405.01592v1 | Gondy Leroy, David Kauchak, Philip Harber, Ankit Pal, Akash Shukla | 2024-04-29T21:00:33Z | cs.CL, cs.AI, H.4 | Text and audio simplification to increase information comprehension are
important in healthcare. With the introduction of ChatGPT, an evaluation of its
simplification performance is needed. We provide a systematic comparison of
human and ChatGPT simplified texts using fourteen metrics indicative of text
difficulty. We briefly introduce our online editor where these simplification
tools, including ChatGPT, are available. We scored twelve corpora using our
metrics: six text, one audio, and five ChatGPT simplified corpora. We then
compare these corpora with texts simplified and verified in a prior user study.
Finally, a medical domain expert evaluated these texts and five, new ChatGPT
simplified versions. We found that simple corpora show higher similarity with
the human simplified texts. ChatGPT simplification moves metrics in the right
direction. The medical domain expert evaluation showed a preference for the
ChatGPT style, but the text itself was rated lower for content retention. | cs.CL | Text and Audio Simplification: Human vs. ChatGPT |
2024-04-29T20:19:25Z | http://arxiv.org/pdf/2404.19094v1 | http://arxiv.org/abs/2404.19094v1 | Matteo Merler, Nicola Dainese, Katsiaryna Haitsiukevich | 2024-04-29T20:19:25Z | cs.CL, cs.LG | Symbolic Regression (SR) is a task which aims to extract the mathematical
expression underlying a set of empirical observations. Transformer-based
methods trained on SR datasets detain the current state-of-the-art in this
task, while the application of Large Language Models (LLMs) to SR remains
unexplored. This work investigates the integration of pre-trained LLMs into the
SR pipeline, utilizing an approach that iteratively refines a functional form
based on the prediction error it achieves on the observation set, until it
reaches convergence. Our method leverages LLMs to propose an initial set of
possible functions based on the observations, exploiting their strong
pre-training prior. These functions are then iteratively refined by the model
itself and by an external optimizer for their coefficients. The process is
repeated until the results are satisfactory. We then analyze Vision-Language
Models in this context, exploring the inclusion of plots as visual inputs to
aid the optimization process. Our findings reveal that LLMs are able to
successfully recover good symbolic equations that fit the given data,
outperforming SR baselines based on Genetic Programming, with the addition of
images in the input showing promising results for the most complex benchmarks. | cs.CL | In-Context Symbolic Regression: Leveraging Language Models for Function Discovery |
2024-04-29T19:28:35Z | http://arxiv.org/pdf/2404.19071v1 | http://arxiv.org/abs/2404.19071v1 | Sanjana Gautam, Mukund Srinath | 2024-04-29T19:28:35Z | cs.HC, cs.CL | With the rapid proliferation of artificial intelligence, there is growing
concern over its potential to exacerbate existing biases and societal
disparities and introduce novel ones. This issue has prompted widespread
attention from academia, policymakers, industry, and civil society. While
evidence suggests that integrating human perspectives can mitigate bias-related
issues in AI systems, it also introduces challenges associated with cognitive
biases inherent in human decision-making. Our research focuses on reviewing
existing methodologies and ongoing investigations aimed at understanding
annotation attributes that contribute to bias. | cs.HC | Blind Spots and Biases: Exploring the Role of Annotator Cognitive Biases in NLP |
2024-04-29T19:12:42Z | http://arxiv.org/pdf/2404.19065v1 | http://arxiv.org/abs/2404.19065v1 | Gabriel Sarch, Sahil Somani, Raghav Kapoor, Michael J. Tarr, Katerina Fragkiadaki | 2024-04-29T19:12:42Z | cs.AI, cs.CL, cs.CV, cs.LG | Recent research on instructable agents has used memory-augmented Large
Language Models (LLMs) as task planners, a technique that retrieves
language-program examples relevant to the input instruction and uses them as
in-context examples in the LLM prompt to improve the performance of the LLM in
inferring the correct action and task plans. In this technical report, we
extend the capabilities of HELPER, by expanding its memory with a wider array
of examples and prompts, and by integrating additional APIs for asking
questions. This simple expansion of HELPER into a shared memory enables the
agent to work across the domains of executing plans from dialogue, natural
language instruction following, active question asking, and commonsense room
reorganization. We evaluate the agent on four diverse interactive
visual-language embodied agent benchmarks: ALFRED, TEACh, DialFRED, and the
Tidy Task. HELPER-X achieves few-shot, state-of-the-art performance across
these benchmarks using a single agent, without requiring in-domain training,
and remains competitive with agents that have undergone in-domain training. | cs.AI | HELPER-X: A Unified Instructable Embodied Agent to Tackle Four Interactive Vision-Language Domains with Memory-Augmented Language Models |
2024-04-29T19:04:35Z | http://arxiv.org/pdf/2404.19063v1 | http://arxiv.org/abs/2404.19063v1 | Liang Xu, Lei Zhu, Yaotong Wu, Hang Xue | 2024-04-29T19:04:35Z | cs.CL | The SuperCLUE-Fin (SC-Fin) benchmark is a pioneering evaluation framework
tailored for Chinese-native financial large language models (FLMs). It assesses
FLMs across six financial application domains and twenty-five specialized
tasks, encompassing theoretical knowledge and practical applications such as
compliance, risk management, and investment analysis. Using multi-turn,
open-ended conversations that mimic real-life scenarios, SC-Fin measures models
on a range of criteria, including accurate financial understanding, logical
reasoning, clarity, computational efficiency, business acumen, risk perception,
and compliance with Chinese regulations.
In a rigorous evaluation involving over a thousand questions, SC-Fin
identifies a performance hierarchy where domestic models like GLM-4 and
MoonShot-v1-128k outperform others with an A-grade, highlighting the potential
for further development in transforming theoretical knowledge into pragmatic
financial solutions. This benchmark serves as a critical tool for refining FLMs
in the Chinese context, directing improvements in financial knowledge
databases, standardizing financial interpretations, and promoting models that
prioritize compliance, risk management, and secure practices.
We create a contextually relevant and comprehensive benchmark that drives the
development of AI in the Chinese financial sector. SC-Fin facilitates the
advancement and responsible deployment of FLMs, offering valuable insights for
enhancing model performance and usability for both individual and institutional
users in the Chinese market..~\footnote{Our benchmark can be found at
\url{https://www.CLUEbenchmarks.com}}. | cs.CL | SuperCLUE-Fin: Graded Fine-Grained Analysis of Chinese LLMs on Diverse Financial Tasks and Applications |
2024-04-29T18:51:17Z | http://arxiv.org/pdf/2404.19055v1 | http://arxiv.org/abs/2404.19055v1 | Houjun Liu | 2024-04-29T18:51:17Z | cs.CL | While language models (LMs) offer significant capability in zero-shot
reasoning tasks across a wide range of domains, they do not perform
satisfactorily in problems which requires multi-step reasoning. Previous
approaches to mitigate this involves breaking a larger, multi-step task into
sub-tasks and asking the language model to generate proposals ("thoughts") for
each sub-task and using exhaustive planning approaches such as DFS to compose a
solution. In this work, we leverage this idea to introduce two new
contributions: first, we formalize a planning-based approach to perform
multi-step problem solving with LMs via Partially Observable Markov Decision
Processes (POMDPs), with the LM's own reflections about the value of a state
used as a search heuristic; second, leveraging the online POMDP solver POMCP,
we demonstrate a superior success rate of 89.4% on the Game of 24 task as
compared to existing approaches while also offering better anytime performance
characteristics than fixed tree-search which is used previously. Taken
together, these contributions allow modern LMs to decompose and solve
larger-scale reasoning tasks more effectively. | cs.CL | Plan of Thoughts: Heuristic-Guided Problem Solving with Large Language Models |
2024-04-29T18:40:01Z | http://arxiv.org/pdf/2404.19048v2 | http://arxiv.org/abs/2404.19048v2 | Ximing Dong, Dayi Lin, Shaowei Wang, Ahmed E. Hassan | 2024-05-01T19:53:12Z | cs.CL, cs.AI | Large Language Models (LLMs) have significantly advanced natural language
processing (NLP) tasks but also pose ethical and societal risks due to their
propensity to generate harmful content. To address this, various approaches
have been developed to safeguard LLMs from producing unsafe content. However,
existing methods have limitations, including the need for training specific
control models and proactive intervention during text generation, that lead to
quality degradation and increased computational overhead. To mitigate those
limitations, we propose LLMSafeGuard, a lightweight framework to safeguard LLM
text generation in real-time. LLMSafeGuard integrates an external validator
into the beam search algorithm during decoding, rejecting candidates that
violate safety constraints while allowing valid ones to proceed. We introduce a
similarity based validation approach, simplifying constraint introduction and
eliminating the need for control model training. Additionally, LLMSafeGuard
employs a context-wise timing selection strategy, intervening LLMs only when
necessary. We evaluate LLMSafeGuard on two tasks, detoxification and copyright
safeguarding, and demonstrate its superior performance over SOTA baselines. For
instance, LLMSafeGuard reduces the average toxic score of. LLM output by 29.7%
compared to the best baseline meanwhile preserving similar linguistic quality
as natural output in detoxification task. Similarly, in the copyright task,
LLMSafeGuard decreases the Longest Common Subsequence (LCS) by 56.2% compared
to baselines. Moreover, our context-wise timing selection strategy reduces
inference time by at least 24% meanwhile maintaining comparable effectiveness
as validating each time step. LLMSafeGuard also offers tunable parameters to
balance its effectiveness and efficiency. | cs.CL | A Framework for Real-time Safeguarding the Text Generation of Large Language Model |
2024-04-29T18:00:03Z | http://arxiv.org/pdf/2404.19007v1 | http://arxiv.org/abs/2404.19007v1 | Yilun Hua, Nicholas Chernogor, Yuzhe Gu, Seoyeon Julie Jeong, Miranda Luo, Cristian Danescu-Niculescu-Mizil | 2024-04-29T18:00:03Z | cs.CL, cs.AI, cs.CY | Throughout a conversation, the way participants interact with each other is
in constant flux: their tones may change, they may resort to different
strategies to convey their points, or they might alter their interaction
patterns. An understanding of these dynamics can complement that of the actual
facts and opinions discussed, offering a more holistic view of the trajectory
of the conversation: how it arrived at its current state and where it is likely
heading.
In this work, we introduce the task of summarizing the dynamics of
conversations, by constructing a dataset of human-written summaries, and
exploring several automated baselines. We evaluate whether such summaries can
capture the trajectory of conversations via an established downstream task:
forecasting whether an ongoing conversation will eventually derail into toxic
behavior. We show that they help both humans and automated systems with this
forecasting task. Humans make predictions three times faster, and with greater
confidence, when reading the summaries than when reading the transcripts.
Furthermore, automated forecasting systems are more accurate when constructing,
and then predicting based on, summaries of conversation dynamics, compared to
directly predicting on the transcripts. | cs.CL | How Did We Get Here? Summarizing Conversation Dynamics |
2024-04-29T17:59:16Z | http://arxiv.org/pdf/2404.18928v1 | http://arxiv.org/abs/2404.18928v1 | Michael Luo, Justin Wong, Brandon Trabucco, Yanping Huang, Joseph E. Gonzalez, Zhifeng Chen, Ruslan Salakhutdinov, Ion Stoica | 2024-04-29T17:59:16Z | cs.CV, cs.AI, cs.CL, cs.GR, cs.LG | Beyond scaling base models with more data or parameters, fine-tuned adapters
provide an alternative way to generate high fidelity, custom images at reduced
costs. As such, adapters have been widely adopted by open-source communities,
accumulating a database of over 100K adapters-most of which are highly
customized with insufficient descriptions. This paper explores the problem of
matching the prompt to a set of relevant adapters, built on recent work that
highlight the performance gains of composing adapters. We introduce Stylus,
which efficiently selects and automatically composes task-specific adapters
based on a prompt's keywords. Stylus outlines a three-stage approach that first
summarizes adapters with improved descriptions and embeddings, retrieves
relevant adapters, and then further assembles adapters based on prompts'
keywords by checking how well they fit the prompt. To evaluate Stylus, we
developed StylusDocs, a curated dataset featuring 75K adapters with
pre-computed adapter embeddings. In our evaluation on popular Stable Diffusion
checkpoints, Stylus achieves greater CLIP-FID Pareto efficiency and is twice as
preferred, with humans and multimodal models as evaluators, over the base
model. See stylus-diffusion.github.io for more. | cs.CV | Stylus: Automatic Adapter Selection for Diffusion Models |
2024-04-29T17:58:36Z | http://arxiv.org/pdf/2404.18923v2 | http://arxiv.org/abs/2404.18923v2 | Andreas Waldis, Yotam Perlitz, Leshem Choshen, Yufang Hou, Iryna Gurevych | 2024-05-22T08:36:49Z | cs.CL | We introduce Holmes, a benchmark to assess the linguistic competence of
language models (LMs) - their ability to grasp linguistic phenomena. Unlike
prior prompting-based evaluations, Holmes assesses the linguistic competence of
LMs via their internal representations using classifier-based probing. In doing
so, we disentangle specific phenomena (e.g., part-of-speech of words) from
other cognitive abilities, like following textual instructions, and meet recent
calls to assess LMs' linguistic competence in isolation. Composing Holmes, we
review over 250 probing studies and feature more than 200 datasets to assess
syntax, morphology, semantics, reasoning, and discourse phenomena. Analyzing
over 50 LMs reveals that, aligned with known trends, their linguistic
competence correlates with model size. However, surprisingly, model
architecture and instruction tuning also significantly influence performance,
particularly in morphology and syntax. Finally, we propose FlashHolmes, a
streamlined version of Holmes designed to lower the high computation load while
maintaining high-ranking precision. | cs.CL | Holmes: Benchmark the Linguistic Competence of Language Models |
2024-04-29T17:58:30Z | http://arxiv.org/pdf/2404.18922v1 | http://arxiv.org/abs/2404.18922v1 | Han Zhong, Guhao Feng, Wei Xiong, Li Zhao, Di He, Jiang Bian, Liwei Wang | 2024-04-29T17:58:30Z | cs.LG, cs.AI, cs.CL, stat.ML | In the classical Reinforcement Learning from Human Feedback (RLHF) framework,
Proximal Policy Optimization (PPO) is employed to learn from sparse,
sentence-level rewards -- a challenging scenario in traditional deep
reinforcement learning. Despite the great successes of PPO in the alignment of
state-of-the-art closed-source large language models (LLMs), its open-source
implementation is still largely sub-optimal, as widely reported by numerous
research studies. To address these issues, we introduce a framework that models
RLHF problems as a Markov decision process (MDP), enabling the capture of
fine-grained token-wise information. Furthermore, we provide theoretical
insights that demonstrate the superiority of our MDP framework over the
previous sentence-level bandit formulation. Under this framework, we introduce
an algorithm, dubbed as Reinforced Token Optimization (\texttt{RTO}), which
learns the token-wise reward function from preference data and performs policy
optimization based on this learned token-wise reward signal. Theoretically,
\texttt{RTO} is proven to have the capability of finding the near-optimal
policy sample-efficiently. For its practical implementation, \texttt{RTO}
innovatively integrates Direct Preference Optimization (DPO) and PPO. DPO,
originally derived from sparse sentence rewards, surprisingly provides us with
a token-wise characterization of response quality, which is seamlessly
incorporated into our subsequent PPO training stage. Extensive real-world
alignment experiments verify the effectiveness of the proposed approach. | cs.LG | DPO Meets PPO: Reinforced Token Optimization for RLHF |
2024-04-29T17:53:54Z | http://arxiv.org/pdf/2404.18911v1 | http://arxiv.org/abs/2404.18911v1 | Fangcheng Liu, Yehui Tang, Zhenhua Liu, Yunsheng Ni, Kai Han, Yunhe Wang | 2024-04-29T17:53:54Z | cs.CL, cs.LG | Speculative decoding has demonstrated its effectiveness in accelerating the
inference of large language models while maintaining a consistent sampling
distribution. However, the conventional approach of training a separate draft
model to achieve a satisfactory token acceptance rate can be costly. Drawing
inspiration from early exiting, we propose a novel self-speculative decoding
framework \emph{Kangaroo}, which uses a fixed shallow sub-network as a
self-draft model, with the remaining layers serving as the larger target model.
