publication_date
stringlengths 20
20
| doc_url
stringlengths 33
34
| id
stringlengths 33
34
| authors
stringlengths 5
20.1k
| update_date
stringlengths 20
20
| category_all
stringlengths 5
138
| abstract
stringlengths 123
3.09k
| category_primary
stringclasses 78
values | title
stringlengths 10
211
|
|---|---|---|---|---|---|---|---|---|
2023-10-12T12:36:46Z | http://arxiv.org/pdf/2310.08280v1 | http://arxiv.org/abs/2310.08280v1 | Monnie Parida, Manjira Sinha, Anupam Basu, Pabitra Mitra | 2023-10-12T12:36:46Z | cs.CL | This study aims to conduct an extensive detailed analysis of the Odia Braille
reading comprehension among students with visual disability. Specifically, the
study explores their reading speed and hand or finger movements. The study also
aims to investigate any comprehension difficulties and reading errors they may
encounter. Six students from the 9th and 10th grades, aged between 14 and 16,
participated in the study. We observed participants hand movements to
understand how reading errors were connected to hand movement and identify the
students reading difficulties. We also evaluated the participants Odia Braille
reading skills, including their reading speed (in words per minute), errors,
and comprehension. The average speed of Odia Braille reader is 17.64wpm.
According to the study, there was a noticeable correlation between reading
speed and reading errors. As reading speed decreased, the number of reading
errors tended to increase. Moreover, the study established a link between
reduced Braille reading errors and improved reading comprehension. In contrast,
the study found that better comprehension was associated with increased reading
speed. The researchers concluded with some interesting findings about preferred
Braille reading patterns. These findings have important theoretical,
developmental, and methodological implications for instruction. | cs.CL | Optimizing Odia Braille Literacy: The Influence of Speed on Error Reduction and Enhanced Comprehension |
2023-10-12T12:31:23Z | http://arxiv.org/pdf/2310.08279v3 | http://arxiv.org/abs/2310.08279v3 | Rui Yang, Jiahao Zhu, Jianping Man, Li Fang, Yi Zhou | 2024-06-27T04:55:28Z | cs.CL, cs.AI | The design and development of text-based knowledge graph completion (KGC)
methods leveraging textual entity descriptions are at the forefront of
research. These methods involve advanced optimization techniques such as soft
prompts and contrastive learning to enhance KGC models. The effectiveness of
text-based methods largely hinges on the quality and richness of the training
data. Large language models (LLMs) can utilize straightforward prompts to alter
text data, thereby enabling data augmentation for KGC. Nevertheless, LLMs
typically demand substantial computational resources. To address these issues,
we introduce a framework termed constrained prompts for KGC (CP-KGC). This
CP-KGC framework designs prompts that adapt to different datasets to enhance
semantic richness. Additionally, CP-KGC employs a context constraint strategy
to effectively identify polysemous entities within KGC datasets. Through
extensive experimentation, we have verified the effectiveness of this
framework. Even after quantization, the LLM (Qwen-7B-Chat-int4) still enhances
the performance of text-based KGC methods \footnote{Code and datasets are
available at
\href{https://github.com/sjlmg/CP-KGC}{https://github.com/sjlmg/CP-KGC}}. This
study extends the performance limits of existing models and promotes further
integration of KGC with LLMs. | cs.CL | Enhancing Text-based Knowledge Graph Completion with Zero-Shot Large Language Models: A Focus on Semantic Enhancement |
2023-10-12T12:01:32Z | http://arxiv.org/pdf/2310.08256v1 | http://arxiv.org/abs/2310.08256v1 | Cheongwoong Kang, Jaesik Choi | 2023-10-12T12:01:32Z | cs.CL, cs.AI, cs.LG | Large language models (LLMs) often make factually incorrect responses despite
their success in various applications. In this paper, we hypothesize that
relying heavily on simple co-occurrence statistics of the pre-training corpora
is one of the main factors that cause factual errors. Our results reveal that
LLMs are vulnerable to the co-occurrence bias, defined as preferring frequently
co-occurred words over the correct answer. Consequently, LLMs struggle to
recall facts whose subject and object rarely co-occur in the pre-training
dataset although they are seen during finetuning. We show that co-occurrence
bias remains despite scaling up model sizes or finetuning. Therefore, we
suggest finetuning on a debiased dataset to mitigate the bias by filtering out
biased samples whose subject-object co-occurrence count is high. Although
debiased finetuning allows LLMs to memorize rare facts in the training set, it
is not effective in recalling rare facts unseen during finetuning. Further
research in mitigation will help build reliable language models by preventing
potential errors. The code is available at
\url{https://github.com/CheongWoong/impact_of_cooccurrence}. | cs.CL | Impact of Co-occurrence on Factual Knowledge of Large Language Models |
2023-10-12T11:35:24Z | http://arxiv.org/pdf/2310.08240v1 | http://arxiv.org/abs/2310.08240v1 | Wanyun Cui, Linqiu Zhang, Qianle Wang, Shuyang Cai | 2023-10-12T11:35:24Z | cs.CL | AI-generated text has proliferated across various online platforms, offering
both transformative prospects and posing significant risks related to
misinformation and manipulation. Addressing these challenges, this paper
introduces SAID (Social media AI Detection), a novel benchmark developed to
assess AI-text detection models' capabilities in real social media platforms.
It incorporates real AI-generate text from popular social media platforms like
Zhihu and Quora. Unlike existing benchmarks, SAID deals with content that
reflects the sophisticated strategies employed by real AI users on the Internet
which may evade detection or gain visibility, providing a more realistic and
challenging evaluation landscape. A notable finding of our study, based on the
Zhihu dataset, reveals that annotators can distinguish between AI-generated and
human-generated texts with an average accuracy rate of 96.5%. This finding
necessitates a re-evaluation of human capability in recognizing AI-generated
text in today's widely AI-influenced environment. Furthermore, we present a new
user-oriented AI-text detection challenge focusing on the practicality and
effectiveness of identifying AI-generated text based on user information and
multiple responses. The experimental results demonstrate that conducting
detection tasks on actual social media platforms proves to be more challenging
compared to traditional simulated AI-text detection, resulting in a decreased
accuracy. On the other hand, user-oriented AI-generated text detection
significantly improve the accuracy of detection. | cs.CL | Who Said That? Benchmarking Social Media AI Detection |
2023-10-12T11:25:46Z | http://arxiv.org/pdf/2310.08232v1 | http://arxiv.org/abs/2310.08232v1 | Xin Zhang, Zehan Li, Yanzhao Zhang, Dingkun Long, Pengjun Xie, Meishan Zhang, Min Zhang | 2023-10-12T11:25:46Z | cs.CL | In the large language model (LLM) revolution, embedding is a key component of
various systems. For example, it is used to retrieve knowledge or memories for
LLMs, to build content moderation filters, etc. As such cases span from English
to other natural or programming languages, from retrieval to classification and
beyond, it is desirable to build a unified embedding model rather than
dedicated ones for each scenario. In this work, we make an initial step towards
this goal, demonstrating that multiple languages (both natural and programming)
pre-trained transformer decoders can embed universally when finetuned on
limited English data. We provide a comprehensive practice with thorough
evaluations. On English MTEB, our models achieve competitive performance on
different embedding tasks by minimal training data. On other benchmarks, such
as multilingual classification and code search, our models (without any
supervision) perform comparably to, or even surpass heavily supervised
baselines and/or APIs. These results provide evidence of a promising path
towards building powerful unified embedders that can be applied across tasks
and languages. | cs.CL | Language Models are Universal Embedders |
2023-10-12T11:17:40Z | http://arxiv.org/pdf/2310.08225v1 | http://arxiv.org/abs/2310.08225v1 | Chanho Park, Chengsong Lu, Mingjie Chen, Thomas Hain | 2023-10-12T11:17:40Z | eess.AS, cs.CL, cs.SD | The quality of automatic speech recognition (ASR) is typically measured by
word error rate (WER). WER estimation is a task aiming to predict the WER of an
ASR system, given a speech utterance and a transcription. This task has gained
increasing attention while advanced ASR systems are trained on large amounts of
data. In this case, WER estimation becomes necessary in many scenarios, for
example, selecting training data with unknown transcription quality or
estimating the testing performance of an ASR system without ground truth
transcriptions. Facing large amounts of data, the computation efficiency of a
WER estimator becomes essential in practical applications. However, previous
works usually did not consider it as a priority. In this paper, a Fast WER
estimator (Fe-WER) using self-supervised learning representation (SSLR) is
introduced. The estimator is built upon SSLR aggregated by average pooling. The
results show that Fe-WER outperformed the e-WER3 baseline relatively by 19.69%
and 7.16% on Ted-Lium3 in both evaluation metrics of root mean square error and
Pearson correlation coefficient, respectively. Moreover, the estimation
weighted by duration was 10.43% when the target was 10.88%. Lastly, the
inference speed was about 4x in terms of a real-time factor. | eess.AS | Fast Word Error Rate Estimation Using Self-Supervised Representations For Speech And Text |
2023-10-12T11:17:23Z | http://arxiv.org/pdf/2310.10679v1 | http://arxiv.org/abs/2310.10679v1 | Paweł Niszczota, Mateusz Janczak | 2023-10-12T11:17:23Z | cs.CL, cs.AI, cs.CY, I.2.7; K.4.2; J.4 | We use a large-scale experiment (N=8000) to determine whether GPT-4 can
replicate cross-cultural differences in the Big Five, measured using the
Ten-Item Personality Inventory. We used the US and South Korea as the cultural
pair, given that prior research suggests substantial personality differences
between people from these two countries. We manipulated the target of the
simulation (US vs. Korean), the language of the inventory (English vs. Korean),
and the language model (GPT-4 vs. GPT-3.5). Our results show that GPT-4
replicated the cross-cultural differences for each factor. However, mean
ratings had an upward bias and exhibited lower variation than in the human
samples, as well as lower structural validity. Overall, we provide preliminary
evidence that LLMs can aid cross-cultural psychological research. | cs.CL | Large language models can replicate cross-cultural differences in personality |
2023-10-12T11:11:54Z | http://arxiv.org/pdf/2310.08221v1 | http://arxiv.org/abs/2310.08221v1 | Minseok Choi, Chaeheon Gwak, Seho Kim, Si Hyeong Kim, Jaegul Choo | 2023-10-12T11:11:54Z | cs.CL, cs.AI, cs.LG | Keyphrase generation (KG) aims to generate a set of summarizing words or
phrases given a source document, while keyphrase extraction (KE) aims to
identify them from the text. Because the search space is much smaller in KE, it
is often combined with KG to predict keyphrases that may or may not exist in
the corresponding document. However, current unified approaches adopt sequence
labeling and maximization-based generation that primarily operate at a token
level, falling short in observing and scoring keyphrases as a whole. In this
work, we propose SimCKP, a simple contrastive learning framework that consists
of two stages: 1) An extractor-generator that extracts keyphrases by learning
context-aware phrase-level representations in a contrastive manner while also
generating keyphrases that do not appear in the document; 2) A reranker that
adapts scores for each generated phrase by likewise aligning their
representations with the corresponding document. Experimental results on
multiple benchmark datasets demonstrate the effectiveness of our proposed
approach, which outperforms the state-of-the-art models by a significant
margin. | cs.CL | SimCKP: Simple Contrastive Learning of Keyphrase Representations |
2023-10-12T10:26:26Z | http://arxiv.org/pdf/2310.08187v1 | http://arxiv.org/abs/2310.08187v1 | Mahmud Hasan, Labiba Islam, Jannatul Ferdous Ruma, Tasmiah Tahsin Mayeesha, Rashedur M. Rahman | 2023-10-12T10:26:26Z | cs.CL | The task of Visual Question Generation (VQG) is to generate human-like
questions relevant to the given image. As VQG is an emerging research field,
existing works tend to focus only on resource-rich language such as English due
to the availability of datasets. In this paper, we propose the first Bengali
Visual Question Generation task and develop a novel transformer-based
encoder-decoder architecture that generates questions in Bengali when given an
image. We propose multiple variants of models - (i) image-only: baseline model
of generating questions from images without additional information, (ii)
image-category and image-answer-category: guided VQG where we condition the
model to generate questions based on the answer and the category of expected
question. These models are trained and evaluated on the translated VQAv2.0
dataset. Our quantitative and qualitative results establish the first state of
the art models for VQG task in Bengali and demonstrate that our models are
capable of generating grammatically correct and relevant questions. Our
quantitative results show that our image-cat model achieves a BLUE-1 score of
33.12 and BLEU-3 score of 7.56 which is the highest of the other two variants.
We also perform a human evaluation to assess the quality of the generation
tasks. Human evaluation suggests that image-cat model is capable of generating
goal-driven and attribute-specific questions and also stays relevant to the
corresponding image. | cs.CL | Visual Question Generation in Bengali |
2023-10-12T10:21:37Z | http://arxiv.org/pdf/2310.08185v1 | http://arxiv.org/abs/2310.08185v1 | Wang You, Wenshan Wu, Yaobo Liang, Shaoguang Mao, Chenfei Wu, Maosong Cao, Yuzhe Cai, Yiduo Guo, Yan Xia, Furu Wei, Nan Duan | 2023-10-12T10:21:37Z | cs.CL, cs.AI | Plan-and-Write is a common hierarchical approach in long-form narrative text
generation, which first creates a plan to guide the narrative writing.
