publication_date
stringlengths 20
20
| doc_url
stringlengths 33
34
| id
stringlengths 33
34
| authors
stringlengths 5
20.1k
| update_date
stringlengths 20
20
| category_all
stringlengths 5
138
| abstract
stringlengths 123
3.09k
| category_primary
stringclasses 78
values | title
stringlengths 10
211
|
|---|---|---|---|---|---|---|---|---|
2024-02-27T22:06:55Z | http://arxiv.org/pdf/2402.17914v2 | http://arxiv.org/abs/2402.17914v2 | Roy Xie, Orevaoghene Ahia, Yulia Tsvetkov, Antonios Anastasopoulos | 2024-03-23T20:21:04Z | cs.CL, cs.AI | Identifying linguistic differences between dialects of a language often
requires expert knowledge and meticulous human analysis. This is largely due to
the complexity and nuance involved in studying various dialects. We present a
novel approach to extract distinguishing lexical features of dialects by
utilizing interpretable dialect classifiers, even in the absence of human
experts. We explore both post-hoc and intrinsic approaches to interpretability,
conduct experiments on Mandarin, Italian, and Low Saxon, and experimentally
demonstrate that our method successfully identifies key language-specific
lexical features that contribute to dialectal variations. | cs.CL | Extracting Lexical Features from Dialects via Interpretable Dialect Classifiers |
2024-02-27T21:27:35Z | http://arxiv.org/pdf/2402.17897v2 | http://arxiv.org/abs/2402.17897v2 | Hang Dong, Jiaoyan Chen, Yuan He, Yongsheng Gao, Ian Horrocks | 2024-03-04T14:06:42Z | cs.CL, cs.IR, I.2.7; I.2.4 | We investigate the task of inserting new concepts extracted from texts into
an ontology using language models. We explore an approach with three steps:
edge search which is to find a set of candidate locations to insert (i.e.,
subsumptions between concepts), edge formation and enrichment which leverages
the ontological structure to produce and enhance the edge candidates, and edge
selection which eventually locates the edge to be placed into. In all steps, we
propose to leverage neural methods, where we apply embedding-based methods and
contrastive learning with Pre-trained Language Models (PLMs) such as BERT for
edge search, and adapt a BERT fine-tuning-based multi-label Edge-Cross-encoder,
and Large Language Models (LLMs) such as GPT series, FLAN-T5, and Llama 2, for
edge selection. We evaluate the methods on recent datasets created using the
SNOMED CT ontology and the MedMentions entity linking benchmark. The best
settings in our framework use fine-tuned PLM for search and a multi-label
Cross-encoder for selection. Zero-shot prompting of LLMs is still not adequate
for the task, and we propose explainable instruction tuning of LLMs for
improved performance. Our study shows the advantages of PLMs and highlights the
encouraging performance of LLMs that motivates future studies. | cs.CL | A Language Model based Framework for New Concept Placement in Ontologies |
2024-02-27T21:27:16Z | http://arxiv.org/pdf/2402.17896v1 | http://arxiv.org/abs/2402.17896v1 | Corby Rosset, Ho-Lam Chung, Guanghui Qin, Ethan C. Chau, Zhuo Feng, Ahmed Awadallah, Jennifer Neville, Nikhil Rao | 2024-02-27T21:27:16Z | cs.CL, cs.AI | Existing question answering (QA) datasets are no longer challenging to most
powerful Large Language Models (LLMs). Traditional QA benchmarks like TriviaQA,
NaturalQuestions, ELI5 and HotpotQA mainly study ``known unknowns'' with clear
indications of both what information is missing, and how to find it to answer
the question. Hence, good performance on these benchmarks provides a false
sense of security. A yet unmet need of the NLP community is a bank of
non-factoid, multi-perspective questions involving a great deal of unclear
information needs, i.e. ``unknown uknowns''. We claim we can find such
questions in search engine logs, which is surprising because most
question-intent queries are indeed factoid. We present Researchy Questions, a
dataset of search engine queries tediously filtered to be non-factoid,
``decompositional'' and multi-perspective. We show that users spend a lot of
``effort'' on these questions in terms of signals like clicks and session
length, and that they are also challenging for GPT-4. We also show that ``slow
thinking'' answering techniques, like decomposition into sub-questions shows
benefit over answering directly. We release $\sim$ 100k Researchy Questions,
along with the Clueweb22 URLs that were clicked. | cs.CL | Researchy Questions: A Dataset of Multi-Perspective, Decompositional Questions for LLM Web Agents |
2024-02-27T21:01:41Z | http://arxiv.org/pdf/2402.17887v4 | http://arxiv.org/abs/2402.17887v4 | Junda Wang, Zhichao Yang, Zonghai Yao, Hong Yu | 2024-06-28T13:23:31Z | cs.CL, cs.IR | Large Language Models (LLMs) have demonstrated a remarkable potential in
medical knowledge acquisition and question-answering. However, LLMs can
potentially hallucinate and yield factually incorrect outcomes, even with
domain-specific pretraining. Previously, retrieval augmented generation (RAG)
has limited success in addressing hallucinations. Unlike previous methods in
RAG where the retrieval model was trained separately from the LLM, we introduce
JMLR (for Jointly trains LLM and information Retrieval) during the fine-tuning
phase. The synchronized training mechanism enhances JMLR's ability to retrieve
clinical guidelines and leverage medical knowledge to reason and answer
questions and reduces the demand for computational resources. We evaluated JMLR
on the important medical question-answering application. Our experimental
results demonstrate that JMLR-13B (70.5%) outperforms a previous
state-of-the-art open-source model using conventional pre-training and
fine-tuning Meditron-70B (68.9%) and Llama2-13B with RAG (67.7%) on a medical
question-answering dataset. Comprehensive evaluations reveal JMLR-13B enhances
reasoning quality and reduces hallucinations better than Claude3-Opus.
Additionally, JMLR-13B (148 GPU hours) also trains much faster than
Meditron-70B (42630 GPU hours). Through this work, we provide a new and
efficient knowledge enhancement method for healthcare, demonstrating the
potential of integrating retrieval and LLM training for medical
question-answering systems. | cs.CL | JMLR: Joint Medical LLM and Retrieval Training for Enhancing Reasoning and Professional Question Answering Capability |
2024-02-27T20:48:24Z | http://arxiv.org/pdf/2402.17882v2 | http://arxiv.org/abs/2402.17882v2 | Parker Glenn, Parag Pravin Dakle, Liang Wang, Preethi Raghavan | 2024-06-10T20:27:28Z | cs.CL | Many existing end-to-end systems for hybrid question answering tasks can
often be boiled down to a "prompt-and-pray" paradigm, where the user has
limited control and insight into the intermediate reasoning steps used to
achieve the final result. Additionally, due to the context size limitation of
many transformer-based LLMs, it is often not reasonable to expect that the full
structured and unstructured context will fit into a given prompt in a zero-shot
setting, let alone a few-shot setting. We introduce BlendSQL, a superset of
SQLite to act as a unified dialect for orchestrating reasoning across both
unstructured and structured data. For hybrid question answering tasks involving
multi-hop reasoning, we encode the full decomposed reasoning roadmap into a
single interpretable BlendSQL query. Notably, we show that BlendSQL can scale
to massive datasets and improve the performance of end-to-end systems while
using 35% fewer tokens. Our code is available and installable as a package at
https://github.com/parkervg/blendsql. | cs.CL | BlendSQL: A Scalable Dialect for Unifying Hybrid Question Answering in Relational Algebra |
2024-02-27T20:33:22Z | http://arxiv.org/pdf/2402.17879v2 | http://arxiv.org/abs/2402.17879v2 | Michael Y. Li, Emily B. Fox, Noah D. Goodman | 2024-06-22T05:08:30Z | cs.LG, cs.CL | Statistical model discovery is a challenging search over a vast space of
models subject to domain-specific constraints. Efficiently searching over this
space requires expertise in modeling and the problem domain. Motivated by the
domain knowledge and programming capabilities of large language models (LMs),
we introduce a method for language model driven automated statistical model
discovery. We cast our automated procedure within the principled framework of
Box's Loop: the LM iterates between proposing statistical models represented as
probabilistic programs, acting as a modeler, and critiquing those models,
acting as a domain expert. By leveraging LMs, we do not have to define a
domain-specific language of models or design a handcrafted search procedure,
which are key restrictions of previous systems. We evaluate our method in three
settings in probabilistic modeling: searching within a restricted space of
models, searching over an open-ended space, and improving expert models under
natural language constraints (e.g., this model should be interpretable to an
ecologist). Our method identifies models on par with human expert designed
models and extends classic models in interpretable ways. Our results highlight
the promise of LM-driven model discovery. | cs.LG | Automated Statistical Model Discovery with Language Models |
2024-02-27T19:16:39Z | http://arxiv.org/pdf/2403.00828v1 | http://arxiv.org/abs/2403.00828v1 | Bushra Alhijawi, Rawan Jarrar, Aseel AbuAlRub, Arwa Bader | 2024-02-27T19:16:39Z | cs.CL, cs.AI, cs.LG | Large Language Models (LLMs), such as GPT-3 and BERT, reshape how textual
content is written and communicated. These models have the potential to
generate scientific content that is indistinguishable from that written by
humans. Hence, LLMs carry severe consequences for the scientific community,
which relies on the integrity and reliability of publications. This research
paper presents a novel ChatGPT-generated scientific text detection method,
AI-Catcher. AI-Catcher integrates two deep learning models, multilayer
perceptron (MLP) and convolutional neural networks (CNN). The MLP learns the
feature representations of the linguistic and statistical features. The CNN
extracts high-level representations of the sequential patterns from the textual
content. AI-Catcher is a multimodal model that fuses hidden patterns derived
from MLP and CNN. In addition, a new ChatGPT-Generated scientific text dataset
is collected to enhance AI-generated text detection tools, AIGTxt. AIGTxt
contains 3000 records collected from published academic articles across ten
domains and divided into three classes: Human-written, ChatGPT-generated, and
Mixed text. Several experiments are conducted to evaluate the performance of
AI-Catcher. The comparative results demonstrate the capability of AI-Catcher to
distinguish between human-written and ChatGPT-generated scientific text more
accurately than alternative methods. On average, AI-Catcher improved accuracy
by 37.4%. | cs.CL | Deep Learning Detection Method for Large Language Models-Generated Scientific Content |
2024-02-27T19:13:01Z | http://arxiv.org/pdf/2403.00827v1 | http://arxiv.org/abs/2403.00827v1 | Keshav Ramji, Young-Suk Lee, Ramón Fernandez Astudillo, Md Arafat Sultan, Tahira Naseem, Asim Munawar, Radu Florian, Salim Roukos | 2024-02-27T19:13:01Z | cs.CL, cs.AI, cs.LG | It is often desirable for Large Language Models (LLMs) to capture multiple
objectives when providing a response. In document-grounded response generation,
for example, agent responses are expected to be relevant to a user's query
while also being grounded in a given document. In this paper, we introduce
Proxy Metric-based Self-Refinement (ProMiSe), which enables an LLM to refine
its own initial response along key dimensions of quality guided by external
metrics feedback, yielding an overall better final response. ProMiSe leverages
feedback on response quality through principle-specific proxy metrics, and
iteratively refines its response one principle at a time. We apply ProMiSe to
open source language models Flan-T5-XXL and Llama-2-13B-Chat, to evaluate its
performance on document-grounded question answering datasets, MultiDoc2Dial and
QuAC, demonstrating that self-refinement improves response quality. We further
show that fine-tuning Llama-2-13B-Chat on the synthetic dialogue data generated
by ProMiSe yields significant performance improvements over the zero-shot
baseline as well as a supervised fine-tuned model on human annotated data. | cs.CL | Self-Refinement of Language Models from External Proxy Metrics Feedback |
2024-02-27T19:08:05Z | http://arxiv.org/pdf/2402.17840v2 | http://arxiv.org/abs/2402.17840v2 | Zhenting Qi, Hanlin Zhang, Eric Xing, Sham Kakade, Himabindu Lakkaraju | 2024-06-20T21:32:35Z | cs.CL, cs.AI, cs.CR, cs.LG | Retrieval-Augmented Generation (RAG) improves pre-trained models by
incorporating external knowledge at test time to enable customized adaptation.
We study the risk of datastore leakage in Retrieval-In-Context RAG Language
Models (LMs). We show that an adversary can exploit LMs' instruction-following
capabilities to easily extract text data verbatim from the datastore of RAG
systems built with instruction-tuned LMs via prompt injection. The
vulnerability exists for a wide range of modern LMs that span Llama2,
Mistral/Mixtral, Vicuna, SOLAR, WizardLM, Qwen1.5, and Platypus2, and the
exploitability exacerbates as the model size scales up. Extending our study to
production RAG models GPTs, we design an attack that can cause datastore
leakage with a 100% success rate on 25 randomly selected customized GPTs with
at most 2 queries, and we extract text data verbatim at a rate of 41% from a
book of 77,000 words and 3% from a corpus of 1,569,000 words by prompting the
GPTs with only 100 queries generated by themselves. | cs.CL | Follow My Instruction and Spill the Beans: Scalable Data Extraction from Retrieval-Augmented Generation Systems |
2024-02-27T19:00:07Z | http://arxiv.org/pdf/2402.17834v1 | http://arxiv.org/abs/2402.17834v1 | Marco Bellagente, Jonathan Tow, Dakota Mahan, Duy Phung, Maksym Zhuravinskyi, Reshinth Adithyan, James Baicoianu, Ben Brooks, Nathan Cooper, Ashish Datta, Meng Lee, Emad Mostaque, Michael Pieler, Nikhil Pinnaparju, Paulo Rocha, Harry Saini, Hannah Teufel, Niccolo Zanichelli, Carlos Riquelme | 2024-02-27T19:00:07Z | cs.CL, stat.ML | We introduce StableLM 2 1.6B, the first in a new generation of our language
model series. In this technical report, we present in detail the data and
training procedure leading to the base and instruction-tuned versions of
StableLM 2 1.6B. The weights for both models are available via Hugging Face for
anyone to download and use. The report contains thorough evaluations of these
models, including zero- and few-shot benchmarks, multilingual benchmarks, and
the MT benchmark focusing on multi-turn dialogues. At the time of publishing
this report, StableLM 2 1.6B was the state-of-the-art open model under 2B
parameters by a significant margin. Given its appealing small size, we also
provide throughput measurements on a number of edge devices. In addition, we
open source several quantized checkpoints and provide their performance metrics
compared to the original model. | cs.CL | Stable LM 2 1.6B Technical Report |
2024-02-27T19:00:01Z | http://arxiv.org/pdf/2402.17826v2 | http://arxiv.org/abs/2402.17826v2 | Ivi Chatzi, Eleni Straitouri, Suhas Thejaswi, Manuel Gomez Rodriguez | 2024-05-23T16:50:21Z | cs.LG, cs.AI, cs.CL, cs.CY, cs.HC, stat.ML | Large language models are often ranked according to their level of alignment
with human preferences -- a model is better than other models if its outputs
are more frequently preferred by humans. One of the popular ways to elicit
human preferences utilizes pairwise comparisons between the outputs provided by
different models to the same inputs. However, since gathering pairwise
comparisons by humans is costly and time-consuming, it has become a common
practice to gather pairwise comparisons by a strong large language model -- a
model strongly aligned with human preferences. Surprisingly, practitioners
cannot currently measure the uncertainty that any mismatch between human and
model preferences may introduce in the constructed rankings. In this work, we
develop a statistical framework to bridge this gap. Given a (small) set of
pairwise comparisons by humans and a large set of pairwise comparisons by a
model, our framework provides a rank-set -- a set of possible ranking positions
-- for each of the models under comparison. Moreover, it guarantees that, with
a probability greater than or equal to a user-specified value, the rank-sets
cover the true ranking consistent with the distribution of human pairwise
preferences asymptotically. Using pairwise comparisons made by humans in the
LMSYS Chatbot Arena platform and pairwise comparisons made by three strong
large language models, we empirically demonstrate the effectivity of our
framework and show that the rank-sets constructed using only pairwise
comparisons by the strong large language models are often inconsistent with
(the distribution of) human pairwise preferences. | cs.LG | Prediction-Powered Ranking of Large Language Models |
2024-02-27T18:56:19Z | http://arxiv.org/pdf/2402.17764v1 | http://arxiv.org/abs/2402.17764v1 | Shuming Ma, Hongyu Wang, Lingxiao Ma, Lei Wang, Wenhui Wang, Shaohan Huang, Li Dong, Ruiping Wang, Jilong Xue, Furu Wei | 2024-02-27T18:56:19Z | cs.CL, cs.LG | Recent research, such as BitNet, is paving the way for a new era of 1-bit
Large Language Models (LLMs). In this work, we introduce a 1-bit LLM variant,
namely BitNet b1.58, in which every single parameter (or weight) of the LLM is
ternary {-1, 0, 1}. It matches the full-precision (i.e., FP16 or BF16)
Transformer LLM with the same model size and training tokens in terms of both
perplexity and end-task performance, while being significantly more
cost-effective in terms of latency, memory, throughput, and energy consumption.