We train a lightweight and efficient adapter module on top of the sub-network
to bridge the gap between the sub-network and the full model's representation
ability. It is noteworthy that the inference latency of the self-draft model
may no longer be negligible compared to the large model, necessitating
strategies to increase the token acceptance rate while minimizing the drafting
steps of the small model. To address this challenge, we introduce an additional
early exiting mechanism for generating draft tokens. Specifically, we halt the
small model's subsequent prediction during the drafting phase once the
confidence level for the current token falls below a certain threshold.
Extensive experiments on the Spec-Bench demonstrate the effectiveness of
Kangaroo. Under single-sequence verification, Kangaroo achieves speedups up to
$1.68\times$ on Spec-Bench, outperforming Medusa-1 with 88.7\% fewer additional
parameters (67M compared to 591M). The code for Kangaroo is available at
https://github.com/Equationliu/Kangaroo. | cs.CL | Kangaroo: Lossless Self-Speculative Decoding via Double Early Exiting |
2024-04-29T17:36:58Z | http://arxiv.org/pdf/2404.18988v2 | http://arxiv.org/abs/2404.18988v2 | Scott Viteri, Max Lamparth, Peter Chatain, Clark Barrett | 2024-05-23T00:00:23Z | cs.CL | Chain-of-Thought (CoT) reasoning could in principle enable a deeper
understanding of a language model's (LM) internal reasoning. However, prior
work suggests that LMs can answer questions similarly despite changes in their
CoT, suggesting that those models are not truly using the CoT. We propose an
reinforcement learning technique to produce CoTs that are sufficient alone for
predicting future text, independent of other context. This methodology ensures
that if the LM can predict future tokens, then it must have used the CoT to
understand its context. We formalize the informativeness of a sender to a
receiver LM as the degree to which the sender helps the receiver predict their
future observations, and we define a "Markovian" LM as one which predicts
future text given only a CoT as context. We derive a "Markovian training"
procedure by applying our definition of informativeness to a Markovian LM and
optimizing via policy gradient and Proximal Policy Optimization (PPO). We
demonstrate our training algorithm's effectiveness on fifteen-term arithmetic
problems, show the model utilizes the CoT, and externally validate that the
generated CoT is meaningful and usable by another model. | cs.CL | Markovian Agents for Informative Language Modeling |
2024-04-29T17:16:22Z | http://arxiv.org/pdf/2404.18880v1 | http://arxiv.org/abs/2404.18880v1 | Aman Saini, Artem Chernodub, Vipul Raheja, Vivek Kulkarni | 2024-04-29T17:16:22Z | cs.CL | We introduce Spivavtor, a dataset, and instruction-tuned models for text
editing focused on the Ukrainian language. Spivavtor is the Ukrainian-focused
adaptation of the English-only CoEdIT model. Similar to CoEdIT, Spivavtor
performs text editing tasks by following instructions in Ukrainian. This paper
describes the details of the Spivavtor-Instruct dataset and Spivavtor models.
We evaluate Spivavtor on a variety of text editing tasks in Ukrainian, such as
Grammatical Error Correction (GEC), Text Simplification, Coherence, and
Paraphrasing, and demonstrate its superior performance on all of them. We
publicly release our best-performing models and data as resources to the
community to advance further research in this space. | cs.CL | Spivavtor: An Instruction Tuned Ukrainian Text Editing Model |
2024-04-29T17:00:53Z | http://arxiv.org/pdf/2404.18870v1 | http://arxiv.org/abs/2404.18870v1 | Aaron J. Li, Satyapriya Krishna, Himabindu Lakkaraju | 2024-04-29T17:00:53Z | cs.CL, cs.AI | The surge in Large Language Models (LLMs) development has led to improved
performance on cognitive tasks as well as an urgent need to align these models
with human values in order to safely exploit their power. Despite the
effectiveness of preference learning algorithms like Reinforcement Learning
From Human Feedback (RLHF) in aligning human preferences, their assumed
improvements on model trustworthiness haven't been thoroughly testified. Toward
this end, this study investigates how models that have been aligned with
general-purpose preference data on helpfulness and harmlessness perform across
five trustworthiness verticals: toxicity, stereotypical bias, machine ethics,
truthfulness, and privacy. For model alignment, we focus on three widely used
RLHF variants: Supervised Finetuning (SFT), Proximal Policy Optimization (PPO),
and Direct Preference Optimization (DPO). Through extensive empirical
investigations, we discover that the improvement in trustworthiness by RLHF is
far from guaranteed, and there exists a complex interplay between preference
data, alignment algorithms, and specific trustworthiness aspects. Together, our
results underscore the need for more nuanced approaches for model alignment. By
shedding light on the intricate dynamics of these components within model
alignment, we hope this research will guide the community towards developing
language models that are both capable and trustworthy. | cs.CL | More RLHF, More Trust? On The Impact of Human Preference Alignment On Language Model Trustworthiness |
2024-04-29T16:52:57Z | http://arxiv.org/pdf/2404.18865v1 | http://arxiv.org/abs/2404.18865v1 | Stefan F. Schouten, Peter Bloem, Ilia Markov, Piek Vossen | 2024-04-29T16:52:57Z | cs.CL | Recent work has demonstrated that the latent spaces of large language models
(LLMs) contain directions predictive of the truth of sentences. Multiple
methods recover such directions and build probes that are described as getting
at a model's "knowledge" or "beliefs". We investigate this phenomenon, looking
closely at the impact of context on the probes. Our experiments establish where
in the LLM the probe's predictions can be described as being conditional on the
preceding (related) sentences. Specifically, we quantify the responsiveness of
the probes to the presence of (negated) supporting and contradicting sentences,
and score the probes on their consistency. We also perform a causal
intervention experiment, investigating whether moving the representation of a
premise along these belief directions influences the position of the hypothesis
along that same direction. We find that the probes we test are generally
context sensitive, but that contexts which should not affect the truth often
still impact the probe outputs. Our experiments show that the type of errors
depend on the layer, the (type of) model, and the kind of data. Finally, our
results suggest that belief directions are (one of the) causal mediators in the
inference process that incorporates in-context information. | cs.CL | Truth-value judgment in language models: belief directions are context sensitive |
2024-04-29T16:44:27Z | http://arxiv.org/pdf/2404.18851v1 | http://arxiv.org/abs/2404.18851v1 | Mario Corrales-Astorgano, David Escudero-Mancebo, Lourdes Aguilar, Valle Flores-Lucas, Valentín Cardeñoso-Payo, Carlos Vivaracho-Pascual, César González-Ferreras | 2024-04-29T16:44:27Z | cs.CL | Rubrics are a commonly used tool for labeling voice corpora in speech quality
assessment, although their application in the context of pathological speech
remains relatively limited. In this study, we introduce a comprehensive rubric
based on various dimensions of speech quality, including phonetics, fluency,
and prosody. The objective is to establish standardized criteria for
identifying errors within the speech of individuals with Down syndrome, thereby
enabling the development of automated assessment systems. To achieve this
objective, we utilized the Prautocal corpus. To assess the quality of
annotations using our rubric, two experiments were conducted, focusing on
phonetics and fluency. For phonetic evaluation, we employed the Goodness of
Pronunciation (GoP) metric, utilizing automatic segmentation systems and
correlating the results with evaluations conducted by a specialized speech
therapist. While the obtained correlation values were not notably high, a
positive trend was observed. In terms of fluency assessment, deep learning
models like wav2vec were used to extract audio features, and we employed an SVM
classifier trained on a corpus focused on identifying fluency issues to
categorize Prautocal corpus samples. The outcomes highlight the complexities of
evaluating such phenomena, with variability depending on the specific type of
disfluency detected. | cs.CL | A Comprehensive Rubric for Annotating Pathological Speech |
2024-04-29T16:42:26Z | http://arxiv.org/pdf/2404.18848v3 | http://arxiv.org/abs/2404.18848v3 | Yuxuan Yan, Qianqian Yang, Shunpu Tang, Zhiguo Shi | 2024-05-25T06:55:19Z | cs.LG, cs.AI, cs.CL | Despite their exceptional performance on various tasks after fine-tuning,
pre-trained language models (PLMs) face significant challenges due to growing
privacy concerns with data in centralized training methods. We consider
federated learning (FL) to fine-tune PLMs in this paper. However, the
substantial number of parameters in PLMs poses significant difficulties for
client devices with limited communication and computational resources. One
promising solution is to exploit parameter-efficient fine-tuning (PEFT) into
FL, which trains a much smaller set of parameters than full parameter
fine-tuning (FFT). Although remarkably improving training efficiency, PEFT
methods may lead to degraded performance especially when data across different
clients are non i.i.d, as revealed by experimental results. To overcome this,
we propose FeDeRA, which extends and improves a widely used PEFT method, i.e.,
low-rank adaption (LoRA). FeDeRA follows LoRA by decomposing the weight
matrices of the PLMs into low-rank matrices, which allows for more efficient
computation and parameter updates during fine-tuning. Different from LoRA which
simply initializes these low-rank matrices by random sampling or zeros, the
proposed FeDeRA initializes these matrices by the results of performing
singular value decomposition (SVD) on the pre-trained weight matrices.
Extensive experiments across various tasks and datasets show that FeDeRA
outperforms the considered PEFT baselines and is comparable to or even
surpasses FFT method within the FL setting in terms of task performance.
Moreover, FeDeRA requires only 1% trainable paramentes compared to FFT,
significantly reducing training time costs by more than 90% to achieve the same
task performance level. The experimental results also highlight the robustness
of FeDeRA against data heterogeneity, as it maintains stable task performance
even as data heterogeneity increases. | cs.LG | FeDeRA:Efficient Fine-tuning of Language Models in Federated Learning Leveraging Weight Decomposition |
2024-04-29T16:19:47Z | http://arxiv.org/pdf/2404.18832v1 | http://arxiv.org/abs/2404.18832v1 | Petter Mæhlum, David Samuel, Rebecka Maria Norman, Elma Jelin, Øyvind Andresen Bjertnæs, Lilja Øvrelid, Erik Velldal | 2024-04-29T16:19:47Z | cs.CL | Sentiment analysis is an important tool for aggregating patient voices, in
order to provide targeted improvements in healthcare services. A prerequisite
for this is the availability of in-domain data annotated for sentiment. This
article documents an effort to add sentiment annotations to free-text comments
in patient surveys collected by the Norwegian Institute of Public Health
(NIPH). However, annotation can be a time-consuming and resource-intensive
process, particularly when it requires domain expertise. We therefore also
evaluate a possible alternative to human annotation, using large language
models (LLMs) as annotators. We perform an extensive evaluation of the approach
for two openly available pretrained LLMs for Norwegian, experimenting with
different configurations of prompts and in-context learning, comparing their
performance to human annotators. We find that even for zero-shot runs, models
perform well above the baseline for binary sentiment, but still cannot compete
with human annotators on the full dataset. | cs.CL | It's Difficult to be Neutral -- Human and LLM-based Sentiment Annotation of Patient Comments |
2024-04-29T16:05:36Z | http://arxiv.org/pdf/2404.18824v1 | http://arxiv.org/abs/2404.18824v1 | Ruijie Xu, Zengzhi Wang, Run-Ze Fan, Pengfei Liu | 2024-04-29T16:05:36Z | cs.CL, cs.AI, cs.LG | Amid the expanding use of pre-training data, the phenomenon of benchmark
dataset leakage has become increasingly prominent, exacerbated by opaque
training processes and the often undisclosed inclusion of supervised data in
contemporary Large Language Models (LLMs). This issue skews benchmark
effectiveness and fosters potentially unfair comparisons, impeding the field's
healthy development. To address this, we introduce a detection pipeline
utilizing Perplexity and N-gram accuracy, two simple and scalable metrics that
gauge a model's prediction precision on benchmark, to identify potential data
leakages. By analyzing 31 LLMs under the context of mathematical reasoning, we
reveal substantial instances of training even test set misuse, resulting in
potentially unfair comparisons. These findings prompt us to offer several
recommendations regarding model documentation, benchmark setup, and future
evaluations. Notably, we propose the "Benchmark Transparency Card" to encourage
clear documentation of benchmark utilization, promoting transparency and
healthy developments of LLMs. we have made our leaderboard, pipeline
implementation, and model predictions publicly available, fostering future
research. | cs.CL | Benchmarking Benchmark Leakage in Large Language Models |
2024-04-29T15:48:01Z | http://arxiv.org/pdf/2404.18810v2 | http://arxiv.org/abs/2404.18810v2 | Wondimagegnhue Tsegaye Tufa, Ilia Markov, Piek Vossen | 2024-05-07T12:23:39Z | cs.CL | Cross-lingual transfer has become an effective way of transferring knowledge
between languages. In this paper, we explore an often overlooked aspect in this
domain: the influence of the source language of a language model on language
transfer performance. We consider a case where the target language and its
script are not part of the pre-trained model. We conduct a series of
experiments on monolingual and multilingual models that are pre-trained on
different tokenization methods to determine factors that affect cross-lingual
transfer to a new language with a unique script. Our findings reveal the
importance of the tokenizer as a stronger factor than the shared script,
language similarity, and model size. | cs.CL | Unknown Script: Impact of Script on Cross-Lingual Transfer |
2024-04-29T15:33:23Z | http://arxiv.org/pdf/2404.18796v2 | http://arxiv.org/abs/2404.18796v2 | Pat Verga, Sebastian Hofstatter, Sophia Althammer, Yixuan Su, Aleksandra Piktus, Arkady Arkhangorodsky, Minjie Xu, Naomi White, Patrick Lewis | 2024-05-01T15:37:11Z | cs.CL, cs.AI | As Large Language Models (LLMs) have become more advanced, they have outpaced
our abilities to accurately evaluate their quality. Not only is finding data to
adequately probe particular model properties difficult, but evaluating the
correctness of a model's freeform generation alone is a challenge. To address
this, many evaluations now rely on using LLMs themselves as judges to score the
quality of outputs from other LLMs. Evaluations most commonly use a single
large model like GPT4. While this method has grown in popularity, it is costly,
has been shown to introduce intramodel bias, and in this work, we find that
very large models are often unnecessary. We propose instead to evaluate models
using a Panel of LLm evaluators (PoLL). Across three distinct judge settings
and spanning six different datasets, we find that using a PoLL composed of a
larger number of smaller models outperforms a single large judge, exhibits less
intra-model bias due to its composition of disjoint model families, and does so
while being over seven times less expensive. | cs.CL | Replacing Judges with Juries: Evaluating LLM Generations with a Panel of Diverse Models |
2024-04-29T15:18:33Z | http://arxiv.org/pdf/2404.18784v1 | http://arxiv.org/abs/2404.18784v1 | Tessa Masis, Brendan O'Connor | 2024-04-29T15:18:33Z | cs.CL, cs.AI | Geo-entity linking is the task of linking a location mention to the
real-world geographic location. In this paper we explore the challenging task
of geo-entity linking for noisy, multilingual social media data. There are few
open-source multilingual geo-entity linking tools available and existing ones
are often rule-based, which break easily in social media settings, or
LLM-based, which are too expensive for large-scale datasets. We present a
method which represents real-world locations as averaged embeddings from
labeled user-input location names and allows for selective prediction via an
interpretable confidence score. We show that our approach improves geo-entity
linking on a global and multilingual social media dataset, and discuss progress
and problems with evaluating at different geographic granularities. | cs.CL | Where on Earth Do Users Say They Are?: Geo-Entity Linking for Noisy Multilingual User Input |
2024-04-29T14:56:47Z | http://arxiv.org/pdf/2404.18759v2 | http://arxiv.org/abs/2404.18759v2 | Juraj Vladika, Stephen Meisenbacher, Martina Preis, Alexandra Klymenko, Florian Matthes | 2024-05-02T09:56:51Z | cs.CL, cs.CY | In recent years, the field of Legal Tech has risen in prevalence, as the
Natural Language Processing (NLP) and legal disciplines have combined forces to
digitalize legal processes. Amidst the steady flow of research solutions
stemming from the NLP domain, the study of use cases has fallen behind, leading
to a number of innovative technical methods without a place in practice. In
this work, we aim to build a structured overview of Legal Tech use cases,
grounded in NLP literature, but also supplemented by voices from legal practice
in Germany. Based upon a Systematic Literature Review, we identify seven
categories of NLP technologies for the legal domain, which are then studied in
juxtaposition to 22 legal use cases. In the investigation of these use cases,
we identify 15 ethical, legal, and social aspects (ELSA), shedding light on the
potential concerns of digitally transforming the legal domain. | cs.CL | Towards A Structured Overview of Use Cases for Natural Language Processing in the Legal Domain: A German Perspective |
2024-04-29T14:52:38Z | http://arxiv.org/pdf/2404.18977v1 | http://arxiv.org/abs/2404.18977v1 | Mike Zhang | 2024-04-29T14:52:38Z | cs.CL | [Abridged Abstract]
Recent technological advances underscore labor market dynamics, yielding
significant consequences for employment prospects and increasing job vacancy
data across platforms and languages. Aggregating such data holds potential for
valuable insights into labor market demands, new skills emergence, and
facilitating job matching for various stakeholders. However, despite prevalent
insights in the private sector, transparent language technology systems and
data for this domain are lacking. This thesis investigates Natural Language
Processing (NLP) technology for extracting relevant information from job
descriptions, identifying challenges including scarcity of training data, lack
of standardized annotation guidelines, and shortage of effective extraction
methods from job ads. We frame the problem, obtaining annotated data, and
introducing extraction methodologies. Our contributions include job description
datasets, a de-identification dataset, and a novel active learning algorithm
for efficient model training. We propose skill extraction using weak
supervision, a taxonomy-aware pre-training methodology adapting multilingual
language models to the job market domain, and a retrieval-augmented model
leveraging multiple skill extraction datasets to enhance overall performance.