Following this approach, several studies rely on simply prompting large
language models for planning, which often yields suboptimal results. In this
paper, we propose a new framework called Evaluation-guided Iterative Plan
Extraction for long-form narrative text generation (EIPE-text), which extracts
plans from the corpus of narratives and utilizes the extracted plans to
construct a better planner. EIPE-text has three stages: plan extraction,
learning, and inference. In the plan extraction stage, it iteratively extracts
and improves plans from the narrative corpus and constructs a plan corpus. We
propose a question answer (QA) based evaluation mechanism to automatically
evaluate the plans and generate detailed plan refinement instructions to guide
the iterative improvement. In the learning stage, we build a better planner by
fine-tuning with the plan corpus or in-context learning with examples in the
plan corpus. Finally, we leverage a hierarchical approach to generate long-form
narratives. We evaluate the effectiveness of EIPE-text in the domains of novels
and storytelling. Both GPT-4-based evaluations and human evaluations
demonstrate that our method can generate more coherent and relevant long-form
narratives. Our code will be released in the future. | cs.CL | EIPE-text: Evaluation-Guided Iterative Plan Extraction for Long-Form Narrative Text Generation |
2023-10-12T09:55:45Z | http://arxiv.org/pdf/2310.08172v2 | http://arxiv.org/abs/2310.08172v2 | Zheyuan Zhang, Jifan Yu, Juanzi Li, Lei Hou | 2023-10-18T11:37:43Z | cs.CL | Large Language Models (LLMs) have not only exhibited exceptional performance
across various tasks, but also demonstrated sparks of intelligence. Recent
studies have focused on assessing their capabilities on human exams and
revealed their impressive competence in different domains. However, cognitive
research on the overall knowledge structure of LLMs is still lacking. In this
paper, based on educational diagnostic assessment method, we conduct an
evaluation using MoocRadar, a meticulously annotated human test dataset based
on Bloom Taxonomy. We aim to reveal the knowledge structures of LLMs and gain
insights of their cognitive capabilities. This research emphasizes the
significance of investigating LLMs' knowledge and understanding the disparate
cognitive patterns of LLMs. By shedding light on models' knowledge, researchers
can advance development and utilization of LLMs in a more informed and
effective manner. | cs.CL | Exploring the Cognitive Knowledge Structure of Large Language Models: An Educational Diagnostic Assessment Approach |
2023-10-12T09:49:10Z | http://arxiv.org/pdf/2310.08170v1 | http://arxiv.org/abs/2310.08170v1 | Liam Cripwell, Joël Legrand, Claire Gardent | 2023-10-12T09:49:10Z | cs.CL | Automatic evaluation for sentence simplification remains a challenging
problem. Most popular evaluation metrics require multiple high-quality
references -- something not readily available for simplification -- which makes
it difficult to test performance on unseen domains. Furthermore, most existing
metrics conflate simplicity with correlated attributes such as fluency or
meaning preservation. We propose a new learned evaluation metric (SLE) which
focuses on simplicity, outperforming almost all existing metrics in terms of
correlation with human judgements. | cs.CL | Simplicity Level Estimate (SLE): A Learned Reference-Less Metric for Sentence Simplification |
2023-10-12T09:41:22Z | http://arxiv.org/pdf/2310.08167v1 | http://arxiv.org/abs/2310.08167v1 | Erkan Gunes, Christoffer Koch Florczak | 2023-10-12T09:41:22Z | cs.CL | Classifying policy documents into policy issue topics has been a long-time
effort in political science and communication disciplines. Efforts to automate
text classification processes for social science research purposes have so far
achieved remarkable results, but there is still a large room for progress. In
this work, we test the prediction performance of an alternative strategy, which
requires human involvement much less than full manual coding. We use the GPT
3.5 and GPT 4 models of the OpenAI, which are pre-trained instruction-tuned
Large Language Models (LLM), to classify congressional bills and congressional
hearings into Comparative Agendas Project's 21 major policy issue topics. We
propose three use-case scenarios and estimate overall accuracies ranging from
%58-83 depending on scenario and GPT model employed. The three scenarios aims
at minimal, moderate, and major human interference, respectively. Overall, our
results point towards the insufficiency of complete reliance on GPT with
minimal human intervention, an increasing accuracy along with the human effort
exerted, and a surprisingly high accuracy achieved in the most humanly
demanding use-case. However, the superior use-case achieved the %83 accuracy on
the %65 of the data in which the two models agreed, suggesting that a similar
approach to ours can be relatively easily implemented and allow for mostly
automated coding of a majority of a given dataset. This could free up resources
allowing manual human coding of the remaining %35 of the data to achieve an
overall higher level of accuracy while reducing costs significantly. | cs.CL | Multiclass Classification of Policy Documents with Large Language Models |
2023-10-12T09:39:17Z | http://arxiv.org/pdf/2310.08166v3 | http://arxiv.org/abs/2310.08166v3 | Junyu Lu, Dixiang Zhang, Xiaojun Wu, Xinyu Gao, Ruyi Gan, Jiaxing Zhang, Yan Song, Pingjian Zhang | 2023-10-31T17:51:51Z | cs.CL | Recent advancements enlarge the capabilities of large language models (LLMs)
in zero-shot image-to-text generation and understanding by integrating
multi-modal inputs. However, such success is typically limited to English
scenarios due to the lack of large-scale and high-quality non-English
multi-modal resources, making it extremely difficult to establish competitive
counterparts in other languages. In this paper, we introduce the Ziya-Visual
series, a set of bilingual large-scale vision-language models (LVLMs) designed
to incorporate visual semantics into LLM for multi-modal dialogue. Composed of
Ziya-Visual-Base and Ziya-Visual-Chat, our models adopt the Querying
Transformer from BLIP-2, further exploring the assistance of optimization
schemes such as instruction tuning, multi-stage training and low-rank
adaptation module for visual-language alignment. In addition, we stimulate the
understanding ability of GPT-4 in multi-modal scenarios, translating our
gathered English image-text datasets into Chinese and generating
instruction-response through the in-context learning method. The experiment
results demonstrate that compared to the existing LVLMs, Ziya-Visual achieves
competitive performance across a wide range of English-only tasks including
zero-shot image-text retrieval, image captioning, and visual question
answering. The evaluation leaderboard accessed by GPT-4 also indicates that our
models possess satisfactory image-text understanding and generation
capabilities in Chinese multi-modal scenario dialogues. Code, demo and models
are available at
~\url{https://huggingface.co/IDEA-CCNL/Ziya-BLIP2-14B-Visual-v1}. | cs.CL | Ziya-Visual: Bilingual Large Vision-Language Model via Multi-Task Instruction Tuning |
2023-10-12T09:18:19Z | http://arxiv.org/pdf/2310.08152v2 | http://arxiv.org/abs/2310.08152v2 | Siyu Ren, Qi Jia, Kenny Q. Zhu | 2023-10-15T09:15:02Z | cs.CL | The quadratic complexity of the attention module makes it gradually become
the bulk of compute in Transformer-based LLMs during generation. Moreover, the
excessive key-value cache that arises when dealing with long inputs also brings
severe issues on memory footprint and inference latency. In this work, we
propose a plug-and-play approach that is able to incrementally compress the
intermediate activation of a specified span of tokens into compact ones,
thereby reducing both memory and computational cost when processing subsequent
context. Experiments on both in-domain language modeling and zero-shot
open-ended document generation demonstrate the advantage of our approach over
sparse attention baselines in terms of fluency, n-gram matching, and semantic
similarity. At last, we comprehensively profile the benefit of context
compression on improving the system throughout. Code is available at
https://github.com/DRSY/KV_Compression. | cs.CL | Context Compression for Auto-regressive Transformers with Sentinel Tokens |
2023-10-12T08:45:21Z | http://arxiv.org/pdf/2310.08132v1 | http://arxiv.org/abs/2310.08132v1 | Nick Rossenbach, Benedikt Hilmes, Ralf Schlüter | 2023-10-12T08:45:21Z | cs.CL, cs.SD, eess.AS | Synthetic data generated by text-to-speech (TTS) systems can be used to
improve automatic speech recognition (ASR) systems in low-resource or domain
mismatch tasks. It has been shown that TTS-generated outputs still do not have
the same qualities as real data. In this work we focus on the temporal
structure of synthetic data and its relation to ASR training. By using a novel
oracle setup we show how much the degradation of synthetic data quality is
influenced by duration modeling in non-autoregressive (NAR) TTS. To get
reference phoneme durations we use two common alignment methods, a hidden
Markov Gaussian-mixture model (HMM-GMM) aligner and a neural connectionist
temporal classification (CTC) aligner. Using a simple algorithm based on random
walks we shift phoneme duration distributions of the TTS system closer to real
durations, resulting in an improvement of an ASR system using synthetic data in
a semi-supervised setting. | cs.CL | On the Relevance of Phoneme Duration Variability of Synthesized Training Data for Automatic Speech Recognition |
2023-10-12T08:38:12Z | http://arxiv.org/pdf/2310.08130v4 | http://arxiv.org/abs/2310.08130v4 | Yuxuan Yao, Han Wu, Qiling Xu, Linqi Song | 2023-11-15T03:59:42Z | cs.CL | General-purpose text decoding approaches are usually adopted for dialogue
response generation. Although the quality of the generated responses can be
improved with dialogue-specific encoding methods, conversational decoding
methods are still under-explored. Inspired by \citet{wu2023learning} that a
good dialogue feature space should follow the rules of locality and isotropy,
we present a fine-grained conversational decoding method, termed
\textit{isotropic and proximal search (IPS)}. Our method is designed to
generate the semantic-concentrated response, while still maintaining
informativeness and discrimination against the context. Experiments show that
our approach outperforms existing decoding strategies in the dialogue field
across both automatic and human evaluation metrics. More in-depth analyses
further confirm the effectiveness of our approach. | cs.CL | Fine-grained Conversational Decoding via Isotropic and Proximal Search |
2023-10-12T08:24:15Z | http://arxiv.org/pdf/2310.08123v1 | http://arxiv.org/abs/2310.08123v1 | Chia-Yu Hung, Zhiqiang Hu, Yujia Hu, Roy Ka-Wei Lee | 2023-10-12T08:24:15Z | cs.CL | Authorship verification (AV) is a fundamental task in natural language
processing (NLP) and computational linguistics, with applications in forensic
analysis, plagiarism detection, and identification of deceptive content.
Existing AV techniques, including traditional stylometric and deep learning
approaches, face limitations in terms of data requirements and lack of
explainability. To address these limitations, this paper proposes PromptAV, a
novel technique that leverages Large-Language Models (LLMs) for AV by providing
step-by-step stylometric explanation prompts. PromptAV outperforms
state-of-the-art baselines, operates effectively with limited training data,
and enhances interpretability through intuitive explanations, showcasing its
potential as an effective and interpretable solution for the AV task. | cs.CL | Who Wrote it and Why? Prompting Large-Language Models for Authorship Verification |
2023-10-12T08:00:25Z | http://arxiv.org/pdf/2310.08104v1 | http://arxiv.org/abs/2310.08104v1 | Matthew Baas, Herman Kamper | 2023-10-12T08:00:25Z | eess.AS, cs.CL, cs.SD | Voice conversion aims to convert source speech into a target voice using
recordings of the target speaker as a reference. Newer models are producing
increasingly realistic output. But what happens when models are fed with
non-standard data, such as speech from a user with a speech impairment? We
investigate how a recent voice conversion model performs on non-standard
downstream voice conversion tasks. We use a simple but robust approach called
k-nearest neighbors voice conversion (kNN-VC). We look at four non-standard
applications: stuttered voice conversion, cross-lingual voice conversion,
musical instrument conversion, and text-to-voice conversion. The latter
involves converting to a target voice specified through a text description,
e.g. "a young man with a high-pitched voice". Compared to an established
baseline, we find that kNN-VC retains high performance in stuttered and
cross-lingual voice conversion. Results are more mixed for the musical
instrument and text-to-voice conversion tasks. E.g., kNN-VC works well on some
instruments like drums but not on others. Nevertheless, this shows that voice
conversion models - and kNN-VC in particular - are increasingly applicable in a
range of non-standard downstream tasks. But there are still limitations when
samples are very far from the training distribution. Code, samples, trained
models: https://rf5.github.io/sacair2023-knnvc-demo/. | eess.AS | Voice Conversion for Stuttered Speech, Instruments, Unseen Languages and Textually Described Voices |
2023-10-12T07:52:19Z | http://arxiv.org/pdf/2310.08102v1 | http://arxiv.org/abs/2310.08102v1 | Muhammad Razif Rizqullah, Ayu Purwarianti, Alham Fikri Aji | 2023-10-12T07:52:19Z | cs.CL | Nowadays, Question Answering (QA) tasks receive significant research focus,
particularly with the development of Large Language Model (LLM) such as Chat
GPT [1]. LLM can be applied to various domains, but it contradicts the
principles of information transmission when applied to the Islamic domain. In
Islam we strictly regulates the sources of information and who can give
interpretations or tafseer for that sources [2]. The approach used by LLM to
generate answers based on its own interpretation is similar to the concept of
tafseer, LLM is neither an Islamic expert nor a human which is not permitted in
Islam. Indonesia is the country with the largest Islamic believer population in
the world [3]. With the high influence of LLM, we need to make evaluation of
LLM in religious domain. Currently, there is only few religious QA dataset
available and none of them using Sirah Nabawiyah especially in Indonesian
Language. In this paper, we propose the Question Answering Sirah Nabawiyah
(QASiNa) dataset, a novel dataset compiled from Sirah Nabawiyah literatures in
Indonesian language. We demonstrate our dataset by using mBERT [4], XLM-R [5],
and IndoBERT [6] which fine-tuned with Indonesian translation of SQuAD v2.0
[7]. XLM-R model returned the best performance on QASiNa with EM of 61.20,
F1-Score of 75.94, and Substring Match of 70.00. We compare XLM-R performance
with Chat GPT-3.5 and GPT-4 [1]. Both Chat GPT version returned lower EM and
F1-Score with higher Substring Match, the gap of EM and Substring Match get
wider in GPT-4. The experiment indicate that Chat GPT tends to give excessive
interpretations as evidenced by its higher Substring Match scores compared to
EM and F1-Score, even after providing instruction and context. This concludes
Chat GPT is unsuitable for question answering task in religious domain
especially for Islamic religion. | cs.CL | QASiNa: Religious Domain Question Answering using Sirah Nabawiyah |
2023-10-12T07:51:43Z | http://arxiv.org/pdf/2310.08101v2 | http://arxiv.org/abs/2310.08101v2 | Junxiao Shen, John J. Dudley, Jingyao Zheng, Bill Byrne, Per Ola Kristensson | 2023-10-15T06:59:36Z | cs.CL, cs.AI | Text entry is an essential task in our day-to-day digital interactions.
Numerous intelligent features have been developed to streamline this process,
making text entry more effective, efficient, and fluid. These improvements
include sentence prediction and user personalization. However, as deep
learning-based language models become the norm for these advanced features, the
necessity for data collection and model fine-tuning increases. These challenges
can be mitigated by harnessing the in-context learning capability of large
language models such as GPT-3.5. This unique feature allows the language model
to acquire new skills through prompts, eliminating the need for data collection
and fine-tuning. Consequently, large language models can learn various text
prediction techniques. We initially showed that, for a sentence prediction
task, merely prompting GPT-3.5 surpassed a GPT-2 backed system and is
comparable with a fine-tuned GPT-3.5 model, with the latter two methods
requiring costly data collection, fine-tuning and post-processing. However, the
task of prompting large language models to specialize in specific text
prediction tasks can be challenging, particularly for designers without
expertise in prompt engineering. To address this, we introduce Promptor, a
conversational prompt generation agent designed to engage proactively with
designers. Promptor can automatically generate complex prompts tailored to meet
specific needs, thus offering a solution to this challenge. We conducted a user
study involving 24 participants creating prompts for three intelligent text
entry tasks, half of the participants used Promptor while the other half
designed prompts themselves. The results show that Promptor-designed prompts
result in a 35% increase in similarity and 22% in coherence over those by
designers. | cs.CL | Promptor: A Conversational and Autonomous Prompt Generation Agent for Intelligent Text Entry Techniques |
2023-10-12T07:48:50Z | http://arxiv.org/pdf/2310.08099v2 | http://arxiv.org/abs/2310.08099v2 | Ajay Krishnan, V. S. Anoop | 2023-10-19T16:07:41Z | cs.CL | Climate change's impact on human health poses unprecedented and diverse
challenges. Unless proactive measures based on solid evidence are implemented,
these threats will likely escalate and continue to endanger human well-being.