More profoundly, the 1.58-bit LLM defines a new scaling law and recipe for
training new generations of LLMs that are both high-performance and
cost-effective. Furthermore, it enables a new computation paradigm and opens
the door for designing specific hardware optimized for 1-bit LLMs. | cs.CL | The Era of 1-bit LLMs: All Large Language Models are in 1.58 Bits |
2024-02-27T18:55:17Z | http://arxiv.org/pdf/2402.17762v1 | http://arxiv.org/abs/2402.17762v1 | Mingjie Sun, Xinlei Chen, J. Zico Kolter, Zhuang Liu | 2024-02-27T18:55:17Z | cs.CL, cs.LG | We observe an empirical phenomenon in Large Language Models (LLMs) -- very
few activations exhibit significantly larger values than others (e.g., 100,000
times larger). We call them massive activations. First, we demonstrate the
widespread existence of massive activations across various LLMs and
characterize their locations. Second, we find their values largely stay
constant regardless of the input, and they function as indispensable bias terms
in LLMs. Third, these massive activations lead to the concentration of
attention probabilities to their corresponding tokens, and further, implicit
bias terms in the self-attention output. Last, we also study massive
activations in Vision Transformers. | cs.CL | Massive Activations in Large Language Models |
2024-02-27T18:52:19Z | http://arxiv.org/pdf/2402.17759v2 | http://arxiv.org/abs/2402.17759v2 | Yuxian Gu, Li Dong, Yaru Hao, Qingxiu Dong, Minlie Huang, Furu Wei | 2024-03-03T15:07:50Z | cs.CL | This work studies the general principles of improving the learning of
language models (LMs), which aims at reducing the necessary training steps for
achieving superior performance. Specifically, we present a theory for the
optimal learning of LMs. We first propose an objective that optimizes LM
learning by maximizing the data compression ratio in an
"LM-training-as-lossless-compression" view. Then, we derive a theorem, named
Learning Law, to reveal the properties of the dynamics in the optimal learning
process under our objective. The theorem is then validated by experiments on a
linear classification and a real-world language modeling task. Finally, we
empirically verify that the optimal learning of LMs essentially stems from the
improvement of the coefficients in the scaling law of LMs, indicating great
promise and significance for designing practical learning acceleration methods.
Our code can be found at https://aka.ms/LearningLaw. | cs.CL | Towards Optimal Learning of Language Models |
2024-02-27T18:42:31Z | http://arxiv.org/pdf/2402.17753v1 | http://arxiv.org/abs/2402.17753v1 | Adyasha Maharana, Dong-Ho Lee, Sergey Tulyakov, Mohit Bansal, Francesco Barbieri, Yuwei Fang | 2024-02-27T18:42:31Z | cs.CL, cs.AI, cs.LG | Existing works on long-term open-domain dialogues focus on evaluating model
responses within contexts spanning no more than five chat sessions. Despite
advancements in long-context large language models (LLMs) and retrieval
augmented generation (RAG) techniques, their efficacy in very long-term
dialogues remains unexplored. To address this research gap, we introduce a
machine-human pipeline to generate high-quality, very long-term dialogues by
leveraging LLM-based agent architectures and grounding their dialogues on
personas and temporal event graphs. Moreover, we equip each agent with the
capability of sharing and reacting to images. The generated conversations are
verified and edited by human annotators for long-range consistency and
grounding to the event graphs. Using this pipeline, we collect LoCoMo, a
dataset of very long-term conversations, each encompassing 300 turns and 9K
tokens on avg., over up to 35 sessions. Based on LoCoMo, we present a
comprehensive evaluation benchmark to measure long-term memory in models,
encompassing question answering, event summarization, and multi-modal dialogue
generation tasks. Our experimental results indicate that LLMs exhibit
challenges in understanding lengthy conversations and comprehending long-range
temporal and causal dynamics within dialogues. Employing strategies like
long-context LLMs or RAG can offer improvements but these models still
substantially lag behind human performance. | cs.CL | Evaluating Very Long-Term Conversational Memory of LLM Agents |
2024-02-27T18:09:36Z | http://arxiv.org/pdf/2402.17733v1 | http://arxiv.org/abs/2402.17733v1 | Duarte M. Alves, José Pombal, Nuno M. Guerreiro, Pedro H. Martins, João Alves, Amin Farajian, Ben Peters, Ricardo Rei, Patrick Fernandes, Sweta Agrawal, Pierre Colombo, José G. C. de Souza, André F. T. Martins | 2024-02-27T18:09:36Z | cs.CL | While general-purpose large language models (LLMs) demonstrate proficiency on
multiple tasks within the domain of translation, approaches based on open LLMs
are competitive only when specializing on a single task. In this paper, we
propose a recipe for tailoring LLMs to multiple tasks present in translation
workflows. We perform continued pretraining on a multilingual mixture of
monolingual and parallel data, creating TowerBase, followed by finetuning on
instructions relevant for translation processes, creating TowerInstruct. Our
final model surpasses open alternatives on several tasks relevant to
translation workflows and is competitive with general-purpose closed LLMs. To
facilitate future research, we release the Tower models, our specialization
dataset, an evaluation framework for LLMs focusing on the translation
ecosystem, and a collection of model generations, including ours, on our
benchmark. | cs.CL | Tower: An Open Multilingual Large Language Model for Translation-Related Tasks |
2024-02-27T17:52:33Z | http://arxiv.org/pdf/2402.17717v1 | http://arxiv.org/abs/2402.17717v1 | Ayana Niwa, Hayate Iso | 2024-02-27T17:52:33Z | cs.CL | In this study, we introduce AmbigNLG, a new task designed to tackle the
challenge of task ambiguity in instructions for Natural Language Generation
(NLG) tasks. Despite the impressive capabilities of Large Language Models
(LLMs) in understanding and executing a wide range of tasks through natural
language interaction, their performance is significantly hindered by the
ambiguity present in real-world instructions. To address this, AmbigNLG seeks
to identify and mitigate such ambiguities, aiming to refine instructions to
match user expectations better. We introduce a dataset, AmbigSNI-NLG,
consisting of 2,500 instances, and develop an ambiguity taxonomy for
categorizing and annotating instruction ambiguities. Our approach demonstrates
substantial improvements in text generation quality, highlighting the critical
role of clear and specific instructions in enhancing LLM performance in NLG
tasks. | cs.CL | AmbigNLG: Addressing Task Ambiguity in Instruction for NLG |
2024-02-27T17:41:58Z | http://arxiv.org/pdf/2402.17709v2 | http://arxiv.org/abs/2402.17709v2 | Yi Hu, Xiaojuan Tang, Haotong Yang, Muhan Zhang | 2024-06-26T09:25:07Z | cs.AI, cs.CL | Despite the impressive performance in a variety of complex tasks, modern
large language models (LLMs) still have trouble dealing with some math problems
that are simple and intuitive for humans, such as addition. While we can easily
learn basic rules of addition and apply them to new problems of any length,
LLMs struggle to do the same. Instead, they may rely on similar cases seen in
the training corpus for help. We define these two different reasoning
mechanisms as "rule-based reasoning" and "case-based reasoning". Since
rule-based reasoning is essential for acquiring systematic generalization
ability, we aim to explore exactly whether transformers use rule-based or
case-based reasoning for math problems. Through carefully designed intervention
experiments on five math tasks, we confirm that transformers are performing
case-based reasoning, no matter whether scratchpad is used, which aligns with
the previous observations that transformers use subgraph matching/shortcut
learning to reason. To mitigate such problems, we propose a Rule-Following
Fine-Tuning (RFFT) technique to teach transformers to perform rule-based
reasoning. Specifically, we provide explicit rules in the input and then
instruct transformers to recite and follow the rules step by step. Through
RFFT, we successfully enable LLMs fine-tuned on 1-5 digit addition to
generalize to up to 12-digit addition with over 95% accuracy, which is over 40%
higher than scratchpad. The significant improvement demonstrates that teaching
LLMs to use rules explicitly helps them learn rule-based reasoning and
generalize better in length. | cs.AI | Case-Based or Rule-Based: How Do Transformers Do the Math? |
2024-02-27T17:25:37Z | http://arxiv.org/pdf/2402.17700v1 | http://arxiv.org/abs/2402.17700v1 | Jing Huang, Zhengxuan Wu, Christopher Potts, Mor Geva, Atticus Geiger | 2024-02-27T17:25:37Z | cs.CL, cs.LG | Individual neurons participate in the representation of multiple high-level
concepts. To what extent can different interpretability methods successfully
disentangle these roles? To help address this question, we introduce RAVEL
(Resolving Attribute-Value Entanglements in Language Models), a dataset that
enables tightly controlled, quantitative comparisons between a variety of
existing interpretability methods. We use the resulting conceptual framework to
define the new method of Multi-task Distributed Alignment Search (MDAS), which
allows us to find distributed representations satisfying multiple causal
criteria. With Llama2-7B as the target language model, MDAS achieves
state-of-the-art results on RAVEL, demonstrating the importance of going beyond
neuron-level analyses to identify features distributed across activations. We
release our benchmark at https://github.com/explanare/ravel. | cs.CL | RAVEL: Evaluating Interpretability Methods on Disentangling Language Model Representations |
2024-02-27T16:56:30Z | http://arxiv.org/pdf/2402.17682v2 | http://arxiv.org/abs/2402.17682v2 | Tamara Czinczoll, Christoph Hönes, Maximilian Schall, Gerard de Melo | 2024-06-13T10:21:03Z | cs.CL | While (large) language models have significantly improved over the last
years, they still struggle to sensibly process long sequences found, e.g., in
books, due to the quadratic scaling of the underlying attention mechanism. To
address this, we propose NextLevelBERT, a Masked Language Model operating not
on tokens, but on higher-level semantic representations in the form of text
embeddings. We pretrain NextLevelBERT to predict the vector representation of
entire masked text chunks and evaluate the effectiveness of the resulting
document vectors on three types of tasks: 1) Semantic Textual Similarity via
zero-shot document embeddings, 2) Long document classification, 3)
Multiple-choice question answering. We find that next-level Masked Language
Modeling is an effective technique to tackle long-document use cases and can
outperfor much larger embedding models as long as the required level of detail
of semantic information is not too fine. Our models and code are publicly
available online. | cs.CL | NextLevelBERT: Masked Language Modeling with Higher-Level Representations for Long Documents |
2024-02-27T16:19:37Z | http://arxiv.org/pdf/2402.17649v2 | http://arxiv.org/abs/2402.17649v2 | Tanise Ceron, Neele Falk, Ana Barić, Dmitry Nikolaev, Sebastian Padó | 2024-06-03T22:25:08Z | cs.CL, cs.CY | Due to the widespread use of large language models (LLMs) in ubiquitous
systems, we need to understand whether they embed a specific worldview and what
these views reflect. Recent studies report that, prompted with political
questionnaires, LLMs show left-liberal leanings (Feng et al., 2023; Motoki et
al., 2024). However, it is as yet unclear whether these leanings are reliable
(robust to prompt variations) and whether the leaning is consistent across
policies and political leaning. We propose a series of tests which assess the
reliability and consistency of LLMs' stances on political statements based on a
dataset of voting-advice questionnaires collected from seven EU countries and
annotated for policy domains. We study LLMs ranging in size from 7B to 70B
parameters and find that their reliability increases with parameter count.
Larger models show overall stronger alignment with left-leaning parties but
differ among policy programs: They evince a (left-wing) positive stance towards
environment protection, social welfare state and liberal society but also
(right-wing) law and order, with no consistent preferences in foreign policy
and migration. | cs.CL | Beyond prompt brittleness: Evaluating the reliability and consistency of political worldviews in LLMs |
2024-02-27T16:15:28Z | http://arxiv.org/pdf/2402.17645v1 | http://arxiv.org/abs/2402.17645v1 | Shuangrui Ding, Zihan Liu, Xiaoyi Dong, Pan Zhang, Rui Qian, Conghui He, Dahua Lin, Jiaqi Wang | 2024-02-27T16:15:28Z | cs.SD, cs.AI, cs.CL, eess.AS | We present SongComposer, an innovative LLM designed for song composition. It
could understand and generate melodies and lyrics in symbolic song
representations, by leveraging the capability of LLM. Existing music-related
LLM treated the music as quantized audio signals, while such implicit encoding
leads to inefficient encoding and poor flexibility. In contrast, we resort to
symbolic song representation, the mature and efficient way humans designed for
music, and enable LLM to explicitly compose songs like humans. In practice, we
design a novel tuple design to format lyric and three note attributes (pitch,
duration, and rest duration) in the melody, which guarantees the correct LLM
understanding of musical symbols and realizes precise alignment between lyrics
and melody. To impart basic music understanding to LLM, we carefully collected
SongCompose-PT, a large-scale song pretraining dataset that includes lyrics,
melodies, and paired lyrics-melodies in either Chinese or English. After
adequate pre-training, 10K carefully crafted QA pairs are used to empower the
LLM with the instruction-following capability and solve diverse tasks. With
extensive experiments, SongComposer demonstrates superior performance in
lyric-to-melody generation, melody-to-lyric generation, song continuation, and
text-to-song creation, outperforming advanced LLMs like GPT-4. | cs.SD | SongComposer: A Large Language Model for Lyric and Melody Composition in Song Generation |
2024-02-27T16:15:03Z | http://arxiv.org/pdf/2402.17644v2 | http://arxiv.org/abs/2402.17644v2 | Xiao Liu, Zirui Wu, Xueqing Wu, Pan Lu, Kai-Wei Chang, Yansong Feng | 2024-06-09T13:54:09Z | cs.CL, cs.AI | Quantitative reasoning is a critical skill to analyze data, yet the
assessment of such ability remains limited. To address this gap, we introduce
the Quantitative Reasoning with Data (QRData) benchmark, aiming to evaluate
Large Language Models' capability in statistical and causal reasoning with
real-world data. The benchmark comprises a carefully constructed dataset of 411
questions accompanied by data sheets from textbooks, online learning materials,
and academic papers. To compare models' quantitative reasoning abilities on
data and text, we enrich the benchmark with an auxiliary set of 290 text-only
questions, namely QRText. We evaluate natural language reasoning, program-based
reasoning, and agent reasoning methods including Chain-of-Thought,
Program-of-Thoughts, ReAct, and code interpreter assistants on diverse models.
The strongest model GPT-4 achieves an accuracy of 58%, which has much room for
improvement. Among open-source models, Deepseek-coder-instruct, a code LLM
pretrained on 2T tokens, gets the highest accuracy of 37%. Analysis reveals
that models encounter difficulties in data analysis and causal reasoning, and
struggle in using causal knowledge and provided data simultaneously. Code and
data are in https://github.com/xxxiaol/QRData. | cs.CL | Are LLMs Capable of Data-based Statistical and Causal Reasoning? Benchmarking Advanced Quantitative Reasoning with Data |
2024-02-27T16:11:05Z | http://arxiv.org/pdf/2402.17641v2 | http://arxiv.org/abs/2402.17641v2 | Yuesong Shen, Nico Daheim, Bai Cong, Peter Nickl, Gian Maria Marconi, Clement Bazan, Rio Yokota, Iryna Gurevych, Daniel Cremers, Mohammad Emtiyaz Khan, Thomas Möllenhoff | 2024-06-06T04:31:43Z | cs.LG, cs.AI, cs.CL, math.OC, stat.ML | We give extensive empirical evidence against the common belief that
variational learning is ineffective for large neural networks. We show that an
optimizer called Improved Variational Online Newton (IVON) consistently matches
or outperforms Adam for training large networks such as GPT-2 and ResNets from
scratch. IVON's computational costs are nearly identical to Adam but its
predictive uncertainty is better. We show several new use cases of IVON where
we improve finetuning and model merging in Large Language Models, accurately
predict generalization error, and faithfully estimate sensitivity to data. We
find overwhelming evidence that variational learning is effective. | cs.LG | Variational Learning is Effective for Large Deep Networks |
2024-02-27T15:59:37Z | http://arxiv.org/pdf/2402.17633v1 | http://arxiv.org/abs/2402.17633v1 | Fabian Retkowski, Alexander Waibel | 2024-02-27T15:59:37Z | cs.CL | Text segmentation is a fundamental task in natural language processing, where
documents are split into contiguous sections. However, prior research in this
area has been constrained by limited datasets, which are either small in scale,
synthesized, or only contain well-structured documents. In this paper, we
address these limitations by introducing a novel benchmark YTSeg focusing on
spoken content that is inherently more unstructured and both topically and
structurally diverse. As part of this work, we introduce an efficient
hierarchical segmentation model MiniSeg, that outperforms state-of-the-art
baselines. Lastly, we expand the notion of text segmentation to a more
practical "smart chaptering" task that involves the segmentation of
unstructured content, the generation of meaningful segment titles, and a
potential real-time application of the models. | cs.CL | From Text Segmentation to Smart Chaptering: A Novel Benchmark for Structuring Video Transcriptions |
2024-02-27T15:57:11Z | http://arxiv.org/pdf/2402.17630v1 | http://arxiv.org/abs/2402.17630v1 | Huajian Zhang, Yumo Xu, Laura Perez-Beltrachini | 2024-02-27T15:57:11Z | cs.CL | We study existing approaches to leverage off-the-shelf Natural Language
Inference (NLI) models for the evaluation of summary faithfulness and argue
that these are sub-optimal due to the granularity level considered for premises
and hypotheses. That is, the smaller content unit considered as hypothesis is a
sentence and premises are made up of a fixed number of document sentences. We
propose a novel approach, namely InFusE, that uses a variable premise size and
simplifies summary sentences into shorter hypotheses. Departing from previous
studies which focus on single short document summarisation, we analyse NLI
based faithfulness evaluation for diverse summarisation tasks. We introduce
DiverSumm, a new benchmark comprising long form summarisation (long documents
and summaries) and diverse summarisation tasks (e.g., meeting and
multi-document summarisation). In experiments, InFusE obtains superior
performance across the different summarisation tasks. Our code and data are
available at https://github.com/HJZnlp/infuse. | cs.CL | Fine-Grained Natural Language Inference Based Faithfulness Evaluation for Diverse Summarisation Tasks |
2024-02-27T15:42:33Z | http://arxiv.org/pdf/2402.17613v2 | http://arxiv.org/abs/2402.17613v2 | Izia Xiaoxiao Wang, Xihan Wu, Edith Coates, Min Zeng, Jiexin Kuang, Siliang Liu, Mengyang Qiu, Jungyeul Park | 2024-05-06T10:02:08Z | cs.CL | The utilization of technology in second language learning and teaching has
become ubiquitous. For the assessment of writing specifically, automated
writing evaluation (AWE) and grammatical error correction (GEC) have become
immensely popular and effective methods for enhancing writing proficiency and
delivering instant and individualized feedback to learners. By leveraging the
power of natural language processing (NLP) and machine learning algorithms, AWE
and GEC systems have been developed separately to provide language learners
with automated corrective feedback and more accurate and unbiased scoring that
would otherwise be subject to examiners. In this paper, we propose an
integrated system for automated writing evaluation with corrective feedback as
a means of bridging the gap between AWE and GEC results for second language
learners. This system enables language learners to simulate the essay writing
tests: a student writes and submits an essay, and the system returns the
assessment of the writing along with suggested grammatical error corrections.