Finally, we ground extracted information within a designated taxonomy. | cs.CL | Computational Job Market Analysis with Natural Language Processing |
2024-04-29T14:45:28Z | http://arxiv.org/pdf/2404.18976v1 | http://arxiv.org/abs/2404.18976v1 | Paul Pu Liang | 2024-04-29T14:45:28Z | cs.LG, cs.AI, cs.CL, cs.CV, cs.MM | Building multisensory AI systems that learn from multiple sensory inputs such
as text, speech, video, real-world sensors, wearable devices, and medical data
holds great promise for impact in many scientific areas with practical
benefits, such as in supporting human health and well-being, enabling
multimedia content processing, and enhancing real-world autonomous agents. By
synthesizing a range of theoretical frameworks and application domains, this
thesis aims to advance the machine learning foundations of multisensory AI. In
the first part, we present a theoretical framework formalizing how modalities
interact with each other to give rise to new information for a task. These
interactions are the basic building blocks in all multimodal problems, and
their quantification enables users to understand their multimodal datasets,
design principled approaches to learn these interactions, and analyze whether
their model has succeeded in learning. In the second part, we study the design
of practical multimodal foundation models that generalize over many modalities
and tasks, which presents a step toward grounding large language models to
real-world sensory modalities. We introduce MultiBench, a unified large-scale
benchmark across a wide range of modalities, tasks, and research areas,
followed by the cross-modal attention and multimodal transformer architectures
that now underpin many of today's multimodal foundation models. Scaling these
architectures on MultiBench enables the creation of general-purpose
multisensory AI systems, and we discuss our collaborative efforts in applying
these models for real-world impact in affective computing, mental health,
cancer prognosis, and robotics. Finally, we conclude this thesis by discussing
how future work can leverage these ideas toward more general, interactive, and
safe multisensory AI. | cs.LG | Foundations of Multisensory Artificial Intelligence |
2024-04-29T14:41:59Z | http://arxiv.org/pdf/2404.18739v1 | http://arxiv.org/abs/2404.18739v1 | Artem Abzaliev, Humberto Pérez Espinosa, Rada Mihalcea | 2024-04-29T14:41:59Z | cs.CL | Similar to humans, animals make extensive use of verbal and non-verbal forms
of communication, including a large range of audio signals. In this paper, we
address dog vocalizations and explore the use of self-supervised speech
representation models pre-trained on human speech to address dog bark
classification tasks that find parallels in human-centered tasks in speech
recognition. We specifically address four tasks: dog recognition, breed
identification, gender classification, and context grounding. We show that
using speech embedding representations significantly improves over simpler
classification baselines. Further, we also find that models pre-trained on
large human speech acoustics can provide additional performance boosts on
several tasks. | cs.CL | Towards Dog Bark Decoding: Leveraging Human Speech Processing for Automated Bark Classification |
2024-04-29T14:14:33Z | http://arxiv.org/pdf/2404.18726v1 | http://arxiv.org/abs/2404.18726v1 | Wondimagegnhue Tsegaye Tufa, Ilia Markov, Piek Vossen | 2024-04-29T14:14:33Z | cs.CL | Toxic language remains an ongoing challenge on social media platforms,
presenting significant issues for users and communities. This paper provides a
cross-topic and cross-lingual analysis of toxicity in Reddit conversations. We
collect 1.5 million comment threads from 481 communities in six languages:
English, German, Spanish, Turkish,Arabic, and Dutch, covering 80 topics such as
Culture, Politics, and News. We thoroughly analyze how toxicity spikes within
different communities in relation to specific topics. We observe consistent
patterns of increased toxicity across languages for certain topics, while also
noting significant variations within specific language communities. | cs.CL | The Constant in HATE: Analyzing Toxicity in Reddit across Topics and Languages |
2024-04-29T14:11:16Z | http://arxiv.org/pdf/2404.18722v1 | http://arxiv.org/abs/2404.18722v1 | Solène Tarride, Yoann Schneider, Marie Generali-Lince, Mélodie Boillet, Bastien Abadie, Christopher Kermorvant | 2024-04-29T14:11:16Z | cs.CV, cs.CL | PyLaia is one of the most popular open-source software for Automatic Text
Recognition (ATR), delivering strong performance in terms of speed and
accuracy. In this paper, we outline our recent contributions to the PyLaia
library, focusing on the incorporation of reliable confidence scores and the
integration of statistical language modeling during decoding. Our
implementation provides an easy way to combine PyLaia with n-grams language
models at different levels. One of the highlights of this work is that language
models are completely auto-tuned: they can be built and used easily without any
expert knowledge, and without requiring any additional data. To demonstrate the
significance of our contribution, we evaluate PyLaia's performance on twelve
datasets, both with and without language modelling. The results show that
decoding with small language models improves the Word Error Rate by 13% and the
Character Error Rate by 12% in average. Additionally, we conduct an analysis of
confidence scores and highlight the importance of calibration techniques. Our
implementation is publicly available in the official PyLaia repository at
https://gitlab.teklia.com/atr/pylaia, and twelve open-source models are
released on Hugging Face. | cs.CV | Improving Automatic Text Recognition with Language Models in the PyLaia Open-Source Library |
2024-04-29T13:58:03Z | http://arxiv.org/pdf/2404.18708v1 | http://arxiv.org/abs/2404.18708v1 | Andy Lücking, Alexander Henlein, Alexander Mehler | 2024-04-29T13:58:03Z | cs.CL | The "meaning" of an iconic gesture is conditioned on its informational
evaluation. Only informational evaluation lifts a gesture to a quasi-linguistic
level that can interact with verbal content. Interaction is either vacuous or
regimented by usual lexicon-driven inferences. Informational evaluation is
spelled out as extended exemplification (extemplification) in terms of
perceptual classification of a gesture's visual iconic model. The iconic model
is derived from Frege/Montague-like truth-functional evaluation of a gesture's
form within spatially extended domains. We further argue that the perceptual
classification of instances of visual communication requires a notion of
meaning different from Frege/Montague frameworks. Therefore, a heuristic for
gesture interpretation is provided that can guide the working semanticist. In
sum, an iconic gesture semantics is introduced which covers the full range from
kinematic gesture representations over model-theoretic evaluation to
inferential interpretation in dynamic semantic frameworks. | cs.CL | Iconic Gesture Semantics |
2024-04-29T13:47:04Z | http://arxiv.org/pdf/2404.18971v1 | http://arxiv.org/abs/2404.18971v1 | Zacharias Chrysidis, Stefanos-Iordanis Papadopoulos, Symeon Papadopoulos, Panagiotis C. Petrantonakis | 2024-04-29T13:47:04Z | cs.CL, cs.CY, cs.IR, cs.SI | Automated fact-checking (AFC) is garnering increasing attention by
researchers aiming to help fact-checkers combat the increasing spread of
misinformation online. While many existing AFC methods incorporate external
information from the Web to help examine the veracity of claims, they often
overlook the importance of verifying the source and quality of collected
"evidence". One overlooked challenge involves the reliance on "leaked
evidence", information gathered directly from fact-checking websites and used
to train AFC systems, resulting in an unrealistic setting for early
misinformation detection. Similarly, the inclusion of information from
unreliable sources can undermine the effectiveness of AFC systems. To address
these challenges, we present a comprehensive approach to evidence verification
and filtering. We create the "CREDible, Unreliable or LEaked" (CREDULE)
dataset, which consists of 91,632 articles classified as Credible, Unreliable
and Fact checked (Leaked). Additionally, we introduce the EVidence VERification
Network (EVVER-Net), trained on CREDULE to detect leaked and unreliable
evidence in both short and long texts. EVVER-Net can be used to filter evidence
collected from the Web, thus enhancing the robustness of end-to-end AFC
systems. We experiment with various language models and show that EVVER-Net can
demonstrate impressive performance of up to 91.5% and 94.4% accuracy, while
leveraging domain credibility scores along with short or long texts,
respectively. Finally, we assess the evidence provided by widely-used
fact-checking datasets including LIAR-PLUS, MOCHEG, FACTIFY, NewsCLIPpings+ and
VERITE, some of which exhibit concerning rates of leaked and unreliable
evidence. | cs.CL | Credible, Unreliable or Leaked?: Evidence Verification for Enhanced Automated Fact-checking |
2024-04-29T13:30:27Z | http://arxiv.org/pdf/2404.18684v2 | http://arxiv.org/abs/2404.18684v2 | Sidharth Ranjan, Titus von der Malsburg | 2024-05-11T03:51:38Z | cs.CL, econ.TH, math.OC | Dependency length minimization is a universally observed quantitative
property of natural languages. However, the extent of dependency length
minimization, and the cognitive mechanisms through which the language processor
achieves this minimization remain unclear. This research offers mechanistic
insights by postulating that moving a short preverbal constituent next to the
main verb explains preverbal constituent ordering decisions better than global
minimization of dependency length in SOV languages. This approach constitutes a
least-effort strategy because it's just one operation but simultaneously
reduces the length of all preverbal dependencies linked to the main verb. We
corroborate this strategy using large-scale corpus evidence across all seven
SOV languages that are prominently represented in the Universal Dependency
Treebank. These findings align with the concept of bounded rationality, where
decision-making is influenced by 'quick-yet-economical' heuristics rather than
exhaustive searches for optimal solutions. Overall, this work sheds light on
the role of bounded rationality in linguistic decision-making and language
evolution. | cs.CL | Work Smarter...Not Harder: Efficient Minimization of Dependency Length in SOV Languages |
2024-04-29T13:23:33Z | http://arxiv.org/pdf/2405.01591v1 | http://arxiv.org/abs/2405.01591v1 | Seonhee Cho, Choonghan Kim, Jiho Lee, Chetan Chilkunda, Sujin Choi, Joo Heung Yoon | 2024-04-29T13:23:33Z | cs.CL, cs.AI, eess.IV | Recent advancements in Large Multimodal Models (LMMs) have attracted interest
in their generalization capability with only a few samples in the prompt. This
progress is particularly relevant to the medical domain, where the quality and
sensitivity of data pose unique challenges for model training and application.
However, the dependency on high-quality data for effective in-context learning
raises questions about the feasibility of these models when encountering with
the inevitable variations and errors inherent in real-world medical data. In
this paper, we introduce MID-M, a novel framework that leverages the in-context
learning capabilities of a general-domain Large Language Model (LLM) to process
multimodal data via image descriptions. MID-M achieves a comparable or superior
performance to task-specific fine-tuned LMMs and other general-domain ones,
without the extensive domain-specific training or pre-training on multimodal
data, with significantly fewer parameters. This highlights the potential of
leveraging general-domain LLMs for domain-specific tasks and offers a
sustainable and cost-effective alternative to traditional LMM developments.
Moreover, the robustness of MID-M against data quality issues demonstrates its
practical utility in real-world medical domain applications. | cs.CL | Simplifying Multimodality: Unimodal Approach to Multimodal Challenges in Radiology with General-Domain Large Language Model |
2024-04-29T13:15:03Z | http://arxiv.org/pdf/2405.01590v1 | http://arxiv.org/abs/2405.01590v1 | Manel Aloui, Hasna Chouikhi, Ghaith Chaabane, Haithem Kchaou, Chehir Dhaouadi | 2024-04-29T13:15:03Z | cs.CL | In recent years, Large Language Models have revolutionized the field of
natural language processing, showcasing an impressive rise predominantly in
English-centric domains. These advancements have set a global benchmark,
inspiring significant efforts toward developing Arabic LLMs capable of
understanding and generating the Arabic language with remarkable accuracy.
