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
|
|---|---|---|---|---|---|---|---|---|
2024-03-21T17:06:17Z | http://arxiv.org/pdf/2403.14562v1 | http://arxiv.org/abs/2403.14562v1 | Maxime Peyrard, Martin Josifoski, Robert West | 2024-03-21T17:06:17Z | cs.CL, cs.AI, cs.HC, cs.MA | Recent work demonstrated great promise in the idea of orchestrating
collaborations between LLMs, human input, and various tools to address the
inherent limitations of LLMs. We propose a novel perspective called semantic
decoding, which frames these collaborative processes as optimization procedures
in semantic space. Specifically, we conceptualize LLMs as semantic processors
that manipulate meaningful pieces of information that we call semantic tokens
(known thoughts). LLMs are among a large pool of other semantic processors,
including humans and tools, such as search engines or code executors.
Collectively, semantic processors engage in dynamic exchanges of semantic
tokens to progressively construct high-utility outputs. We refer to these
orchestrated interactions among semantic processors, optimizing and searching
in semantic space, as semantic decoding algorithms. This concept draws a direct
parallel to the well-studied problem of syntactic decoding, which involves
crafting algorithms to best exploit auto-regressive language models for
extracting high-utility sequences of syntactic tokens. By focusing on the
semantic level and disregarding syntactic details, we gain a fresh perspective
on the engineering of AI systems, enabling us to imagine systems with much
greater complexity and capabilities. In this position paper, we formalize the
transition from syntactic to semantic tokens as well as the analogy between
syntactic and semantic decoding. Subsequently, we explore the possibilities of
optimizing within the space of semantic tokens via semantic decoding
algorithms. We conclude with a list of research opportunities and questions
arising from this fresh perspective. The semantic decoding perspective offers a
powerful abstraction for search and optimization directly in the space of
meaningful concepts, with semantic tokens as the fundamental units of a new
type of computation. | cs.CL | The Era of Semantic Decoding |
2024-03-21T16:52:01Z | http://arxiv.org/pdf/2403.14551v1 | http://arxiv.org/abs/2403.14551v1 | Chengxu Zhuang, Evelina Fedorenko, Jacob Andreas | 2024-03-21T16:52:01Z | cs.CL, cs.AI, cs.LG | Today's most accurate language models are trained on orders of magnitude more
language data than human language learners receive - but with no supervision
from other sensory modalities that play a crucial role in human learning. Can
we make LMs' representations and predictions more accurate (and more
human-like) with more ecologically plausible supervision? This paper describes
LexiContrastive Grounding (LCG), a grounded language learning procedure that
leverages visual supervision to improve textual representations.
LexiContrastive Grounding combines a next token prediction strategy with a
contrastive visual grounding objective, focusing on early-layer representations
that encode lexical information. Across multiple word-learning and
sentence-understanding benchmarks, LexiContrastive Grounding not only
outperforms standard language-only models in learning efficiency, but also
improves upon vision-and-language learning procedures including CLIP, GIT,
Flamingo, and Vokenization. Moreover, LexiContrastive Grounding improves
perplexity by around 5% on multiple language modeling tasks. This work
underscores the potential of incorporating visual grounding into language
models, aligning more closely with the multimodal nature of human language
acquisition. | cs.CL | Lexicon-Level Contrastive Visual-Grounding Improves Language Modeling |
2024-03-21T16:41:12Z | http://arxiv.org/pdf/2403.14541v2 | http://arxiv.org/abs/2403.14541v2 | Shimao Zhang, Yu Bao, Shujian Huang | 2024-04-03T16:09:22Z | cs.CL | Recently, Large Language Models (LLMs) have demonstrated outstanding
performance across a wide range of downstream language tasks. Temperature
sampling is a commonly used decoding strategy for LLMs' generation process.
However, a fixed temperature parameter is used in most cases, which may not
always be an optimal choice for balancing generation quality and diversity. In
this paper, we propose an effective Entropy-based Dynamic Temperature (EDT)
Sampling method, to achieve a more balanced performance in terms of both
generation quality and diversity by dynamically selecting the temperature
parameter. Additionally, we also show model performance and comprehensive
analyses for 4 different generation benchmarks. Our experiments show that EDT
significantly outperforms the existing strategies across different tasks. | cs.CL | EDT: Improving Large Language Models' Generation by Entropy-based Dynamic Temperature Sampling |
2024-03-21T16:11:44Z | http://arxiv.org/pdf/2403.14515v1 | http://arxiv.org/abs/2403.14515v1 | Gustavo Polleti | 2024-03-21T16:11:44Z | cs.CL | In this paper we discuss a first attempt to build a language learning game
for brazilian indigenous languages and the challenges around it. We present a
design for the tool with gamification aspects. Then we describe a process to
automatically generate language exercises and questions from a dependency
treebank and a lexical database for Tupian languages. We discuss the
limitations of our prototype highlighting ethical and practical implementation
concerns. Finally, we conclude that new data gathering processes should be
established in partnership with indigenous communities and oriented for
educational purposes. | cs.CL | Building a Language-Learning Game for Brazilian Indigenous Languages: A Case of Study |
2024-03-21T15:41:02Z | http://arxiv.org/pdf/2403.15491v1 | http://arxiv.org/abs/2403.15491v1 | Javier Conde, Miguel González, Nina Melero, Raquel Ferrando, Gonzalo Martínez, Elena Merino-Gómez, José Alberto Hernández, Pedro Reviriego | 2024-03-21T15:41:02Z | cs.CL | The growing interest in Large Language Models (LLMs) and in particular in
conversational models with which users can interact has led to the development
of a large number of open-source chat LLMs. These models are evaluated on a
wide range of benchmarks to assess their capabilities in answering questions or
solving problems on almost any possible topic or to test their ability to
reason or interpret texts. Instead, the evaluation of the knowledge that these
models have of the languages has received much less attention. For example, the
words that they can recognize and use in different languages. In this paper, we
evaluate the knowledge that open-source chat LLMs have of Spanish words by
testing a sample of words in a reference dictionary. The results show that
open-source chat LLMs produce incorrect meanings for an important fraction of
the words and are not able to use most of the words correctly to write
sentences with context. These results show how Spanish is left behind in the
open-source LLM race and highlight the need to push for linguistic fairness in
conversational LLMs ensuring that they provide similar performance across
languages. | cs.CL | Open Source Conversational LLMs do not know most Spanish words |
2024-03-21T15:18:30Z | http://arxiv.org/pdf/2403.14472v5 | http://arxiv.org/abs/2403.14472v5 | Mengru Wang, Ningyu Zhang, Ziwen Xu, Zekun Xi, Shumin Deng, Yunzhi Yao, Qishen Zhang, Linyi Yang, Jindong Wang, Huajun Chen | 2024-05-28T09:11:25Z | cs.CL, cs.AI, cs.CV, cs.HC, cs.LG | This paper investigates using knowledge editing techniques to detoxify Large
Language Models (LLMs). We construct a benchmark, SafeEdit, which covers nine
unsafe categories with various powerful attack prompts and equips comprehensive
metrics for systematic evaluation. We conduct experiments with several
knowledge editing approaches, indicating that knowledge editing has the
potential to detoxify LLMs with a limited impact on general performance
efficiently. Then, we propose a simple yet effective baseline, dubbed
Detoxifying with Intraoperative Neural Monitoring (DINM), to diminish the
toxicity of LLMs within a few tuning steps via only one instance. We further
provide an in-depth analysis of the internal mechanism for various detoxifying
approaches, demonstrating that previous methods like SFT and DPO may merely
suppress the activations of toxic parameters, while DINM mitigates the toxicity
of the toxic parameters to a certain extent, making permanent adjustments. We
hope that these insights could shed light on future work of developing
detoxifying approaches and the underlying knowledge mechanisms of LLMs. Code
and benchmark are available at https://github.com/zjunlp/EasyEdit. | cs.CL | Detoxifying Large Language Models via Knowledge Editing |
2024-03-21T15:16:50Z | http://arxiv.org/pdf/2403.14469v1 | http://arxiv.org/abs/2403.14469v1 | Hanieh Alipour, Nick Pendar, Kohinoor Roy | 2024-03-21T15:16:50Z | cs.CL, cs.AI | In recent times, the grandeur of Large Language Models (LLMs) has not only
shone in the realm of natural language processing but has also cast its
brilliance across a vast array of applications. This remarkable display of LLM
capabilities has ignited a surge in research contributions within this domain,
spanning a diverse spectrum of topics. These contributions encompass
advancements in neural network architecture, context length enhancements, model
alignment, training datasets, benchmarking, efficiency improvements, and more.
Recent years have witnessed a dynamic synergy between academia and industry,
propelling the field of LLM research to new heights. A notable milestone in
this journey is the introduction of ChatGPT, a powerful AI chatbot grounded in
LLMs, which has garnered widespread societal attention. The evolving technology
of LLMs has begun to reshape the landscape of the entire AI community,
promising a revolutionary shift in the way we create and employ AI algorithms.
Given this swift-paced technical evolution, our survey embarks on a journey to
encapsulate the recent strides made in the world of LLMs. Through an
exploration of the background, key discoveries, and prevailing methodologies,
we offer an up-to-the-minute review of the literature. By examining multiple
LLM models, our paper not only presents a comprehensive overview but also
charts a course that identifies existing challenges and points toward potential
future research trajectories. This survey furnishes a well-rounded perspective
on the current state of generative AI, shedding light on opportunities for
further exploration, enhancement, and innovation. | cs.CL | ChatGPT Alternative Solutions: Large Language Models Survey |
2024-03-21T15:14:25Z | http://arxiv.org/pdf/2403.14467v2 | http://arxiv.org/abs/2403.14467v2 | Jennifer Chien, Kevin R. McKee, Jackie Kay, William Isaac | 2024-04-21T10:59:51Z | cs.HC, cs.CL, cs.CY | Researchers and developers increasingly rely on toxicity scoring to moderate
generative language model outputs, in settings such as customer service,
information retrieval, and content generation. However, toxicity scoring may
render pertinent information inaccessible, rigidify or "value-lock" cultural
norms, and prevent language reclamation processes, particularly for
marginalized people. In this work, we extend the concept of algorithmic
recourse to generative language models: we provide users a novel mechanism to
achieve their desired prediction by dynamically setting thresholds for toxicity
filtering. Users thereby exercise increased agency relative to interactions
with the baseline system. A pilot study ($n = 30$) supports the potential of
our proposed recourse mechanism, indicating improvements in usability compared
to fixed-threshold toxicity-filtering of model outputs. Future work should
explore the intersection of toxicity scoring, model controllability, user
agency, and language reclamation processes -- particularly with regard to the
bias that many communities encounter when interacting with generative language
models. | cs.HC | Recourse for reclamation: Chatting with generative language models |
2024-03-21T15:07:57Z | http://arxiv.org/pdf/2403.14460v1 | http://arxiv.org/abs/2403.14460v1 | Krzysztof Lebioda, Viktor Vorobev, Nenad Petrovic, Fengjunjie Pan, Vahid Zolfaghari, Alois Knoll | 2024-03-21T15:07:57Z | cs.SE, cs.AI, cs.CL, D.2.1; D.2.2; D.2.4; I.2.7; I.2.2; I.7.0 | We propose a novel model- and feature-based approach to development of
vehicle software systems, where the end architecture is not explicitly defined.
Instead, it emerges from an iterative process of search and optimization given
certain constraints, requirements and hardware architecture, while retaining
the property of single-system illusion, where applications run in a logically
uniform environment. One of the key points of the presented approach is the
inclusion of modern generative AI, specifically Large Language Models (LLMs),
in the loop. With the recent advances in the field, we expect that the LLMs
will be able to assist in processing of requirements, generation of formal
system models, as well as generation of software deployment specification and
test code. The resulting pipeline is automated to a large extent, with feedback
being generated at each step. | cs.SE | Towards Single-System Illusion in Software-Defined Vehicles -- Automated, AI-Powered Workflow |
2024-03-21T15:06:14Z | http://arxiv.org/pdf/2403.14459v1 | http://arxiv.org/abs/2403.14459v1 | Lucas Monteiro Paes, Dennis Wei, Hyo Jin Do, Hendrik Strobelt, Ronny Luss, Amit Dhurandhar, Manish Nagireddy, Karthikeyan Natesan Ramamurthy, Prasanna Sattigeri, Werner Geyer, Soumya Ghosh | 2024-03-21T15:06:14Z | cs.CL, cs.AI | Perturbation-based explanation methods such as LIME and SHAP are commonly
applied to text classification. This work focuses on their extension to
generative language models. To address the challenges of text as output and
long text inputs, we propose a general framework called MExGen that can be
instantiated with different attribution algorithms. To handle text output, we
introduce the notion of scalarizers for mapping text to real numbers and
investigate multiple possibilities. To handle long inputs, we take a
multi-level approach, proceeding from coarser levels of granularity to finer
ones, and focus on algorithms with linear scaling in model queries. We conduct
a systematic evaluation, both automated and human, of perturbation-based
attribution methods for summarization and context-grounded question answering.
The results show that our framework can provide more locally faithful
explanations of generated outputs. | cs.CL | Multi-Level Explanations for Generative Language Models |
2024-03-21T15:04:32Z | http://arxiv.org/pdf/2403.14457v1 | http://arxiv.org/abs/2403.14457v1 | Anirudh Sundar, Christopher Richardson, Larry Heck | 2024-03-21T15:04:32Z | cs.CL, cs.IR, cs.LG | Distilling large, unstructured text into a structured, condensed form such as
tables is an open research problem. One of the primary challenges in
automatically generating tables is ensuring their syntactic validity. Prior
approaches address this challenge by including additional parameters in the
Transformer's attention mechanism to attend to specific rows and column
headers. In contrast to this single-stage method, this paper presents a
two-stage approach called Generative Tables (gTBLS). The first stage infers
table structure (row and column headers) from the text. The second stage
formulates questions using these headers and fine-tunes a causal language model
to answer them. Furthermore, the gTBLS approach is amenable to the utilization
of pre-trained Large Language Models in a zero-shot configuration, presenting a
solution for table generation in situations where fine-tuning is not feasible.
gTBLS improves prior approaches by up to 10% in BERTScore on the table
construction task and up to 20% on the table content generation task of the
E2E, WikiTableText, WikiBio, and RotoWire datasets. | cs.CL | gTBLS: Generating Tables from Text by Conditional Question Answering |
2024-03-21T15:02:03Z | http://arxiv.org/pdf/2403.14454v1 | http://arxiv.org/abs/2403.14454v1 | Fan Zhou, Vincent Vandeghinste | 2024-03-21T15:02:03Z | cs.CL | Machine translation (MT) encompasses a variety of methodologies aimed at
enhancing the accuracy of translations. In contrast, the process of
human-generated translation relies on a wide range of translation techniques,
which are crucial for ensuring linguistic adequacy and fluency. This study
suggests that these translation techniques could further optimize machine
translation if they are automatically identified before being applied to guide
the translation process effectively. The study differentiates between two
scenarios of the translation process: from-scratch translation and
post-editing. For each scenario, a specific set of experiments has been
designed to forecast the most appropriate translation techniques. The findings
indicate that the predictive accuracy for from-scratch translation reaches 82%,
while the post-editing process exhibits even greater potential, achieving an
accuracy rate of 93%. | cs.CL | Prediction of Translation Techniques for the Translation Process |
2024-03-21T14:51:51Z | http://arxiv.org/pdf/2403.14444v1 | http://arxiv.org/abs/2403.14444v1 | Ashvini Varatharaj, Simon Todd | 2024-03-21T14:51:51Z | cs.CL | Non-M\=aori-speaking New Zealanders (NMS)are able to segment M\=aori words in
a highlysimilar way to fluent speakers (Panther et al.,2024). This ability is
assumed to derive through the identification and extraction of statistically
recurrent forms. We examine this assumption by asking how NMS segmentations
compare to those produced by Morfessor, an unsupervised machine learning model
that operates based on statistical recurrence, across words formed by a variety
of morphological processes. Both NMS and Morfessor succeed in segmenting words
formed by concatenative processes (compounding and affixation without
allomorphy), but NMS also succeed for words that invoke templates
(reduplication and allomorphy) and other cues to morphological structure,
implying that their learning process is sensitive to more than just statistical
recurrence. | cs.CL | More than Just Statistical Recurrence: Human and Machine Unsupervised Learning of Māori Word Segmentation across Morphological Processes |
2024-03-21T14:48:37Z | http://arxiv.org/pdf/2403.14443v1 | http://arxiv.org/abs/2403.14443v1 | Nasim Rahaman, Martin Weiss, Manuel Wüthrich, Yoshua Bengio, Li Erran Li, Chris Pal, Bernhard Schölkopf | 2024-03-21T14:48:37Z | cs.AI, cs.CL, cs.GT, cs.LG, cs.MA, cs.SI | This work addresses the buyer's inspection paradox for information markets.