The escalating advancements in information and communication technologies have
facilitated the widespread availability and utilization of social media
platforms. Individuals utilize platforms such as Twitter and Facebook to
express their opinions, thoughts, and critiques on diverse subjects,
encompassing the pressing issue of climate change. The proliferation of climate
change-related content on social media necessitates comprehensive analysis to
glean meaningful insights. This paper employs natural language processing (NLP)
techniques to analyze climate change discourse and quantify the sentiment of
climate change-related tweets. We use ClimateBERT, a pretrained model
fine-tuned specifically for the climate change domain. The objective is to
discern the sentiment individuals express and uncover patterns in public
opinion concerning climate change. Analyzing tweet sentiments allows a deeper
comprehension of public perceptions, concerns, and emotions about this critical
global challenge. The findings from this experiment unearth valuable insights
into public sentiment and the entities associated with climate change
discourse. Policymakers, researchers, and organizations can leverage such
analyses to understand public perceptions, identify influential actors, and
devise informed strategies to address climate change challenges. | cs.CL | ClimateNLP: Analyzing Public Sentiment Towards Climate Change Using Natural Language Processing |
2023-10-12T07:17:17Z | http://arxiv.org/pdf/2310.08085v1 | http://arxiv.org/abs/2310.08085v1 | Ni Putu Intan Maharani, Ayu Purwarianti, Alham Fikri Aji | 2023-10-12T07:17:17Z | cs.CL, cs.AI | Clickbait spoiling aims to generate a short text to satisfy the curiosity
induced by a clickbait post. As it is a newly introduced task, the dataset is
only available in English so far. Our contributions include the construction of
manually labeled clickbait spoiling corpus in Indonesian and an evaluation on
using cross-lingual zero-shot question answering-based models to tackle
clikcbait spoiling for low-resource language like Indonesian. We utilize
selection of multilingual language models. The experimental results suggest
that XLM-RoBERTa (large) model outperforms other models for phrase and passage
spoilers, meanwhile, mDeBERTa (base) model outperforms other models for
multipart spoilers. | cs.CL | Low-Resource Clickbait Spoiling for Indonesian via Question Answering |
2023-10-12T06:59:10Z | http://arxiv.org/pdf/2310.08078v1 | http://arxiv.org/abs/2310.08078v1 | Md Mushfiqur Rahman, Fardin Ahsan Sakib, Fahim Faisal, Antonios Anastasopoulos | 2023-10-12T06:59:10Z | cs.CL, cs.LG | Choosing an appropriate tokenization scheme is often a bottleneck in
low-resource cross-lingual transfer. To understand the downstream implications
of text representation choices, we perform a comparative analysis on language
models having diverse text representation modalities including 2
segmentation-based models (\texttt{BERT}, \texttt{mBERT}), 1 image-based model
(\texttt{PIXEL}), and 1 character-level model (\texttt{CANINE}). First, we
propose a scoring Language Quotient (LQ) metric capable of providing a weighted
representation of both zero-shot and few-shot evaluation combined. Utilizing
this metric, we perform experiments comprising 19 source languages and 133
target languages on three tasks (POS tagging, Dependency parsing, and NER). Our
analysis reveals that image-based models excel in cross-lingual transfer when
languages are closely related and share visually similar scripts. However, for
tasks biased toward word meaning (POS, NER), segmentation-based models prove to
be superior. Furthermore, in dependency parsing tasks where word relationships
play a crucial role, models with their character-level focus, outperform
others. Finally, we propose a recommendation scheme based on our findings to
guide model selection according to task and language requirements. | cs.CL | To token or not to token: A Comparative Study of Text Representations for Cross-Lingual Transfer |
2023-10-12T06:46:07Z | http://arxiv.org/pdf/2310.08072v2 | http://arxiv.org/abs/2310.08072v2 | Kosuke Takahashi, Takahiro Omi, Kosuke Arima, Tatsuya Ishigaki | 2023-10-13T00:40:29Z | cs.CL, 68T50 | This paper presents a simple and cost-effective method for synthesizing data
to train question-answering systems. For training, fine-tuning GPT models is a
common practice in resource-rich languages like English, however, it becomes
challenging for non-English languages due to the scarcity of sufficient
question-answer (QA) pairs. Existing approaches use question and answer
generators trained on human-authored QA pairs, which involves substantial human
expenses. In contrast, we use an instruct-tuned model to generate QA pairs in a
zero-shot or few-shot manner. We conduct experiments to compare various
strategies for obtaining QA pairs from the instruct-tuned model. The results
demonstrate that a model trained on our proposed synthetic data achieves
comparable performance to a model trained on manually curated datasets, without
incurring human costs. | cs.CL | Training Generative Question-Answering on Synthetic Data Obtained from an Instruct-tuned Model |
2023-10-12T06:32:42Z | http://arxiv.org/pdf/2310.08069v1 | http://arxiv.org/abs/2310.08069v1 | Haochen Li, Xin Zhou, Luu Anh Tuan, Chunyan Miao | 2023-10-12T06:32:42Z | cs.SE, cs.CL, cs.IR, cs.LG | Recently, contrastive learning has become a key component in fine-tuning code
search models for software development efficiency and effectiveness. It pulls
together positive code snippets while pushing negative samples away given
search queries. Among contrastive learning, InfoNCE is the most widely used
loss function due to its better performance. However, the following problems in
negative samples of InfoNCE may deteriorate its representation learning: 1) The
existence of false negative samples in large code corpora due to duplications.
2). The failure to explicitly differentiate between the potential relevance of
negative samples. As an example, a bubble sorting algorithm example is less
``negative'' than a file saving function for the quick sorting algorithm query.
In this paper, we tackle the above problems by proposing a simple yet effective
Soft-InfoNCE loss that inserts weight terms into InfoNCE. In our proposed loss
function, we apply three methods to estimate the weights of negative pairs and
show that the vanilla InfoNCE loss is a special case of Soft-InfoNCE.
Theoretically, we analyze the effects of Soft-InfoNCE on controlling the
distribution of learnt code representations and on deducing a more precise
mutual information estimation. We furthermore discuss the superiority of
proposed loss functions with other design alternatives. Extensive experiments
demonstrate the effectiveness of Soft-InfoNCE and weights estimation methods
under state-of-the-art code search models on a large-scale public dataset
consisting of six programming languages. Source code is available at
\url{https://github.com/Alex-HaochenLi/Soft-InfoNCE}. | cs.SE | Rethinking Negative Pairs in Code Search |
2023-10-12T05:25:49Z | http://arxiv.org/pdf/2310.08041v3 | http://arxiv.org/abs/2310.08041v3 | Jing Liu, Ruihao Gong, Xiuying Wei, Zhiwei Dong, Jianfei Cai, Bohan Zhuang | 2024-04-06T10:22:57Z | cs.CL, cs.AI, cs.LG | Large Language Models (LLMs) excel in NLP, but their demands hinder their
widespread deployment. While Quantization-Aware Training (QAT) offers a
solution, its extensive training costs make Post-Training Quantization (PTQ) a
more practical approach for LLMs. In existing studies, activation outliers in
particular channels are identified as the bottleneck to PTQ accuracy. They
propose to transform the magnitudes from activations to weights, which however
offers limited alleviation or suffers from unstable gradients, resulting in a
severe performance drop at low-bitwidth. In this paper, we propose QLLM, an
accurate and efficient low-bitwidth PTQ method designed for LLMs. QLLM
introduces an adaptive channel reassembly technique that reallocates the
magnitude of outliers to other channels, thereby mitigating their impact on the
quantization range. This is achieved by channel disassembly and channel
assembly, which first breaks down the outlier channels into several
sub-channels to ensure a more balanced distribution of activation magnitudes.
Then similar channels are merged to maintain the original channel number for
efficiency. Additionally, an adaptive strategy is designed to autonomously
determine the optimal number of sub-channels for channel disassembly. To
further compensate for the performance loss caused by quantization, we propose
an efficient tuning method that only learns a small number of low-rank weights
while freezing the pre-trained quantized model. After training, these low-rank
parameters can be fused into the frozen weights without affecting inference.
Extensive experiments on LLaMA-1 and LLaMA-2 show that QLLM can obtain accurate
quantized models efficiently. For example, QLLM quantizes the 4-bit LLaMA-2-70B
within 10 hours on a single A100-80G GPU, outperforming the previous
state-of-the-art method by 7.89% on the average accuracy across five zero-shot
tasks. | cs.CL | QLLM: Accurate and Efficient Low-Bitwidth Quantization for Large Language Models |
2023-10-12T04:14:28Z | http://arxiv.org/pdf/2310.08027v1 | http://arxiv.org/abs/2310.08027v1 | Yi Dai, Hao Lang, Kaisheng Zeng, Fei Huang, Yongbin Li | 2023-10-12T04:14:28Z | cs.CL, cs.CV | Out-of-distribution (OOD) detection is essential for reliable and trustworthy
machine learning. Recent multi-modal OOD detection leverages textual
information from in-distribution (ID) class names for visual OOD detection, yet
it currently neglects the rich contextual information of ID classes. Large
language models (LLMs) encode a wealth of world knowledge and can be prompted
to generate descriptive features for each class. Indiscriminately using such
knowledge causes catastrophic damage to OOD detection due to LLMs'
hallucinations, as is observed by our analysis. In this paper, we propose to
apply world knowledge to enhance OOD detection performance through selective
generation from LLMs. Specifically, we introduce a consistency-based
uncertainty calibration method to estimate the confidence score of each
generation. We further extract visual objects from each image to fully
capitalize on the aforementioned world knowledge. Extensive experiments
demonstrate that our method consistently outperforms the state-of-the-art. | cs.CL | Exploring Large Language Models for Multi-Modal Out-of-Distribution Detection |
2023-10-12T03:33:06Z | http://arxiv.org/pdf/2310.08017v1 | http://arxiv.org/abs/2310.08017v1 | Siyuan Brandon Loh, Aravind Sesagiri Raamkumar | 2023-10-12T03:33:06Z | cs.CL, I.2 | Large Language Models (LLMs) have demonstrated remarkable performance across
various information-seeking and reasoning tasks. These computational systems
drive state-of-the-art dialogue systems, such as ChatGPT and Bard. They also
carry substantial promise in meeting the growing demands of mental health care,
albeit relatively unexplored. As such, this study sought to examine LLMs'
capability to generate empathetic responses in conversations that emulate those
in a mental health counselling setting. We selected five LLMs: version 3.5 and
version 4 of the Generative Pre-training (GPT), Vicuna FastChat-T5, Pathways
Language Model (PaLM) version 2, and Falcon-7B-Instruct. Based on a simple
instructional prompt, these models responded to utterances derived from the
EmpatheticDialogues (ED) dataset. Using three empathy-related metrics, we
compared their responses to those from traditional response generation dialogue
systems, which were fine-tuned on the ED dataset, along with human-generated
responses. Notably, we discovered that responses from the LLMs were remarkably
more empathetic in most scenarios. We position our findings in light of
catapulting advancements in creating empathetic conversational systems. | cs.CL | Harnessing Large Language Models' Empathetic Response Generation Capabilities for Online Mental Health Counselling Support |
2023-10-12T01:17:56Z | http://arxiv.org/pdf/2310.07968v4 | http://arxiv.org/abs/2310.07968v4 | Yinpei Dai, Run Peng, Sikai Li, Joyce Chai | 2024-05-29T21:06:18Z | cs.RO, cs.CL, cs.HC | Zero-Shot Object Navigation (ZSON) enables agents to navigate towards
open-vocabulary objects in unknown environments. The existing works of ZSON
mainly focus on following individual instructions to find generic object
classes, neglecting the utilization of natural language interaction and the
complexities of identifying user-specific objects. To address these
limitations, we introduce Zero-shot Interactive Personalized Object Navigation
(ZIPON), where robots need to navigate to personalized goal objects while
engaging in conversations with users. To solve ZIPON, we propose a new
framework termed Open-woRld Interactive persOnalized Navigation (ORION), which
uses Large Language Models (LLMs) to make sequential decisions to manipulate
different modules for perception, navigation and communication. Experimental
results show that the performance of interactive agents that can leverage user
feedback exhibits significant improvement. However, obtaining a good balance
between task completion and the efficiency of navigation and interaction
remains challenging for all methods. We further provide more findings on the
impact of diverse user feedback forms on the agents' performance. Code is
available at https://github.com/sled-group/navchat. | cs.RO | Think, Act, and Ask: Open-World Interactive Personalized Robot Navigation |
2023-10-12T00:57:32Z | http://arxiv.org/pdf/2310.07962v1 | http://arxiv.org/abs/2310.07962v1 | Prathamesh Pawar | 2023-10-12T00:57:32Z | cs.CL | One of the prominent problems with processing and operating on text data is
the non uniformity of it. Due to the change in the dialects and languages, the
caliber of translation is low. This creates a unique problem while using NLP in
text data; which is the spell variation arising from the inconsistent
translations and transliterations. This problem can also be further aggravated
by the human error arising from the various ways to write a Proper Noun from an
Indian language into its English equivalent. Translating proper nouns
originating from Indian languages can be complicated as some proper nouns are
also used as common nouns which might be taken literally. Applications of NLP
that require addresses, names and other proper nouns face this problem
frequently. We propose a method to cluster these spell variations for proper
nouns using ML techniques and mathematical similarity equations. We aimed to
use Affinity Propagation to determine relative similarity between the tokens.
The results are augmented by filtering the token-variation pair by a similarity
threshold. We were able to reduce the spell variations by a considerable
amount. This application can significantly reduce the amount of human
annotation efforts needed for data cleansing and formatting. | cs.CL | Clustering of Spell Variations for Proper Nouns Transliterated from the other languages |
2023-10-12T00:50:24Z | http://arxiv.org/pdf/2310.07957v2 | http://arxiv.org/abs/2310.07957v2 | Nilay Patel, Rahul Saha, Jeffrey Flanigan | 2023-10-19T18:08:20Z | cs.CL, cs.AI | Verifying mathematical proofs is difficult, but can be automated with the
assistance of a computer. Autoformalization is the task of automatically
translating natural language mathematics into a formal language that can be
verified by a program. This is a challenging task, and especially for
higher-level mathematics found in research papers. Research paper mathematics
requires large amounts of background and context. In this paper, we propose an
avenue towards tackling autoformalization for research-level mathematics, by
breaking the task into easier and more approachable subtasks: unlinked
formalization (formalization with unlinked definitions and theorems), entity
linking (linking to the proper theorems and definitions), and finally adjusting
types so it passes the type checker. In addition, we present arXiv2Formal, a
benchmark dataset for unlinked formalization consisting of 50 theorems
formalized for the Lean theorem prover sampled from papers on arXiv.org. We
welcome any contributions from the community to future versions of this
dataset. | cs.CL | A New Approach Towards Autoformalization |
2023-10-11T23:01:29Z | http://arxiv.org/pdf/2310.07931v1 | http://arxiv.org/abs/2310.07931v1 | Adyasha Maharana, Prateek Yadav, Mohit Bansal | 2023-10-11T23:01:29Z | cs.LG, cs.AI, cs.CL, cs.CV | Analytical theories suggest that higher-quality data can lead to lower test
errors in models trained on a fixed data budget. Moreover, a model can be
trained on a lower compute budget without compromising performance if a dataset
can be stripped of its redundancies. Coreset selection (or data pruning) seeks
to select a subset of the training data so as to maximize the performance of
models trained on this subset, also referred to as coreset. There are two
dominant approaches: (1) geometry-based data selection for maximizing data
diversity in the coreset, and (2) functions that assign difficulty scores to
samples based on training dynamics. Optimizing for data diversity leads to a
coreset that is biased towards easier samples, whereas, selection by difficulty
ranking omits easy samples that are necessary for the training of deep learning
models. This demonstrates that data diversity and importance scores are two
complementary factors that need to be jointly considered during coreset
selection. We represent a dataset as an undirected graph and propose a novel
pruning algorithm, D2 Pruning, that uses forward and reverse message passing
over this dataset graph for coreset selection. D2 Pruning updates the
difficulty scores of each example by incorporating the difficulty of its
neighboring examples in the dataset graph. Then, these updated difficulty
scores direct a graph-based sampling method to select a coreset that
encapsulates both diverse and difficult regions of the dataset space. We
evaluate supervised and self-supervised versions of our method on various
vision and language datasets. Results show that D2 Pruning improves coreset
selection over previous state-of-the-art methods for up to 70% pruning rates.
Additionally, we find that using D2 Pruning for filtering large multimodal
datasets leads to increased diversity in the dataset and improved
generalization of pretrained models. | cs.LG | D2 Pruning: Message Passing for Balancing Diversity and Difficulty in Data Pruning |
2023-10-11T22:57:03Z | http://arxiv.org/pdf/2310.07929v1 | http://arxiv.org/abs/2310.07929v1 | Catherine Arnett, Tyler A. Chang, James A. Michaelov, Benjamin K. Bergen | 2023-10-11T22:57:03Z | cs.CL | Do multilingual language models share abstract grammatical representations
across languages, and if so, when do these develop? Following Sinclair et al.