Given that automated scoring and grammatical correction are more efficient and
cost-effective than human grading, this integrated system would also alleviate
the burden of manually correcting innumerable essays. | cs.CL | Neural Automated Writing Evaluation with Corrective Feedback |
2024-02-27T15:34:15Z | http://arxiv.org/pdf/2402.17608v1 | http://arxiv.org/abs/2402.17608v1 | Alessio Miaschi, Felice Dell'Orletta, Giulia Venturi | 2024-02-27T15:34:15Z | cs.CL | In this paper, we explore the impact of augmenting pre-trained
Encoder-Decoder models, specifically T5, with linguistic knowledge for the
prediction of a target task. In particular, we investigate whether fine-tuning
a T5 model on an intermediate task that predicts structural linguistic
properties of sentences modifies its performance in the target task of
predicting sentence-level complexity. Our study encompasses diverse experiments
conducted on Italian and English datasets, employing both monolingual and
multilingual T5 models at various sizes. Results obtained for both languages
and in cross-lingual configurations show that linguistically motivated
intermediate fine-tuning has generally a positive impact on target task
performance, especially when applied to smaller models and in scenarios with
limited data availability. | cs.CL | Linguistic Knowledge Can Enhance Encoder-Decoder Models (If You Let It) |
2024-02-27T15:14:19Z | http://arxiv.org/pdf/2402.17583v1 | http://arxiv.org/abs/2402.17583v1 | Junjie Huang, Jinyang Liu, Zhuangbin Chen, Zhihan Jiang, Yichen LI, Jiazhen Gu, Cong Feng, Zengyin Yang, Yongqiang Yang, Michael R. Lyu | 2024-02-27T15:14:19Z | cs.SE, cs.CL, cs.LG | Postmortem analysis is essential in the management of incidents within cloud
systems, which provides valuable insights to improve system's reliability and
robustness. At CloudA, fault pattern profiling is performed during the
postmortem phase, which involves the classification of incidents' faults into
unique categories, referred to as fault pattern. By aggregating and analyzing
these fault patterns, engineers can discern common faults, vulnerable
components and emerging fault trends. However, this process is currently
conducted by manual labeling, which has inherent drawbacks. On the one hand,
the sheer volume of incidents means only the most severe ones are analyzed,
causing a skewed overview of fault patterns. On the other hand, the complexity
of the task demands extensive domain knowledge, which leads to errors and
inconsistencies. To address these limitations, we propose an automated
approach, named FaultProfIT, for Fault pattern Profiling of Incident Tickets.
It leverages hierarchy-guided contrastive learning to train a hierarchy-aware
incident encoder and predicts fault patterns with enhanced incident
representations. We evaluate FaultProfIT using the production incidents from
CloudA. The results demonstrate that FaultProfIT outperforms state-of-the-art
methods. Our ablation study and analysis also verify the effectiveness of
hierarchy-guided contrastive learning. Additionally, we have deployed
FaultProfIT at CloudA for six months. To date, FaultProfIT has analyzed 10,000+
incidents from 30+ cloud services, successfully revealing several fault trends
that have informed system improvements. | cs.SE | FaultProfIT: Hierarchical Fault Profiling of Incident Tickets in Large-scale Cloud Systems |
2024-02-27T15:09:20Z | http://arxiv.org/pdf/2402.17574v3 | http://arxiv.org/abs/2402.17574v3 | Wenqi Zhang, Ke Tang, Hai Wu, Mengna Wang, Yongliang Shen, Guiyang Hou, Zeqi Tan, Peng Li, Yueting Zhuang, Weiming Lu | 2024-06-06T18:40:47Z | cs.AI, cs.CL | Large Language Models (LLMs) exhibit robust problem-solving capabilities for
diverse tasks. However, most LLM-based agents are designed as specific task
solvers with sophisticated prompt engineering, rather than agents capable of
learning and evolving through interactions. These task solvers necessitate
manually crafted prompts to inform task rules and regulate LLM behaviors,
inherently incapacitating to address complex dynamic scenarios e.g., large
interactive games. In light of this, we propose Agent-Pro: an LLM-based Agent
with Policy-level Reflection and Optimization that can learn a wealth of
expertise from interactive experiences and progressively elevate its behavioral
policy. Specifically, it involves a dynamic belief generation and reflection
process for policy evolution. Rather than action-level reflection, Agent-Pro
iteratively reflects on past trajectories and beliefs, fine-tuning its
irrational beliefs for a better policy. Moreover, a depth-first search is
employed for policy optimization, ensuring continual enhancement in policy
payoffs. Agent-Pro is evaluated across two games: Blackjack and Texas Hold'em,
outperforming vanilla LLM and specialized models. Our results show Agent-Pro
can learn and evolve in complex and dynamic scenes, which also benefits
numerous LLM-based applications. | cs.AI | Agent-Pro: Learning to Evolve via Policy-Level Reflection and Optimization |
2024-02-27T15:05:32Z | http://arxiv.org/pdf/2402.17564v2 | http://arxiv.org/abs/2402.17564v2 | Xinyu Tang, Xiaolei Wang, Wayne Xin Zhao, Siyuan Lu, Yaliang Li, Ji-Rong Wen | 2024-04-17T03:17:58Z | cs.CL | Automatic prompt optimization is an important approach to improving the
performance of large language models (LLMs). Recent research demonstrates the
potential of using LLMs as prompt optimizers, which can generate improved task
prompts via iterative refinement. In this paper, we propose a novel perspective
to investigate the design of LLM-based prompt optimizers, by drawing an analogy
with gradient-based model optimizers. To connect these two approaches, we
identify two pivotal factors in model parameter learning: update direction and
update method. Focused on the two aspects, we borrow the theoretical framework
and learning methods from gradient-based optimization to design improved
strategies for LLM-based prompt optimizers. By systematically analyzing a rich
set of improvement strategies, we further develop a capable Gradient-inspired
LLM-based Prompt Optimizer called GPO. At each step, it first retrieves
relevant prompts from the optimization trajectory as the update direction.
Then, it utilizes the generation-based refinement strategy to perform the
update, while controlling the edit distance through a cosine-based decay
strategy. Extensive experiments demonstrate the effectiveness and efficiency of
GPO. In particular, GPO brings an additional improvement of up to 56.8% on
Big-Bench Hard and 55.3% on MMLU compared to baseline methods. | cs.CL | Unleashing the Potential of Large Language Models as Prompt Optimizers: An Analogical Analysis with Gradient-based Model Optimizers |
2024-02-27T14:51:11Z | http://arxiv.org/pdf/2402.17812v1 | http://arxiv.org/abs/2402.17812v1 | Sunghyeon Woo, Baeseong Park, Byeongwook Kim, Minjung Jo, Sejung Kwon, Dongsuk Jeon, Dongsoo Lee | 2024-02-27T14:51:11Z | cs.LG, cs.CL | Training deep neural networks typically involves substantial computational
costs during both forward and backward propagation. The conventional layer
dropping techniques drop certain layers during training for reducing the
computations burden. However, dropping layers during forward propagation
adversely affects the training process by degrading accuracy. In this paper, we
propose Dropping Backward Propagation (DropBP), a novel approach designed to
reduce computational costs while maintaining accuracy. DropBP randomly drops
layers during the backward propagation, which does not deviate forward
propagation. Moreover, DropBP calculates the sensitivity of each layer to
assign appropriate drop rate, thereby stabilizing the training process. DropBP
is designed to enhance the efficiency of the training process with
backpropagation, thereby enabling the acceleration of both full fine-tuning and
parameter-efficient fine-tuning using backpropagation. Specifically, utilizing
DropBP in QLoRA reduces training time by 44%, increases the convergence speed
to the identical loss level by 1.5$\times$, and enables training with a
6.2$\times$ larger sequence length on a single NVIDIA-A100 80GiB GPU in
LLaMA2-70B. The code is available at https://github.com/WooSunghyeon/dropbp. | cs.LG | DropBP: Accelerating Fine-Tuning of Large Language Models by Dropping Backward Propagation |
2024-02-27T14:47:53Z | http://arxiv.org/pdf/2402.17553v2 | http://arxiv.org/abs/2402.17553v2 | Raghav Kapoor, Yash Parag Butala, Melisa Russak, Jing Yu Koh, Kiran Kamble, Waseem Alshikh, Ruslan Salakhutdinov | 2024-02-28T17:27:39Z | cs.AI, cs.CL, cs.CV, cs.HC | For decades, human-computer interaction has fundamentally been manual. Even
today, almost all productive work done on the computer necessitates human input
at every step. Autonomous virtual agents represent an exciting step in
automating many of these menial tasks. Virtual agents would empower users with
limited technical proficiency to harness the full possibilities of computer
systems. They could also enable the efficient streamlining of numerous computer
tasks, ranging from calendar management to complex travel bookings, with
minimal human intervention. In this paper, we introduce OmniACT, the
first-of-a-kind dataset and benchmark for assessing an agent's capability to
generate executable programs to accomplish computer tasks. Our scope extends
beyond traditional web automation, covering a diverse range of desktop
applications. The dataset consists of fundamental tasks such as "Play the next
song", as well as longer horizon tasks such as "Send an email to John Doe
mentioning the time and place to meet". Specifically, given a pair of screen
image and a visually-grounded natural language task, the goal is to generate a
script capable of fully executing the task. We run several strong baseline
language model agents on our benchmark. The strongest baseline, GPT-4, performs
the best on our benchmark However, its performance level still reaches only 15%
of the human proficiency in generating executable scripts capable of completing
the task, demonstrating the challenge of our task for conventional web agents.
Our benchmark provides a platform to measure and evaluate the progress of
language model agents in automating computer tasks and motivates future work
towards building multimodal models that bridge large language models and the
visual grounding of computer screens. | cs.AI | OmniACT: A Dataset and Benchmark for Enabling Multimodal Generalist Autonomous Agents for Desktop and Web |
2024-02-27T14:45:04Z | http://arxiv.org/pdf/2402.17811v2 | http://arxiv.org/abs/2402.17811v2 | Shaolei Zhang, Tian Yu, Yang Feng | 2024-06-05T11:15:04Z | cs.CL, cs.AI, cs.LG | Large Language Models (LLMs) sometimes suffer from producing hallucinations,
especially LLMs may generate untruthful responses despite knowing the correct
knowledge. Activating the truthfulness within LLM is the key to fully unlocking
LLM's knowledge potential. In this paper, we propose TruthX, an inference-time
intervention method to activate the truthfulness of LLM by identifying and
editing the features within LLM's internal representations that govern the
truthfulness. TruthX employs an auto-encoder to map LLM's representations into
semantic and truthful latent spaces respectively, and applies contrastive
learning to identify a truthful editing direction within the truthful space.
During inference, by editing LLM's internal representations in truthful space,
TruthX effectively enhances the truthfulness of LLM. Experiments show that
TruthX improves the truthfulness of 13 advanced LLMs by an average of 20% on
TruthfulQA benchmark. Further analyses suggest that TruthX can control LLM to
produce truthful or hallucinatory responses via editing only one vector in
LLM's internal representations. | cs.CL | TruthX: Alleviating Hallucinations by Editing Large Language Models in Truthful Space |
2024-02-27T14:38:47Z | http://arxiv.org/pdf/2402.17546v1 | http://arxiv.org/abs/2402.17546v1 | Suyeon Lee, Jieun Kang, Harim Kim, Kyoung-Mee Chung, Dongha Lee, Jinyoung Yeo | 2024-02-27T14:38:47Z | cs.AI, cs.CL | The demand for conversational agents that provide mental health care is
consistently increasing. In this work, we develop a psychological counseling
agent, referred to as CoCoA, that applies Cognitive Behavioral Therapy (CBT)
techniques to identify and address cognitive distortions inherent in the
client's statements. Specifically, we construct a memory system to efficiently
manage information necessary for counseling while extracting high-level
insights about the client from their utterances. Additionally, to ensure that
the counseling agent generates appropriate responses, we introduce dynamic
prompting to flexibly apply CBT techniques and facilitate the appropriate
retrieval of information. We conducted dialogues between CoCoA and characters
from Character.ai, creating a dataset for evaluation. Then, we asked GPT to
evaluate the constructed counseling dataset, and our model demonstrated a
statistically significant difference from other models. | cs.AI | COCOA: CBT-based Conversational Counseling Agent using Memory Specialized in Cognitive Distortions and Dynamic Prompt |
2024-02-27T14:16:19Z | http://arxiv.org/pdf/2402.17532v3 | http://arxiv.org/abs/2402.17532v3 | Bowen Cao, Deng Cai, Leyang Cui, Xuxin Cheng, Wei Bi, Yuexian Zou, Shuming Shi | 2024-03-16T04:31:47Z | cs.CL | Standard language models generate text by selecting tokens from a fixed,
finite, and standalone vocabulary. We introduce a novel method that selects
context-aware phrases from a collection of supporting documents. One of the
most significant challenges for this paradigm shift is determining the training
oracles, because a string of text can be segmented in various ways and each
segment can be retrieved from numerous possible documents. To address this, we
propose to initialize the training oracles using linguistic heuristics and,
more importantly, bootstrap the oracles through iterative self-reinforcement.
Extensive experiments show that our model not only outperforms standard
language models on a variety of knowledge-intensive tasks but also demonstrates
improved generation quality in open-ended text generation. For instance,
compared to the standard language model counterpart, our model raises the
accuracy from 23.47% to 36.27% on OpenbookQA, and improves the MAUVE score from
42.61% to 81.58% in open-ended text generation. Remarkably, our model also
achieves the best performance and the lowest latency among several
retrieval-augmented baselines. In conclusion, we assert that retrieval is more
accurate generation and hope that our work will encourage further research on
this new paradigm shift. | cs.CL | Retrieval is Accurate Generation |
2024-02-27T14:14:23Z | http://arxiv.org/pdf/2402.17531v2 | http://arxiv.org/abs/2402.17531v2 | Kaikai An, Fangkai Yang, Junting Lu, Liqun Li, Zhixing Ren, Hao Huang, Lu Wang, Pu Zhao, Yu Kang, Hua Ding, Qingwei Lin, Saravan Rajmohan, Dongmei Zhang, Qi Zhang | 2024-05-10T11:57:46Z | cs.SE, cs.AI, cs.CL | Effective incident management is pivotal for the smooth operation of
enterprises-level cloud services. In order to expedite incident mitigation,
service teams compile troubleshooting knowledge into Troubleshooting Guides
(TSGs) accessible to on-call engineers (OCEs). While automated pipelines are
enabled to resolve the most frequent and easy incidents, there still exist
complex incidents that require OCEs' intervention. However, TSGs are often
unstructured and incomplete, which requires manual interpretation by OCEs,
leading to on-call fatigue and decreased productivity, especially among
new-hire OCEs. In this work, we propose Nissist which leverages TSGs and
incident mitigation histories to provide proactive suggestions, reducing human
intervention. Leveraging Large Language Models (LLM), Nissist extracts insights
from unstructured TSGs and historical incident mitigation discussions, forming
a comprehensive knowledge base. Its multi-agent system design enhances
proficiency in precisely discerning user queries, retrieving relevant
information, and delivering systematic plans consecutively. Through our user
case and experiment, we demonstrate that Nissist significant reduce Time to
Mitigate (TTM) in incident mitigation, alleviating operational burdens on OCEs
and improving service reliability. Our demo is available at
https://aka.ms/nissist_demo. | cs.SE | Nissist: An Incident Mitigation Copilot based on Troubleshooting Guides |
2024-02-27T14:11:32Z | http://arxiv.org/pdf/2402.17527v2 | http://arxiv.org/abs/2402.17527v2 | Evgenia Ilia, Wilker Aziz | 2024-03-18T16:21:24Z | cs.CL, cs.AI | Language models (LMs) are statistical models trained to assign probability to
human-generated text. As such, it is reasonable to question whether they
approximate linguistic variability exhibited by humans well. This form of
statistical assessment is difficult to perform at the passage level, for it
requires acceptability judgements (i.e., human evaluation) or a robust
automated proxy (which is non-trivial). At the word level, however, given some
context, samples from an LM can be assessed via exact matching against a
prerecorded dataset of alternative single-word continuations of the available
context. We exploit this fact and evaluate the LM's ability to reproduce
variability that humans (in particular, a population of English speakers)
exhibit in the 'next word prediction' task. This can be seen as assessing a
form of calibration, which, in the context of text classification, Baan et al.