The paradox is that buyers need to access information to determine its value,
while sellers need to limit access to prevent theft. To study this, we
introduce an open-source simulated digital marketplace where intelligent
agents, powered by language models, buy and sell information on behalf of
external participants. The central mechanism enabling this marketplace is the
agents' dual capabilities: they not only have the capacity to assess the
quality of privileged information but also come equipped with the ability to
forget. This ability to induce amnesia allows vendors to grant temporary access
to proprietary information, significantly reducing the risk of unauthorized
retention while enabling agents to accurately gauge the information's relevance
to specific queries or tasks. To perform well, agents must make rational
decisions, strategically explore the marketplace through generated sub-queries,
and synthesize answers from purchased information. Concretely, our experiments
(a) uncover biases in language models leading to irrational behavior and
evaluate techniques to mitigate these biases, (b) investigate how price affects
demand in the context of informational goods, and (c) show that inspection and
higher budgets both lead to higher quality outcomes. | cs.AI | Language Models Can Reduce Asymmetry in Information Markets |
2024-03-21T14:44:03Z | http://arxiv.org/pdf/2403.14438v2 | http://arxiv.org/abs/2403.14438v2 | Dominik Wagner, Alexander Churchill, Siddharth Sigtia, Panayiotis Georgiou, Matt Mirsamadi, Aarshee Mishra, Erik Marchi | 2024-03-26T11:02:32Z | cs.CL, cs.LG, eess.AS | Interactions with virtual assistants typically start with a predefined
trigger phrase followed by the user command. To make interactions with the
assistant more intuitive, we explore whether it is feasible to drop the
requirement that users must begin each command with a trigger phrase. We
explore this task in three ways: First, we train classifiers using only
acoustic information obtained from the audio waveform. Second, we take the
decoder outputs of an automatic speech recognition (ASR) system, such as 1-best
hypotheses, as input features to a large language model (LLM). Finally, we
explore a multimodal system that combines acoustic and lexical features, as
well as ASR decoder signals in an LLM. Using multimodal information yields
relative equal-error-rate improvements over text-only and audio-only models of
up to 39% and 61%. Increasing the size of the LLM and training with low-rank
adaption leads to further relative EER reductions of up to 18% on our dataset. | cs.CL | A Multimodal Approach to Device-Directed Speech Detection with Large Language Models |
2024-03-21T14:33:34Z | http://arxiv.org/pdf/2403.14427v1 | http://arxiv.org/abs/2403.14427v1 | Lukas Galke, Limor Raviv | 2024-03-21T14:33:34Z | cs.CL, I.2.7 | Language models and humans are two types of learning systems. Finding or
facilitating commonalities could enable major breakthroughs in our
understanding of the acquisition and evolution of language. Many theories of
language evolution rely heavily on learning biases and learning pressures. Yet
due to substantial differences in learning pressures, it is questionable
whether the similarity between humans and machines is sufficient for insights
to carry over and to be worth testing with human participants. Here, we review
the emergent communication literature, a subfield of multi-agent reinforcement
learning, from a language evolution perspective. We find that the emergent
communication literature excels at designing and adapting models to recover
initially absent linguistic phenomena of natural languages. Based on a short
literature review, we identify key pressures that have recovered initially
absent human patterns in emergent communication models: communicative success,
efficiency, learnability, and other psycho-/sociolinguistic factors. We argue
that this may serve as inspiration for how to design language models for
language acquisition and language evolution research. | cs.CL | Emergent communication and learning pressures in language models: a language evolution perspective |
2024-03-21T13:57:43Z | http://arxiv.org/pdf/2403.14409v1 | http://arxiv.org/abs/2403.14409v1 | Yuchen Cai, Ding Cao, Rongxi Guo, Yaqin Wen, Guiquan Liu, Enhong Chen | 2024-03-21T13:57:43Z | cs.CL, cs.AI | Large language models(LLM) are pre-trained on extensive corpora to learn
facts and human cognition which contain human preferences. However, this
process can inadvertently lead to these models acquiring biases and stereotypes
prevalent in society. Prior research has typically tackled the issue of bias
through a one-dimensional perspective, concentrating either on locating or
mitigating it. This limited perspective has created obstacles in facilitating
research on bias to synergistically complement and progressively build upon one
another. In this study, we integrate the processes of locating and mitigating
bias within a unified framework. Initially, we use causal mediation analysis to
trace the causal effects of different components' activation within a large
language model. Building on this, we propose the LSDM (Least Square Debias
Method), a knowledge-editing based method for mitigating gender bias in
occupational pronouns, and compare it against two baselines on three gender
bias datasets and seven knowledge competency test datasets. The experimental
results indicate that the primary contributors to gender bias are the bottom
MLP modules acting on the last token of occupational pronouns and the top
attention module acting on the final word in the sentence. Furthermore, LSDM
mitigates gender bias in the model more effectively than the other baselines,
while fully preserving the model's capabilities in all other aspects. | cs.CL | Locating and Mitigating Gender Bias in Large Language Models |
2024-03-21T13:52:30Z | http://arxiv.org/pdf/2403.14403v2 | http://arxiv.org/abs/2403.14403v2 | Soyeong Jeong, Jinheon Baek, Sukmin Cho, Sung Ju Hwang, Jong C. Park | 2024-03-28T06:45:11Z | cs.CL, cs.AI | Retrieval-Augmented Large Language Models (LLMs), which incorporate the
non-parametric knowledge from external knowledge bases into LLMs, have emerged
as a promising approach to enhancing response accuracy in several tasks, such
as Question-Answering (QA). However, even though there are various approaches
dealing with queries of different complexities, they either handle simple
queries with unnecessary computational overhead or fail to adequately address
complex multi-step queries; yet, not all user requests fall into only one of
the simple or complex categories. In this work, we propose a novel adaptive QA
framework, that can dynamically select the most suitable strategy for
(retrieval-augmented) LLMs from the simplest to the most sophisticated ones
based on the query complexity. Also, this selection process is operationalized
with a classifier, which is a smaller LM trained to predict the complexity
level of incoming queries with automatically collected labels, obtained from
actual predicted outcomes of models and inherent inductive biases in datasets.
This approach offers a balanced strategy, seamlessly adapting between the
iterative and single-step retrieval-augmented LLMs, as well as the no-retrieval
methods, in response to a range of query complexities. We validate our model on
a set of open-domain QA datasets, covering multiple query complexities, and
show that ours enhances the overall efficiency and accuracy of QA systems,
compared to relevant baselines including the adaptive retrieval approaches.
Code is available at: https://github.com/starsuzi/Adaptive-RAG. | cs.CL | Adaptive-RAG: Learning to Adapt Retrieval-Augmented Large Language Models through Question Complexity |
2024-03-21T13:52:17Z | http://arxiv.org/pdf/2403.14402v1 | http://arxiv.org/abs/2403.14402v1 | HyoJung Han, Mohamed Anwar, Juan Pino, Wei-Ning Hsu, Marine Carpuat, Bowen Shi, Changhan Wang | 2024-03-21T13:52:17Z | cs.SD, cs.CL, eess.AS | Speech recognition and translation systems perform poorly on noisy inputs,
which are frequent in realistic environments. Augmenting these systems with
visual signals has the potential to improve robustness to noise. However,
audio-visual (AV) data is only available in limited amounts and for fewer
languages than audio-only resources. To address this gap, we present XLAVS-R, a
cross-lingual audio-visual speech representation model for noise-robust speech
recognition and translation in over 100 languages. It is designed to maximize
the benefits of limited multilingual AV pre-training data, by building on top
of audio-only multilingual pre-training and simplifying existing pre-training
schemes. Extensive evaluation on the MuAViC benchmark shows the strength of
XLAVS-R on downstream audio-visual speech recognition and translation tasks,
where it outperforms the previous state of the art by up to 18.5% WER and 4.7
BLEU given noisy AV inputs, and enables strong zero-shot audio-visual ability
with audio-only fine-tuning. | cs.SD | XLAVS-R: Cross-Lingual Audio-Visual Speech Representation Learning for Noise-Robust Speech Perception |
2024-03-21T13:47:40Z | http://arxiv.org/pdf/2403.14399v1 | http://arxiv.org/abs/2403.14399v1 | Changtong Zan, Liang Ding, Li Shen, Yibing Zhen, Weifeng Liu, Dacheng Tao | 2024-03-21T13:47:40Z | cs.CL, cs.AI | Translation-tailored Large language models (LLMs) exhibit remarkable
translation capabilities, even competing with supervised-trained commercial
translation systems. However, off-target translation remains an unsolved
problem, especially for low-resource languages, hindering us from developing
accurate LLMs-based translation models. To mitigate the off-target translation
problem and enhance the performance of LLMs on translation, recent works have
either designed advanced prompting strategies to highlight the functionality of
translation instructions or exploited the in-context learning ability of LLMs
by feeding few-shot demonstrations. However, these methods essentially do not
improve LLM's ability to follow translation instructions, especially the
language direction information. In this work, we design a two-stage fine-tuning
algorithm to improve the instruction-following ability (especially the
translation direction) of LLMs. Specifically, we first tune LLMs with the
maximum likelihood estimation loss on the translation dataset to elicit the
basic translation capabilities. In the second stage, we construct
instruction-conflicting samples by randomly replacing the translation
directions with a wrong one within the instruction, and then introduce an extra
unlikelihood loss to learn those samples. Experiments on IWSLT and WMT
benchmarks upon the LLaMA model spanning 16 zero-shot directions show that,
compared to the competitive baseline -- translation-finetuned LLama, our method
could effectively reduce the off-target translation ratio (averagely -53.3\%),
thus improving translation quality with average +5.7 SacreBLEU and +16.4
BLEURT. Analysis shows that our method could preserve the model's general task
performance on AlpacaEval. Code and models will be released at
\url{https://github.com/alphadl/LanguageAware_Tuning}. | cs.CL | Building Accurate Translation-Tailored LLMs with Language Aware Instruction Tuning |
2024-03-21T13:29:54Z | http://arxiv.org/pdf/2403.14390v1 | http://arxiv.org/abs/2403.14390v1 | Qingwen Lin, Boyan Xu, Zhengting Huang, Ruichu Cai | 2024-03-21T13:29:54Z | cs.CL | Addressing the challenge of high annotation costs in solving Math Word
Problems (MWPs) through full supervision with intermediate equations, recent
works have proposed weakly supervised task settings that rely solely on the
final answer as a supervised signal. Existing leading approaches typically
employ various search techniques to infer intermediate equations, but cannot
ensure their semantic consistency with natural language descriptions. The rise
of Large Language Models (LLMs) like ChatGPT has opened up new possibilities
for addressing MWPs directly. However, the computational demands of LLMs make
them less than ideal for use in settings where resources are tight. In light of
these challenges, we introduce an innovative two-stage framework that adeptly
transfers mathematical Expertise from large to tiny language models. In
\emph{Distillation Stage}, we propose a series of extraction processes that
satisfy the properties of MWPs to distill mathematical knowledge from LLMs to
construct problem-equation pairs required for supervised training. In
\emph{Refinement Stage}, Due to Knowledge distilling method cannot guarantee
the full utilization of all data, we further utilize the unsuccessfully
searched data effectively by Knowledge Refine method. Finally, We train a small
model using distilled data generated through two-stage methods. As our method
fully leverages the semantic understanding capabilities during the searching
'problem-equation' pair, it demonstrates significantly improved performance on
the Math23K and Weak12K datasets compared to existing small model methods,
while maintaining a much lower computational cost than ChatGPT. | cs.CL | From Large to Tiny: Distilling and Refining Mathematical Expertise for Math Word Problems with Weakly Supervision |
2024-03-21T13:15:25Z | http://arxiv.org/pdf/2403.14381v2 | http://arxiv.org/abs/2403.14381v2 | Yuchen Cai, Ding Cao, Rongxi Guo, Yaqin Wen, Guiquan Liu, Enhong Chen | 2024-05-11T13:51:32Z | cs.CL, cs.AI | Neural language models (LMs) have been extensively trained on vast corpora to
store factual knowledge about various aspects of the world described in texts.
Current technologies typically employ knowledge editing methods or specific
prompts to modify LM outputs. However, existing knowledge editing methods are
costly and inefficient, struggling to produce appropriate text. Additionally,
prompt engineering is opaque and requires significant effort to find suitable
prompts. To address these issues, we introduce a new method called PSPEM
(Prefix Soft Prompt Editing Method), that can be used for a lifetime with just
one training. It resolves the inefficiencies and generalizability issues in
knowledge editing methods and overcomes the opacity of prompt engineering by
automatically seeking optimal soft prompts. Specifically, PSPEM utilizes a
prompt encoder and an encoding converter to refine key information in prompts
and uses prompt alignment techniques to guide model generation, ensuring text
consistency and adherence to the intended structure and content, thereby
maintaining an optimal balance between efficiency and accuracy. We have
validated the effectiveness of PSPEM through knowledge editing and attribute
inserting. On the COUNTERFACT dataset, PSPEM achieved nearly 100\% editing
accuracy and demonstrated the highest level of fluency. We further analyzed the
similarities between PSPEM and original prompts and their impact on the model's
internals. The results indicate that PSPEM can serve as an alternative to
original prompts, supporting the model in effective editing. | cs.CL | Editing Knowledge Representation of Language Model via Rephrased Prefix Prompts |
2024-03-21T13:05:18Z | http://arxiv.org/pdf/2403.14374v1 | http://arxiv.org/abs/2403.14374v1 | Yuren Mao, Xuemei Dong, Wenyi Xu, Yunjun Gao, Bin Wei, Ying Zhang | 2024-03-21T13:05:18Z | cs.CL, cs.IR | Due to the extraordinarily large number of parameters, fine-tuning Large
Language Models (LLMs) to update long-tail or out-of-date knowledge is
impractical in lots of applications. To avoid fine-tuning, we can alternatively
treat a LLM as a black-box (i.e., freeze the parameters of the LLM) and augment
it with a Retrieval-Augmented Generation (RAG) system, namely black-box RAG.
Recently, black-box RAG has achieved success in knowledge-intensive tasks and
has gained much attention. Existing black-box RAG methods typically fine-tune
the retriever to cater to LLMs' preferences and concatenate all the retrieved
documents as the input, which suffers from two issues: (1) Ignorance of Factual
Information. The LLM preferred documents may not contain the factual
information for the given question, which can mislead the retriever and hurt
the effectiveness of black-box RAG; (2) Waste of Tokens. Simply concatenating
all the retrieved documents brings large amounts of unnecessary tokens for
LLMs, which degenerates the efficiency of black-box RAG. To address these
issues, this paper proposes a novel black-box RAG framework which utilizes the
factual information in the retrieval and reduces the number of tokens for
augmentation, dubbed FIT-RAG. FIT-RAG utilizes the factual information by
constructing a bi-label document scorer. Besides, it reduces the tokens by
introducing a self-knowledge recognizer and a sub-document-level token reducer.
FIT-RAG achieves both superior effectiveness and efficiency, which is validated
by extensive experiments across three open-domain question-answering datasets:
TriviaQA, NQ and PopQA. FIT-RAG can improve the answering accuracy of
Llama2-13B-Chat by 14.3\% on TriviaQA, 19.9\% on NQ and 27.5\% on PopQA,
respectively. Furthermore, it can save approximately half of the tokens on
average across the three datasets. | cs.CL | FIT-RAG: Black-Box RAG with Factual Information and Token Reduction |
2024-03-21T12:45:12Z | http://arxiv.org/pdf/2403.14364v1 | http://arxiv.org/abs/2403.14364v1 | Hichem Ammar Khodja, Frédéric Béchet, Quentin Brabant, Alexis Nasr, Gwénolé Lecorvé | 2024-03-21T12:45:12Z | cs.CL | The factuality of large language model (LLMs) tends to decay over time since
events posterior to their training are "unknown" to them. One way to keep
models up-to-date could be factual update: the task of inserting, replacing, or
removing certain simple (atomic) facts within the model. To study this task, we
present WikiFactDiff, a dataset that describes the evolution of factual
knowledge between two dates as a collection of simple facts divided into three
categories: new, obsolete, and static. We describe several update scenarios
arising from various combinations of these three types of basic update. The
facts are represented by subject-relation-object triples; indeed, WikiFactDiff
was constructed by comparing the state of the Wikidata knowledge base at 4
January 2021 and 27 February 2023. Those fact are accompanied by verbalization
templates and cloze tests that enable running update algorithms and their
evaluation metrics. Contrary to other datasets, such as zsRE and CounterFact,
WikiFactDiff constitutes a realistic update setting that involves various
update scenarios, including replacements, archival, and new entity insertions.