(2022), we use structural priming to test for abstract grammatical
representations with causal effects on model outputs. We extend the approach to
a Dutch-English bilingual setting, and we evaluate a Dutch-English language
model during pre-training. We find that crosslingual structural priming effects
emerge early after exposure to the second language, with less than 1M tokens of
data in that language. We discuss implications for data contamination,
low-resource transfer, and how abstract grammatical representations emerge in
multilingual models. | cs.CL | Crosslingual Structural Priming and the Pre-Training Dynamics of Bilingual Language Models |
2023-10-11T22:35:18Z | http://arxiv.org/pdf/2310.07923v5 | http://arxiv.org/abs/2310.07923v5 | William Merrill, Ashish Sabharwal | 2024-04-11T18:03:53Z | cs.LG, cs.CC, cs.CL, cs.LO | Recent theoretical work has identified surprisingly simple reasoning
problems, such as checking if two nodes in a graph are connected or simulating
finite-state machines, that are provably unsolvable by standard transformers
that answer immediately after reading their input. However, in practice,
transformers' reasoning can be improved by allowing them to use a "chain of
thought" or "scratchpad", i.e., generate and condition on a sequence of
intermediate tokens before answering. Motivated by this, we ask: Does such
intermediate generation fundamentally extend the computational power of a
decoder-only transformer? We show that the answer is yes, but the amount of
increase depends crucially on the amount of intermediate generation. For
instance, we find that transformer decoders with a logarithmic number of
decoding steps (w.r.t. the input length) push the limits of standard
transformers only slightly, while a linear number of decoding steps, assuming
projected pre-norm (a slight generalization of standard pre-norm), adds a clear
new ability (under standard complexity conjectures): recognizing all regular
languages. Our results also imply that linear steps keep transformer decoders
within context-sensitive languages, and polynomial steps with generalized
pre-norm make them recognize exactly the class of polynomial-time solvable
problems -- the first exact characterization of a type of transformers in terms
of standard complexity classes. Together, this provides a nuanced framework for
understanding how the length of a transformer's chain of thought or scratchpad
impacts its reasoning power. | cs.LG | The Expressive Power of Transformers with Chain of Thought |
2023-10-11T21:38:40Z | http://arxiv.org/pdf/2310.07911v1 | http://arxiv.org/abs/2310.07911v1 | Huiyin Xue, Nikolaos Aletras | 2023-10-11T21:38:40Z | cs.CL | Scaling pre-trained language models has resulted in large performance gains
in various natural language processing tasks but comes with a large cost in
memory requirements. Inspired by the position embeddings in transformers, we
aim to simplify and reduce the memory footprint of the multi-head attention
(MHA) mechanism. We propose an alternative module that uses only a single
shared projection matrix and multiple head embeddings (MHE), i.e. one per head.
We empirically demonstrate that our MHE attention is substantially more memory
efficient compared to alternative attention mechanisms while achieving high
predictive performance retention ratio to vanilla MHA on several downstream
tasks. MHE attention only requires a negligible fraction of additional
parameters ($3nd$, where $n$ is the number of attention heads and $d$ the size
of the head embeddings) compared to a single-head attention, while MHA requires
$(3n^2-3n)d^2-3nd$ additional parameters. | cs.CL | Pit One Against Many: Leveraging Attention-head Embeddings for Parameter-efficient Multi-head Attention |
2023-10-11T20:52:30Z | http://arxiv.org/pdf/2310.07889v2 | http://arxiv.org/abs/2310.07889v2 | Bowen Pan, Rameswar Panda, SouYoung Jin, Rogerio Feris, Aude Oliva, Phillip Isola, Yoon Kim | 2024-03-30T22:00:22Z | cs.CV, cs.AI, cs.CL, cs.RO | We explore the use of language as a perceptual representation for
vision-and-language navigation (VLN), with a focus on low-data settings. Our
approach uses off-the-shelf vision systems for image captioning and object
detection to convert an agent's egocentric panoramic view at each time step
into natural language descriptions. We then finetune a pretrained language
model to select an action, based on the current view and the trajectory
history, that would best fulfill the navigation instructions. In contrast to
the standard setup which adapts a pretrained language model to work directly
with continuous visual features from pretrained vision models, our approach
instead uses (discrete) language as the perceptual representation. We explore
several use cases of our language-based navigation (LangNav) approach on the
R2R VLN benchmark: generating synthetic trajectories from a prompted language
model (GPT-4) with which to finetune a smaller language model; domain transfer
where we transfer a policy learned on one simulated environment (ALFRED) to
another (more realistic) environment (R2R); and combining both vision- and
language-based representations for VLN. Our approach is found to improve upon
baselines that rely on visual features in settings where only a few expert
trajectories (10-100) are available, demonstrating the potential of language as
a perceptual representation for navigation. | cs.CV | LangNav: Language as a Perceptual Representation for Navigation |
2023-10-11T20:34:42Z | http://arxiv.org/pdf/2310.07875v1 | http://arxiv.org/abs/2310.07875v1 | Gus Eggert, Kevin Huo, Mike Biven, Justin Waugh | 2023-10-11T20:34:42Z | cs.CL, cs.AI, cs.DB, cs.LG | It is well-established that large, diverse datasets play a pivotal role in
the performance of modern AI systems for text and image modalities. However,
there are no datasets for tabular data of comparable size and diversity to
those available for text and images. Thus we present "TabLib'', a compilation
of 627 million tables totaling 69 TiB, along with 867B tokens of context.
TabLib was extracted from numerous file formats, including CSV, HTML, SQLite,
PDF, Excel, and others, sourced from GitHub and Common Crawl. The size and
diversity of TabLib offer considerable promise in the table modality,
reminiscent of the original promise of foundational datasets for text and
images, such as The Pile and LAION. | cs.CL | TabLib: A Dataset of 627M Tables with Context |
2023-10-11T19:58:07Z | http://arxiv.org/pdf/2310.07856v1 | http://arxiv.org/abs/2310.07856v1 | Jiho Shin, Hadi Hemmati, Moshi Wei, Song Wang | 2023-10-11T19:58:07Z | cs.CL, cs.SE | In this work, we revisit existing oracle generation studies plus ChatGPT to
empirically investigate the current standing of their performance in both
NLG-based and test adequacy metrics. Specifically, we train and run four
state-of-the-art test oracle generation models on five NLG-based and two test
adequacy metrics for our analysis. We apply two different correlation analyses
between these two different sets of metrics. Surprisingly, we found no
significant correlation between the NLG-based metrics and test adequacy
metrics. For instance, oracles generated from ChatGPT on the project
activemq-artemis had the highest performance on all the NLG-based metrics among
the studied NOGs, however, it had the most number of projects with a decrease
in test adequacy metrics compared to all the studied NOGs. We further conduct a
qualitative analysis to explore the reasons behind our observations, we found
that oracles with high NLG-based metrics but low test adequacy metrics tend to
have complex or multiple chained method invocations within the oracle's
parameters, making it hard for the model to generate completely, affecting the
test adequacy metrics. On the other hand, oracles with low NLG-based metrics
but high test adequacy metrics tend to have to call different assertion types
or a different method that functions similarly to the ones in the ground truth.
Overall, this work complements prior studies on test oracle generation with an
extensive performance evaluation with both NLG and test adequacy metrics and
provides guidelines for better assessment of deep learning applications in
software test generation in the future. | cs.CL | Assessing Evaluation Metrics for Neural Test Oracle Generation |
2023-10-11T19:51:13Z | http://arxiv.org/pdf/2310.07849v2 | http://arxiv.org/abs/2310.07849v2 | Zhuoyan Li, Hangxiao Zhu, Zhuoran Lu, Ming Yin | 2023-10-13T01:31:59Z | cs.CL, cs.AI | The collection and curation of high-quality training data is crucial for
developing text classification models with superior performance, but it is
often associated with significant costs and time investment. Researchers have
recently explored using large language models (LLMs) to generate synthetic
datasets as an alternative approach. However, the effectiveness of the
LLM-generated synthetic data in supporting model training is inconsistent
across different classification tasks. To better understand factors that
moderate the effectiveness of the LLM-generated synthetic data, in this study,
we look into how the performance of models trained on these synthetic data may
vary with the subjectivity of classification. Our results indicate that
subjectivity, at both the task level and instance level, is negatively
associated with the performance of the model trained on synthetic data. We
conclude by discussing the implications of our work on the potential and
limitations of leveraging LLM for synthetic data generation. | cs.CL | Synthetic Data Generation with Large Language Models for Text Classification: Potential and Limitations |
2023-10-11T19:50:59Z | http://arxiv.org/pdf/2310.07848v1 | http://arxiv.org/abs/2310.07848v1 | Hrishikesh Terdalkar, Arnab Bhattacharya | 2023-10-11T19:50:59Z | cs.CL | Sanskrit (sa\d{m}sk\d{r}ta) enjoys one of the largest and most varied
literature in the whole world. Extracting the knowledge from it, however, is a
challenging task due to multiple reasons including complexity of the language
and paucity of standard natural language processing tools. In this paper, we
target the problem of building knowledge graphs for particular types of
relationships from sa\d{m}sk\d{r}ta texts. We build a natural language
question-answering system in sa\d{m}sk\d{r}ta that uses the knowledge graph to
answer factoid questions. We design a framework for the overall system and
implement two separate instances of the system on human relationships from
mah\=abh\=arata and r\=am\=aya\d{n}a, and one instance on synonymous
relationships from bh\=avaprak\=a\'sa nigha\d{n}\d{t}u, a technical text from
\=ayurveda. We show that about 50% of the factoid questions can be answered
correctly by the system. More importantly, we analyse the shortcomings of the
system in detail for each step, and discuss the possible ways forward. | cs.CL | Framework for Question-Answering in Sanskrit through Automated Construction of Knowledge Graphs |
2023-10-11T19:16:09Z | http://arxiv.org/pdf/2310.07830v1 | http://arxiv.org/abs/2310.07830v1 | Sia Gholami, Marwan Omar | 2023-10-11T19:16:09Z | cs.CL, cs.AI, cs.LG | Natural Language Processing (NLP) has undergone transformative changes with
the advent of deep learning methodologies. One challenge persistently
confronting researchers is the scarcity of high-quality, annotated datasets
that drive these models. This paper explores the nuances of synthetic data
generation in NLP, with a focal point on template-based question generation. By
assessing its advantages, including data augmentation potential and the
introduction of structured variety, we juxtapose these benefits against
inherent limitations, such as the risk of overfitting and the constraints posed
by pre-defined templates. Drawing from empirical evaluations, we demonstrate
the impact of template-based synthetic data on the performance of modern
transformer models. We conclude by emphasizing the delicate balance required
between synthetic and real-world data, and the future trajectories of
integrating synthetic data in model training pipelines. The findings aim to
guide NLP practitioners in harnessing synthetic data's potential, ensuring
optimal model performance in diverse applications. | cs.CL | Does Synthetic Data Make Large Language Models More Efficient? |
2023-10-11T19:09:07Z | http://arxiv.org/pdf/2310.07826v1 | http://arxiv.org/abs/2310.07826v1 | Hrishikesh Terdalkar, Arnab Bhattacharya | 2023-10-11T19:09:07Z | cs.CL | One of the primary obstacles in the advancement of Natural Language
Processing (NLP) technologies for low-resource languages is the lack of
annotated datasets for training and testing machine learning models. In this
paper, we present Antarlekhaka, a tool for manual annotation of a comprehensive
set of tasks relevant to NLP. The tool is Unicode-compatible,
language-agnostic, Web-deployable and supports distributed annotation by
multiple simultaneous annotators. The system sports user-friendly interfaces
for 8 categories of annotation tasks. These, in turn, enable the annotation of
a considerably larger set of NLP tasks. The task categories include two
linguistic tasks not handled by any other tool, namely, sentence boundary
detection and deciding canonical word order, which are important tasks for text
that is in the form of poetry. We propose the idea of sequential annotation
based on small text units, where an annotator performs several tasks related to
a single text unit before proceeding to the next unit. The research
applications of the proposed mode of multi-task annotation are also discussed.
Antarlekhaka outperforms other annotation tools in objective evaluation. It has
been also used for two real-life annotation tasks on two different languages,
namely, Sanskrit and Bengali. The tool is available at
https://github.com/Antarlekhaka/code. | cs.CL | Antarlekhaka: A Comprehensive Tool for Multi-task Natural Language Annotation |
2023-10-11T19:02:57Z | http://arxiv.org/pdf/2310.07821v1 | http://arxiv.org/abs/2310.07821v1 | Yu Zhang, Yue Zhang, Leyang Cui, Guohong Fu | 2023-10-11T19:02:57Z | cs.CL | Recent work has witnessed a paradigm shift from Seq2Seq to Seq2Edit in the
field of text editing, with the aim of addressing the slow autoregressive
inference problem posed by the former. Despite promising results, Seq2Edit
approaches still face several challenges such as inflexibility in generation
and difficulty in generalizing to other languages. In this work, we propose a
novel non-autoregressive text editing method to circumvent the above issues, by
modeling the edit process with latent CTC alignments. We make a crucial
extension to CTC by introducing the copy operation into the edit space, thus
enabling more efficient management of textual overlap in editing. We conduct
extensive experiments on GEC and sentence fusion tasks, showing that our
proposed method significantly outperforms existing Seq2Edit models and achieves
similar or even better results than Seq2Seq with over $4\times$ speedup.
Moreover, it demonstrates good generalizability on German and Russian. In-depth
analyses reveal the strengths of our method in terms of the robustness under
various scenarios and generating fluent and flexible outputs. | cs.CL | Non-autoregressive Text Editing with Copy-aware Latent Alignments |
2023-10-11T19:00:40Z | http://arxiv.org/pdf/2310.07819v2 | http://arxiv.org/abs/2310.07819v2 | Andreas Madsen, Siva Reddy, Sarath Chandar | 2024-05-09T18:55:43Z | cs.CL, cs.LG | A common approach to explaining NLP models is to use importance measures that
express which tokens are important for a prediction. Unfortunately, such
explanations are often wrong despite being persuasive. Therefore, it is
essential to measure their faithfulness. One such metric is if tokens are truly
important, then masking them should result in worse model performance. However,
token masking introduces out-of-distribution issues, and existing solutions
that address this are computationally expensive and employ proxy models.