(2022) termed calibration to human uncertainty. We assess GPT2, BLOOM and
ChatGPT and find that they exhibit fairly low calibration to human uncertainty.
We also verify the failure of expected calibration error (ECE) to reflect this,
and as such, advise the community against relying on it in this setting. | cs.CL | Predict the Next Word: Humans exhibit uncertainty in this task and language models _____ |
2024-02-27T13:54:48Z | http://arxiv.org/pdf/2402.17512v2 | http://arxiv.org/abs/2402.17512v2 | Rares Dolga, Marius Cobzarenco, David Barber | 2024-03-04T12:21:52Z | cs.CL, stat.ML | The time complexity of the standard attention mechanism in a transformer
scales quadratically with the length of the sequence. We introduce a method to
reduce this to linear scaling with time, based on defining attention via latent
vectors. The method is readily usable as a drop-in replacement for the standard
attention mechanism. Our "Latte Transformer" model can be implemented for both
bidirectional and unidirectional tasks, with the causal version allowing a
recurrent implementation which is memory and time-efficient during inference of
language generation tasks. Whilst next token prediction scales linearly with
the sequence length for a standard transformer, a Latte Transformer requires
constant time to compute the next token. The empirical performance of our
method is comparable to standard attention, yet allows scaling to context
windows much larger than practical in standard attention. | cs.CL | Latent Attention for Linear Time Transformers |
2024-02-27T13:49:12Z | http://arxiv.org/pdf/2402.17509v2 | http://arxiv.org/abs/2402.17509v2 | Vyas Raina, Samson Tan, Volkan Cevher, Aditya Rawal, Sheng Zha, George Karypis | 2024-05-31T00:09:56Z | cs.CL | Deep learning-based Natural Language Processing (NLP) models are vulnerable
to adversarial attacks, where small perturbations can cause a model to
misclassify. Adversarial Training (AT) is often used to increase model
robustness. However, we have discovered an intriguing phenomenon: deliberately
or accidentally miscalibrating models masks gradients in a way that interferes
with adversarial attack search methods, giving rise to an apparent increase in
robustness. We show that this observed gain in robustness is an illusion of
robustness (IOR), and demonstrate how an adversary can perform various forms of
test-time temperature calibration to nullify the aforementioned interference
and allow the adversarial attack to find adversarial examples. Hence, we urge
the NLP community to incorporate test-time temperature scaling into their
robustness evaluations to ensure that any observed gains are genuine. Finally,
we show how the temperature can be scaled during \textit{training} to improve
genuine robustness. | cs.CL | Extreme Miscalibration and the Illusion of Adversarial Robustness |
2024-02-27T13:44:09Z | http://arxiv.org/pdf/2402.17505v1 | http://arxiv.org/abs/2402.17505v1 | Ruiyang Ren, Peng Qiu, Yingqi Qu, Jing Liu, Wayne Xin Zhao, Hua Wu, Ji-Rong Wen, Haifeng Wang | 2024-02-27T13:44:09Z | cs.IR, cs.CL | Due to the excellent capacities of large language models (LLMs), it becomes
feasible to develop LLM-based agents for reliable user simulation. Considering
the scarcity and limit (e.g., privacy issues) of real user data, in this paper,
we conduct large-scale user simulation for web search, to improve the analysis
and modeling of user search behavior. Specially, we propose BASES, a novel user
simulation framework with LLM-based agents, designed to facilitate
comprehensive simulations of web search user behaviors. Our simulation
framework can generate unique user profiles at scale, which subsequently leads
to diverse search behaviors. To demonstrate the effectiveness of BASES, we
conduct evaluation experiments based on two human benchmarks in both Chinese
and English, demonstrating that BASES can effectively simulate large-scale
human-like search behaviors. To further accommodate the research on web search,
we develop WARRIORS, a new large-scale dataset encompassing web search user
behaviors, including both Chinese and English versions, which can greatly
bolster research in the field of information retrieval. Our code and data will
be publicly released soon. | cs.IR | BASES: Large-scale Web Search User Simulation with Large Language Model based Agents |
2024-02-27T13:22:51Z | http://arxiv.org/pdf/2402.17497v1 | http://arxiv.org/abs/2402.17497v1 | Yuhao Wang, Ruiyang Ren, Junyi Li, Wayne Xin Zhao, Jing Liu, Ji-Rong Wen | 2024-02-27T13:22:51Z | cs.CL, cs.IR | Considering the limited internal parametric knowledge, retrieval-augmented
generation (RAG) has been widely used to extend the knowledge scope of large
language models (LLMs). Despite the extensive efforts on RAG research, in
existing methods, LLMs cannot precisely assess the relevance of retrieved
documents, thus likely leading to misleading or even incorrect utilization of
external knowledge (i.e., retrieved documents). To address this issue, in this
paper, we propose REAR, a RElevance-Aware Retrieval-augmented approach for
open-domain question answering (QA). As the key motivation, we aim to enhance
the self-awareness of source relevance for LLMs, so as to adaptively utilize
external knowledge in RAG systems. Specially, we develop a new architecture for
LLM based RAG system, by incorporating a specially designed rank head that
precisely assesses the relevance of retrieved documents. Furthermore, we
propose an improved training method based on bi-granularity relevance fusion
and noise-resistant training. By combining the improvements in both
architecture and training, our proposed REAR can better utilize external
knowledge by effectively perceiving the relevance of retrieved documents.
Experiments on four open-domain QA tasks show that REAR significantly
outperforms previous a number of competitive RAG approaches. Our code and data
can be accessed at https://github.com/RUCAIBox/REAR. | cs.CL | REAR: A Relevance-Aware Retrieval-Augmented Framework for Open-Domain Question Answering |
2024-02-27T13:22:47Z | http://arxiv.org/pdf/2402.17496v2 | http://arxiv.org/abs/2402.17496v2 | Lucía Gómez Zaragozá, Rocío del Amor, Elena Parra Vargas, Valery Naranjo, Mariano Alcañiz Raya, Javier Marín-Morales | 2024-06-13T13:09:48Z | cs.SD, cs.AI, cs.CL, eess.AS, I.5.1; I.5.4; I.2.7 | Emotional Voice Messages (EMOVOME) is a spontaneous speech dataset containing
999 audio messages from real conversations on a messaging app from 100 Spanish
speakers, gender balanced. Voice messages were produced in-the-wild conditions
before participants were recruited, avoiding any conscious bias due to
laboratory environment. Audios were labeled in valence and arousal dimensions
by three non-experts and two experts, which were then combined to obtain a
final label per dimension. The experts also provided an extra label
corresponding to seven emotion categories. To set a baseline for future
investigations using EMOVOME, we implemented emotion recognition models using
both speech and audio transcriptions. For speech, we used the standard eGeMAPS
feature set and support vector machines, obtaining 49.27% and 44.71% unweighted
accuracy for valence and arousal respectively. For text, we fine-tuned a
multilingual BERT model and achieved 61.15% and 47.43% unweighted accuracy for
valence and arousal respectively. This database will significantly contribute
to research on emotion recognition in the wild, while also providing a unique
natural and freely accessible resource for Spanish. | cs.SD | Emotional Voice Messages (EMOVOME) database: emotion recognition in spontaneous voice messages |
2024-02-27T13:18:00Z | http://arxiv.org/pdf/2402.17493v4 | http://arxiv.org/abs/2402.17493v4 | Bing Xue, Charles Alba, Joanna Abraham, Thomas Kannampallil, Chenyang Lu | 2024-05-05T04:07:44Z | cs.CL, J.3; I.2.7 | Predicting postoperative risk can inform effective care management &
planning. We explored large language models (LLMs) in predicting postoperative
risk through clinical texts using various tuning strategies. Records spanning
84,875 patients from Barnes Jewish Hospital (BJH) between 2018 & 2021, with a
mean duration of follow-up based on the length of postoperative ICU stay less
than 7 days, were utilized. Methods were replicated on the MIMIC-III dataset.
Outcomes included 30-day mortality, pulmonary embolism (PE) & pneumonia. Three
domain adaptation & finetuning strategies were implemented for three LLMs
(BioGPT, ClinicalBERT & BioClinicalBERT): self-supervised objectives;
incorporating labels with semi-supervised fine-tuning; & foundational modelling
through multi-task learning. Model performance was compared using the AUROC &
AUPRC for classification tasks & MSE & R2 for regression tasks. Cohort had a
mean age of 56.9 (sd: 16.8) years; 50.3% male; 74% White. Pre-trained LLMs
outperformed traditional word embeddings, with absolute maximal gains of 38.3%
for AUROC & 14% for AUPRC. Adapting models through self-supervised finetuning
further improved performance by 3.2% for AUROC & 1.5% for AUPRC Incorporating
labels into the finetuning procedure further boosted performances, with
semi-supervised finetuning improving by 1.8% for AUROC & 2% for AUPRC &
foundational modelling improving by 3.6% for AUROC & 2.6% for AUPRC compared to
self-supervised finetuning. Pre-trained clinical LLMs offer opportunities for
postoperative risk predictions with unseen data, & further improvements from
finetuning suggests benefits in adapting pre-trained models to note-specific
perioperative use cases. Incorporating labels can further boost performance.
The superior performance of foundational models suggests the potential of
task-agnostic learning towards the generalizable LLMs in perioperative care. | cs.CL | Predicting postoperative risks using large language models |
2024-02-27T13:02:19Z | http://arxiv.org/pdf/2402.17478v1 | http://arxiv.org/abs/2402.17478v1 | Maram Hasanain, Fatema Ahmed, Firoj Alam | 2024-02-27T13:02:19Z | cs.CL | The use of propaganda has spiked on mainstream and social media, aiming to
manipulate or mislead users. While efforts to automatically detect propaganda
techniques in textual, visual, or multimodal content have increased, most of
them primarily focus on English content. The majority of the recent initiatives
targeting medium to low-resource languages produced relatively small annotated
datasets, with a skewed distribution, posing challenges for the development of
sophisticated propaganda detection models. To address this challenge, we
carefully develop the largest propaganda dataset to date, ArPro, comprised of
8K paragraphs from newspaper articles, labeled at the text span level following
a taxonomy of 23 propagandistic techniques. Furthermore, our work offers the
first attempt to understand the performance of large language models (LLMs),
using GPT-4, for fine-grained propaganda detection from text. Results showed
that GPT-4's performance degrades as the task moves from simply classifying a
paragraph as propagandistic or not, to the fine-grained task of detecting
propaganda techniques and their manifestation in text. Compared to models
fine-tuned on the dataset for propaganda detection at different classification
granularities, GPT-4 is still far behind. Finally, we evaluate GPT-4 on a
dataset consisting of six other languages for span detection, and results
suggest that the model struggles with the task across languages. Our dataset
and resources will be released to the community. | cs.CL | Can GPT-4 Identify Propaganda? Annotation and Detection of Propaganda Spans in News Articles |
2024-02-27T12:39:23Z | http://arxiv.org/pdf/2402.17463v2 | http://arxiv.org/abs/2402.17463v2 | Chenxin An, Fei Huang, Jun Zhang, Shansan Gong, Xipeng Qiu, Chang Zhou, Lingpeng Kong | 2024-05-29T05:44:58Z | cs.CL | The ability of Large Language Models (LLMs) to process and generate coherent
text is markedly weakened when the number of input tokens exceeds their
pretraining length. Given the expensive overhead of finetuning large-scale
models with longer sequences, we propose Dual Chunk Attention (DCA), which
enables Llama2 70B to support context windows of more than 100k tokens without
continual training. By decomposing the attention computation for long sequences
into chunk-based modules, DCA manages to effectively capture the relative
positional information of tokens within the same chunk (Intra-Chunk) and across
distinct chunks (Inter-Chunk), as well as integrates seamlessly with Flash
Attention. In addition to its impressive extrapolation capability, DCA achieves
performance on practical long-context tasks that is comparable to or even
better than that of finetuned models. When compared with proprietary models,
our training-free 70B model attains 94% of the performance of gpt-3.5-16k,
indicating it is a viable open-source alternative. All code and data used in
this work are released at \url{https://github.com/HKUNLP/ChunkLlama}. | cs.CL | Training-Free Long-Context Scaling of Large Language Models |
2024-02-27T12:27:51Z | http://arxiv.org/pdf/2402.17456v1 | http://arxiv.org/abs/2402.17456v1 | Michael A. Hedderich, Natalie N. Bazarova, Wenting Zou, Ryun Shim, Xinda Ma, Qian Yang | 2024-02-27T12:27:51Z | cs.HC, cs.AI, cs.CL | Cyberbullying harms teenagers' mental health, and teaching them upstanding
intervention is crucial. Wizard-of-Oz studies show chatbots can scale up
personalized and interactive cyberbullying education, but implementing such
chatbots is a challenging and delicate task. We created a no-code chatbot
design tool for K-12 teachers. Using large language models and prompt chaining,
our tool allows teachers to prototype bespoke dialogue flows and chatbot
utterances. In offering this tool, we explore teachers' distinctive needs when
designing chatbots to assist their teaching, and how chatbot design tools might
better support them. Our findings reveal that teachers welcome the tool
enthusiastically. Moreover, they see themselves as playwrights guiding both the
students' and the chatbot's behaviors, while allowing for some improvisation.
Their goal is to enable students to rehearse both desirable and undesirable
reactions to cyberbullying in a safe environment. We discuss the design
opportunities LLM-Chains offer for empowering teachers and the research
opportunities this work opens up. | cs.HC | A Piece of Theatre: Investigating How Teachers Design LLM Chatbots to Assist Adolescent Cyberbullying Education |
2024-02-27T12:03:56Z | http://arxiv.org/pdf/2402.17447v2 | http://arxiv.org/abs/2402.17447v2 | Mansi Goel, Ayush Agarwal, Shubham Agrawal, Janak Kapuriya, Akhil Vamshi Konam, Rishabh Gupta, Shrey Rastogi, Niharika, Ganesh Bagler | 2024-06-06T07:41:21Z | cs.CL, cs.AI, cs.IR | Food touches our lives through various endeavors, including flavor,
nourishment, health, and sustainability. Recipes are cultural capsules
transmitted across generations via unstructured text. Automated protocols for
recognizing named entities, the building blocks of recipe text, are of immense
value for various applications ranging from information extraction to novel
recipe generation. Named entity recognition is a technique for extracting
information from unstructured or semi-structured data with known labels.
Starting with manually-annotated data of 6,611 ingredient phrases, we created
an augmented dataset of 26,445 phrases cumulatively. Simultaneously, we
systematically cleaned and analyzed ingredient phrases from RecipeDB, the
gold-standard recipe data repository, and annotated them using the Stanford
NER. Based on the analysis, we sampled a subset of 88,526 phrases using a
clustering-based approach while preserving the diversity to create the
machine-annotated dataset. A thorough investigation of NER approaches on these
three datasets involving statistical, fine-tuning of deep learning-based
language models and few-shot prompting on large language models (LLMs) provides
deep insights. We conclude that few-shot prompting on LLMs has abysmal
performance, whereas the fine-tuned spaCy-transformer emerges as the best model
with macro-F1 scores of 95.9%, 96.04%, and 95.71% for the manually-annotated,
augmented, and machine-annotated datasets, respectively. | cs.CL | Deep Learning Based Named Entity Recognition Models for Recipes |
2024-02-27T11:50:05Z | http://arxiv.org/pdf/2402.17437v1 | http://arxiv.org/abs/2402.17437v1 | Zhou Yang, Zhaochun Ren, Yufeng Wang, Xiaofei Zhu, Zhihao Chen, Tiecheng Cai, Yunbing Wu, Yisong Su, Sibo Ju, Xiangwen Liao | 2024-02-27T11:50:05Z | cs.CL, cs.AI | Empathetic response generation aims to generate empathetic responses by
understanding the speaker's emotional feelings from the language of dialogue.