We also present an evaluation of existing update algorithms on WikiFactDiff. | cs.CL | WikiFactDiff: A Large, Realistic, and Temporally Adaptable Dataset for Atomic Factual Knowledge Update in Causal Language Models |
2024-03-21T12:17:59Z | http://arxiv.org/pdf/2403.14341v1 | http://arxiv.org/abs/2403.14341v1 | Jiaxin Liu, Yi Yang, Kar Yan Tam | 2024-03-21T12:17:59Z | cs.CL | In this paper, we introduce the Financial-STS task, a financial
domain-specific NLP task designed to measure the nuanced semantic similarity
between pairs of financial narratives. These narratives originate from the
financial statements of the same company but correspond to different periods,
such as year-over-year comparisons. Measuring the subtle semantic differences
between these paired narratives enables market stakeholders to gauge changes
over time in the company's financial and operational situations, which is
critical for financial decision-making. We find that existing pretrained
embedding models and LLM embeddings fall short in discerning these subtle
financial narrative shifts. To address this gap, we propose an LLM-augmented
pipeline specifically designed for the Financial-STS task. Evaluation on a
human-annotated dataset demonstrates that our proposed method outperforms
existing methods trained on classic STS tasks and generic LLM embeddings. | cs.CL | Beyond Surface Similarity: Detecting Subtle Semantic Shifts in Financial Narratives |
2024-03-21T12:11:26Z | http://arxiv.org/pdf/2403.14339v1 | http://arxiv.org/abs/2403.14339v1 | Daniel Trippa, Cesare Campagnano, Maria Sofia Bucarelli, Gabriele Tolomei, Fabrizio Silvestri | 2024-03-21T12:11:26Z | cs.LG, cs.AI, cs.CL, cs.CV | Machine Unlearning, the process of selectively eliminating the influence of
certain data examples used during a model's training, has gained significant
attention as a means for practitioners to comply with recent data protection
regulations. However, existing unlearning methods face critical drawbacks,
including their prohibitively high cost, often associated with a large number
of hyperparameters, and the limitation of forgetting only relatively small data
portions. This often makes retraining the model from scratch a quicker and more
effective solution. In this study, we introduce Gradient-based and
Task-Agnostic machine Unlearning ($\nabla \tau$), an optimization framework
designed to remove the influence of a subset of training data efficiently. It
applies adaptive gradient ascent to the data to be forgotten while using
standard gradient descent for the remaining data. $\nabla \tau$ offers multiple
benefits over existing approaches. It enables the unlearning of large sections
of the training dataset (up to 30%). It is versatile, supporting various
unlearning tasks (such as subset forgetting or class removal) and applicable
across different domains (images, text, etc.). Importantly, $\nabla \tau$
requires no hyperparameter adjustments, making it a more appealing option than
retraining the model from scratch. We evaluate our framework's effectiveness
using a set of well-established Membership Inference Attack metrics,
demonstrating up to 10% enhancements in performance compared to
state-of-the-art methods without compromising the original model's accuracy. | cs.LG | $\nabla τ$: Gradient-based and Task-Agnostic machine Unlearning |
2024-03-21T11:34:26Z | http://arxiv.org/pdf/2403.14312v1 | http://arxiv.org/abs/2403.14312v1 | Xiaoxue Cheng, Junyi Li, Wayne Xin Zhao, Ji-Rong Wen | 2024-03-21T11:34:26Z | cs.CL | Chain-of-Thought (CoT) prompting can enhance the reasoning capabilities of
large language models (LLMs), establishing itself as a primary approach to
solving complex reasoning tasks. Existing CoT synthesis approaches usually
focus on simpler reasoning tasks and thus result in low-quality and
inconsistent CoT prompts. In response to this challenge, we present an
empirical investigation of CoT prompting and introduce CoTGenius, a novel
framework designed for the automatic generation of superior CoT prompts.
CoTGenius is developed based on three major evolution strategies, i.e.,
complicate, diversify, and specify-alongside two filtering mechanisms:
evolutionary success judgement and correctness verification. We further employ
CoTGenius to create an extensive CoT dataset, and subsequently fine-tune the
Llama 2-Chat 7B and 13B models on this dataset. We call the resulting model
ChainLM. To deal with the cumulative error issue in reasoning steps, we propose
a step-level debating method, wherein multiple debaters discuss each reasoning
step to arrive at the correct answer. Extensive experiments demonstrate that
our ChainLM models exhibit enhanced proficiency in addressing a spectrum of
complex reasoning problems compared to existing models. In addition, we conduct
an in-depth analysis of the impact of data categories within CoTGenius on the
model performance. We release our dataset and code at
https://github.com/RUCAIBox/ChainLM. | cs.CL | ChainLM: Empowering Large Language Models with Improved Chain-of-Thought Prompting |
2024-03-21T10:31:11Z | http://arxiv.org/pdf/2403.14275v1 | http://arxiv.org/abs/2403.14275v1 | Shuqian Sheng, Yi Xu, Luoyi Fu, Jiaxin Ding, Lei Zhou, Xinbing Wang, Chenghu Zhou | 2024-03-21T10:31:11Z | cs.CL | The majority of automatic metrics for evaluating NLG systems are
reference-based. However, the challenge of collecting human annotation results
in a lack of reliable references in numerous application scenarios. Despite
recent advancements in reference-free metrics, it has not been well understood
when and where they can be used as an alternative to reference-based metrics.
In this study, by employing diverse analytical approaches, we comprehensively
assess the performance of both metrics across a wide range of NLG tasks,
encompassing eight datasets and eight evaluation models. Based on solid
experiments, the results show that reference-free metrics exhibit a higher
correlation with human judgment and greater sensitivity to deficiencies in
language quality. However, their effectiveness varies across tasks and is
influenced by the quality of candidate texts. Therefore, it's important to
assess the performance of reference-free metrics before applying them to a new
task, especially when inputs are in uncommon form or when the answer space is
highly variable. Our study can provide insight into the appropriate application
of automatic metrics and the impact of metric choice on evaluation performance. | cs.CL | Is Reference Necessary in the Evaluation of NLG Systems? When and Where? |
2024-03-21T10:15:57Z | http://arxiv.org/pdf/2403.14270v1 | http://arxiv.org/abs/2403.14270v1 | Tim Salzmann, Markus Ryll, Alex Bewley, Matthias Minderer | 2024-03-21T10:15:57Z | cs.CV, cs.CL, cs.LG, cs.RO | Visual relationship detection aims to identify objects and their
relationships in images. Prior methods approach this task by adding separate
relationship modules or decoders to existing object detection architectures.
This separation increases complexity and hinders end-to-end training, which
limits performance. We propose a simple and highly efficient decoder-free
architecture for open-vocabulary visual relationship detection. Our model
consists of a Transformer-based image encoder that represents objects as tokens
and models their relationships implicitly. To extract relationship information,
we introduce an attention mechanism that selects object pairs likely to form a
relationship. We provide a single-stage recipe to train this model on a mixture
of object and relationship detection data. Our approach achieves
state-of-the-art relationship detection performance on Visual Genome and on the
large-vocabulary GQA benchmark at real-time inference speeds. We provide
analyses of zero-shot performance, ablations, and real-world qualitative
examples. | cs.CV | Scene-Graph ViT: End-to-End Open-Vocabulary Visual Relationship Detection |
2024-03-21T10:01:05Z | http://arxiv.org/pdf/2405.01556v1 | http://arxiv.org/abs/2405.01556v1 | Ananya Singha, Bhavya Chopra, Anirudh Khatry, Sumit Gulwani, Austin Z. Henley, Vu Le, Chris Parnin, Mukul Singh, Gust Verbruggen | 2024-03-21T10:01:05Z | cs.SE, cs.AI, cs.CL | Automated insight generation is a common tactic for helping knowledge
workers, such as data scientists, to quickly understand the potential value of
new and unfamiliar data. Unfortunately, automated insights produced by
large-language models can generate code that does not correctly correspond (or
align) to the insight. In this paper, we leverage the semantic knowledge of
large language models to generate targeted and insightful questions about data
and the corresponding code to answer those questions. Then through an empirical
study on data from Open-WikiTable, we show that embeddings can be effectively
used for filtering out semantically unaligned pairs of question and code.
Additionally, we found that generating questions and code together yields more
diverse questions. | cs.SE | Semantically Aligned Question and Code Generation for Automated Insight Generation |
2024-03-21T09:36:36Z | http://arxiv.org/pdf/2403.14258v1 | http://arxiv.org/abs/2403.14258v1 | Stefan Trapp, Joachim Warschat | 2024-03-21T09:36:36Z | cs.CL, I.2.7; H.3.1 | Already since the 1950s TRIZ shows that patents and the technical
contradictions they solve are an important source of inspiration for the
development of innovative products. However, TRIZ is a heuristic based on a
historic patent analysis and does not make use of the ever-increasing number of
latest technological solutions in current patents. Because of the huge number
of patents, their length, and, last but not least, their complexity there is a
need for modern patent retrieval and patent analysis to go beyond
keyword-oriented methods. Recent advances in patent retrieval and analysis
mainly focus on dense vectors based on neural AI Transformer language models
like Google BERT. They are, for example, used for dense retrieval, question
answering or summarization and key concept extraction. A research focus within
the methods for patent summarization and key concept extraction are generic
inventive concepts respectively TRIZ concepts like problems, solutions,
advantage of invention, parameters, and contradictions. Succeeding rule-based
approaches, finetuned BERT-like language models for sentence-wise
classification represent the state-of-the-art of inventive concept extraction.
While they work comparatively well for basic concepts like problems or
solutions, contradictions - as a more complex abstraction - remain a challenge
for these models. This paper goes one step further, as it presents a method to
extract TRIZ contradictions from patent texts based on Prompt Engineering using
a generative Large Language Model (LLM), namely OpenAI's GPT-4. Contradiction
detection, sentence extraction, contradiction summarization, parameter
extraction and assignment to the 39 abstract TRIZ engineering parameters are
all performed in a single prompt using the LangChain framework. Our results
show that "off-the-shelf" GPT-4 is a serious alternative to existing
approaches. | cs.CL | LLM-based Extraction of Contradictions from Patents |
2024-03-21T09:28:38Z | http://arxiv.org/pdf/2403.14255v1 | http://arxiv.org/abs/2403.14255v1 | Sehee Lim, Yejin Kim, Chi-Hyun Choi, Jy-yong Sohn, Byung-Hoon Kim | 2024-03-21T09:28:38Z | cs.CL, cs.LG | Improving the accessibility of psychotherapy with the aid of Large Language
Models (LLMs) is garnering a significant attention in recent years. Recognizing
cognitive distortions from the interviewee's utterances can be an essential
part of psychotherapy, especially for cognitive behavioral therapy. In this
paper, we propose ERD, which improves LLM-based cognitive distortion
classification performance with the aid of additional modules of (1) extracting
the parts related to cognitive distortion, and (2) debating the reasoning steps
by multiple agents. Our experimental results on a public dataset show that ERD
improves the multi-class F1 score as well as binary specificity score.
Regarding the latter score, it turns out that our method is effective in
debiasing the baseline method which has high false positive rate, especially
when the summary of multi-agent debate is provided to LLMs. | cs.CL | ERD: A Framework for Improving LLM Reasoning for Cognitive Distortion Classification |
2024-03-21T09:26:04Z | http://arxiv.org/pdf/2403.14253v2 | http://arxiv.org/abs/2403.14253v2 | Kyuhee Kim, Surin Lee, Sangah Lee | 2024-03-23T15:53:50Z | cs.CL | In many literary texts, emotions are indirectly conveyed through descriptions
of actions, facial expressions, and appearances, necessitating emotion
inference for narrative understanding. In this paper, we introduce K-Act2Emo, a
Korean commonsense knowledge graph (CSKG) comprising 1,900 indirect emotional
expressions and the emotions inferable from them. We categorize reasoning types
into inferences in positive situations, inferences in negative situations, and
inferences when expressions do not serve as emotional cues. Unlike existing
CSKGs, K-Act2Emo specializes in emotional contexts, and experimental results
validate its effectiveness for training emotion inference models.
Significantly, the BART-based knowledge model fine-tuned with K-Act2Emo
outperforms various existing Korean large language models, achieving
performance levels comparable to GPT-4 Turbo. | cs.CL | K-Act2Emo: Korean Commonsense Knowledge Graph for Indirect Emotional Expression |
2024-03-21T09:25:24Z | http://arxiv.org/pdf/2403.14252v1 | http://arxiv.org/abs/2403.14252v1 | Masato Fujitake | 2024-03-21T09:25:24Z | cs.CL, cs.AI, cs.CV, cs.LG | This paper proposes LayoutLLM, a more flexible document analysis method for
understanding imaged documents. Visually Rich Document Understanding tasks,
such as document image classification and information extraction, have gained
significant attention due to their importance. Existing methods have been
developed to enhance document comprehension by incorporating pre-training
awareness of images, text, and layout structure. However, these methods require
fine-tuning for each task and dataset, and the models are expensive to train
and operate. To overcome this limitation, we propose a new LayoutLLM that
integrates these with large-scale language models (LLMs). By leveraging the
strengths of existing research in document image understanding and LLMs'
superior language understanding capabilities, the proposed model, fine-tuned
with multimodal instruction datasets, performs an understanding of document
images in a single model. Our experiments demonstrate improvement over the
baseline model in various document analysis tasks. | cs.CL | LayoutLLM: Large Language Model Instruction Tuning for Visually Rich Document Understanding |
2024-03-21T09:02:17Z | http://arxiv.org/pdf/2403.14243v1 | http://arxiv.org/abs/2403.14243v1 | Dimitrios P. Panagoulias, Evridiki Tsoureli-Nikita, Maria Virvou, George A. Tsihrintzis | 2024-03-21T09:02:17Z | cs.CL, cs.AI, cs.CV | The rise of Artificial Intelligence creates great promise in the field of
medical discovery, diagnostics and patient management. However, the vast
complexity of all medical domains require a more complex approach that combines
machine learning algorithms, classifiers, segmentation algorithms and, lately,
large language models. In this paper, we describe, implement and assess an
Artificial Intelligence-empowered system and methodology aimed at assisting the
diagnosis process of skin lesions and other skin conditions within the field of
dermatology that aims to holistically address the diagnostic process in this
domain. The workflow integrates large language, transformer-based vision models
and sophisticated machine learning tools. This holistic approach achieves a
nuanced interpretation of dermatological conditions that simulates and
facilitates a dermatologist's workflow. We assess our proposed methodology
through a thorough cross-model validation technique embedded in an evaluation
pipeline that utilizes publicly available medical case studies of skin
conditions and relevant images. To quantitatively score the system performance,
advanced machine learning and natural language processing tools are employed
which focus on similarity comparison and natural language inference.