Despite these advancements, a critical challenge persists: the potential bias
in Arabic LLMs, primarily attributed to their reliance on datasets comprising
English data that has been translated into Arabic. This reliance not only
compromises the authenticity of the generated content but also reflects a
broader issue -the scarcity of original quality Arabic linguistic data. This
study aims to address the data scarcity in the Arab world and to encourage the
development of Arabic Language Models that are true to both the linguistic and
nuances of the region. We undertook a large-scale data mining project,
extracting a substantial volume of text from the Common Crawl WET files,
specifically targeting Arabic content. The extracted data underwent a rigorous
cleaning and deduplication process, using innovative techniques to ensure the
integrity and uniqueness of the dataset. The result is the 101 Billion Arabic
Words Dataset, the largest Arabic dataset available to date, which can
significantly contribute to the development of authentic Arabic LLMs. This
study not only highlights the potential for creating linguistically and
culturally accurate Arabic LLMs but also sets a precedent for future research
in enhancing the authenticity of Arabic language models. | cs.CL | 101 Billion Arabic Words Dataset |
2024-04-29T12:38:26Z | http://arxiv.org/pdf/2404.18655v1 | http://arxiv.org/abs/2404.18655v1 | Haeun Yu, Pepa Atanasova, Isabelle Augenstein | 2024-04-29T12:38:26Z | cs.CL, cs.AI, 68T50, I.2.7 | Language Models (LMs) acquire parametric knowledge from their training
process, embedding it within their weights. The increasing scalability of LMs,
however, poses significant challenges for understanding a model's inner
workings and further for updating or correcting this embedded knowledge without
the significant cost of retraining. This underscores the importance of
unveiling exactly what knowledge is stored and its association with specific
model components. Instance Attribution (IA) and Neuron Attribution (NA) offer
insights into this training-acquired knowledge, though they have not been
compared systematically. Our study introduces a novel evaluation framework to
quantify and compare the knowledge revealed by IA and NA. To align the results
of the methods we introduce the attribution method NA-Instances to apply NA for
retrieving influential training instances, and IA-Neurons to discover important
neurons of influential instances discovered by IA. We further propose a
comprehensive list of faithfulness tests to evaluate the comprehensiveness and
sufficiency of the explanations provided by both methods. Through extensive
experiments and analysis, we demonstrate that NA generally reveals more diverse
and comprehensive information regarding the LM's parametric knowledge compared
to IA. Nevertheless, IA provides unique and valuable insights into the LM's
parametric knowledge, which are not revealed by NA. Our findings further
suggest the potential of a synergistic approach of combining the diverse
findings of IA and NA for a more holistic understanding of an LM's parametric
knowledge. | cs.CL | Revealing the Parametric Knowledge of Language Models: A Unified Framework for Attribution Methods |
2024-04-29T12:16:21Z | http://arxiv.org/pdf/2406.02563v1 | http://arxiv.org/abs/2406.02563v1 | Sunil Kumar Kopparapu, Ashish Panda | 2024-04-29T12:16:21Z | eess.AS, cs.CL, cs.SD | Unlike hybrid speech recognition systems where the use of tokens was
restricted to phones, biphones or triphones the choice of tokens in the
end-to-end ASR systems is derived from the text corpus of the training data.
The use of tokenization algorithms like Byte Pair Encoding (BPE) and WordPiece
is popular in identifying the tokens that are used in the overall training
process of the speech recognition system. Popular toolkits, like ESPNet use a
pre-defined vocabulary size (number of tokens) for these tokenization
algorithms, but there is no discussion on how vocabulary size was derived. In
this paper, we build a cost function, assuming the tokenization process to be a
black-box to enable choosing the number of tokens which might most benefit
building an end-to-end ASR. We show through experiments on LibriSpeech 100 hour
set that the performance of an end-to-end ASR system improves when the number
of tokens are chosen carefully. | eess.AS | A cost minimization approach to fix the vocabulary size in a tokenizer for an End-to-End ASR system |
2024-04-29T12:16:08Z | http://arxiv.org/pdf/2404.18638v1 | http://arxiv.org/abs/2404.18638v1 | Sina Gholamian, Domingo Huh | 2024-04-29T12:16:08Z | cs.AI, cs.CL | Large Language Models (LLMs) encapsulate an extensive amount of world
knowledge, and this has enabled their application in various domains to improve
the performance of a variety of Natural Language Processing (NLP) tasks. This
has also facilitated a more accessible paradigm of conversation-based
interactions between humans and AI systems to solve intended problems. However,
one interesting avenue that shows untapped potential is the use of LLMs as
Reinforcement Learning (RL) agents to enable conversational RL problem solving.
Therefore, in this study, we explore the concept of formulating Markov Decision
Process-based RL problems as LLM prompting tasks. We demonstrate how LLMs can
be iteratively prompted to learn and optimize policies for specific RL tasks.
In addition, we leverage the introduced prompting technique for episode
simulation and Q-Learning, facilitated by LLMs. We then show the practicality
of our approach through two detailed case studies for "Research Scientist" and
"Legal Matter Intake" workflows. | cs.AI | Reinforcement Learning Problem Solving with Large Language Models |
2024-04-29T11:52:20Z | http://arxiv.org/pdf/2404.18624v2 | http://arxiv.org/abs/2404.18624v2 | Letitia Parcalabescu, Anette Frank | 2024-06-10T10:43:20Z | cs.CL, cs.AI, cs.CV, cs.LG, 68Txx, I.2.7; I.2.10 | Vision and language model (VLM) decoders are currently the best-performing
architectures on multimodal tasks. Next to predictions, they can also produce
explanations, either in post-hoc or CoT settings. However, it is not clear how
much they use the vision and text modalities when generating predictions or
explanations. In this work, we investigate if VLMs rely on modalities
differently when they produce explanations as opposed to providing answers. We
also evaluate the self-consistency of VLM decoders in both post-hoc and CoT
explanation settings, by extending existing unimodal tests and measures to VLM
decoders. We find that VLMs are less self-consistent than LLMs. Text
contributions in VL decoders are more important than image contributions in all
examined tasks. Moreover, the contributions of images are significantly
stronger for explanation generation compared to answer generation. This
difference is even larger in CoT compared to post-hoc explanations. Lastly, we
provide an up-to-date benchmarking of state-of-the-art VL decoders on the VALSE
benchmark, which before only covered VL encoders. We find that VL decoders
still struggle with most phenomena tested by VALSE. | cs.CL | Do Vision & Language Decoders use Images and Text equally? How Self-consistent are their Explanations? |
2024-04-29T11:34:06Z | http://arxiv.org/pdf/2404.18615v1 | http://arxiv.org/abs/2404.18615v1 | Bashar Alhafni, Reem Hazim, Juan Piñeros Liberato, Muhamed Al Khalil, Nizar Habash | 2024-04-29T11:34:06Z | cs.CL | We present the SAMER Corpus, the first manually annotated Arabic parallel
corpus for text simplification targeting school-aged learners. Our corpus
comprises texts of 159K words selected from 15 publicly available Arabic
fiction novels most of which were published between 1865 and 1955. Our corpus
includes readability level annotations at both the document and word levels, as
well as two simplified parallel versions for each text targeting learners at
two different readability levels. We describe the corpus selection process, and
outline the guidelines we followed to create the annotations and ensure their
quality. Our corpus is publicly available to support and encourage research on
Arabic text simplification, Arabic automatic readability assessment, and the
development of Arabic pedagogical language technologies. | cs.CL | The SAMER Arabic Text Simplification Corpus |
2024-04-29T10:55:08Z | http://arxiv.org/pdf/2404.18585v1 | http://arxiv.org/abs/2404.18585v1 | Wei Zhou, Mohsen Mesgar, Heike Adel, Annemarie Friedrich | 2024-04-29T10:55:08Z | cs.CL | Table Question Answering (TQA) aims at composing an answer to a question
based on tabular data. While prior research has shown that TQA models lack
robustness, understanding the underlying cause and nature of this issue remains
predominantly unclear, posing a significant obstacle to the development of
robust TQA systems. In this paper, we formalize three major desiderata for a
fine-grained evaluation of robustness of TQA systems. They should (i) answer
questions regardless of alterations in table structure, (ii) base their
responses on the content of relevant cells rather than on biases, and (iii)
demonstrate robust numerical reasoning capabilities. To investigate these
aspects, we create and publish a novel TQA evaluation benchmark in English. Our
extensive experimental analysis reveals that none of the examined
state-of-the-art TQA systems consistently excels in these three aspects. Our
benchmark is a crucial instrument for monitoring the behavior of TQA systems
and paves the way for the development of robust TQA systems. We release our
benchmark publicly. | cs.CL | FREB-TQA: A Fine-Grained Robustness Evaluation Benchmark for Table Question Answering |
2024-04-29T10:30:59Z | http://arxiv.org/pdf/2405.04585v1 | http://arxiv.org/abs/2405.04585v1 | Arpit Aggarwal | 2024-04-29T10:30:59Z | cs.CL, cs.AI, cs.LG | There are several improvements proposed over the baseline Absolute Positional
Encoding (APE) method used in original transformer. In this study, we aim to
investigate the implications of inadequately representing positional encoding
in higher dimensions on crucial aspects of the attention mechanism, the model's
capacity to learn relative positional information, and the convergence of
models, all stemming from the choice of sinusoidal basis functions. Through a
combination of theoretical insights and empirical analyses, we elucidate how
these challenges extend beyond APEs and may adversely affect the performance of
Relative Positional Encoding (RPE) methods, such as Rotatory Positional
Encoding (RoPE).
Subsequently, we introduce an innovative solution termed Orthogonal
Polynomial Based Positional Encoding (PoPE) to address some of the limitations
associated with existing methods. The PoPE method encodes positional
information by leveraging Orthogonal Legendre polynomials. Legendre polynomials
as basis functions offers several desirable properties for positional encoding,
including improved correlation structure, non-periodicity, orthogonality, and
distinct functional forms among polynomials of varying orders. Our experimental
findings demonstrate that transformer models incorporating PoPE outperform
baseline transformer models on the $Multi30k$ English-to-German translation
task, thus establishing a new performance benchmark. Furthermore, PoPE-based
transformers exhibit significantly accelerated convergence rates.
Additionally, we will present novel theoretical perspectives on position
encoding based on the superior performance of PoPE. | cs.CL | PoPE: Legendre Orthogonal Polynomials Based Position Encoding for Large Language Models |
2024-04-29T10:20:41Z | http://arxiv.org/pdf/2404.18570v1 | http://arxiv.org/abs/2404.18570v1 | Francesco Periti, Pierluigi Cassotti, Haim Dubossarsky, Nina Tahmasebi | 2024-04-29T10:20:41Z | cs.CL | Modern language models are capable of contextualizing words based on their
surrounding context. However, this capability is often compromised due to
semantic change that leads to words being used in new, unexpected contexts not
encountered during pre-training. In this paper, we model \textit{semantic
change} by studying the effect of unexpected contexts introduced by
\textit{lexical replacements}. We propose a \textit{replacement schema} where a
target word is substituted with lexical replacements of varying relatedness,
thus simulating different kinds of semantic change. Furthermore, we leverage
the replacement schema as a basis for a novel \textit{interpretable} model for
semantic change. We are also the first to evaluate the use of LLaMa for
semantic change detection. | cs.CL | Analyzing Semantic Change through Lexical Replacements |
2024-04-29T10:12:04Z | http://arxiv.org/pdf/2404.18564v1 | http://arxiv.org/abs/2404.18564v1 | Wen-Yu Chang, Yun-Nung Chen | 2024-04-29T10:12:04Z | cs.CL, cs.AI | Recent research in dialogue systems and corpora has focused on two main
categories: task-oriented (TOD) and open-domain (chit-chat) dialogues. TOD
systems help users accomplish specific tasks, while open-domain systems aim to
create engaging conversations. However, in real-world scenarios, user intents
are often revealed during interactions. A recent study introduced SalesBot,
which simulates dialogues transitioning from chit-chat to task-oriented
scenarios to train sales agents. Unfortunately, the initial data lacked smooth
transitions and coherent long-turn dialogues, resulting in poor naturalness in
sales-customer interactions. To address these issues, this paper presents
SalesBot 2.0, an improved dataset. It leverages commonsense knowledge from
large language models (LLMs) through strategic prompting. Additionally, we
introduce a novel model called SalesAgent, trained on salesperson's
interactions, using chain-of-thought (CoT) reasoning. This model excels in
transitioning topics, understanding user intents, and selecting appropriate
strategies. Experiments using diverse user simulations validate the
effectiveness of our method in controlling dialogue strategies in LLMs.
Furthermore, SalesBot 2.0 enhances coherence and reduces aggression,
facilitating better model learning for sales-customer interactions. | cs.CL | Injecting Salesperson's Dialogue Strategies in Large Language Models with Chain-of-Thought Reasoning |
2024-04-29T10:00:00Z | http://arxiv.org/pdf/2404.18557v1 | http://arxiv.org/abs/2404.18557v1 | Stefano Bannò, Hari Krishna Vydana, Kate M. Knill, Mark J. F. Gales | 2024-04-29T10:00:00Z | cs.CL | Automated essay scoring (AES) to evaluate second language (L2) proficiency
has been a firmly established technology used in educational contexts for
decades. Although holistic scoring has seen advancements in AES that match or
even exceed human performance, analytic scoring still encounters issues as it
inherits flaws and shortcomings from the human scoring process. The recent
introduction of large language models presents new opportunities for automating
the evaluation of specific aspects of L2 writing proficiency. In this paper, we
perform a series of experiments using GPT-4 in a zero-shot fashion on a
publicly available dataset annotated with holistic scores based on the Common
European Framework of Reference and aim to extract detailed information about
their underlying analytic components. We observe significant correlations
between the automatically predicted analytic scores and multiple features
associated with the individual proficiency components. | cs.CL | Can GPT-4 do L2 analytic assessment? |
2024-04-29T09:34:25Z | http://arxiv.org/pdf/2404.18543v1 | http://arxiv.org/abs/2404.18543v1 | Felix Drinkall, Eghbal Rahimikia, Janet B. Pierrehumbert, Stefan Zohren | 2024-04-29T09:34:25Z | cs.CL, cs.CE, cs.LG, I.2.1, I.2.7 | Large language models (LLMs) are often trained on extensive, temporally
indiscriminate text corpora, reflecting the lack of datasets with temporal
metadata. This approach is not aligned with the evolving nature of language.
Conventional methods for creating temporally adapted language models often
depend on further pre-training static models on time-specific data. This paper
presents a new approach: a series of point-in-time LLMs called Time Machine GPT
(TiMaGPT), specifically designed to be nonprognosticative. This ensures they
remain uninformed about future factual information and linguistic changes. This
strategy is beneficial for understanding language evolution and is of critical
importance when applying models in dynamic contexts, such as time-series
forecasting, where foresight of future information can prove problematic. We
provide access to both the models and training datasets. | cs.CL | Time Machine GPT |
2024-04-29T09:22:54Z | http://arxiv.org/pdf/2404.18534v2 | http://arxiv.org/abs/2404.18534v2 | Guoliang Dong, Haoyu Wang, Jun Sun, Xinyu Wang | 2024-05-10T07:09:02Z | cs.CL, cs.AI, cs.CR, cs.SE | By training on text in various languages, large language models (LLMs)
typically possess multilingual support and demonstrate remarkable capabilities
in solving tasks described in different languages. However, LLMs can exhibit
linguistic discrimination due to the uneven distribution of training data
across languages. That is, LLMs are hard to keep the consistency of responses
when faced with the same task but depicted in different languages.
In this study, we first explore the consistency in the LLMs' outputs
responding to queries in various languages from two aspects: safety and
quality. We conduct this analysis with two datasets (AdvBench and NQ) based on
four LLMs (Llama2-13b, Gemma-7b, GPT-3.5-turbo and Gemini-pro). The results
show that LLMs exhibit stronger human alignment capabilities with queries in
English, French, Russian, and Spanish (only 1.04\% of harmful queries
successfully jailbreak on average) compared to queries in Bengali, Georgian,
Nepali and Maithili (27.7\% of harmful queries jailbreak successfully on
average). Moreover, for queries in English, Danish, Czech and Slovenian, LLMs
tend to produce responses with a higher quality (with 0.1494 $F_1$ score on
average) compared to the other languages. Upon these findings, we propose
LDFighter, a similarity-based voting, to mitigate the linguistic discrimination
in LLMs. LDFighter ensures consistent service for different language speakers.