Recent methods capture emotional words in the language of communicators and
construct them as static vectors to perceive nuanced emotions. However,
linguistic research has shown that emotional words in language are dynamic and
have correlations with other grammar semantic roles, i.e., words with semantic
meanings, in grammar. Previous methods overlook these two characteristics,
which easily lead to misunderstandings of emotions and neglect of key
semantics. To address this issue, we propose a dynamical Emotion-Semantic
Correlation Model (ESCM) for empathetic dialogue generation tasks. ESCM
constructs dynamic emotion-semantic vectors through the interaction of context
and emotions. We introduce dependency trees to reflect the correlations between
emotions and semantics. Based on dynamic emotion-semantic vectors and
dependency trees, we propose a dynamic correlation graph convolutional network
to guide the model in learning context meanings in dialogue and generating
empathetic responses. Experimental results on the EMPATHETIC-DIALOGUES dataset
show that ESCM understands semantics and emotions more accurately and expresses
fluent and informative empathetic responses. Our analysis results also indicate
that the correlations between emotions and semantics are frequently used in
dialogues, which is of great significance for empathetic perception and
expression. | cs.CL | Exploiting Emotion-Semantic Correlations for Empathetic Response Generation |
2024-02-27T11:45:21Z | http://arxiv.org/pdf/2402.17433v3 | http://arxiv.org/abs/2402.17433v3 | Jiaqi Wang, Zhenxi Song, Zhengyu Ma, Xipeng Qiu, Min Zhang, Zhiguo Zhang | 2024-06-10T09:51:50Z | cs.CL | Reconstructing natural language from non-invasive electroencephalography
(EEG) holds great promise as a language decoding technology for brain-computer
interfaces (BCIs). However, EEG-based language decoding is still in its nascent
stages, facing several technical issues such as: 1) Absence of a hybrid
strategy that can effectively integrate cross-modality (between EEG and text)
self-learning with intra-modality self-reconstruction of EEG features or
textual sequences; 2) Under-utilization of large language models (LLMs) to
enhance EEG-based language decoding. To address above issues, we propose the
Contrastive EEG-Text Masked Autoencoder (CET-MAE), a novel model that
orchestrates compound self-supervised learning across and within EEG and text
through a dedicated multi-stream encoder. Furthermore, we develop a framework
called E2T-PTR (EEG-to-Text decoding using Pretrained Transferable
Representations), which leverages pre-trained modules alongside the EEG stream
from CET-MAE and further enables an LLM (specifically BART) to decode text from
EEG sequences. Comprehensive experiments conducted on the popular text-evoked
EEG database, ZuCo, demonstrate the superiority of E2T-PTR, which outperforms
the state-of-the-art in ROUGE-1 F1 and BLEU-4 scores by 8.34% and 32.21%,
respectively. These results indicate significant advancements in the field and
underscores the proposed framework's potential to enable more powerful and
widespread BCI applications. | cs.CL | Enhancing EEG-to-Text Decoding through Transferable Representations from Pre-trained Contrastive EEG-Text Masked Autoencoder |
2024-02-27T11:02:12Z | http://arxiv.org/pdf/2402.17411v2 | http://arxiv.org/abs/2402.17411v2 | Fufangchen Zhao, Guoqiang Jin, Jiaheng Huang, Rui Zhao, Fei Tan | 2024-03-02T14:08:06Z | cs.CL | Nowadays both commercial and open-source academic LLM have become the
mainstream models of NLP. However, there is still a lack of research on LLM
consistency, meaning that throughout the various stages of LLM research and
deployment, its internal parameters and capabilities should remain unchanged.
This issue exists in both the industrial and academic sectors. The solution to
this problem is often time-consuming and labor-intensive, and there is also an
additional cost of secondary deployment, resulting in economic and time losses.
To fill this gap, we build an LLM consistency task dataset and design several
baselines. Additionally, we choose models of diverse scales for the main
experiments. Specifically, in the LightGBM experiment, we used traditional NLG
metrics (i.e., ROUGE, BLEU, METEOR) as the features needed for model training.
The final result exceeds the manual evaluation and GPT3.5 as well as other
models in the main experiment, achieving the best performance. In the end, we
use the best performing LightGBM model as the base model to build the
evaluation tool, which can effectively assist in the deployment of business
models. Our code and tool demo are available at
https://github.com/heavenhellchen/Consistency.git | cs.CL | Consistency Matters: Explore LLMs Consistency From a Black-Box Perspective |
2024-02-27T10:57:07Z | http://arxiv.org/pdf/2402.17407v1 | http://arxiv.org/abs/2402.17407v1 | Flavio Petruzzellis, Alberto Testolin, Alessandro Sperduti | 2024-02-27T10:57:07Z | cs.NE, cs.AI, cs.CL | Modern neural network architectures still struggle to learn algorithmic
procedures that require to systematically apply compositional rules to solve
out-of-distribution problem instances. In this work, we propose an original
approach to learn algorithmic tasks inspired by rewriting systems, a classic
framework in symbolic artificial intelligence. We show that a rewriting system
can be implemented as a neural architecture composed by specialized modules:
the Selector identifies the target sub-expression to process, the Solver
simplifies the sub-expression by computing the corresponding result, and the
Combiner produces a new version of the original expression by replacing the
sub-expression with the solution provided. We evaluate our model on three types
of algorithmic tasks that require simplifying symbolic formulas involving
lists, arithmetic, and algebraic expressions. We test the extrapolation
capabilities of the proposed architecture using formulas involving a higher
number of operands and nesting levels than those seen during training, and we
benchmark its performance against the Neural Data Router, a recent model
specialized for systematic generalization, and a state-of-the-art large
language model (GPT-4) probed with advanced prompting strategies. | cs.NE | A Neural Rewriting System to Solve Algorithmic Problems |
2024-02-27T10:47:24Z | http://arxiv.org/pdf/2402.17400v1 | http://arxiv.org/abs/2402.17400v1 | Çağatay Yıldız, Nishaanth Kanna Ravichandran, Prishruit Punia, Matthias Bethge, Beyza Ermis | 2024-02-27T10:47:24Z | cs.CL | This paper studies the evolving domain of Continual Learning (CL) in large
language models (LLMs), with a focus on developing strategies for efficient and
sustainable training. Our primary emphasis is on continual domain-adaptive
pretraining, a process designed to equip LLMs with the ability to integrate new
information from various domains while retaining previously learned knowledge
and enhancing cross-domain knowledge transfer without relying on
domain-specific identification. Unlike previous studies, which mostly
concentrate on a limited selection of tasks or domains and primarily aim to
address the issue of forgetting, our research evaluates the adaptability and
capabilities of LLMs to changing data landscapes in practical scenarios. To
this end, we introduce a new benchmark designed to measure the adaptability of
LLMs to these evolving data environments, offering a comprehensive framework
for evaluation. We examine the impact of model size on learning efficacy and
forgetting, as well as how the progression and similarity of emerging domains
affect the knowledge transfer within these models. Our findings uncover several
key insights: (i) when the sequence of domains shows semantic similarity,
continual pretraining enables LLMs to better specialize in the current domain
compared to stand-alone fine-tuning, (ii) training across a diverse range of
domains enhances both backward and forward knowledge transfer, and (iii)
smaller models are particularly sensitive to continual pretraining, showing the
most significant rates of both forgetting and learning. We posit that our
research marks a shift towards establishing a more realistic benchmark for
investigating CL in LLMs, and has the potential to play a key role in guiding
the direction of future research in the field. | cs.CL | Investigating Continual Pretraining in Large Language Models: Insights and Implications |
2024-02-27T10:44:52Z | http://arxiv.org/pdf/2402.17396v1 | http://arxiv.org/abs/2402.17396v1 | Flavio Petruzzellis, Alberto Testolin, Alessandro Sperduti | 2024-02-27T10:44:52Z | cs.CL, cs.AI, cs.NE | Large Language Models (LLMs) have revolutionized the field of Natural
Language Processing thanks to their ability to reuse knowledge acquired on
massive text corpora on a wide variety of downstream tasks, with minimal (if
any) tuning steps. At the same time, it has been repeatedly shown that LLMs
lack systematic generalization, which allows to extrapolate the learned
statistical regularities outside the training distribution. In this work, we
offer a systematic benchmarking of GPT-4, one of the most advanced LLMs
available, on three algorithmic tasks characterized by the possibility to
control the problem difficulty with two parameters. We compare the performance
of GPT-4 with that of its predecessor (GPT-3.5) and with a variant of the
Transformer-Encoder architecture recently introduced to solve similar tasks,
the Neural Data Router. We find that the deployment of advanced prompting
techniques allows GPT-4 to reach superior accuracy on all tasks, demonstrating
that state-of-the-art LLMs constitute a very strong baseline also in
challenging tasks that require systematic generalization. | cs.CL | Benchmarking GPT-4 on Algorithmic Problems: A Systematic Evaluation of Prompting Strategies |
2024-02-27T10:38:37Z | http://arxiv.org/pdf/2402.17392v1 | http://arxiv.org/abs/2402.17392v1 | Vasilii A. Gromov, Alexandra S. Kogan | 2024-02-27T10:38:37Z | cs.CL | Nowadays, technology is rapidly advancing: bots are writing comments,
articles, and reviews. Due to this fact, it is crucial to know if the text was
written by a human or by a bot. This paper focuses on comparing structures of
the coarse-grained partitions of semantic paths for human-written and
bot-generated texts. We compare the clusterizations of datasets of n-grams from
literary texts and texts generated by several bots. The hypothesis is that the
structures and clusterizations are different. Our research supports the
hypothesis. As the semantic structure may be different for different languages,
we investigate Russian, English, German, and Vietnamese languages. | cs.CL | Spot the bot: Coarse-Grained Partition of Semantic Paths for Bots and Humans |
2024-02-27T10:31:00Z | http://arxiv.org/pdf/2402.17389v1 | http://arxiv.org/abs/2402.17389v1 | Mattia Setzu, Marta Marchiori Manerba, Pasquale Minervini, Debora Nozza | 2024-02-27T10:31:00Z | cs.CL, cs.AI | Language Models (LMs) have been shown to inherit undesired biases that might
hurt minorities and underrepresented groups if such systems were integrated
into real-world applications without careful fairness auditing. This paper
proposes FairBelief, an analytical approach to capture and assess beliefs,
i.e., propositions that an LM may embed with different degrees of confidence
and that covertly influence its predictions. With FairBelief, we leverage
prompting to study the behavior of several state-of-the-art LMs across
different previously neglected axes, such as model scale and likelihood,
assessing predictions on a fairness dataset specifically designed to quantify
LMs' outputs' hurtfulness. Finally, we conclude with an in-depth qualitative
assessment of the beliefs emitted by the models. We apply FairBelief to English
LMs, revealing that, although these architectures enable high performances on
diverse natural language processing tasks, they show hurtful beliefs about
specific genders. Interestingly, training procedure and dataset, model scale,
and architecture induce beliefs of different degrees of hurtfulness. | cs.CL | FairBelief -- Assessing Harmful Beliefs in Language Models |
2024-02-27T10:22:45Z | http://arxiv.org/pdf/2403.00826v1 | http://arxiv.org/abs/2403.00826v1 | Shubh Goyal, Medha Hira, Shubham Mishra, Sukriti Goyal, Arnav Goel, Niharika Dadu, Kirushikesh DB, Sameep Mehta, Nishtha Madaan | 2024-02-27T10:22:45Z | cs.CL, cs.CR, cs.LG | Although the rise of Large Language Models (LLMs) in enterprise settings
brings new opportunities and capabilities, it also brings challenges, such as
the risk of generating inappropriate, biased, or misleading content that
violates regulations and can have legal concerns. To alleviate this, we present
"LLMGuard", a tool that monitors user interactions with an LLM application and
flags content against specific behaviours or conversation topics. To do this
robustly, LLMGuard employs an ensemble of detectors. | cs.CL | LLMGuard: Guarding Against Unsafe LLM Behavior |
2024-02-27T10:14:57Z | http://arxiv.org/pdf/2402.17377v2 | http://arxiv.org/abs/2402.17377v2 | Seongbo Jang, Seonghyeon Lee, Hwanjo Yu | 2024-06-17T05:12:56Z | cs.CL | As language models are often deployed as chatbot assistants, it becomes a
virtue for models to engage in conversations in a user's first language. While
these models are trained on a wide range of languages, a comprehensive
evaluation of their proficiency in low-resource languages such as Korean has
been lacking. In this work, we introduce KoDialogBench, a benchmark designed to
assess language models' conversational capabilities in Korean. To this end, we
collect native Korean dialogues on daily topics from public sources, or
translate dialogues from other languages. We then structure these conversations
into diverse test datasets, spanning from dialogue comprehension to response
selection tasks. Leveraging the proposed benchmark, we conduct extensive
evaluations and analyses of various language models to measure a foundational
understanding of Korean dialogues. Experimental results indicate that there
exists significant room for improvement in models' conversation skills.
Furthermore, our in-depth comparisons across different language models
highlight the effectiveness of recent training techniques in enhancing
conversational proficiency. We anticipate that KoDialogBench will promote the
progress towards conversation-aware Korean language models. | cs.CL | KoDialogBench: Evaluating Conversational Understanding of Language Models with Korean Dialogue Benchmark |
2024-02-27T10:09:40Z | http://arxiv.org/pdf/2402.17371v1 | http://arxiv.org/abs/2402.17371v1 | Michael Toker, Oren Mishali, Ophir Münz-Manor, Benny Kimelfeld, Yonatan Belinkov | 2024-02-27T10:09:40Z | cs.CL, I.2.7 | There is a large volume of late antique and medieval Hebrew texts. They
represent a crucial linguistic and cultural bridge between Biblical and modern
Hebrew. Poetry is prominent in these texts and one of its main haracteristics
is the frequent use of metaphor. Distinguishing figurative and literal language
use is a major task for scholars of the Humanities, especially in the fields of
literature, linguistics, and hermeneutics. This paper presents a new,
challenging dataset of late antique and medieval Hebrew poetry with expert
annotations of metaphor, as well as some baseline results, which we hope will
facilitate further research in this area. | cs.CL | A Dataset for Metaphor Detection in Early Medieval Hebrew Poetry |
2024-02-27T09:52:27Z | http://arxiv.org/pdf/2402.17358v1 | http://arxiv.org/abs/2402.17358v1 | Xinyu Lu, Bowen Yu, Yaojie Lu, Hongyu Lin, Haiyang Yu, Le Sun, Xianpei Han, Yongbin Li | 2024-02-27T09:52:27Z | cs.CL | The alignment problem in Large Language Models (LLMs) involves adapting them
to the broad spectrum of human values. This requirement challenges existing
alignment methods due to diversity of preferences and regulatory standards.
This paper introduces a novel alignment paradigm, priority rule following,
which defines rules as the primary control mechanism in each dialog,
prioritizing them over user instructions. Our preliminary analysis reveals that
even the advanced LLMs, such as GPT-4, exhibit shortcomings in understanding
and prioritizing the rules. Therefore, we present PriorityDistill, a
semi-automated approach for distilling priority following signals from LLM
simulations to ensure robust rule integration and adherence. Our experiments
show that this method not only effectively minimizes misalignments utilizing
only one general rule but also adapts smoothly to various unseen rules,
ensuring they are shielded from hijacking and that the model responds
appropriately. | cs.CL | SoFA: Shielded On-the-fly Alignment via Priority Rule Following |
2024-02-27T09:47:36Z | http://arxiv.org/pdf/2402.17355v1 | http://arxiv.org/abs/2402.17355v1 | Qi Zhang, Yiming Zhang, Haobo Wang, Junbo Zhao | 2024-02-27T09:47:36Z | cs.CL | In the current landscape of large language models (LLMs), the process of
instruction tuning serves as an essential step. Considering the high computing
power overhead, data-efficient instruction tuning was proposed to reduce the
training data size in this process, aiming at selecting high-quality
instructional data. Nevertheless, we argue that most current data-efficient
instruction-tuning methods are highly dependent on the quality of the original
instruction-tuning dataset. When it comes to datasets synthesized by LLMs, a
common scenario in this field, dirty samples will even be selected with a
higher probability than other samples. To address these challenges, we utilized
external knowledge (relevant examples or paragraphs) to evaluate those samples
synthesized by LLMs with an in-context-based relative predictive entropy. Based
on the new metric, we proposed a framework, dubbed as \textbf{RECOST}, which
integrates external-knowledge-base re-ranking and diversity-consistent sampling
into a single pipeline. Through extensive experiments on several synthetic
datasets (Alpaca and Alpaca-gpt4), we demonstrate the effectiveness of our
method and achieve even better results with only \textbf{1\%} of the full
dataset. | cs.CL | RECOST: External Knowledge Guided Data-efficient Instruction Tuning |
2024-02-27T09:10:28Z | http://arxiv.org/pdf/2402.17333v1 | http://arxiv.org/abs/2402.17333v1 | Qin Zhang, Hao Ge, Xiaojun Chen, Meng Fang | 2024-02-27T09:10:28Z | cs.CL | Unsupervised question answering is a promising yet challenging task, which
alleviates the burden of building large-scale annotated data in a new domain.
It motivates us to study the unsupervised multiple-choice question answering
(MCQA) problem. In this paper, we propose a novel framework designed to
generate synthetic MCQA data barely based on contexts from the universal domain
without relying on any form of manual annotation. Possible answers are
extracted and used to produce related questions, then we leverage both named
entities (NE) and knowledge graphs to discover plausible distractors to form
complete synthetic samples. Experiments on multiple MCQA datasets demonstrate
the effectiveness of our method. | cs.CL | Unsupervised multiple choices question answering via universal corpus |
2024-02-27T08:33:31Z | http://arxiv.org/pdf/2402.17311v1 | http://arxiv.org/abs/2402.17311v1 | Huy Quoc To, Hung-Nghiep Tran, Andr'e Greiner-Petter, Felix Beierle, Akiko Aizawa | 2024-02-27T08:33:31Z | cs.CL | Summarization for scientific text has shown significant benefits both for the
research community and human society. Given the fact that the nature of
scientific text is distinctive and the input of the multi-document
summarization task is substantially long, the task requires sufficient
embedding generation and text truncation without losing important information.