Additionally, we incorporate a human expert evaluation process based on a
structured checklist to further validate our results. We implemented the
proposed methodology in a system which achieved approximate (weighted) scores
of 0.87 for both contextual understanding and diagnostic accuracy,
demonstrating the efficacy of our approach in enhancing dermatological
analysis. The proposed methodology is expected to prove useful in the
development of next-generation tele-dermatology applications, enhancing remote
consultation capabilities and access to care, especially in underserved areas. | cs.CL | Dermacen Analytica: A Novel Methodology Integrating Multi-Modal Large Language Models with Machine Learning in tele-dermatology |
2024-03-21T08:57:27Z | http://arxiv.org/pdf/2403.14238v1 | http://arxiv.org/abs/2403.14238v1 | Kyungjae Lee, Dasol Hwang, Sunghyun Park, Youngsoo Jang, Moontae Lee | 2024-03-21T08:57:27Z | cs.CL, cs.AI | Despite the promise of RLHF in aligning LLMs with human preferences, it often
leads to superficial alignment, prioritizing stylistic changes over improving
downstream performance of LLMs. Underspecified preferences could obscure
directions to align the models. Lacking exploration restricts identification of
desirable outputs to improve the models. To overcome these challenges, we
propose a novel framework: Reinforcement Learning from Reflective Feedback
(RLRF), which leverages fine-grained feedback based on detailed criteria to
improve the core capabilities of LLMs. RLRF employs a self-reflection mechanism
to systematically explore and refine LLM responses, then fine-tuning the models
via a RL algorithm along with promising responses. Our experiments across
Just-Eval, Factuality, and Mathematical Reasoning demonstrate the efficacy and
transformative potential of RLRF beyond superficial surface-level adjustment. | cs.CL | Reinforcement Learning from Reflective Feedback (RLRF): Aligning and Improving LLMs via Fine-Grained Self-Reflection |
2024-03-21T08:54:56Z | http://arxiv.org/pdf/2403.14734v2 | http://arxiv.org/abs/2403.14734v2 | Qiushi Sun, Zhirui Chen, Fangzhi Xu, Kanzhi Cheng, Chang Ma, Zhangyue Yin, Jianing Wang, Chengcheng Han, Renyu Zhu, Shuai Yuan, Qipeng Guo, Xipeng Qiu, Pengcheng Yin, Xiaoli Li, Fei Yuan, Lingpeng Kong, Xiang Li, Zhiyong Wu | 2024-06-23T14:31:05Z | cs.SE, cs.AI, cs.CL, cs.PL | Neural Code Intelligence -- leveraging deep learning to understand, generate,
and optimize code -- holds immense potential for transformative impacts on the
whole society. Bridging the gap between Natural Language and Programming
Language, this domain has drawn significant attention from researchers in both
research communities over the past few years. This survey presents a systematic
and chronological review of the advancements in code intelligence, encompassing
over 50 representative models and their variants, more than 20 categories of
tasks, and an extensive coverage of over 680 related works. We follow the
historical progression to trace the paradigm shifts across different research
phases (e.g., from modeling code with recurrent neural networks to the era of
Large Language Models). Concurrently, we highlight the major technical
transitions in models, tasks, and evaluations spanning through different
stages. For applications, we also observe a co-evolving shift. It spans from
initial endeavors to tackling specific scenarios, through exploring a diverse
array of tasks during its rapid expansion, to currently focusing on tackling
increasingly complex and varied real-world challenges. Building on our
examination of the developmental trajectories, we further investigate the
emerging synergies between code intelligence and broader machine intelligence,
uncovering new cross-domain opportunities and illustrating the substantial
influence of code intelligence across various domains. Finally, we delve into
both the opportunities and challenges associated with this field, alongside
elucidating our insights on the most promising research directions. An ongoing,
dynamically updated project and resources associated with this survey have been
released at https://github.com/QiushiSun/NCISurvey. | cs.SE | A Survey of Neural Code Intelligence: Paradigms, Advances and Beyond |
2024-03-21T08:54:24Z | http://arxiv.org/pdf/2403.14236v2 | http://arxiv.org/abs/2403.14236v2 | Akshat Gupta, Dev Sajnani, Gopala Anumanchipalli | 2024-04-22T17:56:13Z | cs.LG, cs.AI, cs.CL | We introduce a unifying framework that brings two leading "locate-and-edit"
model editing techniques -- ROME and MEMIT -- under a single conceptual
umbrella, optimizing for the same goal, which we call the
preservation-memorization objective. ROME uses an equality constraint to
perform one edit at a time, whereas MEMIT employs a more flexible least-square
constraint that allows for batched edits. Following the
preservation-memorization objective, we present Equality-constrained Mass Model
Editing algorithm for Transformers or EMMET, a new batched memory-editing
algorithm that uses a closed-form solution for the equality-constrained version
of the preservation-memorization objective. EMMET is a batched-version of ROME
and is able to perform batched-edits up to a batch-size of 10,000 with very
similar performance to MEMIT across multiple dimensions. With EMMET, we unify
and achieve symmetry within the "locate-and-edit" algorithms, allowing
batched-editing using both objectives. | cs.LG | A Unified Framework for Model Editing |
2024-03-21T08:27:49Z | http://arxiv.org/pdf/2403.15486v1 | http://arxiv.org/abs/2403.15486v1 | Gustave Cortal | 2024-03-21T08:27:49Z | cs.CL, cs.AI | The study of dreams has been central to understanding human
(un)consciousness, cognition, and culture for centuries. Analyzing dreams
quantitatively depends on labor-intensive, manual annotation of dream
narratives. We automate this process through a natural language
sequence-to-sequence generation framework. This paper presents the first study
on character and emotion detection in the English portion of the open DreamBank
corpus of dream narratives. Our results show that language models can
effectively address this complex task. To get insight into prediction
performance, we evaluate the impact of model size, prediction order of
characters, and the consideration of proper names and character traits. We
compare our approach with a large language model using in-context learning. Our
supervised models perform better while having 28 times fewer parameters. Our
model and its generated annotations are made publicly available. | cs.CL | Sequence-to-Sequence Language Models for Character and Emotion Detection in Dream Narratives |
2024-03-21T08:22:44Z | http://arxiv.org/pdf/2403.14222v1 | http://arxiv.org/abs/2403.14222v1 | Jonas Golde, Felix Hamborg, Alan Akbik | 2024-03-21T08:22:44Z | cs.CL | Few-shot named entity recognition (NER) detects named entities within text
using only a few annotated examples. One promising line of research is to
leverage natural language descriptions of each entity type: the common label
PER might, for example, be verbalized as ''person entity.'' In an initial label
interpretation learning phase, the model learns to interpret such verbalized
descriptions of entity types. In a subsequent few-shot tagset extension phase,
this model is then given a description of a previously unseen entity type (such
as ''music album'') and optionally a few training examples to perform few-shot
NER for this type. In this paper, we systematically explore the impact of a
strong semantic prior to interpret verbalizations of new entity types by
massively scaling up the number and granularity of entity types used for label
interpretation learning. To this end, we leverage an entity linking benchmark
to create a dataset with orders of magnitude of more distinct entity types and
descriptions as currently used datasets. We find that this increased signal
yields strong results in zero- and few-shot NER in in-domain, cross-domain, and
even cross-lingual settings. Our findings indicate significant potential for
improving few-shot NER through heuristical data-based optimization. | cs.CL | Large-Scale Label Interpretation Learning for Few-Shot Named Entity Recognition |
2024-03-21T08:21:12Z | http://arxiv.org/pdf/2403.14221v2 | http://arxiv.org/abs/2403.14221v2 | Yukun Zhao, Lingyong Yan, Weiwei Sun, Guoliang Xing, Shuaiqiang Wang, Chong Meng, Zhicong Cheng, Zhaochun Ren, Dawei Yin | 2024-03-22T12:34:47Z | cs.CL | Large language models (LLMs) have shown tremendous success in following user
instructions and generating helpful responses. Nevertheless, their robustness
is still far from optimal, as they may generate significantly inconsistent
responses due to minor changes in the verbalized instructions. Recent
literature has explored this inconsistency issue, highlighting the importance
of continued improvement in the robustness of response generation. However,
systematic analysis and solutions are still lacking. In this paper, we
quantitatively define the inconsistency problem and propose a two-stage
training framework consisting of instruction-augmented supervised fine-tuning
and consistency alignment training. The first stage helps a model generalize on
following instructions via similar instruction augmentations. In the second
stage, we improve the diversity and help the model understand which responses
are more aligned with human expectations by differentiating subtle differences
in similar responses. The training process is accomplished by self-rewards
inferred from the trained model at the first stage without referring to
external human preference resources. We conduct extensive experiments on recent
publicly available LLMs on instruction-following tasks and demonstrate the
effectiveness of our training framework. | cs.CL | Improving the Robustness of Large Language Models via Consistency Alignment |
2024-03-21T08:03:46Z | http://arxiv.org/pdf/2403.14733v1 | http://arxiv.org/abs/2403.14733v1 | Bingchen Liu, Huang Peng, Weixin Zeng, Xiang Zhao, Shijun Liu, Li Pan | 2024-03-21T08:03:46Z | cs.AI, cs.CL, cs.LG | The construction of large open knowledge bases (OKBs) is integral to many
knowledge-driven applications on the world wide web such as web search.
However, noun phrases and relational phrases in OKBs often suffer from
redundancy and ambiguity, which calls for the investigation on OKB
canonicalization. Current solutions address OKB canonicalization by devising
advanced clustering algorithms and using knowledge graph embedding (KGE) to
further facilitate the canonicalization process. Nevertheless, these works fail
to fully exploit the synergy between clustering and KGE learning, and the
methods designed for these subtasks are sub-optimal. To this end, we put
forward a multi-task learning framework, namely MulCanon, to tackle OKB
canonicalization. In addition, diffusion model is used in the soft clustering
process to improve the noun phrase representations with neighboring
information, which can lead to more accurate representations. MulCanon unifies
the learning objectives of these sub-tasks, and adopts a two-stage multi-task
learning paradigm for training. A thorough experimental study on popular OKB
canonicalization benchmarks validates that MulCanon can achieve competitive
canonicalization results. | cs.AI | Open Knowledge Base Canonicalization with Multi-task Learning |
2024-03-21T08:00:05Z | http://arxiv.org/pdf/2403.14208v1 | http://arxiv.org/abs/2403.14208v1 | Mitja Nikolaus, Abhishek Agrawal, Petros Kaklamanis, Alex Warstadt, Abdellah Fourtassi | 2024-03-21T08:00:05Z | cs.CL | The acquisition of grammar has been a central question to adjudicate between
theories of language acquisition. In order to conduct faster, more
reproducible, and larger-scale corpus studies on grammaticality in
child-caregiver conversations, tools for automatic annotation can offer an
effective alternative to tedious manual annotation. We propose a coding scheme
for context-dependent grammaticality in child-caregiver conversations and
annotate more than 4,000 utterances from a large corpus of transcribed
conversations. Based on these annotations, we train and evaluate a range of NLP
models. Our results show that fine-tuned Transformer-based models perform best,
achieving human inter-annotation agreement levels.As a first application and
sanity check of this tool, we use the trained models to annotate a corpus
almost two orders of magnitude larger than the manually annotated data and
verify that children's grammaticality shows a steady increase with age.This
work contributes to the growing literature on applying state-of-the-art NLP
methods to help study child language acquisition at scale. | cs.CL | Automatic Annotation of Grammaticality in Child-Caregiver Conversations |
2024-03-21T07:47:57Z | http://arxiv.org/pdf/2403.14197v1 | http://arxiv.org/abs/2403.14197v1 | Kosuke Akimoto, Kunihiro Takeoka, Masafumi Oyamada | 2024-03-21T07:47:57Z | cs.CL | Retrieval-augmented generation models augment knowledge encoded in a language
model by providing additional relevant external knowledge (context) during
generation. Although it has been shown that the quantity and quality of context
impact the performance of retrieval-augmented generation models during
inference, limited research explores how these characteristics affect model
training. This paper explores how context quantity and quality during model
training affect the performance of Fusion-in-Decoder (FiD), the
state-of-the-art retrieval-augmented generation model, in extractive
open-domain question answering tasks. Experimental results suggest that FiD
models overfit to context quality during training and show suboptimal
performance when evaluated on different context quality. Through the
experimental results, we also reveal FiD models trained with different context
quality have different cross-attention distribution patterns. Specifically, as
context quality during training increases, FiD models tend to attend more
uniformly to each passage in context. Finally, based on these observations, we
propose a method to mitigate overfitting to specific context quality by
introducing bias to the cross-attention distribution, which we demonstrate to
be effective in improving the performance of FiD models on different context
quality. | cs.CL | Context Quality Matters in Training Fusion-in-Decoder for Extractive Open-Domain Question Answering |
2024-03-21T07:45:58Z | http://arxiv.org/pdf/2403.15485v1 | http://arxiv.org/abs/2403.15485v1 | Junyeop Cha, Seoyun Kim, Dongjae Kim, Eunil Park | 2024-03-21T07:45:58Z | cs.CL, cs.AI, cs.LG | Early detection plays a crucial role in the treatment of depression.
Therefore, numerous studies have focused on social media platforms, where
individuals express their emotions, aiming to achieve early detection of
depression. However, the majority of existing approaches often rely on specific
features, leading to limited scalability across different types of social media
datasets, such as text, images, or videos. To overcome this limitation, we
introduce a Multimodal Object-Oriented Graph Attention Model (MOGAM), which can
be applied to diverse types of data, offering a more scalable and versatile
solution. Furthermore, to ensure that our model can capture authentic symptoms
of depression, we only include vlogs from users with a clinical diagnosis. To
leverage the diverse features of vlogs, we adopt a multimodal approach and
collect additional metadata such as the title, description, and duration of the
vlogs. To effectively aggregate these multimodal features, we employed a
cross-attention mechanism. MOGAM achieved an accuracy of 0.871 and an F1-score
of 0.888. Moreover, to validate the scalability of MOGAM, we evaluated its
performance with a benchmark dataset and achieved comparable results with prior
studies (0.61 F1-score). In conclusion, we believe that the proposed model,
MOGAM, is an effective solution for detecting depression in social media,
offering potential benefits in the early detection and treatment of this mental
health condition. | cs.CL | MOGAM: A Multimodal Object-oriented Graph Attention Model for Depression Detection |
2024-03-21T06:56:07Z | http://arxiv.org/pdf/2403.15484v1 | http://arxiv.org/abs/2403.15484v1 | Rakuten Group, Aaron Levine, Connie Huang, Chenguang Wang, Eduardo Batista, Ewa Szymanska, Hongyi Ding, Hou Wei Chou, Jean-François Pessiot, Johanes Effendi, Justin Chiu, Kai Torben Ohlhus, Karan Chopra, Keiji Shinzato, Koji Murakami, Lee Xiong, Lei Chen, Maki Kubota, Maksim Tkachenko, Miroku Lee, Naoki Takahashi, Prathyusha Jwalapuram, Ryutaro Tatsushima, Saurabh Jain, Sunil Kumar Yadav, Ting Cai, Wei-Te Chen, Yandi Xia, Yuki Nakayama, Yutaka Higashiyama | 2024-03-21T06:56:07Z | cs.CL, cs.LG | We introduce RakutenAI-7B, a suite of Japanese-oriented large language models
that achieve the best performance on the Japanese LM Harness benchmarks among
the open 7B models. Along with the foundation model, we release instruction-
and chat-tuned models, RakutenAI-7B-instruct and RakutenAI-7B-chat
respectively, under the Apache 2.0 license. | cs.CL | RakutenAI-7B: Extending Large Language Models for Japanese |
2024-03-21T06:47:28Z | http://arxiv.org/pdf/2403.14171v3 | http://arxiv.org/abs/2403.14171v3 | Longzheng Wang, Xiaohan Xu, Lei Zhang, Jiarui Lu, Yongxiu Xu, Hongbo Xu, Minghao Tang, Chuang Zhang | 2024-04-08T08:30:06Z | cs.CL | Automatic detection of multimodal misinformation has gained a widespread
attention recently. However, the potential of powerful Large Language Models
(LLMs) for multimodal misinformation detection remains underexplored. Besides,
how to teach LLMs to interpret multimodal misinformation in cost-effective and
accessible way is still an open question. To address that, we propose MMIDR, a
framework designed to teach LLMs in providing fluent and high-quality textual
explanations for their decision-making process of multimodal misinformation. To
convert multimodal misinformation into an appropriate instruction-following
format, we present a data augmentation perspective and pipeline. This pipeline
consists of a visual information processing module and an evidence retrieval
module. Subsequently, we prompt the proprietary LLMs with processed contents to
extract rationales for interpreting the authenticity of multimodal
misinformation. Furthermore, we design an efficient knowledge distillation
approach to distill the capability of proprietary LLMs in explaining multimodal
misinformation into open-source LLMs. To explore several research questions
regarding the performance of LLMs in multimodal misinformation detection tasks,
we construct an instruction-following multimodal misinformation dataset and
conduct comprehensive experiments. The experimental findings reveal that our
MMIDR exhibits sufficient detection performance and possesses the capacity to
provide compelling rationales to support its assessments. | cs.CL | MMIDR: Teaching Large Language Model to Interpret Multimodal Misinformation via Knowledge Distillation |
2024-03-21T06:43:59Z | http://arxiv.org/pdf/2403.14168v3 | http://arxiv.org/abs/2403.14168v3 | Zhe Chen, Heyang Liu, Wenyi Yu, Guangzhi Sun, Hongcheng Liu, Ji Wu, Chao Zhang, Yu Wang, Yanfeng Wang | 2024-06-04T04:05:09Z | cs.CL | Publishing open-source academic video recordings is an emergent and prevalent
approach to sharing knowledge online. Such videos carry rich multimodal
information including speech, the facial and body movements of the speakers, as
well as the texts and pictures in the slides and possibly even the papers.