Furthermore, other metrics are very limited in scope. This work proposes an
inherently faithfulness measurable model that addresses these challenges. This
is achieved using a novel fine-tuning method that incorporates masking, such
that masking tokens become in-distribution by design. This differs from
existing approaches, which are completely model-agnostic but are inapplicable
in practice. We demonstrate the generality of our approach by applying it to 16
different datasets and validate it using statistical in-distribution tests. The
faithfulness is then measured with 9 different importance measures. Because
masking is in-distribution, importance measures that themselves use masking
become consistently more faithful. Additionally, because the model makes
faithfulness cheap to measure, we can optimize explanations towards maximal
faithfulness; thus, our model becomes indirectly inherently explainable. | cs.CL | Faithfulness Measurable Masked Language Models |
2023-10-11T18:59:48Z | http://arxiv.org/pdf/2310.07818v3 | http://arxiv.org/abs/2310.07818v3 | Thilini Wijesiriwardene, Ruwan Wickramarachchi, Aishwarya Naresh Reganti, Vinija Jain, Aman Chadha, Amit Sheth, Amitava Das | 2024-02-06T02:24:53Z | cs.CL, cs.AI | The ability of Large Language Models (LLMs) to encode syntactic and semantic
structures of language is well examined in NLP. Additionally, analogy
identification, in the form of word analogies are extensively studied in the
last decade of language modeling literature. In this work we specifically look
at how LLMs' abilities to capture sentence analogies (sentences that convey
analogous meaning to each other) vary with LLMs' abilities to encode syntactic
and semantic structures of sentences. Through our analysis, we find that LLMs'
ability to identify sentence analogies is positively correlated with their
ability to encode syntactic and semantic structures of sentences. Specifically,
we find that the LLMs which capture syntactic structures better, also have
higher abilities in identifying sentence analogies. | cs.CL | On the Relationship between Sentence Analogy Identification and Sentence Structure Encoding in Large Language Models |
2023-10-11T18:56:15Z | http://arxiv.org/pdf/2310.07815v3 | http://arxiv.org/abs/2310.07815v3 | Bowen Jin, Hansi Zeng, Guoyin Wang, Xiusi Chen, Tianxin Wei, Ruirui Li, Zhengyang Wang, Zheng Li, Yang Li, Hanqing Lu, Suhang Wang, Jiawei Han, Xianfeng Tang | 2024-06-12T20:36:22Z | cs.IR, cs.CL, cs.LG | Semantic identifier (ID) is an important concept in information retrieval
that aims to preserve the semantics of objects such as documents and items
inside their IDs. Previous studies typically adopt a two-stage pipeline to
learn semantic IDs by first procuring embeddings using off-the-shelf text
encoders and then deriving IDs based on the embeddings. However, each step
introduces potential information loss, and there is usually an inherent
mismatch between the distribution of embeddings within the latent space
produced by text encoders and the anticipated distribution required for
semantic indexing. It is non-trivial to design a method that can learn the
document's semantic representations and its hierarchical structure
simultaneously, given that semantic IDs are discrete and sequentially
structured, and the semantic supervision is deficient. In this paper, we
introduce LMIndexer, a self-supervised framework to learn semantic IDs with a
generative language model. We tackle the challenge of sequential discrete ID by
introducing a semantic indexer capable of generating neural sequential discrete
representations with progressive training and contrastive learning. In response
to the semantic supervision deficiency, we propose to train the model with a
self-supervised document reconstruction objective. We show the high quality of
the learned IDs and demonstrate their effectiveness on three tasks including
recommendation, product search, and document retrieval on five datasets from
various domains. Code is available at
https://github.com/PeterGriffinJin/LMIndexer. | cs.IR | Language Models As Semantic Indexers |
2023-10-11T18:36:13Z | http://arxiv.org/pdf/2310.07803v1 | http://arxiv.org/abs/2310.07803v1 | Javier Martínez | 2023-10-11T18:36:13Z | cs.CL, cs.AI, I.2.m; J.5 | This article proposes a cognitive mechanism of humour of general
applicability, not restricted to verbal communication. It is indebted to
Raskin's concept of script overlap, and conforms to the incongruity-resolution
theoretical framework, but it is built on the notion of constraint, an abstract
correspondence between sets of data. Under this view, script overlap is an
outcome of a more abstractly described phenomenon, constraint overlap. The
important concept of the overlooked argument is introduced to characterise the
two overlapping constraints -- overt and covert. Their inputs and outputs are
not directly encoded in utterances, but implicated by them, and their overlap
results in another overlap at the level of the communicated utterances, that
the incongruity reveals. Our hypothesis assumes as a given that the evocation
of such constraints is a cognitive effect of the inferential process by which a
hearer interprets utterances. We base this assumption on Hofstadter's theory of
analogy-making as the essence of human thought. By substituting "stimuli" of
any kind for "utterances" in this model, we obtain a mechanism as easily
applicable to non-verbal communication -- slapstick, cartoons -- and we propose
it describes the necessary and sufficient conditions for a communicative act in
any modality to carry humour. | cs.CL | A general mechanism of humor: reformulating the semantic overlap |
2023-10-11T18:30:37Z | http://arxiv.org/pdf/2310.07795v1 | http://arxiv.org/abs/2310.07795v1 | Siru Ouyang, Jiaxin Huang, Pranav Pillai, Yunyi Zhang, Yu Zhang, Jiawei Han | 2023-10-11T18:30:37Z | cs.CL | Fine-grained entity typing (FET) is the task of identifying specific entity
types at a fine-grained level for entity mentions based on their contextual
information. Conventional methods for FET require extensive human annotation,
which is time-consuming and costly. Recent studies have been developing weakly
supervised or zero-shot approaches. We study the setting of zero-shot FET where
only an ontology is provided. However, most existing ontology structures lack
rich supporting information and even contain ambiguous relations, making them
ineffective in guiding FET. Recently developed language models, though
promising in various few-shot and zero-shot NLP tasks, may face challenges in
zero-shot FET due to their lack of interaction with task-specific ontology. In
this study, we propose OnEFET, where we (1) enrich each node in the ontology
structure with two types of extra information: instance information for
training sample augmentation and topic information to relate types to contexts,
and (2) develop a coarse-to-fine typing algorithm that exploits the enriched
information by training an entailment model with contrasting topics and
instance-based augmented training samples. Our experiments show that OnEFET
achieves high-quality fine-grained entity typing without human annotation,
outperforming existing zero-shot methods by a large margin and rivaling
supervised methods. | cs.CL | Ontology Enrichment for Effective Fine-grained Entity Typing |
2023-10-11T18:27:12Z | http://arxiv.org/pdf/2310.07793v5 | http://arxiv.org/abs/2310.07793v5 | Ruotong Liao, Xu Jia, Yangzhe Li, Yunpu Ma, Volker Tresp | 2024-04-16T18:35:30Z | cs.CL, cs.AI, cs.LG | The rapid advancements in large language models (LLMs) have ignited interest
in the temporal knowledge graph (tKG) domain, where conventional
embedding-based and rule-based methods dominate. The question remains open of
whether pre-trained LLMs can understand structured temporal relational data and
replace them as the foundation model for temporal relational forecasting.
Therefore, we bring temporal knowledge forecasting into the generative setting.
However, challenges occur in the huge chasms between complex temporal graph
data structure and sequential natural expressions LLMs can handle, and between
the enormous data sizes of tKGs and heavy computation costs of finetuning LLMs.
To address these challenges, we propose a novel retrieval-augmented generation
framework named GenTKG combining a temporal logical rule-based retrieval
strategy and few-shot parameter-efficient instruction tuning to solve the above
challenges, respectively. Extensive experiments have shown that GenTKG
outperforms conventional methods of temporal relational forecasting with low
computation resources using extremely limited training data as few as 16
samples. GenTKG also highlights remarkable cross-domain generalizability with
outperforming performance on unseen datasets without re-training, and in-domain
generalizability regardless of time split in the same dataset. Our work reveals
the huge potential of LLMs in the tKG domain and opens a new frontier for
generative forecasting on tKGs. Code and data are released here:
https://github.com/mayhugotong/GenTKG. | cs.CL | GenTKG: Generative Forecasting on Temporal Knowledge Graph with Large Language Models |
2023-10-11T17:59:36Z | http://arxiv.org/pdf/2310.07715v1 | http://arxiv.org/abs/2310.07715v1 | Sireesh Gururaja, Amanda Bertsch, Clara Na, David Gray Widder, Emma Strubell | 2023-10-11T17:59:36Z | cs.CL, cs.CY | NLP is in a period of disruptive change that is impacting our methodologies,
funding sources, and public perception. In this work, we seek to understand how
to shape our future by better understanding our past. We study factors that
shape NLP as a field, including culture, incentives, and infrastructure by
conducting long-form interviews with 26 NLP researchers of varying seniority,
research area, institution, and social identity. Our interviewees identify
cyclical patterns in the field, as well as new shifts without historical
parallel, including changes in benchmark culture and software infrastructure.
We complement this discussion with quantitative analysis of citation,
authorship, and language use in the ACL Anthology over time. We conclude by
discussing shared visions, concerns, and hopes for the future of NLP. We hope
that this study of our field's past and present can prompt informed discussion
of our community's implicit norms and more deliberate action to consciously
shape the future. | cs.CL | To Build Our Future, We Must Know Our Past: Contextualizing Paradigm Shifts in Natural Language Processing |
2023-10-11T17:59:05Z | http://arxiv.org/pdf/2310.07713v3 | http://arxiv.org/abs/2310.07713v3 | Boxin Wang, Wei Ping, Lawrence McAfee, Peng Xu, Bo Li, Mohammad Shoeybi, Bryan Catanzaro | 2024-05-29T04:15:39Z | cs.CL, cs.AI, cs.IR, cs.LG | Pretraining auto-regressive large language models~(LLMs) with retrieval
demonstrates better perplexity and factual accuracy by leveraging external
databases. However, the size of existing pretrained retrieval-augmented LLM is
still limited (e.g., Retro has 7.5B parameters), which limits the effectiveness
of instruction tuning and zero-shot generalization. In this work, we introduce
Retro 48B, the largest LLM pretrained with retrieval. Specifically, we continue
to pretrain a 43B GPT model on additional 100 billion tokens using the Retro
augmentation method by retrieving from 1.2 trillion tokens. Notably, the
obtained foundation model, Retro 48B, largely outperforms the counterpart GPT
43B trained on 1.2T tokens in terms of perplexity with only 2.58% additional
GPU hours, demonstrating the significant scaling potential of the method. After
instruction tuning on Retro, InstructRetro demonstrates significant improvement
over the instruction tuned GPT on a wide range of zero-shot tasks.
Specifically, the average improvement of InstructRetro is 7% over its GPT
counterpart across 8 short-form QA and reading comprehension tasks, 10% over
GPT across 4 challenging long-form QA tasks, and 16% over GPT across 3
summarization tasks. Surprisingly, we find that one can ablate the encoder from
InstructRetro architecture and directly use its decoder backbone, while
achieving comparable results. Our results highlight the promising direction to
obtain a better GPT decoder through continued pretraining with retrieval before
instruction tuning. Our code and checkpoints are publicly available at:
https://huggingface.co/nvidia/retro-48b-instruct-4k. | cs.CL | InstructRetro: Instruction Tuning post Retrieval-Augmented Pretraining |
2023-10-11T17:59:02Z | http://arxiv.org/pdf/2310.07712v2 | http://arxiv.org/abs/2310.07712v2 | Raphael Tang, Xinyu Zhang, Xueguang Ma, Jimmy Lin, Ferhan Ture | 2024-04-22T17:53:48Z | cs.CL, cs.LG | Large language models (LLMs) exhibit positional bias in how they use context,
which especially complicates listwise ranking. To address this, we propose
permutation self-consistency, a form of self-consistency over ranking list
outputs of black-box LLMs. Our key idea is to marginalize out different list
orders in the prompt to produce an order-independent ranking with less
positional bias. First, given some input prompt, we repeatedly shuffle the list
in the prompt and pass it through the LLM while holding the instructions the
same. Next, we aggregate the resulting sample of rankings by computing the
central ranking closest in distance to all of them, marginalizing out prompt
order biases in the process. Theoretically, we prove the robustness of our
method, showing convergence to the true ranking in the presence of random
perturbations. Empirically, on five list-ranking datasets in sorting and
passage reranking, our approach improves scores from conventional inference by
up to 7-18% for GPT-3.5 and 8-16% for LLaMA v2 (70B), surpassing the previous
state of the art in passage reranking. Our code is at
https://github.com/castorini/perm-sc. | cs.CL | Found in the Middle: Permutation Self-Consistency Improves Listwise Ranking in Large Language Models |
2023-10-11T17:57:35Z | http://arxiv.org/pdf/2310.07710v2 | http://arxiv.org/abs/2310.07710v2 | Yihan Wu, Zhengmian Hu, Junfeng Guo, Hongyang Zhang, Heng Huang | 2024-06-25T07:08:17Z | cs.CR, cs.CL, cs.LG | Watermarking techniques offer a promising way to identify machine-generated
content via embedding covert information into the contents generated from
language models. A challenge in the domain lies in preserving the distribution
of original generated content after watermarking. Our research extends and
improves upon existing watermarking framework, placing emphasis on the
importance of a \textbf{Di}stribution-\textbf{P}reserving (DiP) watermark.
Contrary to the current strategies, our proposed DiPmark simultaneously
preserves the original token distribution during watermarking
(distribution-preserving), is detectable without access to the language model
API and prompts (accessible), and is provably robust to moderate changes of
tokens (resilient). DiPmark operates by selecting a random set of tokens prior
to the generation of a word, then modifying the token distribution through a
distribution-preserving reweight function to enhance the probability of these
selected tokens during the sampling process. Extensive empirical evaluation on
various language models and tasks demonstrates our approach's
distribution-preserving property, accessibility, and resilience, making it a
effective solution for watermarking tasks that demand impeccable quality
preservation. | cs.CR | A Resilient and Accessible Distribution-Preserving Watermark for Large Language Models |
2023-10-11T17:57:14Z | http://arxiv.org/pdf/2310.07707v1 | http://arxiv.org/abs/2310.07707v1 | Devvrit, Sneha Kudugunta, Aditya Kusupati, Tim Dettmers, Kaifeng Chen, Inderjit Dhillon, Yulia Tsvetkov, Hannaneh Hajishirzi, Sham Kakade, Ali Farhadi, Prateek Jain | 2023-10-11T17:57:14Z | cs.LG, cs.CL, cs.CV | Transformer models are deployed in a wide range of settings, from
multi-accelerator clusters to standalone mobile phones. The diverse inference
constraints in these scenarios necessitate practitioners to train foundation
models such as PaLM 2, Llama, & ViTs as a series of models of varying sizes.
Due to significant training costs, only a select few model sizes are trained
and supported, limiting more fine-grained control over relevant tradeoffs,
including latency, cost, and accuracy. This work introduces MatFormer, a nested
Transformer architecture designed to offer elasticity in a variety of
deployment constraints. Each Feed Forward Network (FFN) block of a MatFormer
model is jointly optimized with a few nested smaller FFN blocks. This training
procedure allows for the Mix'n'Match of model granularities across layers --
i.e., a trained universal MatFormer model enables extraction of hundreds of
accurate smaller models, which were never explicitly optimized. We empirically
demonstrate MatFormer's effectiveness across different model classes (decoders
& encoders), modalities (language & vision), and scales (up to 2.6B
parameters). We find that a 2.6B decoder-only MatFormer language model (MatLM)
allows us to extract smaller models spanning from 1.5B to 2.6B, each exhibiting
comparable validation loss and one-shot downstream evaluations to their
independently trained counterparts. Furthermore, we observe that smaller
encoders extracted from a universal MatFormer-based ViT (MatViT) encoder
preserve the metric-space structure for adaptive large-scale retrieval.
Finally, we showcase that speculative decoding with the accurate and consistent
submodels extracted from MatFormer can further reduce inference latency. | cs.LG | MatFormer: Nested Transformer for Elastic Inference |
2023-10-11T17:55:15Z | http://arxiv.org/pdf/2310.07704v1 | http://arxiv.org/abs/2310.07704v1 | Haoxuan You, Haotian Zhang, Zhe Gan, Xianzhi Du, Bowen Zhang, Zirui Wang, Liangliang Cao, Shih-Fu Chang, Yinfei Yang | 2023-10-11T17:55:15Z | cs.CV, cs.CL | We introduce Ferret, a new Multimodal Large Language Model (MLLM) capable of
understanding spatial referring of any shape or granularity within an image and
accurately grounding open-vocabulary descriptions. To unify referring and
grounding in the LLM paradigm, Ferret employs a novel and powerful hybrid
region representation that integrates discrete coordinates and continuous
features jointly to represent a region in the image. To extract the continuous
features of versatile regions, we propose a spatial-aware visual sampler, adept
at handling varying sparsity across different shapes. Consequently, Ferret can
accept diverse region inputs, such as points, bounding boxes, and free-form
shapes. To bolster the desired capability of Ferret, we curate GRIT, a
comprehensive refer-and-ground instruction tuning dataset including 1.1M
samples that contain rich hierarchical spatial knowledge, with 95K hard
negative data to promote model robustness. The resulting model not only
achieves superior performance in classical referring and grounding tasks, but
also greatly outperforms existing MLLMs in region-based and
localization-demanded multimodal chatting. Our evaluations also reveal a
significantly improved capability of describing image details and a remarkable
alleviation in object hallucination. Code and data will be available at
https://github.com/apple/ml-ferret | cs.CV | Ferret: Refer and Ground Anything Anywhere at Any Granularity |
2023-10-11T17:51:28Z | http://arxiv.org/pdf/2310.07700v1 | http://arxiv.org/abs/2310.07700v1 | Mengzhao Jia, Qianglong Chen, Liqiang Jing, Dawei Fu, Renyu Li | 2023-10-11T17:51:28Z | cs.CL | The prevalence of mental disorders has become a significant issue, leading to
the increased focus on Emotional Support Conversation as an effective
supplement for mental health support. Existing methods have achieved compelling
results, however, they still face three challenges: 1) variability of emotions,
2) practicality of the response, and 3) intricate strategy modeling. To address
these challenges, we propose a novel knowledge-enhanced Memory mODEl for
emotional suppoRt coNversation (MODERN). Specifically, we first devise a
knowledge-enriched dialogue context encoding to perceive the dynamic emotion
change of different periods of the conversation for coherent user state
modeling and select context-related concepts from ConceptNet for practical
response generation. Thereafter, we implement a novel memory-enhanced strategy
modeling module to model the semantic patterns behind the strategy categories.