We evaluate LDFighter with both benign queries and harmful queries. The results
show that LDFighter not only significantly reduces the jailbreak success rate
but also improve the response quality on average, demonstrating its
effectiveness. | cs.CL | Evaluating and Mitigating Linguistic Discrimination in Large Language Models |
2024-04-29T09:19:05Z | http://arxiv.org/pdf/2404.18532v2 | http://arxiv.org/abs/2404.18532v2 | Dingjie Song, Shunian Chen, Guiming Hardy Chen, Fei Yu, Xiang Wan, Benyou Wang | 2024-05-15T05:43:30Z | cs.CL, cs.AI, cs.CV, cs.LG | Despite the advancements and impressive performance of Multimodal Large
Language Models (MLLMs) on benchmarks, their effectiveness in real-world,
long-context, and multi-image tasks is unclear due to the benchmarks' limited
scope. Existing benchmarks often focus on single-image and short-text samples,
and when assessing multi-image tasks, they either limit the image count or
focus on specific task (e.g time-series captioning), potentially obscuring the
performance challenges of MLLMs. To address these limitations, we introduce
MileBench, a pioneering benchmark designed to test the MultImodal Long-contExt
capabilities of MLLMs. This benchmark comprises not only multimodal long
contexts, but also multiple tasks requiring both comprehension and generation.
We establish two distinct evaluation sets, diagnostic and realistic, to
systematically assess MLLMs' long-context adaptation capacity and their ability
to complete tasks in long-context scenarios. Our experimental results, obtained
from testing 22 models, revealed that while the closed-source GPT-4o
outperforms others, most open-source MLLMs struggle in long-context situations.
Interestingly, the performance gap tends to widen with an increase in the
number of images. We strongly encourage an intensification of research efforts
towards enhancing MLLMs' long-context capabilities, especially in scenarios
involving multiple images. | cs.CL | MileBench: Benchmarking MLLMs in Long Context |
2024-04-29T09:08:22Z | http://arxiv.org/pdf/2405.01589v2 | http://arxiv.org/abs/2405.01589v2 | Jakub Pokrywka, Jeremi Kaczmarek, Edward Gorzelańczyk | 2024-05-09T10:21:57Z | cs.CL, cs.AI | Introduction: Recently, the effectiveness of Large Language Models (LLMs) has
increased rapidly, allowing them to be used in a great number of applications.
However, the risks posed by the generation of false information through LLMs
significantly limit their applications in sensitive areas such as healthcare,
highlighting the necessity for rigorous validations to determine their utility
and reliability. To date, no study has extensively compared the performance of
LLMs on Polish medical examinations across a broad spectrum of specialties on a
very large dataset. Objectives: This study evaluated the performance of three
Generative Pretrained Transformer (GPT) models on the Polish Board
Certification Exam (Pa\'nstwowy Egzamin Specjalizacyjny, PES) dataset, which
consists of 297 tests. Methods: We developed a software program to download and
process PES exams and tested the performance of GPT models using OpenAI
Application Programming Interface. Results: Our findings reveal that GPT-3.5
did not pass any of the analyzed exams. In contrast, the GPT-4 models
demonstrated the capability to pass the majority of the exams evaluated, with
the most recent model, gpt-4-0125, successfully passing 222 (75%) of them. The
performance of the GPT models varied significantly, displaying excellence in
exams related to certain specialties while completely failing others.
Conclusions: The significant progress and impressive performance of LLM models
hold great promise for the increased application of AI in the field of medicine
in Poland. For instance, this advancement could lead to the development of
AI-based medical assistants for healthcare professionals, enhancing the
efficiency and accuracy of medical services. | cs.CL | GPT-4 passes most of the 297 written Polish Board Certification Examinations |
2024-04-29T09:03:19Z | http://arxiv.org/pdf/2404.18518v1 | http://arxiv.org/abs/2404.18518v1 | Jiangfeng Liu, Ziyi Wang, Jing Xie, Lei Pei | 2024-04-29T09:03:19Z | cs.DL, cs.AI, cs.CL, cs.CY | Generative large-scale language models create the fifth paradigm of
scientific research, organically combine data science and computational
intelligence, transform the research paradigm of natural language processing
and multimodal information processing, promote the new trend of AI-enabled
social science research, and provide new ideas for digital humanities research
and application. This article profoundly explores the application of
large-scale language models in digital humanities research, revealing their
significant potential in ancient book protection, intelligent processing, and
academic innovation. The article first outlines the importance of ancient book
resources and the necessity of digital preservation, followed by a detailed
introduction to developing large-scale language models, such as ChatGPT, and
their applications in document management, content understanding, and
cross-cultural research. Through specific cases, the article demonstrates how
AI can assist in the organization, classification, and content generation of
ancient books. Then, it explores the prospects of AI applications in artistic
innovation and cultural heritage preservation. Finally, the article explores
the challenges and opportunities in the interaction of technology, information,
and society in the digital humanities triggered by AI technologies. | cs.DL | From ChatGPT, DALL-E 3 to Sora: How has Generative AI Changed Digital Humanities Research and Services? |
2024-04-29T08:52:52Z | http://arxiv.org/pdf/2404.18510v2 | http://arxiv.org/abs/2404.18510v2 | Dana Roemling, Yves Scherrer, Aleksandra Miletic | 2024-07-01T15:58:11Z | cs.CL | Forensic authorship profiling uses linguistic markers to infer
characteristics about an author of a text. This task is paralleled in dialect
classification, where a prediction is made about the linguistic variety of a
text based on the text itself. While there have been significant advances in
recent years in variety classification, forensic linguistics rarely relies on
these approaches due to their lack of transparency, among other reasons. In
this paper we therefore explore the explainability of machine learning
approaches considering the forensic context. We focus on variety classification
as a means of geolinguistic profiling of unknown texts based on social media
data from the German-speaking area. For this, we identify the lexical items
that are the most impactful for the variety classification. We find that the
extracted lexical features are indeed representative of their respective
varieties and note that the trained models also rely on place names for
classifications. | cs.CL | Explainability of machine learning approaches in forensic linguistics: a case study in geolinguistic authorship profiling |
2024-04-29T07:11:39Z | http://arxiv.org/pdf/2404.18470v1 | http://arxiv.org/abs/2404.18470v1 | Yupeng Cao, Zhi Chen, Qingyun Pei, Prashant Kumar, K. P. Subbalakshmi, Papa Momar Ndiaye | 2024-04-29T07:11:39Z | cs.CE, cs.AI, cs.CL, q-fin.RM, q-fin.TR | In the realm of financial analytics, leveraging unstructured data, such as
earnings conference calls (ECCs), to forecast stock performance is a critical
challenge that has attracted both academics and investors. While previous
studies have used deep learning-based models to obtain a general view of ECCs,
they often fail to capture detailed, complex information. Our study introduces
a novel framework: \textbf{ECC Analyzer}, combining Large Language Models
(LLMs) and multi-modal techniques to extract richer, more predictive insights.
The model begins by summarizing the transcript's structure and analyzing the
speakers' mode and confidence level by detecting variations in tone and pitch
for audio. This analysis helps investors form an overview perception of the
ECCs. Moreover, this model uses the Retrieval-Augmented Generation (RAG) based
methods to meticulously extract the focuses that have a significant impact on
stock performance from an expert's perspective, providing a more targeted
analysis. The model goes a step further by enriching these extracted focuses
with additional layers of analysis, such as sentiment and audio segment
features. By integrating these insights, the ECC Analyzer performs multi-task
predictions of stock performance, including volatility, value-at-risk (VaR),
and return for different intervals. The results show that our model outperforms
traditional analytic benchmarks, confirming the effectiveness of using advanced
LLM techniques in financial analytics. | cs.CE | ECC Analyzer: Extract Trading Signal from Earnings Conference Calls using Large Language Model for Stock Performance Prediction |
2024-04-29T07:07:58Z | http://arxiv.org/pdf/2404.18466v1 | http://arxiv.org/abs/2404.18466v1 | Tingfeng Hui, Zhenyu Zhang, Shuohuan Wang, Weiran Xu, Yu Sun, Hua Wu | 2024-04-29T07:07:58Z | cs.CL | Large language models (LLMs) with one or more fine-tuning phases have become
a necessary step to unlock various capabilities, enabling LLMs to follow
natural language instructions or align with human preferences. However, it
carries the risk of catastrophic forgetting during sequential training, the
parametric knowledge or the ability learned in previous stages may be
overwhelmed by incoming training data. In this paper, we find that by regularly
resetting partial parameters, LLMs can restore some of the original knowledge.
Inspired by this, we introduce Half Fine-Tuning (HFT) for LLMs, as a substitute
for full fine-tuning (FFT), to mitigate the forgetting issues, where half of
the parameters are selected to learn new tasks while the other half are frozen
to remain previous knowledge. We provide a feasibility analysis from the
perspective of optimization and interpret the parameter selection operation as
a regularization term. Without changing the model architecture, HFT could be
seamlessly integrated into existing fine-tuning frameworks. Extensive
experiments and analysis on supervised fine-tuning, direct preference
optimization, and continual learning consistently demonstrate the
effectiveness, robustness, and efficiency of HFT. Compared with FFT, HFT not
only significantly alleviates the forgetting problem, but also achieves the
best performance in a series of downstream benchmarks, with an approximately
30% reduction in training time. | cs.CL | HFT: Half Fine-Tuning for Large Language Models |
2024-04-29T07:03:23Z | http://arxiv.org/pdf/2404.18963v1 | http://arxiv.org/abs/2404.18963v1 | Venkatesh C, Harshit Oberoi, Anurag Kumar Pandey, Anil Goyal, Nikhil Sikka | 2024-04-29T07:03:23Z | cs.LG, cs.CL | In recent years, digital platform companies have faced increasing challenges
in managing customer complaints, driven by widespread consumer adoption. This
paper introduces an end-to-end pipeline, named RE-GrievanceAssist, designed
specifically for real estate customer complaint management. The pipeline
consists of three key components: i) response/no-response ML model using TF-IDF
vectorization and XGBoost classifier ; ii) user type classifier using fasttext
classifier; iii) issue/sub-issue classifier using TF-IDF vectorization and
XGBoost classifier. Finally, it has been deployed as a batch job in Databricks,
resulting in a remarkable 40% reduction in overall manual effort with monthly
cost reduction of Rs 1,50,000 since August 2023. | cs.LG | RE-GrievanceAssist: Enhancing Customer Experience through ML-Powered Complaint Management |
2024-04-29T06:42:27Z | http://arxiv.org/pdf/2404.18460v1 | http://arxiv.org/abs/2404.18460v1 | Utkarsh Agarwal, Kumar Tanmay, Aditi Khandelwal, Monojit Choudhury | 2024-04-29T06:42:27Z | cs.CL, cs.AI | Ethical reasoning is a crucial skill for Large Language Models (LLMs).
However, moral values are not universal, but rather influenced by language and
culture. This paper explores how three prominent LLMs -- GPT-4, ChatGPT, and
Llama2-70B-Chat -- perform ethical reasoning in different languages and if
their moral judgement depend on the language in which they are prompted. We
extend the study of ethical reasoning of LLMs by Rao et al. (2023) to a
multilingual setup following their framework of probing LLMs with ethical
dilemmas and policies from three branches of normative ethics: deontology,
virtue, and consequentialism. We experiment with six languages: English,
Spanish, Russian, Chinese, Hindi, and Swahili. We find that GPT-4 is the most
consistent and unbiased ethical reasoner across languages, while ChatGPT and
Llama2-70B-Chat show significant moral value bias when we move to languages
other than English. Interestingly, the nature of this bias significantly vary
across languages for all LLMs, including GPT-4. | cs.CL | Ethical Reasoning and Moral Value Alignment of LLMs Depend on the Language we Prompt them in |
2024-04-29T05:40:08Z | http://arxiv.org/pdf/2404.18443v1 | http://arxiv.org/abs/2404.18443v1 | Ran Xu, Wenqi Shi, Yue Yu, Yuchen Zhuang, Yanqiao Zhu, May D. Wang, Joyce C. Ho, Chao Zhang, Carl Yang | 2024-04-29T05:40:08Z | cs.CL, cs.AI, cs.IR, q-bio.QM | Developing effective biomedical retrieval models is important for excelling
at knowledge-intensive biomedical tasks but still challenging due to the
deficiency of sufficient publicly annotated biomedical data and computational
resources. We present BMRetriever, a series of dense retrievers for enhancing
biomedical retrieval via unsupervised pre-training on large biomedical corpora,
followed by instruction fine-tuning on a combination of labeled datasets and
synthetic pairs. Experiments on 5 biomedical tasks across 11 datasets verify
BMRetriever's efficacy on various biomedical applications. BMRetriever also
exhibits strong parameter efficiency, with the 410M variant outperforming
baselines up to 11.7 times larger, and the 2B variant matching the performance
of models with over 5B parameters. The training data and model checkpoints are
released at \url{https://huggingface.co/BMRetriever} to ensure transparency,
reproducibility, and application to new domains. | cs.CL | BMRetriever: Tuning Large Language Models as Better Biomedical Text Retrievers |
2024-04-29T04:11:28Z | http://arxiv.org/pdf/2404.18416v2 | http://arxiv.org/abs/2404.18416v2 | Khaled Saab, Tao Tu, Wei-Hung Weng, Ryutaro Tanno, David Stutz, Ellery Wulczyn, Fan Zhang, Tim Strother, Chunjong Park, Elahe Vedadi, Juanma Zambrano Chaves, Szu-Yeu Hu, Mike Schaekermann, Aishwarya Kamath, Yong Cheng, David G. T. Barrett, Cathy Cheung, Basil Mustafa, Anil Palepu, Daniel McDuff, Le Hou, Tomer Golany, Luyang Liu, Jean-baptiste Alayrac, Neil Houlsby, Nenad Tomasev, Jan Freyberg, Charles Lau, Jonas Kemp, Jeremy Lai, Shekoofeh Azizi, Kimberly Kanada, SiWai Man, Kavita Kulkarni, Ruoxi Sun, Siamak Shakeri, Luheng He, Ben Caine, Albert Webson, Natasha Latysheva, Melvin Johnson, Philip Mansfield, Jian Lu, Ehud Rivlin, Jesper Anderson, Bradley Green, Renee Wong, Jonathan Krause, Jonathon Shlens, Ewa Dominowska, S. M. Ali Eslami, Katherine Chou, Claire Cui, Oriol Vinyals, Koray Kavukcuoglu, James Manyika, Jeff Dean, Demis Hassabis, Yossi Matias, Dale Webster, Joelle Barral, Greg Corrado, Christopher Semturs, S. Sara Mahdavi, Juraj Gottweis, Alan Karthikesalingam, Vivek Natarajan | 2024-05-01T17:12:10Z | cs.AI, cs.CL, cs.CV, cs.LG | Excellence in a wide variety of medical applications poses considerable
challenges for AI, requiring advanced reasoning, access to up-to-date medical
knowledge and understanding of complex multimodal data. Gemini models, with
strong general capabilities in multimodal and long-context reasoning, offer
exciting possibilities in medicine. Building on these core strengths of Gemini,
we introduce Med-Gemini, a family of highly capable multimodal models that are
specialized in medicine with the ability to seamlessly use web search, and that
can be efficiently tailored to novel modalities using custom encoders. We
evaluate Med-Gemini on 14 medical benchmarks, establishing new state-of-the-art
(SoTA) performance on 10 of them, and surpass the GPT-4 model family on every
benchmark where a direct comparison is viable, often by a wide margin. On the
popular MedQA (USMLE) benchmark, our best-performing Med-Gemini model achieves
SoTA performance of 91.1% accuracy, using a novel uncertainty-guided search
strategy. On 7 multimodal benchmarks including NEJM Image Challenges and MMMU
(health & medicine), Med-Gemini improves over GPT-4V by an average relative
margin of 44.5%. We demonstrate the effectiveness of Med-Gemini's long-context
capabilities through SoTA performance on a needle-in-a-haystack retrieval task
from long de-identified health records and medical video question answering,
surpassing prior bespoke methods using only in-context learning. Finally,
Med-Gemini's performance suggests real-world utility by surpassing human
experts on tasks such as medical text summarization, alongside demonstrations
of promising potential for multimodal medical dialogue, medical research and
education. Taken together, our results offer compelling evidence for
Med-Gemini's potential, although further rigorous evaluation will be crucial
before real-world deployment in this safety-critical domain. | cs.AI | Capabilities of Gemini Models in Medicine |
2024-04-29T04:01:45Z | http://arxiv.org/pdf/2405.00732v1 | http://arxiv.org/abs/2405.00732v1 | Justin Zhao, Timothy Wang, Wael Abid, Geoffrey Angus, Arnav Garg, Jeffery Kinnison, Alex Sherstinsky, Piero Molino, Travis Addair, Devvret Rishi | 2024-04-29T04:01:45Z | cs.CL, cs.AI, cs.LG | Low Rank Adaptation (LoRA) has emerged as one of the most widely adopted
methods for Parameter Efficient Fine-Tuning (PEFT) of Large Language Models
(LLMs). LoRA reduces the number of trainable parameters and memory usage while
achieving comparable performance to full fine-tuning. We aim to assess the
viability of training and serving LLMs fine-tuned with LoRA in real-world
applications. First, we measure the quality of LLMs fine-tuned with quantized
low rank adapters across 10 base models and 31 tasks for a total of 310 models.