Although multiple academic video datasets have been constructed and released,
few of them support both multimodal content recognition and understanding
tasks, which is partially due to the lack of high-quality human annotations. In
this paper, we propose a novel multimodal, multigenre, and multipurpose
audio-visual academic lecture dataset (M$^3$AV), which has almost 367 hours of
videos from five sources covering computer science, mathematics, and medical
and biology topics. With high-quality human annotations of the slide text and
spoken words, in particular high-valued name entities, the dataset can be used
for multiple audio-visual recognition and understanding tasks. Evaluations
performed on contextual speech recognition, speech synthesis, and slide and
script generation tasks demonstrate that the diversity of M$^3$AV makes it a
challenging dataset. | cs.CL | M$^3$AV: A Multimodal, Multigenre, and Multipurpose Audio-Visual Academic Lecture Dataset |
2024-03-21T04:54:31Z | http://arxiv.org/pdf/2403.14731v1 | http://arxiv.org/abs/2403.14731v1 | Mingze Ni, Zhensu Sun, Wei Liu | 2024-03-21T04:54:31Z | cs.CR, cs.CL, cs.LG | Recent studies on adversarial examples expose vulnerabilities of natural
language processing (NLP) models. Existing techniques for generating
adversarial examples are typically driven by deterministic hierarchical rules
that are agnostic to the optimal adversarial examples, a strategy that often
results in adversarial samples with a suboptimal balance between magnitudes of
changes and attack successes. To this end, in this research we propose two
algorithms, Reversible Jump Attack (RJA) and Metropolis-Hasting Modification
Reduction (MMR), to generate highly effective adversarial examples and to
improve the imperceptibility of the examples, respectively. RJA utilizes a
novel randomization mechanism to enlarge the search space and efficiently
adapts to a number of perturbed words for adversarial examples. With these
generated adversarial examples, MMR applies the Metropolis-Hasting sampler to
enhance the imperceptibility of adversarial examples. Extensive experiments
demonstrate that RJA-MMR outperforms current state-of-the-art methods in attack
performance, imperceptibility, fluency and grammar correctness. | cs.CR | Reversible Jump Attack to Textual Classifiers with Modification Reduction |
2024-03-21T04:23:56Z | http://arxiv.org/pdf/2403.15482v1 | http://arxiv.org/abs/2403.15482v1 | Alicja Chaszczewicz, Raj Sanjay Shah, Ryan Louie, Bruce A Arnow, Robert Kraut, Diyi Yang | 2024-03-21T04:23:56Z | cs.CL, cs.HC, cs.LG | Realistic practice and tailored feedback are key processes for training peer
counselors with clinical skills. However, existing mechanisms of providing
feedback largely rely on human supervision. Peer counselors often lack
mechanisms to receive detailed feedback from experienced mentors, making it
difficult for them to support the large number of people with mental health
issues who use peer counseling. Our work aims to leverage large language models
to provide contextualized and multi-level feedback to empower peer counselors,
especially novices, at scale. To achieve this, we co-design with a group of
senior psychotherapy supervisors to develop a multi-level feedback taxonomy,
and then construct a publicly available dataset with comprehensive feedback
annotations of 400 emotional support conversations. We further design a
self-improvement method on top of large language models to enhance the
automatic generation of feedback. Via qualitative and quantitative evaluation
with domain experts, we demonstrate that our method minimizes the risk of
potentially harmful and low-quality feedback generation which is desirable in
such high-stakes scenarios. | cs.CL | Multi-Level Feedback Generation with Large Language Models for Empowering Novice Peer Counselors |
2024-03-21T04:08:29Z | http://arxiv.org/pdf/2403.14119v3 | http://arxiv.org/abs/2403.14119v3 | Hee Suk Yoon, Eunseop Yoon, Joshua Tian Jin Tee, Mark Hasegawa-Johnson, Yingzhen Li, Chang D. Yoo | 2024-03-31T13:36:54Z | cs.CV, cs.AI, cs.CL, cs.LG | In deep learning, test-time adaptation has gained attention as a method for
model fine-tuning without the need for labeled data. A prime exemplification is
the recently proposed test-time prompt tuning for large-scale vision-language
models such as CLIP. Unfortunately, these prompts have been mainly developed to
improve accuracy, overlooking the importance of calibration, which is a crucial
aspect for quantifying prediction uncertainty. However, traditional calibration
methods rely on substantial amounts of labeled data, making them impractical
for test-time scenarios. To this end, this paper explores calibration during
test-time prompt tuning by leveraging the inherent properties of CLIP. Through
a series of observations, we find that the prompt choice significantly affects
the calibration in CLIP, where the prompts leading to higher text feature
dispersion result in better-calibrated predictions. Introducing the Average
Text Feature Dispersion (ATFD), we establish its relationship with calibration
error and present a novel method, Calibrated Test-time Prompt Tuning (C-TPT),
for optimizing prompts during test-time with enhanced calibration. Through
extensive experiments on different CLIP architectures and datasets, we show
that C-TPT can effectively improve the calibration of test-time prompt tuning
without needing labeled data. The code is publicly accessible at
https://github.com/hee-suk-yoon/C-TPT. | cs.CV | C-TPT: Calibrated Test-Time Prompt Tuning for Vision-Language Models via Text Feature Dispersion |
2024-03-21T04:07:40Z | http://arxiv.org/pdf/2403.14118v1 | http://arxiv.org/abs/2403.14118v1 | Haofei Zhao, Yilun Liu, Shimin Tao, Weibin Meng, Yimeng Chen, Xiang Geng, Chang Su, Min Zhang, Hao Yang | 2024-03-21T04:07:40Z | cs.CL | Machine Translation Quality Estimation (MTQE) is the task of estimating the
quality of machine-translated text in real time without the need for reference
translations, which is of great importance for the development of MT. After two
decades of evolution, QE has yielded a wealth of results. This article provides
a comprehensive overview of QE datasets, annotation methods, shared tasks,
methodologies, challenges, and future research directions. It begins with an
introduction to the background and significance of QE, followed by an
explanation of the concepts and evaluation metrics for word-level QE,
sentence-level QE, document-level QE, and explainable QE. The paper categorizes
the methods developed throughout the history of QE into those based on
handcrafted features, deep learning, and Large Language Models (LLMs), with a
further division of deep learning-based methods into classic deep learning and
those incorporating pre-trained language models (LMs). Additionally, the
article details the advantages and limitations of each method and offers a
straightforward comparison of different approaches. Finally, the paper
discusses the current challenges in QE research and provides an outlook on
future research directions. | cs.CL | From Handcrafted Features to LLMs: A Brief Survey for Machine Translation Quality Estimation |
2024-03-21T04:03:16Z | http://arxiv.org/pdf/2403.14117v2 | http://arxiv.org/abs/2403.14117v2 | Mina Lee, Katy Ilonka Gero, John Joon Young Chung, Simon Buckingham Shum, Vipul Raheja, Hua Shen, Subhashini Venugopalan, Thiemo Wambsganss, David Zhou, Emad A. Alghamdi, Tal August, Avinash Bhat, Madiha Zahrah Choksi, Senjuti Dutta, Jin L. C. Guo, Md Naimul Hoque, Yewon Kim, Simon Knight, Seyed Parsa Neshaei, Agnia Sergeyuk, Antonette Shibani, Disha Shrivastava, Lila Shroff, Jessi Stark, Sarah Sterman, Sitong Wang, Antoine Bosselut, Daniel Buschek, Joseph Chee Chang, Sherol Chen, Max Kreminski, Joonsuk Park, Roy Pea, Eugenia H. Rho, Shannon Zejiang Shen, Pao Siangliulue | 2024-03-26T12:53:14Z | cs.HC, cs.CL | In our era of rapid technological advancement, the research landscape for
writing assistants has become increasingly fragmented across various research
communities. We seek to address this challenge by proposing a design space as a
structured way to examine and explore the multidimensional space of intelligent
and interactive writing assistants. Through a large community collaboration, we
explore five aspects of writing assistants: task, user, technology,
interaction, and ecosystem. Within each aspect, we define dimensions (i.e.,
fundamental components of an aspect) and codes (i.e., potential options for
each dimension) by systematically reviewing 115 papers. Our design space aims
to offer researchers and designers a practical tool to navigate, comprehend,
and compare the various possibilities of writing assistants, and aid in the
envisioning and design of new writing assistants. | cs.HC | A Design Space for Intelligent and Interactive Writing Assistants |
2024-03-21T03:52:01Z | http://arxiv.org/pdf/2403.14112v2 | http://arxiv.org/abs/2403.14112v2 | Jiaxing Sun, Weiquan Huang, Jiang Wu, Chenya Gu, Wei Li, Songyang Zhang, Hang Yan, Conghui He | 2024-04-19T06:57:21Z | cs.CL | We introduce CHARM, the first benchmark for comprehensively and in-depth
evaluating the commonsense reasoning ability of large language models (LLMs) in
Chinese, which covers both globally known and Chinese-specific commonsense. We
evaluated 7 English and 12 Chinese-oriented LLMs on CHARM, employing 5
representative prompt strategies for improving LLMs' reasoning ability, such as
Chain-of-Thought. Our findings indicate that the LLM's language orientation and
the task's domain influence the effectiveness of the prompt strategy, which
enriches previous research findings. We built closely-interconnected reasoning
and memorization tasks, and found that some LLMs struggle with memorizing
Chinese commonsense, affecting their reasoning ability, while others show
differences in reasoning despite similar memorization performance. We also
evaluated the LLMs' memorization-independent reasoning abilities and analyzed
the typical errors. Our study precisely identified the LLMs' strengths and
weaknesses, providing the clear direction for optimization. It can also serve
as a reference for studies in other fields. We will release CHARM at
https://github.com/opendatalab/CHARM . | cs.CL | Benchmarking Chinese Commonsense Reasoning of LLMs: From Chinese-Specifics to Reasoning-Memorization Correlations |
2024-03-21T03:24:01Z | http://arxiv.org/pdf/2403.14101v1 | http://arxiv.org/abs/2403.14101v1 | Minh-Tuan Tran, Trung Le, Xuan-May Le, Mehrtash Harandi, Dinh Phung | 2024-03-21T03:24:01Z | cs.CV, cs.CL, cs.LG | Federated Class-Incremental Learning (FCIL) is an underexplored yet pivotal
issue, involving the dynamic addition of new classes in the context of
federated learning. In this field, Data-Free Knowledge Transfer (DFKT) plays a
crucial role in addressing catastrophic forgetting and data privacy problems.
However, prior approaches lack the crucial synergy between DFKT and the model
training phases, causing DFKT to encounter difficulties in generating
high-quality data from a non-anchored latent space of the old task model. In
this paper, we introduce LANDER (Label Text Centered Data-Free Knowledge
Transfer) to address this issue by utilizing label text embeddings (LTE)
produced by pretrained language models. Specifically, during the model training
phase, our approach treats LTE as anchor points and constrains the feature
embeddings of corresponding training samples around them, enriching the
surrounding area with more meaningful information. In the DFKT phase, by using
these LTE anchors, LANDER can synthesize more meaningful samples, thereby
effectively addressing the forgetting problem. Additionally, instead of tightly
constraining embeddings toward the anchor, the Bounding Loss is introduced to
encourage sample embeddings to remain flexible within a defined radius. This
approach preserves the natural differences in sample embeddings and mitigates
the embedding overlap caused by heterogeneous federated settings. Extensive
experiments conducted on CIFAR100, Tiny-ImageNet, and ImageNet demonstrate that
LANDER significantly outperforms previous methods and achieves state-of-the-art
performance in FCIL. The code is available at
https://github.com/tmtuan1307/lander. | cs.CV | Text-Enhanced Data-free Approach for Federated Class-Incremental Learning |
2024-03-21T01:52:07Z | http://arxiv.org/pdf/2403.14074v1 | http://arxiv.org/abs/2403.14074v1 | Yang Bai, Anthony Colas, Christan Grant, Daisy Zhe Wang | 2024-03-21T01:52:07Z | cs.IR, cs.CL, cs.LG | In recent research, contrastive learning has proven to be a highly effective
method for representation learning and is widely used for dense retrieval.
However, we identify that relying solely on contrastive learning can lead to
suboptimal retrieval performance. On the other hand, despite many retrieval
datasets supporting various learning objectives beyond contrastive learning,
combining them efficiently in multi-task learning scenarios can be challenging.
In this paper, we introduce M3, an advanced recursive Multi-hop dense sentence
retrieval system built upon a novel Multi-task Mixed-objective approach for
dense text representation learning, addressing the aforementioned challenges.
Our approach yields state-of-the-art performance on a large-scale open-domain
fact verification benchmark dataset, FEVER. Code and data are available at:
https://github.com/TonyBY/M3 | cs.IR | M3: A Multi-Task Mixed-Objective Learning Framework for Open-Domain Multi-Hop Dense Sentence Retrieval |
2024-03-21T01:47:22Z | http://arxiv.org/pdf/2403.14072v1 | http://arxiv.org/abs/2403.14072v1 | Margaret Y. Li, Alisa Liu, Zhaofeng Wu, Noah A. Smith | 2024-03-21T01:47:22Z | cs.CL | Ambiguity is an critical component of language that allows for more effective
communication between speakers, but is often ignored in NLP. Recent work
suggests that NLP systems may struggle to grasp certain elements of human
language understanding because they may not handle ambiguities at the level
that humans naturally do in communication. Additionally, different types of
ambiguity may serve different purposes and require different approaches for
resolution, and we aim to investigate how language models' abilities vary
across types. We propose a taxonomy of ambiguity types as seen in English to
facilitate NLP analysis. Our taxonomy can help make meaningful splits in
language ambiguity data, allowing for more fine-grained assessments of both
datasets and model performance. | cs.CL | A Taxonomy of Ambiguity Types for NLP |
2024-03-21T00:21:38Z | http://arxiv.org/pdf/2403.14727v1 | http://arxiv.org/abs/2403.14727v1 | Hadas Kotek, David Q. Sun, Zidi Xiu, Margit Bowler, Christopher Klein | 2024-03-21T00:21:38Z | cs.CY, cs.CL, cs.LG | As modern Large Language Models (LLMs) shatter many state-of-the-art
benchmarks in a variety of domains, this paper investigates their behavior in
the domains of ethics and fairness, focusing on protected group bias. We
conduct a two-part study: first, we solicit sentence continuations describing
the occupations of individuals from different protected groups, including
gender, sexuality, religion, and race. Second, we have the model generate
stories about individuals who hold different types of occupations. We collect
>10k sentence completions made by a publicly available LLM, which we subject to
human annotation. We find bias across minoritized groups, but in particular in
the domains of gender and sexuality, as well as Western bias, in model
generations. The model not only reflects societal biases, but appears to
amplify them. The model is additionally overly cautious in replies to queries
relating to minoritized groups, providing responses that strongly emphasize
diversity and equity to an extent that other group characteristics are
overshadowed. This suggests that artificially constraining potentially harmful
outputs may itself lead to harm, and should be applied in a careful and
controlled manner. | cs.CY | Protected group bias and stereotypes in Large Language Models |
2024-03-21T00:20:16Z | http://arxiv.org/pdf/2403.14050v2 | http://arxiv.org/abs/2403.14050v2 | Mahdi Rezapour | 2024-03-23T14:46:59Z | cs.CL, cs.LG, stat.AP | Sentiment analysis is a natural language processing task that aims to
identify and extract the emotional aspects of a text. However, many existing
sentiment analysis methods primarily classify the overall polarity of a text,
overlooking the specific phrases that convey sentiment. In this paper, we
applied an approach to sentiment analysis based on a question-answering
framework. Our approach leverages the power of Bidirectional Autoregressive
Transformer (BART), a pre-trained sequence-to-sequence model, to extract a
phrase from a given text that amplifies a given sentiment polarity. We create a
natural language question that identifies the specific emotion to extract and
then guide BART to pay attention to the relevant emotional cues in the text. We
use a classifier within BART to predict the start and end positions of the
answer span within the text, which helps to identify the precise boundaries of
the extracted emotion phrase. Our approach offers several advantages over most
sentiment analysis studies, including capturing the complete context and
meaning of the text and extracting precise token spans that highlight the
intended sentiment. We achieved an end loss of 87% and Jaccard score of 0.61. | cs.CL | Extracting Emotion Phrases from Tweets using BART |
2024-03-21T00:13:59Z | http://arxiv.org/pdf/2403.14048v1 | http://arxiv.org/abs/2403.14048v1 | Alice Baird, Rachel Manzelli, Panagiotis Tzirakis, Chris Gagne, Haoqi Li, Sadie Allen, Sander Dieleman, Brian Kulis, Shrikanth S. Narayanan, Alan Cowen | 2024-03-21T00:13:59Z | cs.SD, cs.CL, eess.AS | The NeurIPS 2023 Machine Learning for Audio Workshop brings together machine
learning (ML) experts from various audio domains. There are several valuable
audio-driven ML tasks, from speech emotion recognition to audio event
detection, but the community is sparse compared to other ML areas, e.g.,
computer vision or natural language processing. A major limitation with audio
is the available data; with audio being a time-dependent modality, high-quality
data collection is time-consuming and costly, making it challenging for
academic groups to apply their often state-of-the-art strategies to a larger,
more generalizable dataset. In this short white paper, to encourage researchers
with limited access to large-datasets, the organizers first outline several
open-source datasets that are available to the community, and for the duration
of the workshop are making several propriety datasets available. Namely, three
vocal datasets, Hume-Prosody, Hume-VocalBurst, an acted emotional speech
dataset Modulate-Sonata, and an in-game streamer dataset Modulate-Stream. We
outline the current baselines on these datasets but encourage researchers from
across audio to utilize them outside of the initial baseline tasks. | cs.SD | The NeurIPS 2023 Machine Learning for Audio Workshop: Affective Audio Benchmarks and Novel Data |
2024-03-20T23:21:35Z | http://arxiv.org/pdf/2403.14037v1 | http://arxiv.org/abs/2403.14037v1 | Sheetal Harris, Jinshuo Liu, Hassan Jalil Hadi, Yue Cao | 2024-03-20T23:21:35Z | cs.CL, cs.AI | Misinformation can seriously impact society, affecting anything from public
opinion to institutional confidence and the political horizon of a state. Fake
News (FN) proliferation on online websites and Online Social Networks (OSNs)
has increased profusely. Various fact-checking websites include news in English
and barely provide information about FN in regional languages. Thus the Urdu FN
purveyors cannot be discerned using factchecking portals. SOTA approaches for
Fake News Detection (FND) count upon appropriately labelled and large datasets.