To tackle these issues, in this paper, we propose SKT5SciSumm - a hybrid
framework for multi-document scientific summarization (MDSS). We leverage the
Sentence-Transformer version of Scientific Paper Embeddings using
Citation-Informed Transformers (SPECTER) to encode and represent textual
sentences, allowing for efficient extractive summarization using k-means
clustering. We employ the T5 family of models to generate abstractive summaries
using extracted sentences. SKT5SciSumm achieves state-of-the-art performance on
the Multi-XScience dataset. Through extensive experiments and evaluation, we
showcase the benefits of our model by using less complicated models to achieve
remarkable results, thereby highlighting its potential in advancing the field
of multi-document summarization for scientific text. | cs.CL | SKT5SciSumm -- A Hybrid Generative Approach for Multi-Document Scientific Summarization |
2024-02-27T08:27:15Z | http://arxiv.org/pdf/2402.17304v2 | http://arxiv.org/abs/2402.17304v2 | Mingxu Tao, Quzhe Huang, Kun Xu, Liwei Chen, Yansong Feng, Dongyan Zhao | 2024-03-27T02:59:57Z | cs.CL, cs.AI | The advancement of Multimodal Large Language Models (MLLMs) has greatly
accelerated the development of applications in understanding integrated texts
and images. Recent works leverage image-caption datasets to train MLLMs,
achieving state-of-the-art performance on image-to-text tasks. However, there
are few studies exploring which layers of MLLMs make the most effort to the
global image information, which plays vital roles in multimodal comprehension
and generation. In this study, we find that the intermediate layers of models
can encode more global semantic information, whose representation vectors
perform better on visual-language entailment tasks, rather than the topmost
layers. We further probe models regarding local semantic representations
through object recognition tasks. We find that the topmost layers may
excessively focus on local information, leading to a diminished ability to
encode global information. Our code and data are released via
https://github.com/kobayashikanna01/probing_MLLM_rep. | cs.CL | Probing Multimodal Large Language Models for Global and Local Semantic Representations |
2024-02-27T08:24:32Z | http://arxiv.org/pdf/2402.17302v2 | http://arxiv.org/abs/2402.17302v2 | Rifki Afina Putri, Faiz Ghifari Haznitrama, Dea Adhista, Alice Oh | 2024-04-16T07:41:12Z | cs.CL | Large Language Models (LLMs) are increasingly being used to generate
synthetic data for training and evaluating models. However, it is unclear
whether they can generate a good quality of question answering (QA) dataset
that incorporates knowledge and cultural nuance embedded in a language,
especially for low-resource languages. In this study, we investigate the
effectiveness of using LLMs in generating culturally relevant commonsense QA
datasets for Indonesian and Sundanese languages. To do so, we create datasets
for these languages using various methods involving both LLMs and human
annotators, resulting in ~4.5K questions per language (~9K in total), making
our dataset the largest of its kind. Our experiments show that automatic data
adaptation from an existing English dataset is less effective for Sundanese.
Interestingly, using the direct generation method on the target language, GPT-4
Turbo can generate questions with adequate general knowledge in both languages,
albeit not as culturally 'deep' as humans. We also observe a higher occurrence
of fluency errors in the Sundanese dataset, highlighting the discrepancy
between medium- and lower-resource languages. | cs.CL | Can LLM Generate Culturally Relevant Commonsense QA Data? Case Study in Indonesian and Sundanese |
2024-02-27T08:18:02Z | http://arxiv.org/pdf/2403.18926v2 | http://arxiv.org/abs/2403.18926v2 | Yuanhang Yang, Shiyi Qi, Wenchao Gu, Chaozheng Wang, Cuiyun Gao, Zenglin Xu | 2024-05-24T10:14:55Z | cs.LG, cs.CL | Sparse models, including sparse Mixture-of-Experts (MoE) models, have emerged
as an effective approach for scaling Transformer models. However, they often
suffer from computational inefficiency since a significant number of parameters
are unnecessarily involved in computations via multiplying values by zero or
low activation values. To address this issue, we present \tool, a novel MoE
designed to enhance both the efficacy and efficiency of sparse MoE models.
\tool leverages small experts and a threshold-based router to enable tokens to
selectively engage only essential parameters. Our extensive experiments on
language modeling and machine translation tasks demonstrate that \tool can
enhance model performance while decreasing the computation load at MoE layers
by over 50\% without sacrificing performance. Furthermore, we present the
versatility of \tool by applying it to dense models, enabling sparse
computation during inference. We provide a comprehensive analysis and make our
code available at https://github.com/ysngki/XMoE. | cs.LG | XMoE: Sparse Models with Fine-grained and Adaptive Expert Selection |
2024-02-27T07:26:16Z | http://arxiv.org/pdf/2403.00825v1 | http://arxiv.org/abs/2403.00825v1 | Jongga Lee, Jaeseung Yim, Seohee Park, Changwon Lim | 2024-02-27T07:26:16Z | cs.CL | Text classification is the task of assigning a document to a predefined
class. However, it is expensive to acquire enough labeled documents or to label
them. In this paper, we study the regularization methods' effects on various
classification models when only a few labeled data are available. We compare a
simple word embedding-based model, which is simple but effective, with complex
models (CNN and BiLSTM). In supervised learning, adversarial training can
further regularize the model. When an unlabeled dataset is available, we can
regularize the model using semi-supervised learning methods such as the Pi
model and virtual adversarial training. We evaluate the regularization effects
on four text classification datasets (AG news, DBpedia, Yahoo! Answers, Yelp
Polarity), using only 0.1% to 0.5% of the original labeled training documents.
The simple model performs relatively well in fully supervised learning, but
with the help of adversarial training and semi-supervised learning, both simple
and complex models can be regularized, showing better results for complex
models. Although the simple model is robust to overfitting, a complex model
with well-designed prior beliefs can be also robust to overfitting. | cs.CL | Comparing effectiveness of regularization methods on text classification: Simple and complex model in data shortage situation |
2024-02-27T07:14:12Z | http://arxiv.org/pdf/2402.17263v2 | http://arxiv.org/abs/2402.17263v2 | Pengjie Ren, Chengshun Shi, Shiguang Wu, Mengqi Zhang, Zhaochun Ren, Maarten de Rijke, Zhumin Chen, Jiahuan Pei | 2024-06-24T12:45:27Z | cs.CL, 68T50, I.2.7 | Parameter-efficient fine-tuning (PEFT) is a popular method for tailoring
pre-trained large language models (LLMs), especially as the models' scale and
the diversity of tasks increase. Low-rank adaptation (LoRA) is based on the
idea that the adaptation process is intrinsically low-dimensional, i.e.,
significant model changes can be represented with relatively few parameters.
However, decreasing the rank encounters challenges with generalization errors
for specific tasks when compared to full-parameter fine-tuning. We present
MELoRA, a mini-ensemble low-rank adapters that uses fewer trainable parameters
while maintaining a higher rank, thereby offering improved performance
potential. The core idea is to freeze original pretrained weights and train a
group of mini LoRAs with only a small number of parameters. This can capture a
significant degree of diversity among mini LoRAs, thus promoting better
generalization ability. We conduct a theoretical analysis and empirical studies
on various NLP tasks. Our experimental results show that, compared to LoRA,
MELoRA achieves better performance with 8 times fewer trainable parameters on
natural language understanding tasks and 36 times fewer trainable parameters on
instruction following tasks, which demonstrates the effectiveness of MELoRA. | cs.CL | MELoRA: Mini-Ensemble Low-Rank Adapters for Parameter-Efficient Fine-Tuning |
2024-02-27T07:11:59Z | http://arxiv.org/pdf/2402.17262v1 | http://arxiv.org/abs/2402.17262v1 | Zhenhong Zhou, Jiuyang Xiang, Haopeng Chen, Quan Liu, Zherui Li, Sen Su | 2024-02-27T07:11:59Z | cs.CL, cs.AI | Large Language Models (LLMs) have been demonstrated to generate illegal or
unethical responses, particularly when subjected to "jailbreak." Research on
jailbreak has highlighted the safety issues of LLMs. However, prior studies
have predominantly focused on single-turn dialogue, ignoring the potential
complexities and risks presented by multi-turn dialogue, a crucial mode through
which humans derive information from LLMs. In this paper, we argue that humans
could exploit multi-turn dialogue to induce LLMs into generating harmful
information. LLMs may not intend to reject cautionary or borderline unsafe
queries, even if each turn is closely served for one malicious purpose in a
multi-turn dialogue. Therefore, by decomposing an unsafe query into several
sub-queries for multi-turn dialogue, we induced LLMs to answer harmful
sub-questions incrementally, culminating in an overall harmful response. Our
experiments, conducted across a wide range of LLMs, indicate current
inadequacies in the safety mechanisms of LLMs in multi-turn dialogue. Our
findings expose vulnerabilities of LLMs in complex scenarios involving
multi-turn dialogue, presenting new challenges for the safety of LLMs. | cs.CL | Speak Out of Turn: Safety Vulnerability of Large Language Models in Multi-turn Dialogue |
2024-02-27T07:02:10Z | http://arxiv.org/pdf/2402.17256v2 | http://arxiv.org/abs/2402.17256v2 | Pei Wang, Keqing He, Yejie Wang, Xiaoshuai Song, Yutao Mou, Jingang Wang, Yunsen Xian, Xunliang Cai, Weiran Xu | 2024-03-04T06:04:32Z | cs.CL | Out-of-domain (OOD) intent detection aims to examine whether the user's query
falls outside the predefined domain of the system, which is crucial for the
proper functioning of task-oriented dialogue (TOD) systems. Previous methods
address it by fine-tuning discriminative models. Recently, some studies have
been exploring the application of large language models (LLMs) represented by
ChatGPT to various downstream tasks, but it is still unclear for their ability
on OOD detection task.This paper conducts a comprehensive evaluation of LLMs
under various experimental settings, and then outline the strengths and
weaknesses of LLMs. We find that LLMs exhibit strong zero-shot and few-shot
capabilities, but is still at a disadvantage compared to models fine-tuned with
full resource. More deeply, through a series of additional analysis
experiments, we discuss and summarize the challenges faced by LLMs and provide
guidance for future work including injecting domain knowledge, strengthening
knowledge transfer from IND(In-domain) to OOD, and understanding long
instructions. | cs.CL | Beyond the Known: Investigating LLMs Performance on Out-of-Domain Intent Detection |
2024-02-27T06:11:54Z | http://arxiv.org/pdf/2402.17237v1 | http://arxiv.org/abs/2402.17237v1 | Rui Cheng, Wanqing Cui | 2024-02-27T06:11:54Z | cs.CV, cs.CL | Existing two-stream models for image-text matching show good performance
while ensuring retrieval speed and have received extensive attention from
industry and academia. These methods use a single representation to encode
image and text separately and get a matching score with cosine similarity or
the inner product of vectors. However, the performance of the two-stream model
is often sub-optimal. On the one hand, a single representation is challenging
to cover complex content comprehensively. On the other hand, in this framework
of lack of interaction, it is challenging to match multiple meanings which
leads to information being ignored. To address the problems mentioned above and
facilitate the performance of the two-stream model, we propose a multi-view
attention approach for two-stream image-text matching MVAM
(\textbf{M}ulti-\textbf{V}iew \textbf{A}ttention \textbf{M}odel). It first
learns multiple image and text representations by diverse attention heads with
different view codes. And then concatenate these representations into one for
matching. A diversity objective is also used to promote diversity between
attention heads. With this method, models are able to encode images and text
from different views and attend to more key points. So we can get
representations that contain more information. When doing retrieval tasks, the
matching scores between images and texts can be calculated from different
aspects, leading to better matching performance. Experiment results on MSCOCO
and Flickr30K show that our proposed model brings improvements over existing
models. Further case studies show that different attention heads can focus on
different contents and finally obtain a more comprehensive representation. | cs.CV | Image-Text Matching with Multi-View Attention |
2024-02-27T05:50:35Z | http://arxiv.org/pdf/2402.17231v3 | http://arxiv.org/abs/2402.17231v3 | Debrup Das, Debopriyo Banerjee, Somak Aditya, Ashish Kulkarni | 2024-04-03T15:22:35Z | cs.CL | Tool-augmented Large Language Models (TALMs) are known to enhance the
skillset of large language models (LLMs), thereby, leading to their improved
reasoning abilities across many tasks. While, TALMs have been successfully
employed in different question-answering benchmarks, their efficacy on complex
mathematical reasoning benchmarks, and the potential complementary benefits
offered by tools for knowledge retrieval and mathematical equation solving are
open research questions. In this work, we present MathSensei, a tool-augmented
large language model for mathematical reasoning. We study the complementary
benefits of the tools - knowledge retriever (Bing Web Search), program
generator + executor (Python), and symbolic equation solver (Wolfram-Alpha API)
through evaluations on mathematical reasoning datasets. We perform exhaustive
ablations on MATH, a popular dataset for evaluating mathematical reasoning on
diverse mathematical disciplines. We also conduct experiments involving
well-known tool planners to study the impact of tool sequencing on the model
performance. MathSensei achieves 13.5% better accuracy over gpt-3.5-turbo with
Chain-of-Thought on the MATH dataset. We further observe that TALMs are not as
effective for simpler math word problems (in GSM-8K), and the benefit increases
as the complexity and required knowledge increases (progressively over AQuA,
MMLU-Math, and higher level complex questions in MATH). The code and data are
available at https://github.com/Debrup-61/MathSensei. | cs.CL | MATHSENSEI: A Tool-Augmented Large Language Model for Mathematical Reasoning |
2024-02-27T05:37:10Z | http://arxiv.org/pdf/2402.17226v1 | http://arxiv.org/abs/2402.17226v1 | Xiaolong Wang, Yile Wang, Yuanchi Zhang, Fuwen Luo, Peng Li, Maosong Sun, Yang Liu | 2024-02-27T05:37:10Z | cs.CL | Large Language Models (LLMs) have achieved remarkable performance in
objective tasks such as open-domain question answering and mathematical
reasoning, which can often be solved through recalling learned factual
knowledge or chain-of-thought style reasoning. However, we find that the
performance of LLMs in subjective tasks is still unsatisfactory, such as
metaphor recognition, dark humor detection, etc. Compared to objective tasks,
subjective tasks focus more on interpretation or emotional response rather than
a universally accepted reasoning pathway. Based on the characteristics of the
tasks and the strong dialogue-generation capabilities of LLMs, we propose RiC
(Reasoning in Conversation), a method that focuses on solving subjective tasks
through dialogue simulation. The motivation of RiC is to mine useful contextual
information by simulating dialogues instead of supplying chain-of-thought style
rationales, thereby offering potential useful knowledge behind dialogues for
giving the final answers. We evaluate both API-based and open-source LLMs
including GPT-4, ChatGPT, and OpenChat across twelve tasks. Experimental
results show that RiC can yield significant improvement compared with various
baselines. | cs.CL | Reasoning in Conversation: Solving Subjective Tasks through Dialogue Simulation for Large Language Models |
2024-02-27T04:55:03Z | http://arxiv.org/pdf/2402.17205v3 | http://arxiv.org/abs/2402.17205v3 | Jianhao Shen, Ye Yuan, Srbuhi Mirzoyan, Ming Zhang, Chenguang Wang | 2024-05-22T05:11:56Z | cs.CL, cs.AI, cs.LG | We introduce a new challenge to test the STEM skills of neural models. The
problems in the real world often require solutions, combining knowledge from
STEM (science, technology, engineering, and math). Unlike existing datasets,
our dataset requires the understanding of multimodal vision-language
information of STEM. Our dataset features one of the largest and most
comprehensive datasets for the challenge. It includes 448 skills and 1,073,146
questions spanning all STEM subjects. Compared to existing datasets that often
focus on examining expert-level ability, our dataset includes fundamental
skills and questions designed based on the K-12 curriculum. We also add
state-of-the-art foundation models such as CLIP and GPT-3.5-Turbo to our
benchmark. Results show that the recent model advances only help master a very
limited number of lower grade-level skills (2.5% in the third grade) in our
dataset. In fact, these models are still well below (averaging 54.7%) the
performance of elementary students, not to mention near expert-level
performance. To understand and increase the performance on our dataset, we
teach the models on a training split of our dataset. Even though we observe
improved performance, the model performance remains relatively low compared to
average elementary students. To solve STEM problems, we will need novel
algorithmic innovations from the community. | cs.CL | Measuring Vision-Language STEM Skills of Neural Models |
2024-02-27T04:18:49Z | http://arxiv.org/pdf/2402.17193v1 | http://arxiv.org/abs/2402.17193v1 | Biao Zhang, Zhongtao Liu, Colin Cherry, Orhan Firat | 2024-02-27T04:18:49Z | cs.CL, cs.LG | While large language models (LLMs) often adopt finetuning to unlock their
capabilities for downstream applications, our understanding on the inductive
biases (especially the scaling properties) of different finetuning methods is
still limited. To fill this gap, we conduct systematic experiments studying
whether and how different scaling factors, including LLM model size,
pretraining data size, new finetuning parameter size and finetuning data size,
affect the finetuning performance. We consider two types of finetuning --
full-model tuning (FMT) and parameter efficient tuning (PET, including prompt
tuning and LoRA), and explore their scaling behaviors in the data-limited
regime where the LLM model size substantially outweighs the finetuning data
size. Based on two sets of pretrained bilingual LLMs from 1B to 16B and
experiments on bilingual machine translation and multilingual summarization
benchmarks, we find that 1) LLM finetuning follows a powerbased multiplicative
joint scaling law between finetuning data size and each other scaling factor;
2) LLM finetuning benefits more from LLM model scaling than pretraining data
scaling, and PET parameter scaling is generally ineffective; and 3) the optimal
finetuning method is highly task- and finetuning data-dependent. We hope our
findings could shed light on understanding, selecting and developing LLM
finetuning methods. | cs.CL | When Scaling Meets LLM Finetuning: The Effect of Data, Model and Finetuning Method |
2024-02-27T04:09:30Z | http://arxiv.org/pdf/2403.14657v1 | http://arxiv.org/abs/2403.14657v1 | Stanley Chinedu Okoro, Alexander Lopez, Austine Unuriode | 2024-02-27T04:09:30Z | cs.CY, cs.CL | Over the past few years, wildfires have become a worldwide environmental
emergency, resulting in substantial harm to natural habitats and playing a part
in the acceleration of climate change. Wildfire management methods involve
prevention, response, and recovery efforts. Despite improvements in detection
techniques, the rising occurrence of wildfires demands creative solutions for
prompt identification and effective control. This research investigates
proactive methods for detecting and handling wildfires in the United States,
utilizing Artificial Intelligence (AI), Machine Learning (ML), and 5G
technology. The specific objective of this research covers proactive detection
and prevention of wildfires using advanced technology; Active monitoring and
mapping with remote sensing and signaling leveraging on 5G technology; and
Advanced response mechanisms to wildfire using drones and IOT devices. This
study was based on secondary data collected from government databases and
analyzed using descriptive statistics. In addition, past publications were
reviewed through content analysis, and narrative synthesis was used to present
the observations from various studies. The results showed that developing new
technology presents an opportunity to detect and manage wildfires proactively.