Extensive experiments on a widely used large-scale dataset verify the
superiority of our model over cutting-edge baselines. | cs.CL | Knowledge-enhanced Memory Model for Emotional Support Conversation |
2023-10-11T17:21:03Z | http://arxiv.org/pdf/2310.07676v2 | http://arxiv.org/abs/2310.07676v2 | Hai Huang, Zhengyu Zhao, Michael Backes, Yun Shen, Yang Zhang | 2024-03-30T16:09:34Z | cs.CR, cs.CL, cs.LG | Large language models (LLMs) have demonstrated superior performance compared
to previous methods on various tasks, and often serve as the foundation models
for many researches and services. However, the untrustworthy third-party LLMs
may covertly introduce vulnerabilities for downstream tasks. In this paper, we
explore the vulnerability of LLMs through the lens of backdoor attacks.
Different from existing backdoor attacks against LLMs, ours scatters multiple
trigger keys in different prompt components. Such a Composite Backdoor Attack
(CBA) is shown to be stealthier than implanting the same multiple trigger keys
in only a single component. CBA ensures that the backdoor is activated only
when all trigger keys appear. Our experiments demonstrate that CBA is effective
in both natural language processing (NLP) and multimodal tasks. For instance,
with $3\%$ poisoning samples against the LLaMA-7B model on the Emotion dataset,
our attack achieves a $100\%$ Attack Success Rate (ASR) with a False Triggered
Rate (FTR) below $2.06\%$ and negligible model accuracy degradation. Our work
highlights the necessity of increased security research on the trustworthiness
of foundation LLMs. | cs.CR | Composite Backdoor Attacks Against Large Language Models |
2023-10-11T17:00:29Z | http://arxiv.org/pdf/2310.07659v3 | http://arxiv.org/abs/2310.07659v3 | Lang Qin, Yao Zhang, Hongru Liang, Jun Wang, Zhenglu Yang | 2023-10-20T13:43:35Z | cs.CL | Accurate knowledge selection is critical in knowledge-grounded dialogue
systems. Towards a closer look at it, we offer a novel perspective to organize
existing literature, i.e., knowledge selection coupled with, after, and before
generation. We focus on the third under-explored category of study, which can
not only select knowledge accurately in advance, but has the advantage to
reduce the learning, adjustment, and interpretation burden of subsequent
response generation models, especially LLMs. We propose GATE, a
generator-agnostic knowledge selection method, to prepare knowledge for
subsequent response generation models by selecting context-related knowledge
among different knowledge structures and variable knowledge requirements.
Experimental results demonstrate the superiority of GATE, and indicate that
knowledge selection before generation is a lightweight yet effective way to
facilitate LLMs (e.g., ChatGPT) to generate more informative responses. | cs.CL | Well Begun is Half Done: Generator-agnostic Knowledge Pre-Selection for Knowledge-Grounded Dialogue |
2023-10-11T16:54:57Z | http://arxiv.org/pdf/2310.07654v1 | http://arxiv.org/abs/2310.07654v1 | Cheng-I Jeff Lai, Freda Shi, Puyuan Peng, Yoon Kim, Kevin Gimpel, Shiyu Chang, Yung-Sung Chuang, Saurabhchand Bhati, David Cox, David Harwath, Yang Zhang, Karen Livescu, James Glass | 2023-10-11T16:54:57Z | cs.CL, cs.LG, cs.SD, eess.AS | We study phrase structure induction from visually-grounded speech. The core
idea is to first segment the speech waveform into sequences of word segments,
and subsequently induce phrase structure using the inferred segment-level
continuous representations. We present the Audio-Visual Neural Syntax Learner
(AV-NSL) that learns phrase structure by listening to audio and looking at
images, without ever being exposed to text. By training on paired images and
spoken captions, AV-NSL exhibits the capability to infer meaningful phrase
structures that are comparable to those derived by naturally-supervised text
parsers, for both English and German. Our findings extend prior work in
unsupervised language acquisition from speech and grounded grammar induction,
and present one approach to bridge the gap between the two topics. | cs.CL | Audio-Visual Neural Syntax Acquisition |
2023-10-11T16:51:46Z | http://arxiv.org/pdf/2310.07652v2 | http://arxiv.org/abs/2310.07652v2 | Lei Wang, Songheng Zhang, Yun Wang, Ee-Peng Lim, Yong Wang | 2023-10-16T03:34:47Z | cs.HC, cs.CL | Data visualization is a powerful tool for exploring and communicating
insights in various domains. To automate visualization choice for datasets, a
task known as visualization recommendation has been proposed. Various
machine-learning-based approaches have been developed for this purpose, but
they often require a large corpus of dataset-visualization pairs for training
and lack natural explanations for their results. To address this research gap,
we propose LLM4Vis, a novel ChatGPT-based prompting approach to perform
visualization recommendation and return human-like explanations using very few
demonstration examples. Our approach involves feature description,
demonstration example selection, explanation generation, demonstration example
construction, and inference steps. To obtain demonstration examples with
high-quality explanations, we propose a new explanation generation
bootstrapping to iteratively refine generated explanations by considering the
previous generation and template-based hint. Evaluations on the VizML dataset
show that LLM4Vis outperforms or performs similarly to supervised learning
models like Random Forest, Decision Tree, and MLP in both few-shot and
zero-shot settings. The qualitative evaluation also shows the effectiveness of
explanations generated by LLM4Vis. We make our code publicly available at
\href{https://github.com/demoleiwang/LLM4Vis}{https://github.com/demoleiwang/LLM4Vis}. | cs.HC | LLM4Vis: Explainable Visualization Recommendation using ChatGPT |
2023-10-11T16:40:57Z | http://arxiv.org/pdf/2310.07644v2 | http://arxiv.org/abs/2310.07644v2 | Chaoqi Liang, Weiqiang Bai, Lifeng Qiao, Yuchen Ren, Jianle Sun, Peng Ye, Hongliang Yan, Xinzhu Ma, Wangmeng Zuo, Wanli Ouyang | 2023-10-12T03:32:32Z | cs.AI, cs.CL, cs.LG | With the success of large-scale pretraining in NLP, there is an increasing
trend of applying it to the domain of life sciences. In particular, pretraining
methods based on DNA sequences have garnered growing attention due to their
potential to capture generic information about genes. However, existing
pretraining methods for DNA sequences largely rely on direct adoptions of BERT
pretraining from NLP, lacking a comprehensive understanding and a specifically
tailored approach. To address this research gap, we first conducted a series of
exploratory experiments and gained several insightful observations: 1) In the
fine-tuning phase of downstream tasks, when using K-mer overlapping
tokenization instead of K-mer non-overlapping tokenization, both overlapping
and non-overlapping pretraining weights show consistent performance
improvement.2) During the pre-training process, using K-mer overlapping
tokenization quickly produces clear K-mer embeddings and reduces the loss to a
very low level, while using K-mer non-overlapping tokenization results in less
distinct embeddings and continuously decreases the loss. 3) Using overlapping
tokenization causes the self-attention in the intermediate layers of
pre-trained models to tend to overly focus on certain tokens, reflecting that
these layers are not adequately optimized. In summary, overlapping tokenization
can benefit the fine-tuning of downstream tasks but leads to inadequate
pretraining with fast convergence. To unleash the pretraining potential, we
introduce a novel approach called RandomMask, which gradually increases the
task difficulty of BERT-like pretraining by continuously expanding its mask
boundary, forcing the model to learn more knowledge. RandomMask is simple but
effective, achieving top-tier performance across 26 datasets of 28 datasets
spanning 7 downstream tasks. | cs.AI | Rethinking the BERT-like Pretraining for DNA Sequences |
2023-10-11T16:38:11Z | http://arxiv.org/pdf/2310.07641v2 | http://arxiv.org/abs/2310.07641v2 | Zhiyuan Zeng, Jiatong Yu, Tianyu Gao, Yu Meng, Tanya Goyal, Danqi Chen | 2024-04-16T04:50:08Z | cs.CL, cs.LG | As research in large language models (LLMs) continues to accelerate,
LLM-based evaluation has emerged as a scalable and cost-effective alternative
to human evaluations for comparing the ever increasing list of models. This
paper investigates the efficacy of these ``LLM evaluators'', particularly in
using them to assess instruction following, a metric that gauges how closely
generated text adheres to the given instruction. We introduce a challenging
meta-evaluation benchmark, LLMBar, designed to test the ability of an LLM
evaluator in discerning instruction-following outputs. The authors manually
curated 419 pairs of outputs, one adhering to instructions while the other
diverging, yet may possess deceptive qualities that mislead an LLM evaluator,
e.g., a more engaging tone. Contrary to existing meta-evaluation, we discover
that different evaluators (i.e., combinations of LLMs and prompts) exhibit
distinct performance on LLMBar and even the highest-scoring ones have
substantial room for improvement. We also present a novel suite of prompting
strategies that further close the gap between LLM and human evaluators. With
LLMBar, we hope to offer more insight into LLM evaluators and foster future
research in developing better instruction-following models. | cs.CL | Evaluating Large Language Models at Evaluating Instruction Following |
2023-10-11T16:18:13Z | http://arxiv.org/pdf/2310.07629v1 | http://arxiv.org/abs/2310.07629v1 | Hannah Rose Kirk, Andrew M. Bean, Bertie Vidgen, Paul Röttger, Scott A. Hale | 2023-10-11T16:18:13Z | cs.CL, cs.CY | Human feedback is increasingly used to steer the behaviours of Large Language
Models (LLMs). However, it is unclear how to collect and incorporate feedback
in a way that is efficient, effective and unbiased, especially for highly
subjective human preferences and values. In this paper, we survey existing
approaches for learning from human feedback, drawing on 95 papers primarily
from the ACL and arXiv repositories.First, we summarise the past, pre-LLM
trends for integrating human feedback into language models. Second, we give an
overview of present techniques and practices, as well as the motivations for
using feedback; conceptual frameworks for defining values and preferences; and
how feedback is collected and from whom. Finally, we encourage a better future
of feedback learning in LLMs by raising five unresolved conceptual and
practical challenges. | cs.CL | The Past, Present and Better Future of Feedback Learning in Large Language Models for Subjective Human Preferences and Values |
2023-10-11T15:56:00Z | http://arxiv.org/pdf/2310.07611v2 | http://arxiv.org/abs/2310.07611v2 | Sumuk Shashidhar, Abhinav Chinta, Vaibhav Sahai, Zhenhailong Wang, Heng Ji | 2023-10-22T00:37:06Z | cs.CL, cs.AI, cs.PF, 68T50 (Primary), I.2.7; A.2; H.3.4; K.4.1; C.4 | The dominance of proprietary LLMs has led to restricted access and raised
information privacy concerns. High-performing open-source alternatives are
crucial for information-sensitive and high-volume applications but often lag
behind in performance. To address this gap, we propose (1) A untargeted variant
of iterative self-critique and self-refinement devoid of external influence.
(2) A novel ranking metric - Performance, Refinement, and Inference Cost Score
(PeRFICS) - to find the optimal model for a given task considering refined
performance and cost. Our experiments show that SoTA open source models of
varying sizes from 7B - 65B, on average, improve 8.2% from their baseline
performance. Strikingly, even models with extremely small memory footprints,
such as Vicuna-7B, show a 11.74% improvement overall and up to a 25.39%
improvement in high-creativity, open ended tasks on the Vicuna benchmark.
Vicuna-13B takes it a step further and outperforms ChatGPT post-refinement.
This work has profound implications for resource-constrained and
information-sensitive environments seeking to leverage LLMs without incurring
prohibitive costs, compromising on performance and privacy. The domain-agnostic
self-refinement process coupled with our novel ranking metric facilitates
informed decision-making in model selection, thereby reducing costs and
democratizing access to high-performing language models, as evidenced by case
studies. | cs.CL | Democratizing LLMs: An Exploration of Cost-Performance Trade-offs in Self-Refined Open-Source Models |
2023-10-11T15:51:53Z | http://arxiv.org/pdf/2310.07609v1 | http://arxiv.org/abs/2310.07609v1 | Liangming Pan, Xinyuan Lu, Min-Yen Kan, Preslav Nakov | 2023-10-11T15:51:53Z | cs.CL | Fact-checking real-world claims often requires complex, multi-step reasoning
due to the absence of direct evidence to support or refute them. However,
existing fact-checking systems often lack transparency in their
decision-making, making it challenging for users to comprehend their reasoning
process. To address this, we propose the Question-guided Multi-hop
Fact-Checking (QACHECK) system, which guides the model's reasoning process by
asking a series of questions critical for verifying a claim. QACHECK has five
key modules: a claim verifier, a question generator, a question-answering
module, a QA validator, and a reasoner. Users can input a claim into QACHECK,
which then predicts its veracity and provides a comprehensive report detailing
its reasoning process, guided by a sequence of (question, answer) pairs.
QACHECK also provides the source of evidence supporting each question,
fostering a transparent, explainable, and user-friendly fact-checking process.
A recorded video of QACHECK is at https://www.youtube.com/watch?v=ju8kxSldM64 | cs.CL | QACHECK: A Demonstration System for Question-Guided Multi-Hop Fact-Checking |
2023-10-11T15:28:44Z | http://arxiv.org/pdf/2310.07588v1 | http://arxiv.org/abs/2310.07588v1 | Caoyun Fan, Wenqing Chen, Jidong Tian, Yitian Li, Hao He, Yaohui Jin | 2023-10-11T15:28:44Z | cs.CL, cs.AI, cs.LG | Multi-Label Text Classification (MLTC) aims to assign the most relevant
labels to each given text. Existing methods demonstrate that label dependency
can help to improve the model's performance. However, the introduction of label
dependency may cause the model to suffer from unwanted prediction bias. In this
study, we attribute the bias to the model's misuse of label dependency, i.e.,
the model tends to utilize the correlation shortcut in label dependency rather
than fusing text information and label dependency for prediction. Motivated by
causal inference, we propose a CounterFactual Text Classifier (CFTC) to
eliminate the correlation bias, and make causality-based predictions.