FND in regional and resource-constrained languages lags due to the lack of
limited-sized datasets and legitimate lexical resources. The previous datasets
for Urdu FND are limited-sized, domain-restricted, publicly unavailable and not
manually verified where the news is translated from English into Urdu. In this
paper, we curate and contribute the first largest publicly available dataset
for Urdu FND, Ax-to-Grind Urdu, to bridge the identified gaps and limitations
of existing Urdu datasets in the literature. It constitutes 10,083 fake and
real news on fifteen domains collected from leading and authentic Urdu
newspapers and news channel websites in Pakistan and India. FN for the
Ax-to-Grind dataset is collected from websites and crowdsourcing. The dataset
contains news items in Urdu from the year 2017 to the year 2023. Expert
journalists annotated the dataset. We benchmark the dataset with an ensemble
model of mBERT,XLNet, and XLM RoBERTa. The selected models are originally
trained on multilingual large corpora. The results of the proposed model are
based on performance metrics, F1-score, accuracy, precision, recall and MCC
value. | cs.CL | Ax-to-Grind Urdu: Benchmark Dataset for Urdu Fake News Detection |
2024-03-20T22:14:39Z | http://arxiv.org/pdf/2403.14009v1 | http://arxiv.org/abs/2403.14009v1 | Ona de Gibert, Graeme Nail, Nikolay Arefyev, Marta Bañón, Jelmer van der Linde, Shaoxiong Ji, Jaume Zaragoza-Bernabeu, Mikko Aulamo, Gema Ramírez-Sánchez, Andrey Kutuzov, Sampo Pyysalo, Stephan Oepen, Jörg Tiedemann | 2024-03-20T22:14:39Z | cs.CL | We present the HPLT (High Performance Language Technologies) language
resources, a new massive multilingual dataset including both monolingual and
bilingual corpora extracted from CommonCrawl and previously unused web crawls
from the Internet Archive. We describe our methods for data acquisition,
management and processing of large corpora, which rely on open-source software
tools and high-performance computing. Our monolingual collection focuses on
low- to medium-resourced languages and covers 75 languages and a total of ~5.6
trillion word tokens de-duplicated on the document level. Our English-centric
parallel corpus is derived from its monolingual counterpart and covers 18
language pairs and more than 96 million aligned sentence pairs with roughly 1.4
billion English tokens. The HPLT language resources are one of the largest open
text corpora ever released, providing a great resource for language modeling
and machine translation training. We publicly release the corpora, the
software, and the tools used in this work. | cs.CL | A New Massive Multilingual Dataset for High-Performance Language Technologies |
2024-03-20T22:11:01Z | http://arxiv.org/pdf/2403.14006v1 | http://arxiv.org/abs/2403.14006v1 | Mostafa M. Amin, Björn W. Schuller | 2024-03-20T22:11:01Z | cs.CL, cs.AI | Recent studies have demonstrated the emerging capabilities of foundation
models like ChatGPT in several fields, including affective computing. However,
accessing these emerging capabilities is facilitated through prompt
engineering. Despite the existence of some prompting techniques, the field is
still rapidly evolving and many prompting ideas still require investigation. In
this work, we introduce a method to evaluate and investigate the sensitivity of
the performance of foundation models based on different prompts or generation
parameters. We perform our evaluation on ChatGPT within the scope of affective
computing on three major problems, namely sentiment analysis, toxicity
detection, and sarcasm detection. First, we carry out a sensitivity analysis on
pivotal parameters in auto-regressive text generation, specifically the
temperature parameter $T$ and the top-$p$ parameter in Nucleus sampling,
dictating how conservative or creative the model should be during generation.
Furthermore, we explore the efficacy of several prompting ideas, where we
explore how giving different incentives or structures affect the performance.
Our evaluation takes into consideration performance measures on the affective
computing tasks, and the effectiveness of the model to follow the stated
instructions, hence generating easy-to-parse responses to be smoothly used in
downstream applications. | cs.CL | On Prompt Sensitivity of ChatGPT in Affective Computing |
2024-03-20T22:05:18Z | http://arxiv.org/pdf/2403.14003v1 | http://arxiv.org/abs/2403.14003v1 | Alessandro Favero, Luca Zancato, Matthew Trager, Siddharth Choudhary, Pramuditha Perera, Alessandro Achille, Ashwin Swaminathan, Stefano Soatto | 2024-03-20T22:05:18Z | cs.CV, cs.CL, cs.LG | Generative Vision-Language Models (VLMs) are prone to generate
plausible-sounding textual answers that, however, are not always grounded in
the input image. We investigate this phenomenon, usually referred to as
"hallucination" and show that it stems from an excessive reliance on the
language prior. In particular, we show that as more tokens are generated, the
reliance on the visual prompt decreases, and this behavior strongly correlates
with the emergence of hallucinations. To reduce hallucinations, we introduce
Multi-Modal Mutual-Information Decoding (M3ID), a new sampling method for
prompt amplification. M3ID amplifies the influence of the reference image over
the language prior, hence favoring the generation of tokens with higher mutual
information with the visual prompt. M3ID can be applied to any pre-trained
autoregressive VLM at inference time without necessitating further training and
with minimal computational overhead. If training is an option, we show that
M3ID can be paired with Direct Preference Optimization (DPO) to improve the
model's reliance on the prompt image without requiring any labels. Our
empirical findings show that our algorithms maintain the fluency and linguistic
capabilities of pre-trained VLMs while reducing hallucinations by mitigating
visually ungrounded answers. Specifically, for the LLaVA 13B model, M3ID and
M3ID+DPO reduce the percentage of hallucinated objects in captioning tasks by
25% and 28%, respectively, and improve the accuracy on VQA benchmarks such as
POPE by 21% and 24%. | cs.CV | Multi-Modal Hallucination Control by Visual Information Grounding |
2024-03-20T21:58:32Z | http://arxiv.org/pdf/2403.14001v1 | http://arxiv.org/abs/2403.14001v1 | Gaifan Zhang, Yi Zhou, Danushka Bollegala | 2024-03-20T21:58:32Z | cs.CL | Sentence embeddings produced by Pretrained Language Models (PLMs) have
received wide attention from the NLP community due to their superior
performance when representing texts in numerous downstream applications.
However, the high dimensionality of the sentence embeddings produced by PLMs is
problematic when representing large numbers of sentences in memory- or
compute-constrained devices. As a solution, we evaluate unsupervised
dimensionality reduction methods to reduce the dimensionality of sentence
embeddings produced by PLMs. Our experimental results show that simple methods
such as Principal Component Analysis (PCA) can reduce the dimensionality of
sentence embeddings by almost $50\%$, without incurring a significant loss in
performance in multiple downstream tasks. Surprisingly, reducing the
dimensionality further improves performance over the original high-dimensional
versions for the sentence embeddings produced by some PLMs in some tasks. | cs.CL | Evaluating Unsupervised Dimensionality Reduction Methods for Pretrained Sentence Embeddings |
2024-03-20T21:53:56Z | http://arxiv.org/pdf/2403.14725v2 | http://arxiv.org/abs/2403.14725v2 | Taeyoun Kim, Suhas Kotha, Aditi Raghunathan | 2024-06-24T05:01:06Z | cs.CR, cs.CL, cs.LG | The rise of "jailbreak" attacks on language models has led to a flurry of
defenses aimed at preventing undesirable responses. We critically examine the
two stages of the defense pipeline: (i) defining what constitutes unsafe
outputs, and (ii) enforcing the definition via methods such as input processing
or fine-tuning. To test the efficacy of existing enforcement mechanisms, we
consider a simple and well-specified definition of unsafe outputs--outputs that
contain the word "purple". Surprisingly, existing fine-tuning and input
defenses fail on this simple problem, casting doubt on whether enforcement
algorithms can be robust for more complicated definitions. We find that real
safety benchmarks similarly test enforcement for a fixed definition. We hope
that future research can lead to effective/fast enforcement as well as high
quality definitions used for enforcement and evaluation. | cs.CR | Testing the Limits of Jailbreaking Defenses with the Purple Problem |
2024-03-20T21:16:10Z | http://arxiv.org/pdf/2403.15478v1 | http://arxiv.org/abs/2403.15478v1 | Rika Tanaka, Yusuke Fukazawa | 2024-03-20T21:16:10Z | cs.CL | We propose a method that integrates supervised extractive and generative
language models for providing supporting evidence of suicide risk in the
CLPsych 2024 shared task. Our approach comprises three steps. Initially, we
construct a BERT-based model for estimating sentence-level suicide risk and
negative sentiment. Next, we precisely identify high suicide risk sentences by
emphasizing elevated probabilities of both suicide risk and negative sentiment.
Finally, we integrate generative summaries using the MentaLLaMa framework and
extractive summaries from identified high suicide risk sentences and a
specialized dictionary of suicidal risk words. SophiaADS, our team, achieved
1st place for highlight extraction and ranked 10th for summary generation, both
based on recall and consistency metrics, respectively. | cs.CL | Integrating Supervised Extractive and Generative Language Models for Suicide Risk Evidence Summarization |
2024-03-20T21:06:42Z | http://arxiv.org/pdf/2404.00027v3 | http://arxiv.org/abs/2404.00027v3 | Azmine Toushik Wasi, Mst Rafia Islam, Raima Islam | 2024-04-22T08:30:30Z | cs.HC, cs.AI, cs.CL, cs.CY, cs.LG | Sense of ownership in writing confines our investment of thoughts, time, and
contribution, leading to attachment to the output. However, using writing
assistants introduces a mental dilemma, as some content isn't directly our
creation. For instance, we tend to credit Large Language Models (LLMs) more in
creative tasks, even though all tasks are equal for them. Additionally, while
we may not claim complete ownership of LLM-generated content, we freely claim
authorship. We conduct a short survey to examine these issues and understand
underlying cognitive processes in order to gain a better knowledge of
human-computer interaction in writing and improve writing aid systems. | cs.HC | LLMs as Writing Assistants: Exploring Perspectives on Sense of Ownership and Reasoning |
2024-03-20T21:02:16Z | http://arxiv.org/pdf/2404.00026v3 | http://arxiv.org/abs/2404.00026v3 | Azmine Toushik Wasi, Raima Islam, Mst Rafia Islam | 2024-04-22T08:30:28Z | cs.HC, cs.AI, cs.CL, cs.IR, cs.LG | Individuality and personalization comprise the distinctive characteristics
that make each writer unique and influence their words in order to effectively
engage readers while conveying authenticity. However, our growing reliance on
LLM-based writing assistants risks compromising our creativity and
individuality over time. We often overlook the negative impacts of this trend
on our creativity and uniqueness, despite the possible consequences. This study
investigates these concerns by performing a brief survey to explore different
perspectives and concepts, as well as trying to understand people's viewpoints,
in conjunction with past studies in the area. Addressing these issues is
essential for improving human-computer interaction systems and enhancing
writing assistants for personalization and individuality. | cs.HC | Ink and Individuality: Crafting a Personalised Narrative in the Age of LLMs |
2024-03-20T18:59:18Z | http://arxiv.org/pdf/2403.13925v1 | http://arxiv.org/abs/2403.13925v1 | Devam Mondal, Carlo Lipizzi | 2024-03-20T18:59:18Z | cs.CL, cs.AI, cs.LG | Despite the growing capabilities of large language models, there exists
concerns about the biases they develop. In this paper, we propose a novel,
automated mechanism for debiasing through specified dataset augmentation in the
lens of bias producers and in the context of 'restricted industries' with
limited data. We additionally create two new additional metrics, the mb-index
and db-index, to quantify bias, considering the idea that bias occurs due to
both intrinsic model architecture and dataset. | cs.CL | Reducing Large Language Model Bias with Emphasis on 'Restricted Industries': Automated Dataset Augmentation and Prejudice Quantification |
2024-03-20T18:49:59Z | http://arxiv.org/pdf/2403.13922v1 | http://arxiv.org/abs/2403.13922v1 | Leanne Nortje, Dan Oneaţă, Yevgen Matusevych, Herman Kamper | 2024-03-20T18:49:59Z | cs.CL, eess.AS | When children learn new words, they employ constraints such as the mutual
exclusivity (ME) bias: a novel word is mapped to a novel object rather than a
familiar one. This bias has been studied computationally, but only in models
that use discrete word representations as input, ignoring the high variability
of spoken words. We investigate the ME bias in the context of visually grounded
speech models that learn from natural images and continuous speech audio.
Concretely, we train a model on familiar words and test its ME bias by asking
it to select between a novel and a familiar object when queried with a novel
word. To simulate prior acoustic and visual knowledge, we experiment with
several initialisation strategies using pretrained speech and vision networks.
Our findings reveal the ME bias across the different initialisation approaches,
with a stronger bias in models with more prior (in particular, visual)
knowledge. Additional tests confirm the robustness of our results, even when
different loss functions are considered. | cs.CL | Visually Grounded Speech Models have a Mutual Exclusivity Bias |
2024-03-20T18:18:48Z | http://arxiv.org/pdf/2403.13903v1 | http://arxiv.org/abs/2403.13903v1 | Fauzan Farooqui, Thanmay Jayakumar, Pulkit Mathur, Mansi Radke | 2024-03-20T18:18:48Z | cs.CL | Open Information Extraction (OIE) is a structured prediction (SP) task in
Natural Language Processing (NLP) that aims to extract structured $n$-ary
tuples - usually subject-relation-object triples - from free text. The word
embeddings in the input text can be enhanced with linguistic features, usually
Part-of-Speech (PoS) and Syntactic Dependency Parse (SynDP) labels. However,
past enhancement techniques cannot leverage the power of pretrained language
models (PLMs), which themselves have been hardly used for OIE. To bridge this
gap, we are the first to leverage linguistic features with a Seq2Seq PLM for
OIE. We do so by introducing two methods - Weighted Addition and Linearized
Concatenation. Our work can give any neural OIE architecture the key
performance boost from both PLMs and linguistic features in one go. In our
settings, this shows wide improvements of up to 24.9%, 27.3% and 14.9% on
Precision, Recall and F1 scores respectively over the baseline. Beyond this, we
address other important challenges in the field: to reduce compute overheads
with the features, we are the first ones to exploit Semantic Dependency Parse
(SemDP) tags; to address flaws in current datasets, we create a clean synthetic
dataset; finally, we contribute the first known study of OIE behaviour in SP
models. | cs.CL | Leveraging Linguistically Enhanced Embeddings for Open Information Extraction |
2024-03-20T18:13:17Z | http://arxiv.org/pdf/2403.13901v1 | http://arxiv.org/abs/2403.13901v1 | Tyler Loakman, Chen Tang, Chenghua Lin | 2024-03-20T18:13:17Z | cs.CL | Previous work in phonologically and phonetically grounded language generation
has mainly focused on domains such as puns and poetry. In this article, we
present new work on the generation of tongue-twisters - a form of language that
is required to be conditioned on a phoneme level to maximize sound overlap,
whilst maintaining semantic consistency with an input topic and still being
grammatically correct. We present TwisterLister, a pipeline for generating
phonologically informed tongue-twisters from Large Language Models (LLMs) that
we use to generate TwistList 2.0, the largest annotated dataset of
tongue-twisters to date, consisting of 17K+ examples from a combination of
human and LLM authors. Our generation pipeline involves the use of a
phonologically constrained vocabulary alongside LLM prompting to generate
novel, non-derivative tongue-twister examples. We additionally present the
results of automatic and human evaluation of smaller models trained on our
generated dataset to demonstrate the extent to which phonologically motivated
language types can be generated without explicit injection of phonological
knowledge. Additionally, we introduce a Phoneme-Aware Constrained Decoding
module (PACD) that can be integrated into any causal language model and
demonstrate that this method generates good quality tongue-twisters both with
and without fine-tuning the underlying language model. We also design and
implement a range of automatic metrics for the task of tongue-twister
generation that is phonologically motivated and captures the unique essence of
tongue-twisters based on Phonemic Edit Distance (PED). | cs.CL | Train & Constrain: Phonologically Informed Tongue-Twister Generation from Topics and Paraphrases |
2024-03-20T17:59:43Z | http://arxiv.org/pdf/2403.13804v1 | http://arxiv.org/abs/2403.13804v1 | Ruozhen He, Paola Cascante-Bonilla, Ziyan Yang, Alexander C. Berg, Vicente Ordonez | 2024-03-20T17:59:43Z | cs.CV, cs.CL, cs.LG | We introduce SynGround, a novel framework that combines data-driven learning
and knowledge transfer from various large-scale pretrained models to enhance
the visual grounding capabilities of a pretrained vision-and-language model.