We find that 4-bit LoRA fine-tuned models outperform base models by 34 points
and GPT-4 by 10 points on average. Second, we investigate the most effective
base models for fine-tuning and assess the correlative and predictive
capacities of task complexity heuristics in forecasting the outcomes of
fine-tuning. Finally, we evaluate the latency and concurrency capabilities of
LoRAX, an open-source Multi-LoRA inference server that facilitates the
deployment of multiple LoRA fine-tuned models on a single GPU using shared base
model weights and dynamic adapter loading. LoRAX powers LoRA Land, a web
application that hosts 25 LoRA fine-tuned Mistral-7B LLMs on a single NVIDIA
A100 GPU with 80GB memory. LoRA Land highlights the quality and
cost-effectiveness of employing multiple specialized LLMs over a single,
general-purpose LLM. | cs.CL | LoRA Land: 310 Fine-tuned LLMs that Rival GPT-4, A Technical Report |
2024-04-29T03:58:12Z | http://arxiv.org/pdf/2404.18410v1 | http://arxiv.org/abs/2404.18410v1 | Bowen Xu, Shaoyu Wu, Kai Liu, Lulu Hu | 2024-04-29T03:58:12Z | cs.CL | With the proliferation of large language models (LLMs), the comprehensive
alignment of such models across multiple tasks has emerged as a critical area
of research. Existing alignment methodologies primarily address single task,
such as multi-turn dialogue, coding, mathematical problem-solving, and tool
usage. However, AI-driven products that leverage language models usually
necessitate a fusion of these abilities to function effectively in real-world
scenarios. Moreover, the considerable computational resources required for
proper alignment of LLMs underscore the need for a more robust, efficient, and
encompassing approach to multi-task alignment, ensuring improved generative
performance. In response to these challenges, we introduce a novel technique
termed Mixture-of-Instructions (MoI), which employs a strategy of instruction
concatenation combined with diverse system prompts to boost the alignment
efficiency of language models. We have also compiled a diverse set of seven
benchmark datasets to rigorously evaluate the alignment efficacy of the
MoI-enhanced language model. Our methodology was applied to the open-source
Qwen-7B-chat model, culminating in the development of Qwen-SFT-MoI. This
enhanced model demonstrates significant advancements in generative capabilities
across coding, mathematics, and tool use tasks. | cs.CL | Mixture-of-Instructions: Comprehensive Alignment of a Large Language Model through the Mixture of Diverse System Prompting Instructions |
2024-04-29T03:30:06Z | http://arxiv.org/pdf/2404.18400v2 | http://arxiv.org/abs/2404.18400v2 | Parshin Shojaee, Kazem Meidani, Shashank Gupta, Amir Barati Farimani, Chandan K Reddy | 2024-06-02T20:17:59Z | cs.LG, cs.AI, cs.CL, cs.NE | Mathematical equations have been unreasonably effective in describing complex
natural phenomena across various scientific disciplines. However, discovering
such insightful equations from data presents significant challenges due to the
necessity of navigating extremely high-dimensional combinatorial and nonlinear
hypothesis spaces. Traditional methods of equation discovery, commonly known as
symbolic regression, largely focus on extracting equations from data alone,
often neglecting the rich domain-specific prior knowledge that scientists
typically depend on. To bridge this gap, we introduce LLM-SR, a novel approach
that leverages the extensive scientific knowledge and robust code generation
capabilities of Large Language Models (LLMs) to discover scientific equations
from data in an efficient manner. Specifically, LLM-SR treats equations as
programs with mathematical operators and combines LLMs' scientific priors with
evolutionary search over equation programs. The LLM iteratively proposes new
equation skeleton hypotheses, drawing from its physical understanding, which
are then optimized against data to estimate skeleton parameters. We demonstrate
LLM-SR's effectiveness across three diverse scientific domains, where it
discovers physically accurate equations that provide significantly better fits
to in-domain and out-of-domain data compared to the well-established symbolic
regression baselines. Incorporating scientific prior knowledge also enables
LLM-SR to search the equation space more efficiently than baselines. Code is
available at: https://github.com/deep-symbolic-mathematics/LLM-SR | cs.LG | LLM-SR: Scientific Equation Discovery via Programming with Large Language Models |
2024-04-29T03:19:39Z | http://arxiv.org/pdf/2404.18398v1 | http://arxiv.org/abs/2404.18398v1 | Xiang Li, Zhi-Qi Cheng, Jun-Yan He, Xiaojiang Peng, Alexander G. Hauptmann | 2024-04-29T03:19:39Z | cs.CL, cs.MM | Emotional Text-to-Speech (E-TTS) synthesis has gained significant attention
in recent years due to its potential to enhance human-computer interaction.
However, current E-TTS approaches often struggle to capture the complexity of
human emotions, primarily relying on oversimplified emotional labels or
single-modality inputs. To address these limitations, we propose the Multimodal
Emotional Text-to-Speech System (MM-TTS), a unified framework that leverages
emotional cues from multiple modalities to generate highly expressive and
emotionally resonant speech. MM-TTS consists of two key components: (1) the
Emotion Prompt Alignment Module (EP-Align), which employs contrastive learning
to align emotional features across text, audio, and visual modalities, ensuring
a coherent fusion of multimodal information; and (2) the Emotion
Embedding-Induced TTS (EMI-TTS), which integrates the aligned emotional
embeddings with state-of-the-art TTS models to synthesize speech that
accurately reflects the intended emotions. Extensive evaluations across diverse
datasets demonstrate the superior performance of MM-TTS compared to traditional
E-TTS models. Objective metrics, including Word Error Rate (WER) and Character
Error Rate (CER), show significant improvements on ESD dataset, with MM-TTS
achieving scores of 7.35% and 3.07%, respectively. Subjective assessments
further validate that MM-TTS generates speech with emotional fidelity and
naturalness comparable to human speech. Our code and pre-trained models are
publicly available at https://anonymous.4open.science/r/MMTTS-D214 | cs.CL | MM-TTS: A Unified Framework for Multimodal, Prompt-Induced Emotional Text-to-Speech Synthesis |
2024-04-29T02:43:23Z | http://arxiv.org/pdf/2404.18384v1 | http://arxiv.org/abs/2404.18384v1 | Jordan Meadows, Tamsin James, Andre Freitas | 2024-04-29T02:43:23Z | cs.CL | Language models can hallucinate when performing complex and detailed
mathematical reasoning. Physics provides a rich domain for assessing
mathematical reasoning capabilities where physical context imbues the use of
symbols which needs to satisfy complex semantics (\textit{e.g.,} units,
tensorial order), leading to instances where inference may be algebraically
coherent, yet unphysical. In this work, we assess the ability of Language
Models (LMs) to perform fine-grained mathematical and physical reasoning using
a curated dataset encompassing multiple notations and Physics subdomains. We
improve zero-shot scores using synthetic in-context examples, and demonstrate
non-linear degradation of derivation quality with perturbation strength via the
progressive omission of supporting premises. We find that the models'
mathematical reasoning is not physics-informed in this setting, where physical
context is predominantly ignored in favour of reverse-engineering solutions. | cs.CL | Exploring the Limits of Fine-grained LLM-based Physics Inference via Premise Removal Interventions |
2024-04-29T02:26:15Z | http://arxiv.org/pdf/2405.01588v1 | http://arxiv.org/abs/2405.01588v1 | Yongjin Yang, Sihyeon Kim, SangMook Kim, Gyubok Lee, Se-Young Yun, Edward Choi | 2024-04-29T02:26:15Z | cs.CL, cs.AI | Incorporating unanswerable questions into EHR QA systems is crucial for
testing the trustworthiness of a system, as providing non-existent responses
can mislead doctors in their diagnoses. The EHRSQL dataset stands out as a
promising benchmark because it is the only dataset that incorporates
unanswerable questions in the EHR QA system alongside practical questions.
However, in this work, we identify a data bias in these unanswerable questions;
they can often be discerned simply by filtering with specific N-gram patterns.
Such biases jeopardize the authenticity and reliability of QA system
evaluations. To tackle this problem, we propose a simple debiasing method of
adjusting the split between the validation and test sets to neutralize the
undue influence of N-gram filtering. By experimenting on the MIMIC-III dataset,
we demonstrate both the existing data bias in EHRSQL and the effectiveness of
our data split strategy in mitigating this bias. | cs.CL | Towards Unbiased Evaluation of Detecting Unanswerable Questions in EHRSQL |
2024-04-29T02:17:31Z | http://arxiv.org/pdf/2404.18371v1 | http://arxiv.org/abs/2404.18371v1 | Tomoki Fukuma, Koki Noda, Toshihide Ubukata Kousuke Hoso, Yoshiharu Ichikawa, Kyosuke Kambe, Yu Masubuch, Fujio Toriumi | 2024-04-29T02:17:31Z | cs.CL | The proliferation of social media has led to information overload and
increased interest in opinion mining. We propose "Question-Answering Network
Analysis" (QANA), a novel opinion mining framework that utilizes Large Language
Models (LLMs) to generate questions from users' comments, constructs a
bipartite graph based on the comments' answerability to the questions, and
applies centrality measures to examine the importance of opinions. We
investigate the impact of question generation styles, LLM selections, and the
choice of embedding model on the quality of the constructed QA networks by
comparing them with annotated Key Point Analysis datasets. QANA achieves
comparable performance to previous state-of-the-art supervised models in a
zero-shot manner for Key Point Matching task, also reducing the computational
cost from quadratic to linear. For Key Point Generation, questions with high
PageRank or degree centrality align well with manually annotated key points.
Notably, QANA enables analysts to assess the importance of key points from
various aspects according to their selection of centrality measure. QANA's
primary contribution lies in its flexibility to extract key points from a wide
range of perspectives, which enhances the quality and impartiality of opinion
mining. | cs.CL | QANA: LLM-based Question Generation and Network Analysis for Zero-shot Key Point Analysis and Beyond |
2024-04-29T01:49:07Z | http://arxiv.org/pdf/2404.18359v1 | http://arxiv.org/abs/2404.18359v1 | Wei Li, Ren Ma, Jiang Wu, Chenya Gu, Jiahui Peng, Jinyang Len, Songyang Zhang, Hang Yan, Dahua Lin, Conghui He | 2024-04-29T01:49:07Z | cs.CL, cs.AI | In the burgeoning field of large language models (LLMs), the assessment of
fundamental knowledge remains a critical challenge, particularly for models
tailored to Chinese language and culture. This paper introduces FoundaBench, a
pioneering benchmark designed to rigorously evaluate the fundamental knowledge
capabilities of Chinese LLMs. FoundaBench encompasses a diverse array of 3354
multiple-choice questions across common sense and K-12 educational subjects,
meticulously curated to reflect the breadth and depth of everyday and academic
knowledge. We present an extensive evaluation of 12 state-of-the-art LLMs using
FoundaBench, employing both traditional assessment methods and our CircularEval
protocol to mitigate potential biases in model responses. Our results highlight
the superior performance of models pre-trained on Chinese corpora, and reveal a
significant disparity between models' reasoning and memory recall capabilities.
The insights gleaned from FoundaBench evaluations set a new standard for
understanding the fundamental knowledge of LLMs, providing a robust framework
for future advancements in the field. | cs.CL | FoundaBench: Evaluating Chinese Fundamental Knowledge Capabilities of Large Language Models |
2024-04-28T20:44:53Z | http://arxiv.org/pdf/2404.18311v4 | http://arxiv.org/abs/2404.18311v4 | Mladjan Jovanovic, Peter Voss | 2024-05-05T08:46:32Z | cs.LG, cs.AI, cs.CL | Incremental learning is the ability of systems to acquire knowledge over
time, enabling their adaptation and generalization to novel tasks. It is a
critical ability for intelligent, real-world systems, especially when data
changes frequently or is limited. This review provides a comprehensive analysis
of incremental learning in Large Language Models. It synthesizes the
state-of-the-art incremental learning paradigms, including continual learning,
meta-learning, parameter-efficient learning, and mixture-of-experts learning.
We demonstrate their utility for incremental learning by describing specific
achievements from these related topics and their critical factors. An important
finding is that many of these approaches do not update the core model, and none
of them update incrementally in real-time. The paper highlights current
problems and challenges for future research in the field. By consolidating the
latest relevant research developments, this review offers a comprehensive
understanding of incremental learning and its implications for designing and
developing LLM-based learning systems. | cs.LG | Towards Incremental Learning in Large Language Models: A Critical Review |
2024-04-28T19:24:28Z | http://arxiv.org/pdf/2404.18286v2 | http://arxiv.org/abs/2404.18286v2 | David Ifeoluwa Adelani, A. Seza Doğruöz, André Coneglian, Atul Kr. Ojha | 2024-04-30T10:00:44Z | cs.CL | Large Language Models are transforming NLP for a variety of tasks. However,
how LLMs perform NLP tasks for low-resource languages (LRLs) is less explored.
In line with the goals of the AmericasNLP workshop, we focus on 12 LRLs from
Brazil, 2 LRLs from Africa and 2 high-resource languages (HRLs) (e.g., English
and Brazilian Portuguese). Our results indicate that the LLMs perform worse for
the part of speech (POS) labeling of LRLs in comparison to HRLs. We explain the
reasons behind this failure and provide an error analysis through examples
observed in our data set. | cs.CL | Comparing LLM prompting with Cross-lingual transfer performance on Indigenous and Low-resource Brazilian Languages |
2024-04-28T19:11:08Z | http://arxiv.org/pdf/2405.01587v1 | http://arxiv.org/abs/2405.01587v1 | Nidhi Kamal, Saurabh Yadav, Jorawar Singh, Aditi Avasthi | 2024-04-28T19:11:08Z | cs.CL, cs.AI, cs.CV, cs.LG | Providing fast and accurate resolution to the student's query is an essential
solution provided by Edtech organizations. This is generally provided with a
chat-bot like interface to enable students to ask their doubts easily. One
preferred format for student queries is images, as it allows students to
capture and post questions without typing complex equations and information.