Specifically, our CFTC first adopts the predict-then-modify backbone to extract
precise label information embedded in label dependency, then blocks the
correlation shortcut through the counterfactual de-bias technique with the help
of the human causal graph. Experimental results on three datasets demonstrate
that our CFTC significantly outperforms the baselines and effectively
eliminates the correlation bias in datasets. | cs.CL | Accurate Use of Label Dependency in Multi-Label Text Classification Through the Lens of Causality |
2023-10-11T14:18:03Z | http://arxiv.org/pdf/2310.07521v3 | http://arxiv.org/abs/2310.07521v3 | Cunxiang Wang, Xiaoze Liu, Yuanhao Yue, Xiangru Tang, Tianhang Zhang, Cheng Jiayang, Yunzhi Yao, Wenyang Gao, Xuming Hu, Zehan Qi, Yidong Wang, Linyi Yang, Jindong Wang, Xing Xie, Zheng Zhang, Yue Zhang | 2023-12-16T12:47:19Z | cs.CL | This survey addresses the crucial issue of factuality in Large Language
Models (LLMs). As LLMs find applications across diverse domains, the
reliability and accuracy of their outputs become vital. We define the
Factuality Issue as the probability of LLMs to produce content inconsistent
with established facts. We first delve into the implications of these
inaccuracies, highlighting the potential consequences and challenges posed by
factual errors in LLM outputs. Subsequently, we analyze the mechanisms through
which LLMs store and process facts, seeking the primary causes of factual
errors. Our discussion then transitions to methodologies for evaluating LLM
factuality, emphasizing key metrics, benchmarks, and studies. We further
explore strategies for enhancing LLM factuality, including approaches tailored
for specific domains. We focus two primary LLM configurations standalone LLMs
and Retrieval-Augmented LLMs that utilizes external data, we detail their
unique challenges and potential enhancements. Our survey offers a structured
guide for researchers aiming to fortify the factual reliability of LLMs. | cs.CL | Survey on Factuality in Large Language Models: Knowledge, Retrieval and Domain-Specificity |
2023-10-11T13:35:05Z | http://arxiv.org/pdf/2310.07488v2 | http://arxiv.org/abs/2310.07488v2 | Jiayi Fu, Lei Lin, Xiaoyang Gao, Pengli Liu, Zhengzong Chen, Zhirui Yang, Shengnan Zhang, Xue Zheng, Yan Li, Yuliang Liu, Xucheng Ye, Yiqiao Liao, Chao Liao, Bin Chen, Chengru Song, Junchen Wan, Zijia Lin, Fuzheng Zhang, Zhongyuan Wang, Di Zhang, Kun Gai | 2023-10-19T12:34:32Z | cs.CL, cs.AI, cs.LG | Recent advancements in large language models (LLMs) have demonstrated
remarkable abilities in handling a variety of natural language processing (NLP)
downstream tasks, even on mathematical tasks requiring multi-step reasoning. In
this report, we introduce the KwaiYiiMath which enhances the mathematical
reasoning abilities of KwaiYiiBase1, by applying Supervised Fine-Tuning (SFT)
and Reinforced Learning from Human Feedback (RLHF), including on both English
and Chinese mathematical tasks. Meanwhile, we also constructed a small-scale
Chinese primary school mathematics test set (named KMath), consisting of 188
examples to evaluate the correctness of the problem-solving process generated
by the models. Empirical studies demonstrate that KwaiYiiMath can achieve
state-of-the-art (SOTA) performance on GSM8k, CMath, and KMath compared with
the similar size models, respectively. | cs.CL | KwaiYiiMath: Technical Report |
2023-10-11T13:34:22Z | http://arxiv.org/pdf/2310.07487v2 | http://arxiv.org/abs/2310.07487v2 | V. S. D. S. Mahesh Akavarapu, Arnab Bhattacharya | 2023-10-18T06:48:37Z | cs.CL, I.2.7 | Phonological reconstruction is one of the central problems in historical
linguistics where a proto-word of an ancestral language is determined from the
observed cognate words of daughter languages. Computational approaches to
historical linguistics attempt to automate the task by learning models on
available linguistic data. Several ideas and techniques drawn from
computational biology have been successfully applied in the area of
computational historical linguistics. Following these lines, we adapt MSA
Transformer, a protein language model, to the problem of automated phonological
reconstruction. MSA Transformer trains on multiple sequence alignments as input
and is, thus, apt for application on aligned cognate words. We, hence, name our
model as Cognate Transformer. We also apply the model on another associated
task, namely, cognate reflex prediction, where a reflex word in a daughter
language is predicted based on cognate words from other daughter languages. We
show that our model outperforms the existing models on both tasks, especially
when it is pre-trained on masked word prediction task. | cs.CL | Cognate Transformer for Automated Phonological Reconstruction and Cognate Reflex Prediction |
2023-10-11T12:15:24Z | http://arxiv.org/pdf/2310.07423v1 | http://arxiv.org/abs/2310.07423v1 | Atharva Kulkarni, Ajinkya Kulkarni, Miguel Couceiro, Hanan Aldarmaki | 2023-10-11T12:15:24Z | cs.CL, cs.SD, eess.AS | Recently, large pre-trained multilingual speech models have shown potential
in scaling Automatic Speech Recognition (ASR) to many low-resource languages.
Some of these models employ language adapters in their formulation, which helps
to improve monolingual performance and avoids some of the drawbacks of
multi-lingual modeling on resource-rich languages. However, this formulation
restricts the usability of these models on code-switched speech, where two
languages are mixed together in the same utterance. In this work, we propose
ways to effectively fine-tune such models on code-switched speech, by
assimilating information from both language adapters at each language
adaptation point in the network. We also model code-switching as a sequence of
latent binary sequences that can be used to guide the flow of information from
each language adapter at the frame level. The proposed approaches are evaluated
on three code-switched datasets encompassing Arabic, Mandarin, and Hindi
languages paired with English, showing consistent improvements in
code-switching performance with at least 10\% absolute reduction in CER across
all test sets. | cs.CL | Adapting the adapters for code-switching in multilingual ASR |
2023-10-11T11:39:36Z | http://arxiv.org/pdf/2310.07403v1 | http://arxiv.org/abs/2310.07403v1 | Qingkai Fang, Yan Zhou, Yang Feng | 2023-10-11T11:39:36Z | cs.CL, cs.AI, cs.SD, eess.AS, I.2.7 | Direct speech-to-speech translation (S2ST) translates speech from one
language into another using a single model. However, due to the presence of
linguistic and acoustic diversity, the target speech follows a complex
multimodal distribution, posing challenges to achieving both high-quality
translations and fast decoding speeds for S2ST models. In this paper, we
propose DASpeech, a non-autoregressive direct S2ST model which realizes both
fast and high-quality S2ST. To better capture the complex distribution of the
target speech, DASpeech adopts the two-pass architecture to decompose the
generation process into two steps, where a linguistic decoder first generates
the target text, and an acoustic decoder then generates the target speech based
on the hidden states of the linguistic decoder. Specifically, we use the
decoder of DA-Transformer as the linguistic decoder, and use FastSpeech 2 as
the acoustic decoder. DA-Transformer models translations with a directed
acyclic graph (DAG). To consider all potential paths in the DAG during
training, we calculate the expected hidden states for each target token via
dynamic programming, and feed them into the acoustic decoder to predict the
target mel-spectrogram. During inference, we select the most probable path and
take hidden states on that path as input to the acoustic decoder. Experiments
on the CVSS Fr-En benchmark demonstrate that DASpeech can achieve comparable or
even better performance than the state-of-the-art S2ST model Translatotron 2,
while preserving up to 18.53x speedup compared to the autoregressive baseline.
Compared with the previous non-autoregressive S2ST model, DASpeech does not
rely on knowledge distillation and iterative decoding, achieving significant
improvements in both translation quality and decoding speed. Furthermore,
DASpeech shows the ability to preserve the speaker's voice of the source speech
during translation. | cs.CL | DASpeech: Directed Acyclic Transformer for Fast and High-quality Speech-to-Speech Translation |
2023-10-11T11:32:57Z | http://arxiv.org/pdf/2310.07397v2 | http://arxiv.org/abs/2310.07397v2 | Jian Wang, Yi Cheng, Dongding Lin, Chak Tou Leong, Wenjie Li | 2023-10-13T11:16:58Z | cs.CL, cs.AI | Target-oriented dialogue systems, designed to proactively steer conversations
toward predefined targets or accomplish specific system-side goals, are an
exciting area in conversational AI. In this work, by formulating a <dialogue
act, topic> pair as the conversation target, we explore a novel problem of
personalized target-oriented dialogue by considering personalization during the
target accomplishment process. However, there remains an emergent need for
high-quality datasets, and building one from scratch requires tremendous human
effort. To address this, we propose an automatic dataset curation framework
using a role-playing approach. Based on this framework, we construct a
large-scale personalized target-oriented dialogue dataset, TopDial, which
comprises about 18K multi-turn dialogues. The experimental results show that
this dataset is of high quality and could contribute to exploring personalized
target-oriented dialogue. | cs.CL | Target-oriented Proactive Dialogue Systems with Personalization: Problem Formulation and Dataset Curation |
2023-10-11T11:08:20Z | http://arxiv.org/pdf/2310.07387v1 | http://arxiv.org/abs/2310.07387v1 | Stuart Semple, Ramon Ferrer-i-Cancho, Morgan L. Gustison | 2023-10-11T11:08:20Z | cs.CL, physics.bio-ph | Linguistic laws, the common statistical patterns of human language, have been
investigated by quantitative linguists for nearly a century. Recently,
biologists from a range of disciplines have started to explore the prevalence
of these laws beyond language, finding patterns consistent with linguistic laws
across multiple levels of biological organisation, from molecular (genomes,
genes, and proteins) to organismal (animal behaviour) to ecological
(populations and ecosystems). We propose a new conceptual framework for the
study of linguistic laws in biology, comprising and integrating distinct levels
of analysis, from description to prediction to theory building. Adopting this
framework will provide critical new insights into the fundamental rules of
organisation underpinning natural systems, unifying linguistic laws and core
theory in biology. | cs.CL | Linguistic laws in biology |
2023-10-11T09:55:46Z | http://arxiv.org/pdf/2310.07347v1 | http://arxiv.org/abs/2310.07347v1 | Chengyu Dong, Liyuan Liu, Hao Cheng, Jingbo Shang, Jianfeng Gao, Xiaodong Liu | 2023-10-11T09:55:46Z | cs.CL, cs.AI, cs.LG | ELECTRA pre-trains language models by detecting tokens in a sequence that
have been replaced by an auxiliary model. Although ELECTRA offers a significant
boost in efficiency, its potential is constrained by the training cost brought
by the auxiliary model. Notably, this model, which is jointly trained with the
main model, only serves to assist the training of the main model and is
discarded post-training. This results in a substantial amount of training cost
being expended in vain. To mitigate this issue, we propose Fast-ELECTRA, which
leverages an existing language model as the auxiliary model. To construct a
learning curriculum for the main model, we smooth its output distribution via
temperature scaling following a descending schedule. Our approach rivals the
performance of state-of-the-art ELECTRA-style pre-training methods, while
significantly eliminating the computation and memory cost brought by the joint
training of the auxiliary model. Our method also reduces the sensitivity to
hyper-parameters and enhances the pre-training stability. | cs.CL | Fast-ELECTRA for Efficient Pre-training |
2023-10-11T09:53:17Z | http://arxiv.org/pdf/2310.07345v1 | http://arxiv.org/abs/2310.07345v1 | Zijian Yang, Wei Zhou, Ralf Schlüter, Hermann Ney | 2023-10-11T09:53:17Z | cs.CL, cs.SD, eess.AS | In this work, we investigate the effect of language models (LMs) with
different context lengths and label units (phoneme vs. word) used in sequence
discriminative training for phoneme-based neural transducers. Both lattice-free
and N-best-list approaches are examined. For lattice-free methods with
phoneme-level LMs, we propose a method to approximate the context history to
employ LMs with full-context dependency. This approximation can be extended to
arbitrary context length and enables the usage of word-level LMs in
lattice-free methods. Moreover, a systematic comparison is conducted across
lattice-free and N-best-list-based methods. Experimental results on Librispeech
show that using the word-level LM in training outperforms the phoneme-level LM.
Besides, we find that the context size of the LM used for probability
computation has a limited effect on performance. Moreover, our results reveal
the pivotal importance of the hypothesis space quality in sequence
discriminative training. | cs.CL | Investigating the Effect of Language Models in Sequence Discriminative Training for Neural Transducers |
2023-10-11T09:46:32Z | http://arxiv.org/pdf/2310.07343v1 | http://arxiv.org/abs/2310.07343v1 | Zihan Zhang, Meng Fang, Ling Chen, Mohammad-Reza Namazi-Rad, Jun Wang | 2023-10-11T09:46:32Z | cs.CL | Although large language models (LLMs) are impressive in solving various
tasks, they can quickly be outdated after deployment. Maintaining their
up-to-date status is a pressing concern in the current era. This paper provides
a comprehensive review of recent advances in aligning LLMs with the
ever-changing world knowledge without re-training from scratch. We categorize
research works systemically and provide in-depth comparisons and discussion. We
also discuss existing challenges and highlight future directions to facilitate
research in this field. We release the paper list at
https://github.com/hyintell/awesome-refreshing-llms | cs.CL | How Do Large Language Models Capture the Ever-changing World Knowledge? A Review of Recent Advances |
2023-10-11T09:18:09Z | http://arxiv.org/pdf/2310.07328v2 | http://arxiv.org/abs/2310.07328v2 | Qingyi Si, Tong Wang, Zheng Lin, Xu Zhang, Yanan Cao, Weiping Wang | 2023-10-20T08:02:09Z | cs.CL, cs.AI | The success of ChatGPT validates the potential of large language models
(LLMs) in artificial general intelligence (AGI). Subsequently, the release of
LLMs has sparked the open-source community's interest in instruction-tuning,
which is deemed to accelerate ChatGPT's replication process. However, research
on instruction-tuning LLMs in Chinese, the world's most spoken language, is
still in its early stages. Therefore, this paper makes an in-depth empirical
study of instruction-tuning LLMs in Chinese, which can serve as a cookbook that
provides valuable findings for effectively customizing LLMs that can better
respond to Chinese instructions. Specifically, we systematically explore the
impact of LLM bases, parameter-efficient methods, instruction data types, which
are the three most important elements for instruction-tuning. Besides, we also
conduct experiment to study the impact of other factors, e.g., chain-of-thought
data and human-value alignment. We hope that this empirical study can make a
modest contribution to the open Chinese version of ChatGPT. This paper will
release a powerful Chinese LLMs that is comparable to ChatGLM. The code and
data are available at https://github.com/PhoebusSi/Alpaca-CoT. | cs.CL | An Empirical Study of Instruction-tuning Large Language Models in Chinese |
2023-10-11T09:09:55Z | http://arxiv.org/pdf/2310.07321v2 | http://arxiv.org/abs/2310.07321v2 | Amin Dada, Aokun Chen, Cheng Peng, Kaleb E Smith, Ahmad Idrissi-Yaghir, Constantin Marc Seibold, Jianning Li, Lars Heiliger, Xi Yang, Christoph M. Friedrich, Daniel Truhn, Jan Egger, Jiang Bian, Jens Kleesiek, Yonghui Wu | 2023-10-13T14:24:31Z | cs.CL, cs.AI, cs.LG | Traditionally, large language models have been either trained on general web
crawls or domain-specific data. However, recent successes of generative large
language models, have shed light on the benefits of cross-domain datasets. To
examine the significance of prioritizing data diversity over quality, we
present a German dataset comprising texts from five domains, along with another
dataset aimed at containing high-quality data. Through training a series of
models ranging between 122M and 750M parameters on both datasets, we conduct a
comprehensive benchmark on multiple downstream tasks. Our findings demonstrate
that the models trained on the cross-domain dataset outperform those trained on
quality data alone, leading to improvements up to $4.45\%$ over the previous
state-of-the-art. The models are available at
https://huggingface.co/ikim-uk-essen | cs.CL | On the Impact of Cross-Domain Data on German Language Models |
2023-10-11T08:40:06Z | http://arxiv.org/pdf/2310.07306v1 | http://arxiv.org/abs/2310.07306v1 | Aditi Kanwar, Aditi Seetha, Satyendra Singh Chouhan, Rajdeep Niyogi | 2023-10-11T08:40:06Z | cs.LG, cs.CL, ACM-class: F.2.2, I.2.7 | This paper presents a Soft Labeling and Noisy Mixup-based open intent
classification model (SNOiC). Most of the previous works have used
threshold-based methods to identify open intents, which are prone to
overfitting and may produce biased predictions. Additionally, the need for more
available data for an open intent class presents another limitation for these
existing models. SNOiC combines Soft Labeling and Noisy Mixup strategies to
reduce the biasing and generate pseudo-data for open intent class. The
experimental results on four benchmark datasets show that the SNOiC model
achieves a minimum and maximum performance of 68.72\% and 94.71\%,
respectively, in identifying open intents. Moreover, compared to
state-of-the-art models, the SNOiC model improves the performance of
identifying open intents by 0.93\% (minimum) and 12.76\% (maximum). The model's
efficacy is further established by analyzing various parameters used in the
proposed model. An ablation study is also conducted, which involves creating