Utilizing advanced technology could save lives and prevent significant economic
losses caused by wildfires. Various methods, such as AI-enabled remote sensing
and 5G-based active monitoring, can enhance proactive wildfire detection and
management. In addition, super intelligent drones and IOT devices can be used
for safer responses to wildfires. This forms the core of the recommendation to
the fire Management Agencies and the government. | cs.CY | A Synergistic Approach to Wildfire Prevention and Management Using AI, ML, and 5G Technology in the United States |
2024-02-27T04:08:59Z | http://arxiv.org/pdf/2402.17189v1 | http://arxiv.org/abs/2402.17189v1 | Tzu-Ting Yang, Hsin-Wei Wang, Yi-Cheng Wang, Chi-Han Lin, Berlin Chen | 2024-02-27T04:08:59Z | cs.CL, cs.AI, cs.SD, eess.AS | With the massive developments of end-to-end (E2E) neural networks, recent
years have witnessed unprecedented breakthroughs in automatic speech
recognition (ASR). However, the codeswitching phenomenon remains a major
obstacle that hinders ASR from perfection, as the lack of labeled data and the
variations between languages often lead to degradation of ASR performance. In
this paper, we focus exclusively on improving the acoustic encoder of E2E ASR
to tackle the challenge caused by the codeswitching phenomenon. Our main
contributions are threefold: First, we introduce a novel disentanglement loss
to enable the lower-layer of the encoder to capture inter-lingual acoustic
information while mitigating linguistic confusion at the higher-layer of the
encoder. Second, through comprehensive experiments, we verify that our proposed
method outperforms the prior-art methods using pretrained dual-encoders,
meanwhile having access only to the codeswitching corpus and consuming half of
the parameterization. Third, the apparent differentiation of the encoders'
output features also corroborates the complementarity between the
disentanglement loss and the mixture-of-experts (MoE) architecture. | cs.CL | An Effective Mixture-Of-Experts Approach For Code-Switching Speech Recognition Leveraging Encoder Disentanglement |
2024-02-27T03:40:44Z | http://arxiv.org/pdf/2402.17184v1 | http://arxiv.org/abs/2402.17184v1 | Rohit Prabhavalkar, Zhong Meng, Weiran Wang, Adam Stooke, Xingyu Cai, Yanzhang He, Arun Narayanan, Dongseong Hwang, Tara N. Sainath, Pedro J. Moreno | 2024-02-27T03:40:44Z | cs.CL, cs.SD, eess.AS | The accuracy of end-to-end (E2E) automatic speech recognition (ASR) models
continues to improve as they are scaled to larger sizes, with some now reaching
billions of parameters. Widespread deployment and adoption of these models,
however, requires computationally efficient strategies for decoding. In the
present work, we study one such strategy: applying multiple frame reduction
layers in the encoder to compress encoder outputs into a small number of output
frames. While similar techniques have been investigated in previous work, we
achieve dramatically more reduction than has previously been demonstrated
through the use of multiple funnel reduction layers. Through ablations, we
study the impact of various architectural choices in the encoder to identify
the most effective strategies. We demonstrate that we can generate one encoder
output frame for every 2.56 sec of input speech, without significantly
affecting word error rate on a large-scale voice search task, while improving
encoder and decoder latencies by 48% and 92% respectively, relative to a strong
but computationally expensive baseline. | cs.CL | Extreme Encoder Output Frame Rate Reduction: Improving Computational Latencies of Large End-to-End Models |
2024-02-27T03:03:06Z | http://arxiv.org/pdf/2402.17168v1 | http://arxiv.org/abs/2402.17168v1 | Yuge Zhang, Qiyang Jiang, Xingyu Han, Nan Chen, Yuqing Yang, Kan Ren | 2024-02-27T03:03:06Z | cs.AI, cs.CL | In the era of data-driven decision-making, the complexity of data analysis
necessitates advanced expertise and tools of data science, presenting
significant challenges even for specialists. Large Language Models (LLMs) have
emerged as promising aids as data science agents, assisting humans in data
analysis and processing. Yet their practical efficacy remains constrained by
the varied demands of real-world applications and complicated analytical
process. In this paper, we introduce DSEval -- a novel evaluation paradigm, as
well as a series of innovative benchmarks tailored for assessing the
performance of these agents throughout the entire data science lifecycle.
Incorporating a novel bootstrapped annotation method, we streamline dataset
preparation, improve the evaluation coverage, and expand benchmarking
comprehensiveness. Our findings uncover prevalent obstacles and provide
critical insights to inform future advancements in the field. | cs.AI | Benchmarking Data Science Agents |
2024-02-27T02:36:43Z | http://arxiv.org/pdf/2402.17151v2 | http://arxiv.org/abs/2402.17151v2 | Zhengxiang Wang, Owen Rambow | 2024-04-26T20:01:28Z | cs.CL | We propose a novel clustering pipeline to detect and characterize influence
campaigns from documents. This approach clusters parts of document, detects
clusters that likely reflect an influence campaign, and then identifies
documents linked to an influence campaign via their association with the
high-influence clusters. Our approach outperforms both the direct
document-level classification and the direct document-level clustering approach
in predicting if a document is part of an influence campaign. We propose
various novel techniques to enhance our pipeline, including using an existing
event factuality prediction system to obtain document parts, and aggregating
multiple clustering experiments to improve the performance of both cluster and
document classification. Classifying documents after clustering not only
accurately extracts the parts of the documents that are relevant to influence
campaigns, but also captures influence campaigns as a coordinated and holistic
phenomenon. Our approach makes possible more fine-grained and interpretable
characterizations of influence campaigns from documents. | cs.CL | Clustering Document Parts: Detecting and Characterizing Influence Campaigns from Documents |
2024-02-27T02:00:28Z | http://arxiv.org/pdf/2403.07910v2 | http://arxiv.org/abs/2403.07910v2 | Tomáš Horych, Martin Wessel, Jan Philip Wahle, Terry Ruas, Jerome Waßmuth, André Greiner-Petter, Akiko Aizawa, Bela Gipp, Timo Spinde | 2024-03-15T09:30:29Z | cs.CY, cs.CL | Media bias detection poses a complex, multifaceted problem traditionally
tackled using single-task models and small in-domain datasets, consequently
lacking generalizability. To address this, we introduce MAGPIE, the first
large-scale multi-task pre-training approach explicitly tailored for media bias
detection. To enable pre-training at scale, we present Large Bias Mixture
(LBM), a compilation of 59 bias-related tasks. MAGPIE outperforms previous
approaches in media bias detection on the Bias Annotation By Experts (BABE)
dataset, with a relative improvement of 3.3% F1-score. MAGPIE also performs
better than previous models on 5 out of 8 tasks in the Media Bias
Identification Benchmark (MBIB). Using a RoBERTa encoder, MAGPIE needs only 15%
of finetuning steps compared to single-task approaches. Our evaluation shows,
for instance, that tasks like sentiment and emotionality boost all learning,
all tasks enhance fake news detection, and scaling tasks leads to the best
results. MAGPIE confirms that MTL is a promising approach for addressing media
bias detection, enhancing the accuracy and efficiency of existing models.
Furthermore, LBM is the first available resource collection focused on media
bias MTL. | cs.CY | MAGPIE: Multi-Task Media-Bias Analysis Generalization for Pre-Trained Identification of Expressions |
2024-02-27T01:48:19Z | http://arxiv.org/pdf/2402.17128v4 | http://arxiv.org/abs/2402.17128v4 | Nguyen Nguyen, Jing Bi, Ali Vosoughi, Yapeng Tian, Pooyan Fazli, Chenliang Xu | 2024-04-02T23:14:42Z | cs.CV, cs.AI, cs.CL, cs.LG | The capability of intelligent models to extrapolate and comprehend changes in
object states is a crucial yet demanding aspect of AI research, particularly
through the lens of human interaction in real-world settings. This task
involves describing complex visual environments, identifying active objects,
and interpreting their changes as conveyed through language. Traditional
methods, which isolate object captioning and state change detection, offer a
limited view of dynamic environments. Moreover, relying on a small set of
symbolic words to represent changes has restricted the expressiveness of the
language. To address these challenges, in this paper, we introduce the Object
State Captioning and State Change Representation (OSCaR) dataset and benchmark.
OSCaR consists of 14,084 annotated video segments with nearly 1,000 unique
objects from various egocentric video collections. It sets a new testbed for
evaluating multimodal large language models (MLLMs). Our experiments
demonstrate that while MLLMs show some skill, they lack a full understanding of
object state changes. The benchmark includes a fine-tuned model that, despite
initial capabilities, requires significant improvements in accuracy and
generalization ability for effective understanding of these changes. Our code
and dataset are available at https://github.com/nguyennm1024/OSCaR. | cs.CV | OSCaR: Object State Captioning and State Change Representation |
2024-02-27T01:37:23Z | http://arxiv.org/pdf/2402.17124v1 | http://arxiv.org/abs/2402.17124v1 | Xinran Zhao, Hongming Zhang, Xiaoman Pan, Wenlin Yao, Dong Yu, Tongshuang Wu, Jianshu Chen | 2024-02-27T01:37:23Z | cs.CL | For a LLM to be trustworthy, its confidence level should be well-calibrated
with its actual performance. While it is now common sense that LLM performances
are greatly impacted by prompts, the confidence calibration in prompting LLMs
has yet to be thoroughly explored. In this paper, we explore how different
prompting strategies influence LLM confidence calibration and how it could be
improved. We conduct extensive experiments on six prompting methods in the
question-answering context and we observe that, while these methods help
improve the expected LLM calibration, they also trigger LLMs to be
over-confident when responding to some instances. Inspired by human cognition,
we propose Fact-and-Reflection (FaR) prompting, which improves the LLM
calibration in two steps. First, FaR elicits the known "facts" that are
relevant to the input prompt from the LLM. And then it asks the model to
"reflect" over them to generate the final answer. Experiments show that FaR
prompting achieves significantly better calibration; it lowers the Expected
Calibration Error by 23.5% on our multi-purpose QA tasks. Notably, FaR
prompting even elicits the capability of verbally expressing concerns in less
confident scenarios, which helps trigger retrieval augmentation for solving
these harder instances. | cs.CL | Fact-and-Reflection (FaR) Improves Confidence Calibration of Large Language Models |
2024-02-27T01:25:52Z | http://arxiv.org/pdf/2402.17119v1 | http://arxiv.org/abs/2402.17119v1 | Kaige Xie, Mark Riedl | 2024-02-27T01:25:52Z | cs.CL | Automated story generation has been one of the long-standing challenges in
NLP. Among all dimensions of stories, suspense is very common in human-written
stories but relatively under-explored in AI-generated stories. While recent
advances in large language models (LLMs) have greatly promoted language
generation in general, state-of-the-art LLMs are still unreliable when it comes
to suspenseful story generation. We propose a novel iterative-prompting-based
planning method that is grounded in two theoretical foundations of story
suspense from cognitive psychology and narratology. This theory-grounded method
works in a fully zero-shot manner and does not rely on any supervised story
corpora. To the best of our knowledge, this paper is the first attempt at
suspenseful story generation with LLMs. Extensive human evaluations of the
generated suspenseful stories demonstrate the effectiveness of our method. | cs.CL | Creating Suspenseful Stories: Iterative Planning with Large Language Models |
2024-02-27T01:13:58Z | http://arxiv.org/pdf/2402.17110v1 | http://arxiv.org/abs/2402.17110v1 | Xiao Cui, Yulei Qin, Yuting Gao, Enwei Zhang, Zihan Xu, Tong Wu, Ke Li, Xing Sun, Wengang Zhou, Houqiang Li | 2024-02-27T01:13:58Z | cs.LG, cs.CL | Knowledge distillation (KD) has been widely adopted to compress large
language models (LLMs). Existing KD methods investigate various divergence
measures including the Kullback-Leibler (KL), reverse Kullback-Leibler (RKL),
and Jensen-Shannon (JS) divergences. However, due to limitations inherent in
their assumptions and definitions, these measures fail to deliver effective
supervision when few distribution overlap exists between the teacher and the
student. In this paper, we show that the aforementioned KL, RKL, and JS
divergences respectively suffer from issues of mode-averaging, mode-collapsing,
and mode-underestimation, which deteriorates logits-based KD for diverse NLP
tasks. We propose the Sinkhorn Knowledge Distillation (SinKD) that exploits the
Sinkhorn distance to ensure a nuanced and precise assessment of the disparity
between teacher and student distributions. Besides, profit by properties of the
Sinkhorn metric, we can get rid of sample-wise KD that restricts the perception
of divergence in each teacher-student sample pair. Instead, we propose a
batch-wise reformulation to capture geometric intricacies of distributions
across samples in the high-dimensional space. Comprehensive evaluation on GLUE
and SuperGLUE, in terms of comparability, validity, and generalizability,
highlights our superiority over state-of-the-art methods on all kinds of LLMs
with encoder-only, encoder-decoder, and decoder-only architectures. | cs.LG | Sinkhorn Distance Minimization for Knowledge Distillation |
2024-02-27T00:24:42Z | http://arxiv.org/pdf/2403.00824v2 | http://arxiv.org/abs/2403.00824v2 | Javier Ferrando, Elena Voita | 2024-04-16T23:32:38Z | cs.CL, cs.AI | Information flows by routes inside the network via mechanisms implemented in
the model. These routes can be represented as graphs where nodes correspond to
token representations and edges to operations inside the network. We
automatically build these graphs in a top-down manner, for each prediction
leaving only the most important nodes and edges. In contrast to the existing
workflows relying on activation patching, we do this through attribution: this
allows us to efficiently uncover existing circuits with just a single forward
pass. Additionally, the applicability of our method is far beyond patching: we
do not need a human to carefully design prediction templates, and we can
extract information flow routes for any prediction (not just the ones among the
allowed templates). As a result, we can talk about model behavior in general,
for specific types of predictions, or different domains. We experiment with
Llama 2 and show that the role of some attention heads is overall important,
e.g. previous token heads and subword merging heads. Next, we find similarities
in Llama 2 behavior when handling tokens of the same part of speech. Finally,
we show that some model components can be specialized on domains such as coding
or multilingual texts. | cs.CL | Information Flow Routes: Automatically Interpreting Language Models at Scale |
2024-02-27T00:22:18Z | http://arxiv.org/pdf/2402.17097v2 | http://arxiv.org/abs/2402.17097v2 | Juyeon Kim, Jeongeun Lee, Yoonho Chang, Chanyeol Choi, Junseong Kim, Jy-yong Sohn | 2024-04-12T11:37:44Z | cs.CL, cs.AI | Mitigating hallucination issues is a key challenge that must be overcome to
reliably deploy large language models (LLMs) in real-world scenarios. Recently,
various methods have been proposed to detect and revise factual errors in
LLM-generated texts, in order to reduce hallucination. In this paper, we
propose Re-Ex, a method for post-editing LLM-generated responses. Re-Ex
introduces a novel reasoning step dubbed as the factual error explanation step.
Re-Ex revises the initial response of LLMs using 3-steps : first, external
tools are used to retrieve the evidences of the factual errors in the initial
LLM response; next, LLM is instructed to explain the problematic parts of the
response based on the gathered evidence; finally, LLM revises the initial
response using the explanations provided in the previous step. In addition to
the explanation step, Re-Ex also incorporates new prompting techniques to
reduce the token count and inference time required for the response revision
process. Compared with existing methods including FacTool, CoVE, and RARR,
Re-Ex provides better detection and revision performance with less inference
time and fewer tokens in multiple benchmarks. | cs.CL | Re-Ex: Revising after Explanation Reduces the Factual Errors in LLM Responses |
2024-02-26T23:15:07Z | http://arxiv.org/pdf/2403.00823v1 | http://arxiv.org/abs/2403.00823v1 | Christopher Archibald, Spencer Brosnahan | 2024-02-26T23:15:07Z | cs.AI, cs.CL | The game of Codenames has recently emerged as a domain of interest for
intelligent agent design. The game is unique due to the way that language and
coordination between teammates play important roles. Previous approaches to
designing agents for this game have utilized a single internal language model
to determine action choices. This often leads to good performance with some
teammates and inferior performance with other teammates, as the agent cannot
adapt to any specific teammate. In this paper we present the first adaptive
agent for playing Codenames. We adopt an ensemble approach with the goal of
determining, during the course of interacting with a specific teammate, which
of our internal expert agents, each potentially with its own language model, is
the best match. One difficulty faced in this approach is the lack of a single
numerical metric that accurately captures the performance of a Codenames team.