However, this format also presents difficulties, as images may contain multiple
questions or textual noise that lowers the accuracy of existing single-query
answering solutions. In this paper, we propose a method for extracting
questions from text or images using a BERT-based deep learning model and
compare it to the other rule-based and layout-based methods. Our method aims to
improve the accuracy and efficiency of student query resolution in Edtech
organizations. | cs.CL | Improve Academic Query Resolution through BERT-based Question Extraction from Images |
2024-04-28T18:47:14Z | http://arxiv.org/pdf/2404.18276v1 | http://arxiv.org/abs/2404.18276v1 | Malur Narayan, John Pasmore, Elton Sampaio, Vijay Raghavan, Gabriella Waters | 2024-04-28T18:47:14Z | cs.CL, cs.AI, D.1; I.2 | The burgeoning influence of Large Language Models (LLMs) in shaping public
discourse and decision-making underscores the imperative to address inherent
biases within these AI systems. In the wake of AI's expansive integration
across sectors, addressing racial bias in LLMs has never been more critical.
This paper introduces a novel framework called Comprehensive Bias
Neutralization Framework (CBNF) which embodies an innovative approach to
quantifying and mitigating biases within LLMs. Our framework combines the Large
Language Model Bias Index (LLMBI) [Oketunji, A., Anas, M., Saina, D., (2023)]
and Bias removaL with No Demographics (BLIND) [Orgad, H., Belinkov, Y. (2023)]
methodologies to create a new metric called Bias Intelligence Quotient
(BiQ)which detects, measures, and mitigates racial bias in LLMs without
reliance on demographic annotations.
By introducing a new metric called BiQ that enhances LLMBI with additional
fairness metrics, CBNF offers a multi-dimensional metric for bias assessment,
underscoring the necessity of a nuanced approach to fairness in AI [Mehrabi et
al., 2021]. This paper presents a detailed analysis of Latimer AI (a language
model incrementally trained on black history and culture) in comparison to
ChatGPT 3.5, illustrating Latimer AI's efficacy in detecting racial, cultural,
and gender biases through targeted training and refined bias mitigation
strategies [Latimer & Bender, 2023]. | cs.CL | Bias Neutralization Framework: Measuring Fairness in Large Language Models with Bias Intelligence Quotient (BiQ) |
2024-04-28T18:36:59Z | http://arxiv.org/pdf/2404.18271v1 | http://arxiv.org/abs/2404.18271v1 | Qi Zhu, Da Zheng, Xiang Song, Shichang Zhang, Bowen Jin, Yizhou Sun, George Karypis | 2024-04-28T18:36:59Z | cs.CL, cs.LG | Text-rich graphs, which exhibit rich textual information on nodes and edges,
are prevalent across a wide range of real-world business applications. Large
Language Models (LLMs) have demonstrated remarkable abilities in understanding
text, which also introduced the potential for more expressive modeling in
text-rich graphs. Despite these capabilities, efficiently applying LLMs to
representation learning on graphs presents significant challenges. Recently,
parameter-efficient fine-tuning methods for LLMs have enabled efficient new
task generalization with minimal time and memory consumption. Inspired by this,
we introduce Graph-aware Parameter-Efficient Fine-Tuning - GPEFT, a novel
approach for efficient graph representation learning with LLMs on text-rich
graphs. Specifically, we utilize a graph neural network (GNN) to encode
structural information from neighboring nodes into a graph prompt. This prompt
is then inserted at the beginning of the text sequence. To improve the quality
of graph prompts, we pre-trained the GNN to assist the frozen LLM in predicting
the next token in the node text. Compared with existing joint GNN and LMs, our
method directly generate the node embeddings from large language models with an
affordable fine-tuning cost. We validate our approach through comprehensive
experiments conducted on 8 different text-rich graphs, observing an average
improvement of 2% in hit@1 and Mean Reciprocal Rank (MRR) in link prediction
evaluations. Our results demonstrate the efficacy and efficiency of our model,
showing that it can be smoothly integrated with various large language models,
including OPT, LLaMA and Falcon. | cs.CL | Parameter-Efficient Tuning Large Language Models for Graph Representation Learning |
2024-04-28T18:21:31Z | http://arxiv.org/pdf/2405.15784v1 | http://arxiv.org/abs/2405.15784v1 | Yizhou Chi, Jessy Lin, Kevin Lin, Dan Klein | 2024-04-28T18:21:31Z | cs.IR, cs.AI, cs.CL | Users often make ambiguous requests that require clarification. We study the
problem of asking clarification questions in an information retrieval setting,
where systems often face ambiguous search queries and it is challenging to turn
the uncertainty in the retrieval model into a natural language question. We
present CLARINET, a system that asks informative clarification questions by
choosing questions whose answers would maximize certainty in the correct
candidate. Our approach works by augmenting a large language model (LLM) to
condition on a retrieval distribution, finetuning end-to-end to generate the
question that would have maximized the rank of the true candidate at each turn.
When evaluated on a real-world retrieval dataset of users searching for books,
our system outperforms traditional heuristics such as information gain on
retrieval success by 17% and vanilla-prompted LLMs by 39% relative. | cs.IR | CLARINET: Augmenting Language Models to Ask Clarification Questions for Retrieval |
2024-04-28T18:07:13Z | http://arxiv.org/pdf/2404.18264v1 | http://arxiv.org/abs/2404.18264v1 | Pin-Jie Lin, Merel Scholman, Muhammed Saeed, Vera Demberg | 2024-04-28T18:07:13Z | cs.CL, cs.AI | Nigerian Pidgin is an English-derived contact language and is traditionally
an oral language, spoken by approximately 100 million people. No orthographic
standard has yet been adopted, and thus the few available Pidgin datasets that
exist are characterised by noise in the form of orthographic variations. This
contributes to under-performance of models in critical NLP tasks. The current
work is the first to describe various types of orthographic variations commonly
found in Nigerian Pidgin texts, and model this orthographic variation. The
variations identified in the dataset form the basis of a phonetic-theoretic
framework for word editing, which is used to generate orthographic variations
to augment training data. We test the effect of this data augmentation on two
critical NLP tasks: machine translation and sentiment analysis. The proposed
variation generation framework augments the training data with new orthographic
variants which are relevant for the test set but did not occur in the training
set originally. Our results demonstrate the positive effect of augmenting the
training data with a combination of real texts from other corpora as well as
synthesized orthographic variation, resulting in performance improvements of
2.1 points in sentiment analysis and 1.4 BLEU points in translation to English. | cs.CL | Modeling Orthographic Variation Improves NLP Performance for Nigerian Pidgin |
2024-04-28T17:43:24Z | http://arxiv.org/pdf/2404.18257v1 | http://arxiv.org/abs/2404.18257v1 | Nilo Pedrazzini | 2024-04-28T17:43:24Z | cs.CL, cs.IR, 68T50, 68U15, 68T35, (Primary), 86A32, 15A03 (Secondary), I.2.7 | Languages can encode temporal subordination lexically, via subordinating
conjunctions, and morphologically, by marking the relation on the predicate.
Systematic cross-linguistic variation among the former can be studied using
well-established token-based typological approaches to token-aligned parallel
corpora. Variation among different morphological means is instead much harder
to tackle and therefore more poorly understood, despite being predominant in
several language groups. This paper explores variation in the expression of
generic temporal subordination ('when'-clauses) among the languages of Latin
America and the Caribbean, where morphological marking is particularly common.
It presents probabilistic semantic maps computed on the basis of the languages
of the region, thus avoiding bias towards the many world's languages that
exclusively use lexified connectors, incorporating associations between
character $n$-grams and English $when$. The approach allows capturing
morphological clause-linkage devices in addition to lexified connectors, paving
the way for larger-scale, strategy-agnostic analyses of typological variation
in temporal subordination. | cs.CL | Mapping 'when'-clauses in Latin American and Caribbean languages: an experiment in subtoken-based typology |
2024-04-28T17:36:43Z | http://arxiv.org/pdf/2404.18255v5 | http://arxiv.org/abs/2404.18255v5 | Zilong Bai, Ruiji Zhang, Linqing Chen, Qijun Cai, Yuan Zhong, Cong Wang, Yan Fang, Jie Fang, Jing Sun, Weikuan Wang, Lizhi Zhou, Haoran Hua, Tian Qiu, Chaochao Wang, Cheng Sun, Jianping Lu, Yixin Wang, Yubin Xia, Meng Hu, Haowen Liu, Peng Xu, Licong Xu, Fu Bian, Xiaolong Gu, Lisha Zhang, Weilei Wang, Changyang Tu | 2024-06-05T03:02:48Z | cs.CL, cs.AI, I.2.7 | In recent years, large language models(LLMs) have attracted significant
attention due to their exceptional performance across a multitude of natural
language process tasks, and have been widely applied in various fields.
However, the application of large language models in the Intellectual Property
(IP) domain is challenging due to the strong need for specialized knowledge,
privacy protection, processing of extremely long text in this field. In this
technical report, we present for the first time a low-cost, standardized
procedure for training IP-oriented LLMs, meeting the unique requirements of the
IP domain. Using this standard process, we have trained the PatentGPT series
models based on open-source pretrained models. By evaluating them on the
open-source IP-oriented benchmark MOZIP, our domain-specific LLMs outperforms
GPT-4, indicating the effectiveness of the proposed training procedure and the
expertise of the PatentGPT models in the IP domain. Remarkably, our model
surpassed GPT-4 on the 2019 China Patent Agent Qualification Examination,
scoring 65 and matching human expert levels. Additionally, the PatentGPT model,
which utilizes the SMoE architecture, achieves performance comparable to that
of GPT-4 in the IP domain and demonstrates a better cost-performance ratio on
long-text tasks, potentially serving as an alternative to GPT-4 within the IP
domain. | cs.CL | PatentGPT: A Large Language Model for Intellectual Property |
2024-04-28T17:15:07Z | http://arxiv.org/pdf/2405.01586v1 | http://arxiv.org/abs/2405.01586v1 | Tohida Rehman, Raghubir Bose, Samiran Chattopadhyay, Debarshi Kumar Sanyal | 2024-04-28T17:15:07Z | cs.CL | Financial sentiment analysis allows financial institutions like Banks and
Insurance Companies to better manage the credit scoring of their customers in a
better way. Financial domain uses specialized mechanisms which makes sentiment
analysis difficult. In this paper, we propose a pre-trained language model
which can help to solve this problem with fewer labelled data. We extend on the
principles of Transfer learning and Transformation architecture principles and
also take into consideration recent outbreak of pandemics like COVID. We apply
the sentiment analysis to two different sets of data. We also take smaller
training set and fine tune the same as part of the model. | cs.CL | Transfer Learning and Transformer Architecture for Financial Sentiment Analysis |
2024-04-28T16:50:12Z | http://arxiv.org/pdf/2404.18243v1 | http://arxiv.org/abs/2404.18243v1 | Zhili Cheng, Zhitong Wang, Jinyi Hu, Shengding Hu, An Liu, Yuge Tu, Pengkai Li, Lei Shi, Zhiyuan Liu, Maosong Sun | 2024-04-28T16:50:12Z | cs.CL | Despite advancements in Large Language Models (LLMs) and Large Multimodal
Models (LMMs), their integration into language-grounded, human-like embodied
agents remains incomplete, hindering complex real-life task performance in
physical environments. Existing integrations often feature limited open
sourcing, challenging collective progress in this field. We introduce LEGENT,
an open, scalable platform for developing embodied agents using LLMs and LMMs.
LEGENT offers a dual approach: a rich, interactive 3D environment with
communicable and actionable agents, paired with a user-friendly interface, and
a sophisticated data generation pipeline utilizing advanced algorithms to
exploit supervision from simulated worlds at scale. In our experiments, an
embryonic vision-language-action model trained on LEGENT-generated data
surpasses GPT-4V in embodied tasks, showcasing promising generalization
capabilities. | cs.CL | LEGENT: Open Platform for Embodied Agents |
2024-04-28T16:31:32Z | http://arxiv.org/pdf/2404.18239v4 | http://arxiv.org/abs/2404.18239v4 | Jinghan Jia, Yihua Zhang, Yimeng Zhang, Jiancheng Liu, Bharat Runwal, James Diffenderfer, Bhavya Kailkhura, Sijia Liu | 2024-06-24T20:24:53Z | cs.LG, cs.CL | Large Language Models (LLMs) have highlighted the necessity of effective
unlearning mechanisms to comply with data regulations and ethical AI practices.
LLM unlearning aims at removing undesired data influences and associated model
capabilities without compromising utility beyond the scope of unlearning. While
interest in studying LLM unlearning is growing, the impact of the optimizer
choice for LLM unlearning remains unexplored. In this work, we shed light on
the significance of optimizer selection in LLM unlearning for the first time,
establishing a clear connection between second-order optimization and influence
unlearning (a classical approach using influence functions to update the model
for data influence removal). This insight propels us to develop a second-order
optimization-based LLM unlearning framework, termed Second-Order UnLearning
(SOUL), which extends the static, one-shot model update using influence
unlearning to a dynamic, iterative unlearning process. Our extensive
experiments show that SOUL consistently outperforms conventional first-order
methods across various unlearning tasks, models, and metrics, indicating that
second-order optimization offers an effective and broadly applicable solution
for LLM unlearning. Codes are available at https://github.com/OPTML-Group/SOUL. | cs.LG | SOUL: Unlocking the Power of Second-Order Optimization for LLM Unlearning |
2024-04-28T15:56:41Z | http://arxiv.org/pdf/2404.18231v1 | http://arxiv.org/abs/2404.18231v1 | Jiangjie Chen, Xintao Wang, Rui Xu, Siyu Yuan, Yikai Zhang, Wei Shi, Jian Xie, Shuang Li, Ruihan Yang, Tinghui Zhu, Aili Chen, Nianqi Li, Lida Chen, Caiyu Hu, Siye Wu, Scott Ren, Ziquan Fu, Yanghua Xiao | 2024-04-28T15:56:41Z | cs.CL, cs.AI | Recent advancements in large language models (LLMs) have significantly
boosted the rise of Role-Playing Language Agents (RPLAs), i.e., specialized AI
systems designed to simulate assigned personas. By harnessing multiple advanced
abilities of LLMs, including in-context learning, instruction following, and
social intelligence, RPLAs achieve a remarkable sense of human likeness and
vivid role-playing performance. RPLAs can mimic a wide range of personas,
ranging from historical figures and fictional characters to real-life
individuals. Consequently, they have catalyzed numerous AI applications, such
as emotional companions, interactive video games, personalized assistants and
copilots, and digital clones. In this paper, we conduct a comprehensive survey
of this field, illustrating the evolution and recent progress in RPLAs
integrating with cutting-edge LLM technologies. We categorize personas into
three types: 1) Demographic Persona, which leverages statistical stereotypes;
2) Character Persona, focused on well-established figures; and 3)
Individualized Persona, customized through ongoing user interactions for
personalized services. We begin by presenting a comprehensive overview of
current methodologies for RPLAs, followed by the details for each persona type,
covering corresponding data sourcing, agent construction, and evaluation.