The knowledge transfer from the models initiates the generation of image
descriptions through an image description generator. These descriptions serve
dual purposes: they act as prompts for synthesizing images through a
text-to-image generator, and as queries for synthesizing text, from which
phrases are extracted using a large language model. Finally, we leverage an
open-vocabulary object detector to generate synthetic bounding boxes for the
synthetic images and texts. We finetune a pretrained vision-and-language model
on this dataset by optimizing a mask-attention consistency objective that
aligns region annotations with gradient-based model explanations. The resulting
model improves the grounding capabilities of an off-the-shelf
vision-and-language model. Particularly, SynGround improves the pointing game
accuracy of ALBEF on the Flickr30k dataset from 79.38% to 87.26%, and on
RefCOCO+ Test A from 69.35% to 79.06% and on RefCOCO+ Test B from 53.77% to
63.67%. | cs.CV | Learning from Models and Data for Visual Grounding |
2024-03-20T17:59:14Z | http://arxiv.org/pdf/2403.13802v2 | http://arxiv.org/abs/2403.13802v2 | Vincent Tao Hu, Stefan Andreas Baumann, Ming Gui, Olga Grebenkova, Pingchuan Ma, Johannes Fischer, Björn Ommer | 2024-04-01T17:58:02Z | cs.CV, cs.AI, cs.CL, cs.LG | The diffusion model has long been plagued by scalability and quadratic
complexity issues, especially within transformer-based structures. In this
study, we aim to leverage the long sequence modeling capability of a
State-Space Model called Mamba to extend its applicability to visual data
generation. Firstly, we identify a critical oversight in most current
Mamba-based vision methods, namely the lack of consideration for spatial
continuity in the scan scheme of Mamba. Secondly, building upon this insight,
we introduce a simple, plug-and-play, zero-parameter method named Zigzag Mamba,
which outperforms Mamba-based baselines and demonstrates improved speed and
memory utilization compared to transformer-based baselines. Lastly, we
integrate Zigzag Mamba with the Stochastic Interpolant framework to investigate
the scalability of the model on large-resolution visual datasets, such as
FacesHQ $1024\times 1024$ and UCF101, MultiModal-CelebA-HQ, and MS COCO
$256\times 256$ . Code will be released at https://taohu.me/zigma/ | cs.CV | ZigMa: A DiT-style Zigzag Mamba Diffusion Model |
2024-03-20T17:58:12Z | http://arxiv.org/pdf/2403.13801v2 | http://arxiv.org/abs/2403.13801v2 | Yusuke Mikami, Andrew Melnik, Jun Miura, Ville Hautamäki | 2024-04-06T04:12:47Z | cs.RO, cs.AI, cs.CL, I.2.9; I.2.7 | We demonstrate experimental results with LLMs that address robotics task
planning problems. Recently, LLMs have been applied in robotics task planning,
particularly using a code generation approach that converts complex high-level
instructions into mid-level policy codes. In contrast, our approach acquires
text descriptions of the task and scene objects, then formulates task planning
through natural language reasoning, and outputs coordinate level control
commands, thus reducing the necessity for intermediate representation code as
policies with pre-defined APIs. Our approach is evaluated on a multi-modal
prompt simulation benchmark, demonstrating that our prompt engineering
experiments with natural language reasoning significantly enhance success rates
compared to its absence. Furthermore, our approach illustrates the potential
for natural language descriptions to transfer robotics skills from known tasks
to previously unseen tasks. The project website:
https://natural-language-as-policies.github.io/ | cs.RO | Natural Language as Policies: Reasoning for Coordinate-Level Embodied Control with LLMs |
2024-03-20T17:55:35Z | http://arxiv.org/pdf/2403.13799v3 | http://arxiv.org/abs/2403.13799v3 | Olga Golovneva, Zeyuan Allen-Zhu, Jason Weston, Sainbayar Sukhbaatar | 2024-05-07T20:35:15Z | cs.CL, cs.AI | Large language models (LLMs) have a surprising failure: when trained on "A
has a feature B", they do not generalize to "B is a feature of A", which is
termed the Reversal Curse. Even when training with trillions of tokens this
issue still appears due to Zipf's law - hence even if we train on the entire
internet. This work proposes an alternative training scheme, called reverse
training, whereby all words are used twice, doubling the amount of available
tokens. The LLM is trained in both forward and reverse directions by reversing
the training strings while preserving (i.e., not reversing) chosen substrings,
such as entities. We show that data-matched reverse-trained models provide
superior performance to standard models on standard tasks, and compute-matched
reverse-trained models provide far superior performance on reversal tasks,
helping resolve the reversal curse issue. | cs.CL | Reverse Training to Nurse the Reversal Curse |
2024-03-20T17:47:49Z | http://arxiv.org/pdf/2403.13786v2 | http://arxiv.org/abs/2403.13786v2 | Guangzeng Han, Weisi Liu, Xiaolei Huang, Brian Borsari | 2024-03-23T04:02:01Z | cs.CL | Automatic coding patient behaviors is essential to support decision making
for psychotherapists during the motivational interviewing (MI), a collaborative
communication intervention approach to address psychiatric issues, such as
alcohol and drug addiction. While the behavior coding task has rapidly adapted
machine learning to predict patient states during the MI sessions, lacking of
domain-specific knowledge and overlooking patient-therapist interactions are
major challenges in developing and deploying those models in real practice. To
encounter those challenges, we introduce the Chain-of-Interaction (CoI)
prompting method aiming to contextualize large language models (LLMs) for
psychiatric decision support by the dyadic interactions. The CoI prompting
approach systematically breaks down the coding task into three key reasoning
steps, extract patient engagement, learn therapist question strategies, and
integrates dyadic interactions between patients and therapists. This approach
enables large language models to leverage the coding scheme, patient state, and
domain knowledge for patient behavioral coding. Experiments on real-world
datasets can prove the effectiveness and flexibility of our prompting method
with multiple state-of-the-art LLMs over existing prompting baselines. We have
conducted extensive ablation analysis and demonstrate the critical role of
dyadic interactions in applying LLMs for psychotherapy behavior understanding. | cs.CL | Chain-of-Interaction: Enhancing Large Language Models for Psychiatric Behavior Understanding by Dyadic Contexts |
2024-03-20T17:42:08Z | http://arxiv.org/pdf/2403.13780v1 | http://arxiv.org/abs/2403.13780v1 | Jaehun Jung, Ximing Lu, Liwei Jiang, Faeze Brahman, Peter West, Pang Wei Koh, Yejin Choi | 2024-03-20T17:42:08Z | cs.CL, cs.AI | The current winning recipe for automatic summarization is using proprietary
large-scale language models (LLMs) such as ChatGPT as is, or imitation learning
from them as teacher models. While increasingly ubiquitous dependence on such
large-scale language models is convenient, there remains an important question
of whether small-scale models could have achieved competitive results, if we
were to seek an alternative learning method -- that allows for a more
cost-efficient, controllable, yet powerful summarizer. We present InfoSumm, a
novel framework to distill a powerful summarizer based on the
information-theoretic objective for summarization, without relying on either
the LLM's capability or human-written references. To achieve this, we first
propose a novel formulation of the desiderata of summarization (saliency,
faithfulness and brevity) through the lens of mutual information between the
original document and the summary. Based on this formulation, we start off from
Pythia-2.8B as the teacher model, which is not yet capable of summarization,
then self-train the model to optimize for the information-centric measures of
ideal summaries. Distilling from the improved teacher, we arrive at a compact
but powerful summarizer with only 568M parameters that performs competitively
against ChatGPT, without ever relying on ChatGPT's capabilities. Extensive
analysis demonstrates that our approach outperforms in-domain supervised models
in human evaluation, let alone state-of-the-art unsupervised methods, and wins
over ChatGPT in controllable summarization. | cs.CL | Information-Theoretic Distillation for Reference-less Summarization |
2024-03-20T17:01:56Z | http://arxiv.org/pdf/2403.13754v1 | http://arxiv.org/abs/2403.13754v1 | Catherine Arnett, Pamela D. Rivière, Tyler A. Chang, Sean Trott | 2024-03-20T17:01:56Z | cs.CL | The relationship between language model tokenization and performance is an
open area of research. Here, we investigate how different tokenization schemes
impact number agreement in Spanish plurals. We find that
morphologically-aligned tokenization performs similarly to other tokenization
schemes, even when induced artificially for words that would not be tokenized
that way during training. We then present exploratory analyses demonstrating
that language model embeddings for different plural tokenizations have similar
distributions along the embedding space axis that maximally distinguishes
singular and plural nouns. Our results suggest that morphologically-aligned
tokenization is a viable tokenization approach, and existing models already
generalize some morphological patterns to new items. However, our results
indicate that morphological tokenization is not strictly required for
performance. | cs.CL | Different Tokenization Schemes Lead to Comparable Performance in Spanish Number Agreement |
2024-03-20T16:43:42Z | http://arxiv.org/pdf/2403.13737v4 | http://arxiv.org/abs/2403.13737v4 | Atnafu Lambebo Tonja, Israel Abebe Azime, Tadesse Destaw Belay, Mesay Gemeda Yigezu, Moges Ahmed Mehamed, Abinew Ali Ayele, Ebrahim Chekol Jibril, Michael Melese Woldeyohannis, Olga Kolesnikova, Philipp Slusallek, Dietrich Klakow, Shengwu Xiong, Seid Muhie Yimam | 2024-06-23T12:06:27Z | cs.CL | Large language models (LLMs) have gained popularity recently due to their
outstanding performance in various downstream Natural Language Processing (NLP)
tasks. However, low-resource languages are still lagging behind current
state-of-the-art (SOTA) developments in the field of NLP due to insufficient
resources to train LLMs. Ethiopian languages exhibit remarkable linguistic
diversity, encompassing a wide array of scripts, and are imbued with profound
religious and cultural significance. This paper introduces EthioLLM --
multilingual large language models for five Ethiopian languages (Amharic,
Ge'ez, Afan Oromo, Somali, and Tigrinya) and English, and Ethiobenchmark -- a
new benchmark dataset for various downstream NLP tasks. We evaluate the
performance of these models across five downstream NLP tasks. We open-source
our multilingual language models, new benchmark datasets for various downstream
tasks, and task-specific fine-tuned language models and discuss the performance
of the models. Our dataset and models are available at the
https://huggingface.co/EthioNLP repository. | cs.CL | EthioLLM: Multilingual Large Language Models for Ethiopian Languages with Task Evaluation |
2024-03-20T16:24:10Z | http://arxiv.org/pdf/2403.15473v1 | http://arxiv.org/abs/2403.15473v1 | Marcin Pietron, Rafał Olszowski, Jakub Gomułka | 2024-03-20T16:24:10Z | cs.CL | Argument mining (AM) is defined as the task of automatically identifying and
extracting argumentative components (e.g. premises, claims, etc.) and detecting
the existing relations among them (i.e., support, attack, no relations). Deep
learning models enable us to analyze arguments more efficiently than
traditional methods and extract their semantics. This paper presents
comparative studies between a few deep learning-based models in argument
mining. The work concentrates on argument classification. The research was done
on a wide spectrum of datasets (Args.me, UKP, US2016). The main novelty of this
paper is the ensemble model which is based on BERT architecture and ChatGPT-4
as fine tuning model. The presented results show that BERT+ChatGPT-4
outperforms the rest of the models including other Transformer-based and
LSTM-based models. The observed improvement is, in most cases, greater than
10The presented analysis can provide crucial insights into how the models for
argument classification should be further improved. Additionally, it can help
develop a prompt-based algorithm to eliminate argument classification errors. | cs.CL | Efficient argument classification with compact language models and ChatGPT-4 refinements |
2024-03-20T15:39:54Z | http://arxiv.org/pdf/2403.13681v1 | http://arxiv.org/abs/2403.13681v1 | Mitodru Niyogi, Arnab Bhattacharya | 2024-03-20T15:39:54Z | cs.CL, cs.AI, cs.LG | In this paper, we present PARAMANU-AYN, a language model based exclusively on
case documents of the Supreme Court of India, the Constitution of India, and
the Indian Penal Code. The novel Auto Regressive (AR) decoder based model is
pretrained from scratch at a context size of 8192. We evaluated our pretrained
legal model on perplexity metrics. We also instruction-tuned our pretrained
model on a set of 10,763 instructions covering various legal tasks such as
legal reasoning, judgement explanation, legal clause generation, legal
drafting, legal contract drafting, case summarization, constitutional
question-answering, etc. We also evaluated the responses of prompts for
instruction-tuned models by GPT-3.5-Turbo on clarity, relevance, completeness,
and legal reasoning metrics in a scale of 10. Our model can be run on CPU and
achieved 42.46 tokens/sec CPU inference speed. We found that our models,
despite not being pretrained on legal books, various legal contracts, and legal
documents, were able to learn the domain knowledge required for drafting
various legal contracts and legal clauses, and generalize to draft legal
contracts and legal clauses with limited instruction tuning. Hence, we conclude
that for a strong domain-specialized generative language model (such as legal),
very large amounts of data are not required to develop models from scratch. We
believe that this work is the first attempt to make a dedicated generative
legal language model from scratch for Indian Supreme Court jurisdiction or in
legal NLP overall. We plan to release our Paramanu-Ayn model at
https://www.bharatgpts.com. | cs.CL | PARAMANU-AYN: An Efficient Novel Generative and Instruction-tuned Language Model for Indian Legal Case Documents |
2024-03-20T15:38:36Z | http://arxiv.org/pdf/2403.13679v3 | http://arxiv.org/abs/2403.13679v3 | Hongzhan Chen, Hehong Chen, Ming Yan, Wenshen Xu, Xing Gao, Weizhou Shen, Xiaojun Quan, Chenliang Li, Ji Zhang, Fei Huang, Jingren Zhou | 2024-03-22T00:49:59Z | cs.CL | Large language models (LLMs) have advanced the development of various AI
conversational agents, including role-playing conversational agents that mimic
diverse characters and human behaviors. While prior research has predominantly
focused on enhancing the conversational capability, role-specific knowledge,
and stylistic attributes of these agents, there has been a noticeable gap in
assessing their social intelligence. In this paper, we introduce RoleInteract,
the first benchmark designed to systematically evaluate the sociality of
role-playing conversational agents at both individual and group levels of
social interactions. The benchmark is constructed from a variety of sources and
covers a wide range of 500 characters and over 6,000 question prompts and
30,800 multi-turn role-playing utterances. We conduct comprehensive evaluations
on this benchmark using mainstream open-source and closed-source LLMs. We find
that agents excelling in individual level does not imply their proficiency in
group level. Moreover, the behavior of individuals may drift as a result of the
influence exerted by other agents within the group. Experimental results on
RoleInteract confirm its significance as a testbed for assessing the social
interaction of role-playing conversational agents. The benchmark is publicly
accessible at https://github.com/X-PLUG/RoleInteract. | cs.CL | RoleInteract: Evaluating the Social Interaction of Role-Playing Agents |
2024-03-20T15:26:23Z | http://arxiv.org/pdf/2403.14720v1 | http://arxiv.org/abs/2403.14720v1 | Keegan Hines, Gary Lopez, Matthew Hall, Federico Zarfati, Yonatan Zunger, Emre Kiciman | 2024-03-20T15:26:23Z | cs.CR, cs.CL, cs.LG | Large Language Models (LLMs), while powerful, are built and trained to
process a single text input. In common applications, multiple inputs can be
processed by concatenating them together into a single stream of text. However,
the LLM is unable to distinguish which sections of prompt belong to various
input sources. Indirect prompt injection attacks take advantage of this
vulnerability by embedding adversarial instructions into untrusted data being
processed alongside user commands. Often, the LLM will mistake the adversarial
instructions as user commands to be followed, creating a security vulnerability
in the larger system. We introduce spotlighting, a family of prompt engineering
techniques that can be used to improve LLMs' ability to distinguish among
multiple sources of input. The key insight is to utilize transformations of an
input to provide a reliable and continuous signal of its provenance. We
evaluate spotlighting as a defense against indirect prompt injection attacks,
and find that it is a robust defense that has minimal detrimental impact to
underlying NLP tasks. Using GPT-family models, we find that spotlighting
reduces the attack success rate from greater than {50}\% to below {2}\% in our
experiments with minimal impact on task efficacy. | cs.CR | Defending Against Indirect Prompt Injection Attacks With Spotlighting |
2024-03-20T15:20:30Z | http://arxiv.org/pdf/2403.13666v1 | http://arxiv.org/abs/2403.13666v1 | Gorka Azkune, Ander Salaberria, Eneko Agirre | 2024-03-20T15:20:30Z | cs.CL | This paper shows that text-only Language Models (LM) can learn to ground
spatial relations like "left of" or "below" if they are provided with explicit
location information of objects and they are properly trained to leverage those
locations. We perform experiments on a verbalized version of the Visual Spatial
Reasoning (VSR) dataset, where images are coupled with textual statements which
contain real or fake spatial relations between two objects of the image. We
verbalize the images using an off-the-shelf object detector, adding location
tokens to every object label to represent their bounding boxes in textual form.
Given the small size of VSR, we do not observe any improvement when using
locations, but pretraining the LM over a synthetic dataset automatically
derived by us improves results significantly when using location tokens. We
thus show that locations allow LMs to ground spatial relations, with our
text-only LMs outperforming Vision-and-Language Models and setting the new
state-of-the-art for the VSR dataset. Our analysis show that our text-only LMs
can generalize beyond the relations seen in the synthetic dataset to some
extent, learning also more useful information than that encoded in the spatial
rules we used to create the synthetic dataset itself. | cs.CL | Grounding Spatial Relations in Text-Only Language Models |
2024-03-20T14:41:01Z | http://arxiv.org/pdf/2403.13638v2 | http://arxiv.org/abs/2403.13638v2 | Meet Doshi, Raj Dabre, Pushpak Bhattacharyya | 2024-03-21T04:03:59Z | cs.CL | In this paper, we explore the utility of Translationese as synthetic data
created using machine translation for pre-training language models (LMs).