Prior Codenames research has utilized a handful of different metrics to
evaluate agent teams. We propose a novel single metric to evaluate the
performance of a Codenames team, whether playing a single team (solitaire)
game, or a competitive game against another team. We then present and analyze
an ensemble agent which selects an internal expert on each turn in order to
maximize this proposed metric. Experimental analysis shows that this ensemble
approach adapts to individual teammates and often performs nearly as well as
the best internal expert with a teammate. Crucially, this success does not
depend on any previous knowledge about the teammates, the ensemble agents, or
their compatibility. This research represents an important step to making
language-based agents for cooperative language settings like Codenames more
adaptable to individual teammates. | cs.AI | Adapting to Teammates in a Cooperative Language Game |
2024-02-26T20:56:06Z | http://arxiv.org/pdf/2402.17019v4 | http://arxiv.org/abs/2402.17019v4 | Hang Jiang, Xiajie Zhang, Robert Mahari, Daniel Kessler, Eric Ma, Tal August, Irene Li, Alex 'Sandy' Pentland, Yoon Kim, Deb Roy, Jad Kabbara | 2024-07-02T17:50:42Z | cs.CL, cs.HC | Making legal knowledge accessible to non-experts is crucial for enhancing
general legal literacy and encouraging civic participation in democracy.
However, legal documents are often challenging to understand for people without
legal backgrounds. In this paper, we present a novel application of large
language models (LLMs) in legal education to help non-experts learn intricate
legal concepts through storytelling, an effective pedagogical tool in conveying
complex and abstract concepts. We also introduce a new dataset LegalStories,
which consists of 294 complex legal doctrines, each accompanied by a story and
a set of multiple-choice questions generated by LLMs. To construct the dataset,
we experiment with various LLMs to generate legal stories explaining these
concepts. Furthermore, we use an expert-in-the-loop approach to iteratively
design multiple-choice questions. Then, we evaluate the effectiveness of
storytelling with LLMs through randomized controlled trials (RCTs) with legal
novices on 10 samples from the dataset. We find that LLM-generated stories
enhance comprehension of legal concepts and interest in law among non-native
speakers compared to only definitions. Moreover, stories consistently help
participants relate legal concepts to their lives. Finally, we find that
learning with stories shows a higher retention rate for non-native speakers in
the follow-up assessment. Our work has strong implications for using LLMs in
promoting teaching and learning in the legal field and beyond. | cs.CL | Leveraging Large Language Models for Learning Complex Legal Concepts through Storytelling |
2024-02-26T20:53:12Z | http://arxiv.org/pdf/2402.17016v1 | http://arxiv.org/abs/2402.17016v1 | Isabelle Mohr, Markus Krimmel, Saba Sturua, Mohammad Kalim Akram, Andreas Koukounas, Michael Günther, Georgios Mastrapas, Vinit Ravishankar, Joan Fontanals Martínez, Feng Wang, Qi Liu, Ziniu Yu, Jie Fu, Saahil Ognawala, Susana Guzman, Bo Wang, Maximilian Werk, Nan Wang, Han Xiao | 2024-02-26T20:53:12Z | cs.CL, cs.AI, cs.IR, 68T50, I.2.7 | We introduce a novel suite of state-of-the-art bilingual text embedding
models that are designed to support English and another target language. These
models are capable of processing lengthy text inputs with up to 8192 tokens,
making them highly versatile for a range of natural language processing tasks
such as text retrieval, clustering, and semantic textual similarity (STS)
calculations.
By focusing on bilingual models and introducing a unique multi-task learning
objective, we have significantly improved the model performance on STS tasks,
which outperforms the capabilities of existing multilingual models in both
target language understanding and cross-lingual evaluation tasks. Moreover, our
bilingual models are more efficient, requiring fewer parameters and less memory
due to their smaller vocabulary needs. Furthermore, we have expanded the
Massive Text Embedding Benchmark (MTEB) to include benchmarks for German and
Spanish embedding models. This integration aims to stimulate further research
and advancement in text embedding technologies for these languages. | cs.CL | Multi-Task Contrastive Learning for 8192-Token Bilingual Text Embeddings |
2024-02-26T20:43:48Z | http://arxiv.org/pdf/2402.17014v2 | http://arxiv.org/abs/2402.17014v2 | Nikhil Narayan, Mrutyunjay Biswal | 2024-04-14T13:43:39Z | cs.CL | In the digital realm, rich data serves as a crucial source of insights into
the complexities of social, political, and economic landscapes. Addressing the
growing need for high-quality information on events and the imperative to
combat hate speech, this research led to the establishment of the Shared Task
on Climate Activism Stance and Hate Event Detection at CASE 2024. Focused on
climate activists contending with hate speech on social media, our study
contributes to hate speech identification from tweets. Analyzing three
sub-tasks - Hate Speech Detection (Sub-task A), Targets of Hate Speech
Identification (Sub-task B), and Stance Detection (Sub-task C) - Team Z-AGI
Labs evaluated various models, including LSTM, Xgboost, and LGBM based on
Tf-Idf. Results unveiled intriguing variations, with Catboost excelling in
Subtask-B (F1: 0.5604) and Subtask-C (F1: 0.7081), while LGBM emerged as the
top-performing model for Subtask-A (F1: 0.8684). This research provides
valuable insights into the suitability of classical machine learning models for
climate hate speech and stance detection, aiding informed model selection for
robust mechanisms. | cs.CL | Z-AGI Labs at ClimateActivism 2024: Stance and Hate Event Detection on Social Media |
2024-02-26T20:42:40Z | http://arxiv.org/pdf/2402.17013v1 | http://arxiv.org/abs/2402.17013v1 | Santosh T. Y. S. S, Nina Baumgartner, Matthias Stürmer, Matthias Grabmair, Joel Niklaus | 2024-02-26T20:42:40Z | cs.CL, cs.AI, cs.LG, 68T50, I.2 | The assessment of explainability in Legal Judgement Prediction (LJP) systems
is of paramount importance in building trustworthy and transparent systems,
particularly considering the reliance of these systems on factors that may lack
legal relevance or involve sensitive attributes. This study delves into the
realm of explainability and fairness in LJP models, utilizing Swiss Judgement
Prediction (SJP), the only available multilingual LJP dataset. We curate a
comprehensive collection of rationales that `support' and `oppose' judgement
from legal experts for 108 cases in German, French, and Italian. By employing
an occlusion-based explainability approach, we evaluate the explainability
performance of state-of-the-art monolingual and multilingual BERT-based LJP
models, as well as models developed with techniques such as data augmentation
and cross-lingual transfer, which demonstrated prediction performance
improvement. Notably, our findings reveal that improved prediction performance
does not necessarily correspond to enhanced explainability performance,
underscoring the significance of evaluating models from an explainability
perspective. Additionally, we introduce a novel evaluation framework, Lower
Court Insertion (LCI), which allows us to quantify the influence of lower court
information on model predictions, exposing current models' biases. | cs.CL | Towards Explainability and Fairness in Swiss Judgement Prediction: Benchmarking on a Multilingual Dataset |
2024-02-26T20:35:34Z | http://arxiv.org/pdf/2402.17011v1 | http://arxiv.org/abs/2402.17011v1 | Silin Gao, Mete Ismayilzada, Mengjie Zhao, Hiromi Wakaki, Yuki Mitsufuji, Antoine Bosselut | 2024-02-26T20:35:34Z | cs.CL | Inferring contextually-relevant and diverse commonsense to understand
narratives remains challenging for knowledge models. In this work, we develop a
series of knowledge models, DiffuCOMET, that leverage diffusion to learn to
reconstruct the implicit semantic connections between narrative contexts and
relevant commonsense knowledge. Across multiple diffusion steps, our method
progressively refines a representation of commonsense facts that is anchored to
a narrative, producing contextually-relevant and diverse commonsense inferences
for an input context. To evaluate DiffuCOMET, we introduce new metrics for
commonsense inference that more closely measure knowledge diversity and
contextual relevance. Our results on two different benchmarks, ComFact and
WebNLG+, show that knowledge generated by DiffuCOMET achieves a better
trade-off between commonsense diversity, contextual relevance and alignment to
known gold references, compared to baseline knowledge models. | cs.CL | DiffuCOMET: Contextual Commonsense Knowledge Diffusion |
2024-02-26T20:35:32Z | http://arxiv.org/pdf/2402.17010v1 | http://arxiv.org/abs/2402.17010v1 | Ye Wang, Xinrun Xu, Rui Xie, Wenxin Hu, Wei Ye | 2024-02-26T20:35:32Z | cs.CL, cs.AI | When completing knowledge-intensive tasks, humans sometimes need not just an
answer but also a corresponding reference passage for auxiliary reading.
Previous methods required obtaining pre-segmented article chunks through
additional retrieval models. This paper explores leveraging the parameterized
knowledge stored during the pre-training phase of large language models (LLMs)
to independently recall reference passage from any starting position. We
propose a two-stage framework that simulates the scenario of humans recalling
easily forgotten references. Initially, the LLM is prompted to recall document
title identifiers to obtain a coarse-grained document set. Then, based on the
acquired coarse-grained document set, it recalls fine-grained passage. In the
two-stage recall process, we use constrained decoding to ensure that content
outside of the stored documents is not generated. To increase speed, we only
recall a short prefix in the second stage, then locate its position to retrieve
a complete passage. Experiments on KILT knowledge-sensitive tasks have verified
that LLMs can independently recall reference passage location in various task
forms, and the obtained reference significantly assist downstream tasks. | cs.CL | Can Large Language Models Recall Reference Location Like Humans? |
2024-02-26T20:33:50Z | http://arxiv.org/pdf/2402.17008v1 | http://arxiv.org/abs/2402.17008v1 | John Salvador, Naman Bansal, Mousumi Akter, Souvika Sarkar, Anupam Das, Shubhra Kanti Karmaker | 2024-02-26T20:33:50Z | cs.CL | Semantic Overlap Summarization (SOS) is a constrained multi-document
summarization task, where the constraint is to capture the common/overlapping
information between two alternative narratives. While recent advancements in
Large Language Models (LLMs) have achieved superior performance in numerous
summarization tasks, a benchmarking study of the SOS task using LLMs is yet to
be performed. As LLMs' responses are sensitive to slight variations in prompt
design, a major challenge in conducting such a benchmarking study is to
systematically explore a variety of prompts before drawing a reliable
conclusion. Fortunately, very recently, the TELeR taxonomy has been proposed
which can be used to design and explore various prompts for LLMs. Using this
TELeR taxonomy and 15 popular LLMs, this paper comprehensively evaluates LLMs
on the SOS Task, assessing their ability to summarize overlapping information
from multiple alternative narratives. For evaluation, we report
well-established metrics like ROUGE, BERTscore, and SEM-F1$ on two different
datasets of alternative narratives. We conclude the paper by analyzing the
strengths and limitations of various LLMs in terms of their capabilities in
capturing overlapping information The code and datasets used to conduct this
study are available at https://anonymous.4open.science/r/llm_eval-E16D. | cs.CL | Benchmarking LLMs on the Semantic Overlap Summarization Task |
2024-02-26T20:13:58Z | http://arxiv.org/pdf/2402.16998v1 | http://arxiv.org/abs/2402.16998v1 | Jerry Ngo, Yoon Kim | 2024-02-26T20:13:58Z | cs.CL, cs.AI, cs.LG, cs.SD, eess.AS | This work explores whether language models encode meaningfully grounded
representations of sounds of objects. We learn a linear probe that retrieves
the correct text representation of an object given a snippet of audio related
to that object, where the sound representation is given by a pretrained audio
model. This probe is trained via a contrastive loss that pushes the language
representations and sound representations of an object to be close to one
another. After training, the probe is tested on its ability to generalize to
objects that were not seen during training. Across different language models
and audio models, we find that the probe generalization is above chance in many
cases, indicating that despite being trained only on raw text, language models
encode grounded knowledge of sounds for some objects. | cs.CL | What Do Language Models Hear? Probing for Auditory Representations in Language Models |
2024-02-26T19:35:45Z | http://arxiv.org/pdf/2402.16986v1 | http://arxiv.org/abs/2402.16986v1 | Ankan Mullick, Ayan Kumar Bhowmick, Raghav R, Ravi Kokku, Prasenjit Dey, Pawan Goyal, Niloy Ganguly | 2024-02-26T19:35:45Z | cs.CL, cs.IR | Dialog summarization has become increasingly important in managing and
comprehending large-scale conversations across various domains. This task
presents unique challenges in capturing the key points, context, and nuances of
multi-turn long conversations for summarization. It is worth noting that the
summarization techniques may vary based on specific requirements such as in a
shopping-chatbot scenario, the dialog summary helps to learn user preferences,
whereas in the case of a customer call center, the summary may involve the
problem attributes that a user specified, and the final resolution provided.
This work emphasizes the significance of creating coherent and contextually
rich summaries for effective communication in various applications. We explore
current state-of-the-art approaches for long dialog summarization in different
domains and benchmark metrics based evaluations show that one single model does
not perform well across various areas for distinct summarization tasks. | cs.CL | Long Dialog Summarization: An Analysis |
2024-02-26T19:19:47Z | http://arxiv.org/pdf/2402.16977v2 | http://arxiv.org/abs/2402.16977v2 | Smita Ghaisas, Anmol Singhal | 2024-02-28T11:12:07Z | cs.SE, cs.CL | Across the dynamic business landscape today, enterprises face an
ever-increasing range of challenges. These include the constantly evolving
regulatory environment, the growing demand for personalization within software
applications, and the heightened emphasis on governance. In response to these
multifaceted demands, large enterprises have been adopting automation that
spans from the optimization of core business processes to the enhancement of
customer experiences. Indeed, Artificial Intelligence (AI) has emerged as a
pivotal element of modern software systems. In this context, data plays an
indispensable role. AI-centric software systems based on supervised learning
and operating at an industrial scale require large volumes of training data to
perform effectively. Moreover, the incorporation of generative AI has led to a
growing demand for adequate evaluation benchmarks. Our experience in this field
has revealed that the requirement for large datasets for training and
evaluation introduces a host of intricate challenges. This book chapter
explores the evolving landscape of Software Engineering (SE) in general, and
Requirements Engineering (RE) in particular, in this era marked by AI
integration. We discuss challenges that arise while integrating Natural
Language Processing (NLP) and generative AI into enterprise-critical software
systems. The chapter provides practical insights, solutions, and examples to
equip readers with the knowledge and tools necessary for effectively building
solutions with NLP at their cores. We also reflect on how these text
data-centric tasks sit together with the traditional RE process. We also
highlight new RE tasks that may be necessary for handling the increasingly
important text data-centricity involved in developing software systems. | cs.SE | Dealing with Data for RE: Mitigating Challenges while using NLP and Generative AI |
2024-02-26T19:16:04Z | http://arxiv.org/pdf/2402.16973v1 | http://arxiv.org/abs/2402.16973v1 | Lingjun Zhao, Khanh Nguyen, Hal Daumé III | 2024-02-26T19:16:04Z | cs.AI, cs.CL, cs.HC | This paper addresses the challenge of leveraging imperfect language models to
guide human decision-making in the context of a grounded navigation task. We
show that an imperfect instruction generation model can be complemented with an
effective communication mechanism to become more successful at guiding humans.
The communication mechanism we build comprises models that can detect potential
hallucinations in instructions and suggest practical alternatives, and an
intuitive interface to present that information to users. We show that this
approach reduces the human navigation error by up to 29% with no additional
cognitive burden. This result underscores the potential of integrating diverse
communication channels into AI systems to compensate for their imperfections
and enhance their utility for humans. | cs.AI | Successfully Guiding Humans with Imperfect Instructions by Highlighting Potential Errors and Suggesting Corrections |
2024-02-26T18:59:33Z | http://arxiv.org/pdf/2402.16846v2 | http://arxiv.org/abs/2402.16846v2 | Yichi Zhang, Ziqiao Ma, Xiaofeng Gao, Suhaila Shakiah, Qiaozi Gao, Joyce Chai | 2024-04-16T17:59:53Z | cs.CV, cs.AI, cs.CL | Most multimodal large language models (MLLMs) learn language-to-object
grounding through causal language modeling where grounded objects are captured
by bounding boxes as sequences of location tokens. This paradigm lacks
pixel-level representations that are important for fine-grained visual
understanding and diagnosis. In this work, we introduce GROUNDHOG, an MLLM
developed by grounding Large Language Models to holistic segmentation.
GROUNDHOG incorporates a masked feature extractor and converts extracted
features into visual entity tokens for the MLLM backbone, which then connects
groundable phrases to unified grounding masks by retrieving and merging the
entity masks. To train GROUNDHOG, we carefully curated M3G2, a grounded visual
instruction tuning dataset with Multi-Modal Multi-Grained Grounding, by
harvesting a collection of segmentation-grounded datasets with rich
annotations. Our experimental results show that GROUNDHOG achieves superior
performance on various language grounding tasks without task-specific
fine-tuning, and significantly reduces object hallucination. GROUNDHOG also
demonstrates better grounding towards complex forms of visual input and
provides easy-to-understand diagnosis in failure cases. | cs.CV | GROUNDHOG: Grounding Large Language Models to Holistic Segmentation |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.