Afterward, we discuss the fundamental risks, existing limitations, and future
prospects of RPLAs. Additionally, we provide a brief review of RPLAs in AI
applications, which reflects practical user demands that shape and drive RPLA
research. Through this work, we aim to establish a clear taxonomy of RPLA
research and applications, and facilitate future research in this critical and
ever-evolving field, and pave the way for a future where humans and RPLAs
coexist in harmony. | cs.CL | From Persona to Personalization: A Survey on Role-Playing Language Agents |
2024-04-28T15:44:57Z | http://arxiv.org/pdf/2404.18228v1 | http://arxiv.org/abs/2404.18228v1 | Sharayu Hiwarkhedkar, Saloni Mittal, Vidula Magdum, Omkar Dhekane, Raviraj Joshi, Geetanjali Kale, Arnav Ladkat | 2024-04-28T15:44:57Z | cs.CL, cs.LG | For green AI, it is crucial to measure and reduce the carbon footprint
emitted during the training of large language models. In NLP, performing
pre-training on Transformer models requires significant computational
resources. This pre-training involves using a large amount of text data to gain
prior knowledge for performing downstream tasks. Thus, it is important that we
select the correct data in the form of domain-specific data from this vast
corpus to achieve optimum results aligned with our domain-specific tasks. While
training on large unsupervised data is expensive, it can be optimized by
performing a data selection step before pretraining. Selecting important data
reduces the space overhead and the substantial amount of time required to
pre-train the model while maintaining constant accuracy. We investigate the
existing selection strategies and propose our own domain-adaptive data
selection method - TextGram - that effectively selects essential data from
large corpora. We compare and evaluate the results of finetuned models for text
classification task with and without data selection. We show that the proposed
strategy works better compared to other selection methods. | cs.CL | TextGram: Towards a better domain-adaptive pretraining |
2024-04-28T15:20:45Z | http://arxiv.org/pdf/2404.18216v1 | http://arxiv.org/abs/2404.18216v1 | Saloni Mittal, Vidula Magdum, Omkar Dhekane, Sharayu Hiwarkhedkar, Raviraj Joshi | 2024-04-28T15:20:45Z | cs.CL, cs.LG | The availability of text or topic classification datasets in the low-resource
Marathi language is limited, typically consisting of fewer than 4 target
labels, with some achieving nearly perfect accuracy. In this work, we introduce
L3Cube-MahaNews, a Marathi text classification corpus that focuses on News
headlines and articles. This corpus stands out as the largest supervised
Marathi Corpus, containing over 1.05L records classified into a diverse range
of 12 categories. To accommodate different document lengths, MahaNews comprises
three supervised datasets specifically designed for short text, long documents,
and medium paragraphs. The consistent labeling across these datasets
facilitates document length-based analysis. We provide detailed data statistics
and baseline results on these datasets using state-of-the-art pre-trained BERT
models. We conduct a comparative analysis between monolingual and multilingual
BERT models, including MahaBERT, IndicBERT, and MuRIL. The monolingual MahaBERT
model outperforms all others on every dataset. These resources also serve as
Marathi topic classification datasets or models and are publicly available at
https://github.com/l3cube-pune/MarathiNLP . | cs.CL | L3Cube-MahaNews: News-based Short Text and Long Document Classification Datasets in Marathi |
2024-04-28T14:58:55Z | http://arxiv.org/pdf/2405.01585v1 | http://arxiv.org/abs/2405.01585v1 | Sujit Khanna, Shishir Subedi | 2024-04-28T14:58:55Z | cs.AI, cs.CL, cs.IR | In recent times Large Language Models have exhibited tremendous capabilities,
especially in the areas of mathematics, code generation and general-purpose
reasoning. However for specialized domains especially in applications that
require parsing and analyzing large chunks of numeric or tabular data even
state-of-the-art (SOTA) models struggle. In this paper, we introduce a new
approach to solving domain-specific tabular data analysis tasks by presenting a
unique RAG workflow that mitigates the scalability issues of existing tabular
LLM solutions. Specifically, we present Tabular Embedding Model (TEM), a novel
approach to fine-tune embedding models for tabular Retrieval-Augmentation
Generation (RAG) applications. Embedding models form a crucial component in the
RAG workflow and even current SOTA embedding models struggle as they are
predominantly trained on textual datasets and thus underperform in scenarios
involving complex tabular data. The evaluation results showcase that our
approach not only outperforms current SOTA embedding models in this domain but
also does so with a notably smaller and more efficient model structure. | cs.AI | Tabular Embedding Model (TEM): Finetuning Embedding Models For Tabular RAG Applications |
2024-04-28T14:45:53Z | http://arxiv.org/pdf/2405.01509v1 | http://arxiv.org/abs/2405.01509v1 | Minhao Bai, Kaiyi Pang, Yongfeng Huang | 2024-04-28T14:45:53Z | cs.CR, cs.AI, cs.CL | In the rapidly evolving domain of artificial intelligence, safeguarding the
intellectual property of Large Language Models (LLMs) is increasingly crucial.
Current watermarking techniques against model extraction attacks, which rely on
signal insertion in model logits or post-processing of generated text, remain
largely heuristic. We propose a novel method for embedding learnable linguistic
watermarks in LLMs, aimed at tracing and preventing model extraction attacks.
Our approach subtly modifies the LLM's output distribution by introducing
controlled noise into token frequency distributions, embedding an statistically
identifiable controllable watermark.We leverage statistical hypothesis testing
and information theory, particularly focusing on Kullback-Leibler Divergence,
to differentiate between original and modified distributions effectively. Our
watermarking method strikes a delicate well balance between robustness and
output quality, maintaining low false positive/negative rates and preserving
the LLM's original performance. | cs.CR | Learnable Linguistic Watermarks for Tracing Model Extraction Attacks on Large Language Models |
2024-04-28T14:05:23Z | http://arxiv.org/pdf/2404.18191v2 | http://arxiv.org/abs/2404.18191v2 | Chen Cheng, Xinzhi Yu, Haodong Wen, Jingsong Sun, Guanzhang Yue, Yihao Zhang, Zeming Wei | 2024-05-01T09:15:16Z | cs.CL, cs.AI, cs.CR, cs.LG, math.OC | Recently, the mysterious In-Context Learning (ICL) ability exhibited by
Transformer architectures, especially in large language models (LLMs), has
sparked significant research interest. However, the resilience of Transformers'
in-context learning capabilities in the presence of noisy samples, prevalent in
both training corpora and prompt demonstrations, remains underexplored. In this
paper, inspired by prior research that studies ICL ability using simple
function classes, we take a closer look at this problem by investigating the
robustness of Transformers against noisy labels. Specifically, we first conduct
a thorough evaluation and analysis of the robustness of Transformers against
noisy labels during in-context learning and show that they exhibit notable
resilience against diverse types of noise in demonstration labels. Furthermore,
we delve deeper into this problem by exploring whether introducing noise into
the training set, akin to a form of data augmentation, enhances such robustness
during inference, and find that such noise can indeed improve the robustness of
ICL. Overall, our fruitful analysis and findings provide a comprehensive
understanding of the resilience of Transformer models against label noises
during ICL and provide valuable insights into the research on Transformers in
natural language processing. Our code is available at
https://github.com/InezYu0928/in-context-learning. | cs.CL | Exploring the Robustness of In-Context Learning with Noisy Labels |
2024-04-28T13:39:33Z | http://arxiv.org/pdf/2404.18185v1 | http://arxiv.org/abs/2404.18185v1 | Chuan Meng, Negar Arabzadeh, Arian Askari, Mohammad Aliannejadi, Maarten de Rijke | 2024-04-28T13:39:33Z | cs.IR, cs.AI, cs.CL, cs.LG, H.3.3 | We study ranked list truncation (RLT) from a novel "retrieve-then-re-rank"
perspective, where we optimize re-ranking by truncating the retrieved list
(i.e., trim re-ranking candidates). RLT is crucial for re-ranking as it can
improve re-ranking efficiency by sending variable-length candidate lists to a
re-ranker on a per-query basis. It also has the potential to improve re-ranking
effectiveness. Despite its importance, there is limited research into applying
RLT methods to this new perspective. To address this research gap, we reproduce
existing RLT methods in the context of re-ranking, especially newly emerged
large language model (LLM)-based re-ranking. In particular, we examine to what
extent established findings on RLT for retrieval are generalizable to the
"retrieve-then-re-rank" setup from three perspectives: (i) assessing RLT
methods in the context of LLM-based re-ranking with lexical first-stage
retrieval, (ii) investigating the impact of different types of first-stage
retrievers on RLT methods, and (iii) investigating the impact of different
types of re-rankers on RLT methods. We perform experiments on the TREC 2019 and
2020 deep learning tracks, investigating 8 RLT methods for pipelines involving
3 retrievers and 2 re-rankers. We reach new insights into RLT methods in the
context of re-ranking. | cs.IR | Ranked List Truncation for Large Language Model-based Re-Ranking |
2024-04-28T13:25:11Z | http://arxiv.org/pdf/2404.18180v1 | http://arxiv.org/abs/2404.18180v1 | Comfort Eseohen Ilevbare, Jesujoba O. Alabi, David Ifeoluwa Adelani, Firdous Damilola Bakare, Oluwatoyin Bunmi Abiola, Oluwaseyi Adesina Adeyemo | 2024-04-28T13:25:11Z | cs.CL | Nigerians have a notable online presence and actively discuss political and
topical matters. This was particularly evident throughout the 2023 general
election, where Twitter was used for campaigning, fact-checking and
verification, and even positive and negative discourse. However, little or none
has been done in the detection of abusive language and hate speech in Nigeria.
In this paper, we curated code-switched Twitter data directed at three
musketeers of the governorship election on the most populous and economically
vibrant state in Nigeria; Lagos state, with the view to detect offensive speech
in political discussions. We developed EkoHate -- an abusive language and hate
speech dataset for political discussions between the three candidates and their
followers using a binary (normal vs offensive) and fine-grained four-label
annotation scheme. We analysed our dataset and provided an empirical evaluation
of state-of-the-art methods across both supervised and cross-lingual transfer
learning settings. In the supervised setting, our evaluation results in both
binary and four-label annotation schemes show that we can achieve 95.1 and 70.3
F1 points respectively. Furthermore, we show that our dataset adequately
transfers very well to three publicly available offensive datasets (OLID,
HateUS2020, and FountaHate), generalizing to political discussions in other
regions like the US. | cs.CL | EkoHate: Abusive Language and Hate Speech Detection for Code-switched Political Discussions on Nigerian Twitter |
2024-04-28T12:11:34Z | http://arxiv.org/pdf/2404.18154v1 | http://arxiv.org/abs/2404.18154v1 | Paul Égré, Benjamin Spector | 2024-04-28T12:11:34Z | cs.CL, cs.GT, cs.IT, math.IT, 91A86, I.2.7 | Why is language vague? Vagueness may be explained and rationalized if it can
be shown that vague language is more useful to speaker and hearer than precise
language. In a well-known paper, Lipman proposes a game-theoretic account of
vagueness in terms of mixed strategy that leads to a puzzle: vagueness cannot
be strictly better than precision at equilibrium. More recently, \'Egr\'e,
Spector, Mortier and Verheyen have put forward a Bayesian account of vagueness
establishing that using vague words can be strictly more informative than using
precise words. This paper proposes to compare both results and to explain why
they are not in contradiction. Lipman's definition of vagueness relies
exclusively on a property of signaling strategies, without making any
assumptions about the lexicon, whereas \'Egr\'e et al.'s involves a layer of
semantic content. We argue that the semantic account of vagueness is needed,
and more adequate and explanatory of vagueness. | cs.CL | Explaining vague language |
2024-04-28T10:11:52Z | http://arxiv.org/pdf/2405.01584v1 | http://arxiv.org/abs/2405.01584v1 | Li Wan, Tansu Alpcan, Margreta Kuijper, Emanuele Viterbo | 2024-04-28T10:11:52Z | cs.CL, cs.LG, eess.SP | We propose a novel, lightweight supervised dictionary learning framework for
text classification based on data compression and representation. This
two-phase algorithm initially employs the Lempel-Ziv-Welch (LZW) algorithm to
construct a dictionary from text datasets, focusing on the conceptual
significance of dictionary elements. Subsequently, dictionaries are refined
considering label data, optimizing dictionary atoms to enhance discriminative
power based on mutual information and class distribution. This process
generates discriminative numerical representations, facilitating the training
of simple classifiers such as SVMs and neural networks. We evaluate our
algorithm's information-theoretic performance using information bottleneck
principles and introduce the information plane area rank (IPAR) as a novel
metric to quantify the information-theoretic performance. Tested on six
benchmark text datasets, our algorithm competes closely with top models,
especially in limited-vocabulary contexts, using significantly fewer
parameters. \review{Our algorithm closely matches top-performing models,
deviating by only ~2\% on limited-vocabulary datasets, using just 10\% of their
parameters. However, it falls short on diverse-vocabulary datasets, likely due
to the LZW algorithm's constraints with low-repetition data. This contrast
highlights its efficiency and limitations across different dataset types. | cs.CL | Lightweight Conceptual Dictionary Learning for Text Classification Using Information Compression |
2024-04-28T10:02:28Z | http://arxiv.org/pdf/2404.18130v1 | http://arxiv.org/abs/2404.18130v1 | Hanmeng Liu, Zhiyang Teng, Chaoli Zhang, Yue Zhang | 2024-04-28T10:02:28Z | cs.AI, cs.CL | Chain-of-Thought (CoT) prompting has emerged as a pivotal technique for
augmenting the inferential capabilities of language models during reasoning
tasks. Despite its advancements, CoT often grapples with challenges in
validating reasoning validity and ensuring informativeness. Addressing these
limitations, this paper introduces the Logic Agent (LA), an agent-based
framework aimed at enhancing the validity of reasoning processes in Large
Language Models (LLMs) through strategic logic rule invocation. Unlike
conventional approaches, LA transforms LLMs into logic agents that dynamically
apply propositional logic rules, initiating the reasoning process by converting
natural language inputs into structured logic forms. The logic agent leverages
a comprehensive set of predefined functions to systematically navigate the
reasoning process. This methodology not only promotes the structured and
coherent generation of reasoning constructs but also significantly improves
their interpretability and logical coherence. Through extensive
experimentation, we demonstrate LA's capacity to scale effectively across
various model sizes, markedly improving the precision of complex reasoning
across diverse tasks. | cs.AI | Logic Agent: Enhancing Validity with Logic Rule Invocation |
2024-04-28T06:50:55Z | http://arxiv.org/pdf/2404.18094v1 | http://arxiv.org/abs/2404.18094v1 | Wenbin Wang, Yang Song, Sanjay Jha | 2024-04-28T06:50:55Z | cs.SD, cs.AI, cs.CL, eess.AS | Conventional text-to-speech (TTS) research has predominantly focused on
enhancing the quality of synthesized speech for speakers in the training
dataset. The challenge of synthesizing lifelike speech for unseen,
out-of-dataset speakers, especially those with limited reference data, remains
a significant and unresolved problem. While zero-shot or few-shot
speaker-adaptive TTS approaches have been explored, they have many limitations.
Zero-shot approaches tend to suffer from insufficient generalization
performance to reproduce the voice of speakers with heavy accents. While
few-shot methods can reproduce highly varying accents, they bring a significant
storage burden and the risk of overfitting and catastrophic forgetting. In
addition, prior approaches only provide either zero-shot or few-shot
adaptation, constraining their utility across varied real-world scenarios with
different demands. Besides, most current evaluations of speaker-adaptive TTS
are conducted only on datasets of native speakers, inadvertently neglecting a
vast portion of non-native speakers with diverse accents. Our proposed
framework unifies both zero-shot and few-shot speaker adaptation strategies,
which we term as "instant" and "fine-grained" adaptations based on their
merits. To alleviate the insufficient generalization performance observed in
zero-shot speaker adaptation, we designed two innovative discriminators and
introduced a memory mechanism for the speech decoder. To prevent catastrophic
forgetting and reduce storage implications for few-shot speaker adaptation, we
designed two adapters and a unique adaptation procedure. | cs.SD | USAT: A Universal Speaker-Adaptive Text-to-Speech Approach |
Subsets and Splits
No saved queries yet
Save your SQL queries to embed, download, and access them later. Queries will appear here once saved.