Pre-training requires vast amounts of monolingual data, which is mostly
unavailable for languages other than English. Recently, there has been a
growing interest in using synthetic data to address this data scarcity. We take
the case of English and Indic languages and translate web-crawled monolingual
documents (clean) into the target language. Then, we train language models
containing 28M and 85M parameters on this translationese data (synthetic). We
show that their performance on downstream natural language understanding and
generative tasks is only 3.56% poorer on NLU tasks and 1.51% on NLG tasks than
LMs pre-trained on clean data. Further, we propose the use of lightweight
TinyLMs pre-trained on clean data to filter synthetic data efficiently which
significantly improves the performance of our models. We also find that LMs
trained on synthetic data strongly benefit from extended pretraining on a tiny
fraction (10%) of clean data. We release the data we collected and created as a
part of this work, IndicMonoDoc, the largest collection of monolingual
document-level corpora, which we hope will help bridge the gap between English
and non-English performance for large language models. | cs.CL | Do Not Worry if You Do Not Have Data: Building Pretrained Language Models Using Translationese |
2024-03-20T13:42:57Z | http://arxiv.org/pdf/2403.13592v2 | http://arxiv.org/abs/2403.13592v2 | Ilias Chalkidis, Stephanie Brandl | 2024-03-22T13:37:28Z | cs.CL | Instruction-finetuned Large Language Models inherit clear political leanings
that have been shown to influence downstream task performance. We expand this
line of research beyond the two-party system in the US and audit Llama Chat in
the context of EU politics in various settings to analyze the model's political
knowledge and its ability to reason in context. We adapt, i.e., further
fine-tune, Llama Chat on speeches of individual euro-parties from debates in
the European Parliament to reevaluate its political leaning based on the EUandI
questionnaire. Llama Chat shows considerable knowledge of national parties'
positions and is capable of reasoning in context. The adapted, party-specific,
models are substantially re-aligned towards respective positions which we see
as a starting point for using chat-based LLMs as data-driven conversational
engines to assist research in political science. | cs.CL | Llama meets EU: Investigating the European Political Spectrum through the Lens of LLMs |
2024-03-20T13:38:07Z | http://arxiv.org/pdf/2403.13590v1 | http://arxiv.org/abs/2403.13590v1 | Adian Liusie, Yassir Fathullah, Mark J. F. Gales | 2024-03-20T13:38:07Z | cs.CL | Large Language Models (LLMs) have demonstrated impressive zero-shot
capabilities and versatility in NLP tasks, however they sometimes fail to
maintain crucial invariances for specific tasks. One example is permutation
sensitivity, where LLMs' outputs may significantly vary depending on the order
of the input options. While debiasing techniques can mitigate these issues, and
yield better performance and reliability, they often come with a high
computational cost at inference. This paper addresses this inefficiency at
inference time. The aim is to distill the capabilities of a computationally
intensive, debiased, teacher model into a more compact student model. We
explore two variants of student models: one based on pure distillation, and the
other on an error-correction approach for more complex tasks, where the student
corrects a single biased decision from the teacher to achieve a debiased
output. Our approach is general and can be applied to both black-box and
white-box LLMs. Furthermore, we demonstrate that our compact, encoder-only
student models can outperform their larger, biased teacher counterparts,
achieving better results with significantly fewer parameters. | cs.CL | Teacher-Student Training for Debiasing: General Permutation Debiasing for Large Language Models |
2024-03-20T13:37:00Z | http://arxiv.org/pdf/2403.13588v1 | http://arxiv.org/abs/2403.13588v1 | Chengzhe Feng, Yanan Sun, Ke Li, Pan Zhou, Jiancheng Lv, Aojun Lu | 2024-03-20T13:37:00Z | cs.SE, cs.CL | As Pre-trained Language Models (PLMs), a popular approach for code
intelligence, continue to grow in size, the computational cost of their usage
has become prohibitively expensive. Prompt learning, a recent development in
the field of natural language processing, emerges as a potential solution to
address this challenge. In this paper, we investigate the effectiveness of
prompt learning in code intelligence tasks. We unveil its reliance on manually
designed prompts, which often require significant human effort and expertise.
Moreover, we discover existing automatic prompt design methods are very limited
to code intelligence tasks due to factors including gradient dependence, high
computational demands, and limited applicability. To effectively address both
issues, we propose Genetic Auto Prompt (GenAP), which utilizes an elaborate
genetic algorithm to automatically design prompts. With GenAP, non-experts can
effortlessly generate superior prompts compared to meticulously manual-designed
ones. GenAP operates without the need for gradients or additional computational
costs, rendering it gradient-free and cost-effective. Moreover, GenAP supports
both understanding and generation types of code intelligence tasks, exhibiting
great applicability. We conduct GenAP on three popular code intelligence PLMs
with three canonical code intelligence tasks including defect prediction, code
summarization, and code translation. The results suggest that GenAP can
effectively automate the process of designing prompts. Specifically, GenAP
outperforms all other methods across all three tasks (e.g., improving accuracy
by an average of 2.13% for defect prediction). To the best of our knowledge,
GenAP is the first work to automatically design prompts for code intelligence
PLMs. | cs.SE | Genetic Auto-prompt Learning for Pre-trained Code Intelligence Language Models |
2024-03-20T13:33:55Z | http://arxiv.org/pdf/2403.13583v2 | http://arxiv.org/abs/2403.13583v2 | Xinyi He, Jiaru Zou, Yun Lin, Mengyu Zhou, Shi Han, Zejian Yuan, Dongmei Zhang | 2024-07-01T09:59:47Z | cs.SE, cs.CL, cs.LG | Large Language Models have revolutionized code generation ability by
converting natural language descriptions into executable code. However,
generating complex code within real-world scenarios remains challenging due to
intricate structures, subtle bugs, understanding of advanced data types, and
lack of supplementary contents. To address these challenges, we introduce the
CoCoST framework, which enhances complex code generation by online searching
for more information with planned queries and correctness testing for code
refinement. Moreover, CoCoST serializes the complex inputs and outputs to
improve comprehension and generates test cases to ensure the adaptability for
real-world applications. CoCoST is validated through rigorous experiments on
the DS-1000 and ClassEval datasets. Experimental results show that CoCoST
substantially improves the quality of complex code generation, highlighting its
potential to enhance the practicality of LLMs in generating complex code. | cs.SE | CoCoST: Automatic Complex Code Generation with Online Searching and Correctness Testing |
2024-03-20T13:24:41Z | http://arxiv.org/pdf/2403.13578v1 | http://arxiv.org/abs/2403.13578v1 | Do June Min, Veronica Perez-Rosas, Kenneth Resnicow, Rada Mihalcea | 2024-03-20T13:24:41Z | cs.CL, cs.LG | In this paper, we study the problem of multi-reward reinforcement learning to
jointly optimize for multiple text qualities for natural language generation.
We focus on the task of counselor reflection generation, where we optimize the
generators to simultaneously improve the fluency, coherence, and reflection
quality of generated counselor responses. We introduce two novel bandit
methods, DynaOpt and C-DynaOpt, which rely on the broad strategy of combining
rewards into a single value and optimizing them simultaneously. Specifically,
we employ non-contextual and contextual multi-arm bandits to dynamically adjust
multiple reward weights during training. Through automatic and manual
evaluations, we show that our proposed techniques, DynaOpt and C-DynaOpt,
outperform existing naive and bandit baselines, showcasing their potential for
enhancing language models. | cs.CL | Dynamic Reward Adjustment in Multi-Reward Reinforcement Learning for Counselor Reflection Generation |
2024-03-20T12:52:38Z | http://arxiv.org/pdf/2403.13560v1 | http://arxiv.org/abs/2403.13560v1 | Amir Zeldes, Tatsuya Aoyama, Yang Janet Liu, Siyao Peng, Debopam Das, Luke Gessler | 2024-03-20T12:52:38Z | cs.CL | In this article we present Enhanced Rhetorical Structure Theory (eRST), a new
theoretical framework for computational discourse analysis, based on an
expansion of Rhetorical Structure Theory (RST). The framework encompasses
discourse relation graphs with tree-breaking, nonprojective and concurrent
relations, as well as implicit and explicit signals which give explainable
rationales to our analyses. We survey shortcomings of RST and other existing
frameworks, such as Segmented Discourse Representation Theory (SDRT), the Penn
Discourse Treebank (PDTB) and Discourse Dependencies, and address these using
constructs in the proposed theory. We provide annotation, search and
visualization tools for data, and present and evaluate a freely available
corpus of English annotated according to our framework, encompassing 12 spoken
and written genres with over 200K tokens. Finally, we discuss automatic
parsing, evaluation metrics and applications for data in our framework. | cs.CL | eRST: A Signaled Graph Theory of Discourse Relations and Organization |
2024-03-20T12:14:54Z | http://arxiv.org/pdf/2403.13537v1 | http://arxiv.org/abs/2403.13537v1 | Paloma García-de-Herreros, Vagrant Gautam, Philipp Slusallek, Dietrich Klakow, Marius Mosbach | 2024-03-20T12:14:54Z | cs.CL, cs.AI, cs.CV, cs.LG | ORCA (Shen et al., 2023) is a recent technique for cross-modal fine-tuning,
i.e., applying pre-trained transformer models to modalities beyond their
training data. The technique consists primarily of training an embedder and
fine-tuning the embedder and model. Despite its high performance on a variety
of downstream tasks, we do not understand precisely how each of these
components contribute to ORCA's success. Therefore, we run a series of
ablations and find that embedder training does not help 2D tasks at all,
contrary to what the original paper posits. In 1D tasks, some amount of
embedder training is necessary but more is not better. In 4 out of 6 datasets
we experiment with, it is model fine-tuning that makes the biggest difference.
Through our ablations and baselines, we contribute a better understanding of
the individual components of ORCA. | cs.CL | What explains the success of cross-modal fine-tuning with ORCA? |
2024-03-20T11:38:30Z | http://arxiv.org/pdf/2403.13518v2 | http://arxiv.org/abs/2403.13518v2 | Kunhang Li, Yansong Feng | 2024-03-26T11:16:47Z | cs.AI, cs.CL, cs.CV, cs.RO | The task of text2motion is to generate human motion sequences from given
textual descriptions, where the model explores diverse mappings from natural
language instructions to human body movements. While most existing works are
confined to coarse-grained motion descriptions, e.g., "A man squats.",
fine-grained descriptions specifying movements of relevant body parts are
barely explored. Models trained with coarse-grained texts may not be able to
learn mappings from fine-grained motion-related words to motion primitives,
resulting in the failure to generate motions from unseen descriptions. In this
paper, we build a large-scale language-motion dataset specializing in
fine-grained textual descriptions, FineHumanML3D, by feeding GPT-3.5-turbo with
step-by-step instructions with pseudo-code compulsory checks. Accordingly, we
design a new text2motion model, FineMotionDiffuse, making full use of
fine-grained textual information. Our quantitative evaluation shows that
FineMotionDiffuse trained on FineHumanML3D improves FID by a large margin of
0.38, compared with competitive baselines. According to the qualitative
evaluation and case study, our model outperforms MotionDiffuse in generating
spatially or chronologically composite motions, by learning the implicit
mappings from fine-grained descriptions to the corresponding basic motions. We
release our data at https://github.com/KunhangL/finemotiondiffuse. | cs.AI | Motion Generation from Fine-grained Textual Descriptions |
2024-03-20T11:30:45Z | http://arxiv.org/pdf/2403.13514v1 | http://arxiv.org/abs/2403.13514v1 | Adnan Al Ali, Jindřich Libovický | 2024-03-20T11:30:45Z | cs.CL, cs.CY | Neural language models, which reach state-of-the-art results on most natural
language processing tasks, are trained on large text corpora that inevitably
contain value-burdened content and often capture undesirable biases, which the
models reflect. This case study focuses on the political biases of pre-trained
encoders in Czech and compares them with a representative value survey. Because
Czech is a gendered language, we also measure how the grammatical gender
coincides with responses to men and women in the survey. We introduce a novel
method for measuring the model's perceived political values. We find that the
models do not assign statement probability following value-driven reasoning,
and there is no systematic difference between feminine and masculine sentences.
We conclude that BERT-sized models do not manifest systematic alignment with
political values and that the biases observed in the models are rather due to
superficial imitation of training data patterns than systematic value beliefs
encoded in the models. | cs.CL | How Gender Interacts with Political Values: A Case Study on Czech BERT Models |
2024-03-20T11:27:20Z | http://arxiv.org/pdf/2403.13513v2 | http://arxiv.org/abs/2403.13513v2 | Junho Kim, Yeon Ju Kim, Yong Man Ro | 2024-06-21T06:11:25Z | cs.CV, cs.AI, cs.CL | This paper presents a way of enhancing the reliability of Large Multi-modal
Models (LMMs) in addressing hallucination, where the models generate
cross-modal inconsistent responses. Without additional training, we propose
Counterfactual Inception, a novel method that implants counterfactual thinking
into LMMs using self-generated counterfactual keywords. Our method is grounded
in the concept of counterfactual thinking, a cognitive process where human
considers alternative realities, enabling more extensive context exploration.
Bridging the human cognition mechanism into LMMs, we aim for the models to
engage with and generate responses that span a wider contextual scene
understanding, mitigating hallucinatory outputs. We further introduce
Plausibility Verification Process (PVP), a simple yet robust keyword constraint
that effectively filters out sub-optimal keywords to enable the consistent
triggering of counterfactual thinking in the model responses. Comprehensive
analyses across various LMMs, including both open-source and proprietary
models, corroborate that counterfactual thinking significantly reduces
hallucination and helps to broaden contextual understanding based on true
visual clues. | cs.CV | What if...?: Thinking Counterfactual Keywords Helps to Mitigate Hallucination in Large Multi-modal Models |
2024-03-20T10:40:01Z | http://arxiv.org/pdf/2403.13485v4 | http://arxiv.org/abs/2403.13485v4 | Yijian Lu, Aiwei Liu, Dianzhi Yu, Jingjing Li, Irwin King | 2024-06-09T06:51:57Z | cs.CL | Text watermarking algorithms for large language models (LLMs) can effectively
identify machine-generated texts by embedding and detecting hidden features in
the text. Although the current text watermarking algorithms perform well in
most high-entropy scenarios, its performance in low-entropy scenarios still
needs to be improved. In this work, we opine that the influence of token
entropy should be fully considered in the watermark detection process, $i.e.$,
the weight of each token during watermark detection should be customized
according to its entropy, rather than setting the weights of all tokens to the
same value as in previous methods. Specifically, we propose
\textbf{E}ntropy-based Text \textbf{W}atermarking \textbf{D}etection
(\textbf{EWD}) that gives higher-entropy tokens higher influence weights during
watermark detection, so as to better reflect the degree of watermarking.
Furthermore, the proposed detection process is training-free and fully
automated. From the experiments, we demonstrate that our EWD can achieve better
detection performance in low-entropy scenarios, and our method is also general
and can be applied to texts with different entropy distributions. Our code and
data is available\footnote{\url{https://github.com/luyijian3/EWD}}.
Additionally, our algorithm could be accessed through MarkLLM
\cite{pan2024markllm}\footnote{\url{https://github.com/THU-BPM/MarkLLM}}. | cs.CL | An Entropy-based Text Watermarking Detection Method |
2024-03-20T10:14:13Z | http://arxiv.org/pdf/2403.15470v1 | http://arxiv.org/abs/2403.15470v1 | James Vo | 2024-03-20T10:14:13Z | cs.CL | The advancement of Large Language Models (LLMs) has significantly transformed
the field of natural language processing, although the focus on English-centric
models has created a noticeable research gap for specific languages, including
Vietnamese. To address this issue, this paper presents vi-mistral-x, an
innovative Large Language Model designed expressly for the Vietnamese language.
It utilizes a unique method of continual pre-training, based on the Mistral
architecture, which incorporates grouped-query attention and sliding window
attention techniques. This model, vi-Mistral-X, marks a significant step
forward in improving the understanding and generation of the Vietnamese
language. It introduces an additional phase of continual pre-training,
specifically adapted for Vietnamese, enhancing the model's capability in
understanding complex language nuances and generating accurate, context-aware
Vietnamese text. Through comprehensive testing on various benchmarks,
vi-mistral-x has shown to outperform existing Vietnamese LLMs in several key
areas, including text classification, question answering, and text generation.
Particularly, in the Vietnamese Multitask Language Understanding (VMLU)
benchmark, vi-mistral-x sets a new standard, outperforming other available
models significantly. This paper highlights the critical role of continual
pre-training in advancing language-specific LLMs and opens new avenues for the
development of multilingual models. We aim for vi-mistral-x to not just be an
important asset for processing the Vietnamese language but also to encourage
more advancements in creating large language models for languages that are less
represented. | cs.CL | Vi-Mistral-X: Building a Vietnamese Language Model with Advanced Continual Pre-training |
2024-03-20T09:42:43Z | http://arxiv.org/pdf/2403.13447v1 | http://arxiv.org/abs/2403.13447v1 | Wenqiao Zhang, Tianwei Lin, Jiang Liu, Fangxun Shu, Haoyuan Li, Lei Zhang, He Wanggui, Hao Zhou, Zheqi Lv, Hao Jiang, Juncheng Li, Siliang Tang, Yueting Zhuang | 2024-03-20T09:42:43Z | cs.AI, cs.CL, cs.CV | Recent advancements indicate that scaling up Multimodal Large Language Models
(MLLMs) effectively enhances performance on downstream multimodal tasks. The
prevailing MLLM paradigm, \emph{e.g.}, LLaVA, transforms visual features into
text-like tokens using a \emph{static} vision-language mapper, thereby enabling
\emph{static} LLMs to develop the capability to comprehend visual information
through visual instruction tuning. Although promising, the \emph{static} tuning
strategy~\footnote{The static tuning refers to the trained model with static
parameters.} that shares the same parameters may constrain performance across
different downstream multimodal tasks. In light of this, we introduce
HyperLLaVA, which involves adaptive tuning of the projector and LLM parameters,
in conjunction with a dynamic visual expert and language expert, respectively.
These experts are derived from HyperNetworks, which generates adaptive
parameter shifts through visual and language guidance, enabling dynamic
projector and LLM modeling in two-stage training.
Our experiments demonstrate that our solution significantly surpasses LLaVA
on existing MLLM benchmarks, including MME, MMBench, SEED-Bench, and
LLaVA-Bench. ~\footnote{Our project is available on the link
https://github.com/DCDmllm/HyperLLaVA}. | cs.AI | HyperLLaVA: Dynamic Visual and Language Expert Tuning for Multimodal Large Language Models |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.