three model variants to validate the effectiveness of the SNOiC model. | cs.LG | SNOiC: Soft Labeling and Noisy Mixup based Open Intent Classification Model |
2023-10-11T08:36:43Z | http://arxiv.org/pdf/2310.07301v2 | http://arxiv.org/abs/2310.07301v2 | Yuchong Sun, Che Liu, Kun Zhou, Jinwen Huang, Ruihua Song, Wayne Xin Zhao, Fuzheng Zhang, Di Zhang, Kun Gai | 2024-05-23T09:32:52Z | cs.CL | Humans often interact with large language models (LLMs) in multi-turn
interaction to obtain desired answers or more information. However, most
existing studies overlook the multi-turn instruction following ability of LLMs,
in terms of training dataset, training method, and evaluation benchmark. In
this paper, we introduce Parrot, a solution aiming to enhance multi-turn
instruction following for LLMs. First, we introduce an efficient but effective
method for collecting multi-turn instructions that feature human-like queries,
such as anaphora and ellipsis. Second, we propose a context-aware preference
optimization strategy to further enhance LLMs for complex queries in multi-turn
interaction. Moreover, to quantitatively evaluate LLMs in multi-turn
instruction following, we manually build a multi-turn benchmark derived from
existing ones. Extensive experiments show that Parrot improves current LLMs by
up to 7.2% in multi-turn instruction following. Our dataset and codes will be
open-sourced to facilitate future research. | cs.CL | Parrot: Enhancing Multi-Turn Instruction Following for Large Language Models |
2023-10-11T08:33:23Z | http://arxiv.org/pdf/2310.07299v1 | http://arxiv.org/abs/2310.07299v1 | Yue Zhang, Leyang Cui, Enbo Zhao, Wei Bi, Shuming Shi | 2023-10-11T08:33:23Z | cs.CL, cs.AI | Grammatical Error Correction (GEC) systems play a vital role in assisting
people with their daily writing tasks. However, users may sometimes come across
a GEC system that initially performs well but fails to correct errors when the
inputs are slightly modified. To ensure an ideal user experience, a reliable
GEC system should have the ability to provide consistent and accurate
suggestions when encountering irrelevant context perturbations, which we refer
to as context robustness. In this paper, we introduce RobustGEC, a benchmark
designed to evaluate the context robustness of GEC systems. RobustGEC comprises
5,000 GEC cases, each with one original error-correct sentence pair and five
variants carefully devised by human annotators. Utilizing RobustGEC, we reveal
that state-of-the-art GEC systems still lack sufficient robustness against
context perturbations. In addition, we propose a simple yet effective method
for remitting this issue. | cs.CL | RobustGEC: Robust Grammatical Error Correction Against Subtle Context Perturbation |
2023-10-11T08:22:37Z | http://arxiv.org/pdf/2310.07289v1 | http://arxiv.org/abs/2310.07289v1 | Liang Chen, Yang Deng, Yatao Bian, Zeyu Qin, Bingzhe Wu, Tat-Seng Chua, Kam-Fai Wong | 2023-10-11T08:22:37Z | cs.CL | Large language models (LLMs) outperform information retrieval techniques for
downstream knowledge-intensive tasks when being prompted to generate world
knowledge. However, community concerns abound regarding the factuality and
potential implications of using this uncensored knowledge. In light of this, we
introduce CONNER, a COmpreheNsive kNowledge Evaluation fRamework, designed to
systematically and automatically evaluate generated knowledge from six
important perspectives -- Factuality, Relevance, Coherence, Informativeness,
Helpfulness and Validity. We conduct an extensive empirical analysis of the
generated knowledge from three different types of LLMs on two widely studied
knowledge-intensive tasks, i.e., open-domain question answering and
knowledge-grounded dialogue. Surprisingly, our study reveals that the
factuality of generated knowledge, even if lower, does not significantly hinder
downstream tasks. Instead, the relevance and coherence of the outputs are more
important than small factual mistakes. Further, we show how to use CONNER to
improve knowledge-intensive tasks by designing two strategies: Prompt
Engineering and Knowledge Selection. Our evaluation code and LLM-generated
knowledge with human annotations will be released to facilitate future
research. | cs.CL | Beyond Factuality: A Comprehensive Evaluation of Large Language Models as Knowledge Generators |
2023-10-11T08:17:54Z | http://arxiv.org/pdf/2310.07284v3 | http://arxiv.org/abs/2310.07284v3 | Xiang Hao, Jibin Wu, Jianwei Yu, Chenglin Xu, Kay Chen Tan | 2023-10-15T03:58:29Z | eess.AS, cs.CL | Humans possess an extraordinary ability to selectively focus on the sound
source of interest amidst complex acoustic environments, commonly referred to
as cocktail party scenarios. In an attempt to replicate this remarkable
auditory attention capability in machines, target speaker extraction (TSE)
models have been developed. These models leverage the pre-registered cues of
the target speaker to extract the sound source of interest. However, the
effectiveness of these models is hindered in real-world scenarios due to the
unreliable or even absence of pre-registered cues. To address this limitation,
this study investigates the integration of natural language description to
enhance the feasibility, controllability, and performance of existing TSE
models. Specifically, we propose a model named LLM-TSE, wherein a large
language model (LLM) extracts useful semantic cues from the user's typed text
input. These cues can serve as independent extraction cues, task selectors to
control the TSE process or complement the pre-registered cues. Our experimental
results demonstrate competitive performance when only text-based cues are
presented, the effectiveness of using input text as a task selector, and a new
state-of-the-art when combining text-based cues with pre-registered cues. To
our knowledge, this is the first study to successfully incorporate LLMs to
guide target speaker extraction, which can be a cornerstone for cocktail party
problem research. | eess.AS | Typing to Listen at the Cocktail Party: Text-Guided Target Speaker Extraction |
2023-10-11T08:16:35Z | http://arxiv.org/pdf/2310.07282v2 | http://arxiv.org/abs/2310.07282v2 | Shyni Sharaf, V. S. Anoop | 2023-10-12T07:53:53Z | cs.AI, cs.CL | This paper conducts a comprehensive investigation into applying large
language models, particularly on BioBERT, in healthcare. It begins with
thoroughly examining previous natural language processing (NLP) approaches in
healthcare, shedding light on the limitations and challenges these methods
face. Following that, this research explores the path that led to the
incorporation of BioBERT into healthcare applications, highlighting its
suitability for addressing the specific requirements of tasks related to
biomedical text mining. The analysis outlines a systematic methodology for
fine-tuning BioBERT to meet the unique needs of the healthcare domain. This
approach includes various components, including the gathering of data from a
wide range of healthcare sources, data annotation for tasks like identifying
medical entities and categorizing them, and the application of specialized
preprocessing techniques tailored to handle the complexities found in
biomedical texts. Additionally, the paper covers aspects related to model
evaluation, with a focus on healthcare benchmarks and functions like processing
of natural language in biomedical, question-answering, clinical document
classification, and medical entity recognition. It explores techniques to
improve the model's interpretability and validates its performance compared to
existing healthcare-focused language models. The paper thoroughly examines
ethical considerations, particularly patient privacy and data security. It
highlights the benefits of incorporating BioBERT into healthcare contexts,
including enhanced clinical decision support and more efficient information
retrieval. Nevertheless, it acknowledges the impediments and complexities of
this integration, encompassing concerns regarding data privacy, transparency,
resource-intensive requirements, and the necessity for model customization to
align with diverse healthcare domains. | cs.AI | An Analysis on Large Language Models in Healthcare: A Case Study of BioBERT |
2023-10-11T08:07:22Z | http://arxiv.org/pdf/2310.07279v1 | http://arxiv.org/abs/2310.07279v1 | Jarod Duret, Benjamin O'Brien, Yannick Estève, Titouan Parcollet | 2023-10-11T08:07:22Z | cs.SD, cs.CL, eess.AS | Textless speech-to-speech translation systems are rapidly advancing, thanks
to the integration of self-supervised learning techniques. However, existing
state-of-the-art systems fall short when it comes to capturing and transferring
expressivity accurately across different languages. Expressivity plays a vital
role in conveying emotions, nuances, and cultural subtleties, thereby enhancing
communication across diverse languages. To address this issue this study
presents a novel method that operates at the discrete speech unit level and
leverages multilingual emotion embeddings to capture language-agnostic
information. Specifically, we demonstrate how these embeddings can be used to
effectively predict the pitch and duration of speech units in the target
language. Through objective and subjective experiments conducted on a
French-to-English translation task, our findings highlight the superior
expressivity transfer achieved by our approach compared to current
state-of-the-art systems. | cs.SD | Enhancing expressivity transfer in textless speech-to-speech translation |
2023-10-11T07:57:08Z | http://arxiv.org/pdf/2310.07276v3 | http://arxiv.org/abs/2310.07276v3 | Qizhi Pei, Wei Zhang, Jinhua Zhu, Kehan Wu, Kaiyuan Gao, Lijun Wu, Yingce Xia, Rui Yan | 2024-01-29T03:34:14Z | cs.CL, cs.AI, cs.LG, q-bio.BM | Recent advancements in biological research leverage the integration of
molecules, proteins, and natural language to enhance drug discovery. However,
current models exhibit several limitations, such as the generation of invalid
molecular SMILES, underutilization of contextual information, and equal
treatment of structured and unstructured knowledge. To address these issues, we
propose $\mathbf{BioT5}$, a comprehensive pre-training framework that enriches
cross-modal integration in biology with chemical knowledge and natural language
associations. $\mathbf{BioT5}$ utilizes SELFIES for $100%$ robust molecular
representations and extracts knowledge from the surrounding context of
bio-entities in unstructured biological literature. Furthermore,
$\mathbf{BioT5}$ distinguishes between structured and unstructured knowledge,
leading to more effective utilization of information. After fine-tuning, BioT5
shows superior performance across a wide range of tasks, demonstrating its
strong capability of capturing underlying relations and properties of
bio-entities. Our code is available at
$\href{https://github.com/QizhiPei/BioT5}{Github}$. | cs.CL | BioT5: Enriching Cross-modal Integration in Biology with Chemical Knowledge and Natural Language Associations |
2023-10-11T07:27:34Z | http://arxiv.org/pdf/2310.07251v1 | http://arxiv.org/abs/2310.07251v1 | Abhinav Rao, Aditi Khandelwal, Kumar Tanmay, Utkarsh Agarwal, Monojit Choudhury | 2023-10-11T07:27:34Z | cs.CL, cs.AI | In this position paper, we argue that instead of morally aligning LLMs to
specific set of ethical principles, we should infuse generic ethical reasoning
capabilities into them so that they can handle value pluralism at a global
scale. When provided with an ethical policy, an LLM should be capable of making
decisions that are ethically consistent to the policy. We develop a framework
that integrates moral dilemmas with moral principles pertaining to different
foramlisms of normative ethics, and at different levels of abstractions.
Initial experiments with GPT-x models shows that while GPT-4 is a nearly
perfect ethical reasoner, the models still have bias towards the moral values
of Western and English speaking societies. | cs.CL | Ethical Reasoning over Moral Alignment: A Case and Framework for In-Context Ethical Policies in LLMs |
2023-10-11T06:26:19Z | http://arxiv.org/pdf/2310.07225v2 | http://arxiv.org/abs/2310.07225v2 | Andrew M. Bean, Karolina Korgul, Felix Krones, Robert McCraith, Adam Mahdi | 2024-03-09T23:26:34Z | cs.CL | With the rapid development of new large language models (LLMs), each claiming
to surpass previous models, an overall picture of medical LLM research can be
elusive. To address this challenge, we benchmark a range of top LLMs and
identify consistent patterns which appear across models. We test $8$ well-known
LLMs on $874$ newly collected questions from Polish medical licensing exams.
For each question, we score each model on the top-1 accuracy and the
distribution of probabilities assigned. We then compare with factors including
question difficulty for humans, question length, and the scores of the other
models. LLM accuracies were positively correlated pairwise ($0.29$ to $0.62$).
Model performance was also correlated with human performance ($0.07$ to
$0.16$), but negatively correlated to the difference between the question-level
accuracy of top-scoring and bottom-scoring humans ($-0.16$ to $-0.23$). The top
output probability and question length were positive and negative predictors of
accuracy respectively (p $< 0.05$). The top scoring LLM, GPT-4 Turbo, scored
$82\%$, followed by Med42, PaLM 2, Mixtral and GPT-3.5 around $63\%$. We found
evidence of similarities between models in which questions they answer
correctly, as well as similarities with human test takers. Larger models
typically performed better, but differences in training methods were also
highly impactful. Model accuracy was positively correlated with confidence, but
negatively correlated with question length. We expect that similar training
methods will lead these patterns to persist across future models. These
patterns can therefore aid medical experts in forming expectations about LLMs
as a category to support application research. | cs.CL | Exploring the landscape of large language models in medical question answering |
2023-10-11T04:30:18Z | http://arxiv.org/pdf/2310.07188v1 | http://arxiv.org/abs/2310.07188v1 | Jiamin Li, Qiang Su, Yitao Yang, Yimin Jiang, Cong Wang, Hong Xu | 2023-10-11T04:30:18Z | cs.CL, cs.AI | Large language models, such as OpenAI's ChatGPT, have demonstrated
exceptional language understanding capabilities in various NLP tasks. Sparsely
activated mixture-of-experts (MoE) has emerged as a promising solution for
scaling models while maintaining a constant number of computational operations.
Existing MoE model adopts a fixed gating network where each token is computed
by the same number of experts. However, this approach contradicts our intuition
that the tokens in each sequence vary in terms of their linguistic complexity
and, consequently, require different computational costs. Little is discussed
in prior research on the trade-off between computation per token and model
performance. This paper introduces adaptive gating in MoE, a flexible training
strategy that allows tokens to be processed by a variable number of experts
based on expert probability distribution. The proposed framework preserves
sparsity while improving training efficiency. Additionally, curriculum learning
is leveraged to further reduce training time. Extensive experiments on diverse
NLP tasks show that adaptive gating reduces at most 22.5% training time while
maintaining inference quality. Moreover, we conduct a comprehensive analysis of
the routing decisions and present our insights when adaptive gating is used. | cs.CL | Adaptive Gating in Mixture-of-Experts based Language Models |
2023-10-11T04:03:42Z | http://arxiv.org/pdf/2310.07177v4 | http://arxiv.org/abs/2310.07177v4 | Xiaoxuan Liu, Lanxiang Hu, Peter Bailis, Alvin Cheung, Zhijie Deng, Ion Stoica, Hao Zhang | 2024-06-10T01:36:31Z | cs.AI, cs.CL, cs.LG | Speculative decoding is a pivotal technique to accelerate the inference of
large language models (LLMs) by employing a smaller draft model to predict the
target model's outputs. However, its efficacy can be limited due to the low
predictive accuracy of the draft model, particularly when faced with diverse
text inputs and a significant capability gap between the draft and target
models. We introduce online speculative decoding to address this challenge. The
main idea is to continuously update the (multiple) draft model(s) on observed
user query data. Adapting to query distribution mitigates the shifts between
the training distribution of the draft model and the query distribution,
enabling the draft model to more accurately predict the target model's outputs.
We develop a prototype of online speculative decoding based on knowledge
distillation and evaluate it using both synthetic and real query data. The
results show a substantial increase in the token acceptance rate by 0.1 to
0.65, bringing 1.42x to 2.17x latency reduction. Our code is available at
https://github.com/LiuXiaoxuanPKU/OSD. | cs.AI | Online Speculative Decoding |
2023-10-11T03:39:46Z | http://arxiv.org/pdf/2310.07170v1 | http://arxiv.org/abs/2310.07170v1 | Tatsuya Ide, Eiki Murata, Daisuke Kawahara, Takato Yamazaki, Shengzhe Li, Kenta Shinzato, Toshinori Sato | 2023-10-11T03:39:46Z | cs.CL | Despite the remarkable progress in natural language understanding with
pretrained Transformers, neural language models often do not handle commonsense
knowledge well. Toward commonsense-aware models, there have been attempts to
obtain knowledge, ranging from automatic acquisition to crowdsourcing. However,
it is difficult to obtain a high-quality knowledge base at a low cost,
especially from scratch. In this paper, we propose PHALM, a method of building
a knowledge graph from scratch, by prompting both crowdworkers and a large
language model (LLM). We used this method to build a Japanese event knowledge
graph and trained Japanese commonsense generation models. Experimental results
revealed the acceptability of the built graph and inferences generated by the
trained models. We also report the difference in prompting humans and an LLM.
Our code, data, and models are available at
github.com/nlp-waseda/comet-atomic-ja. | cs.CL | PHALM: Building a Knowledge Graph from Scratch by Prompting Humans and a Language Model |
2023-10-11T03:19:22Z | http://arxiv.org/pdf/2310.07161v2 | http://arxiv.org/abs/2310.07161v2 | Joseph Konan, Ojas Bhargave, Shikhar Agnihotri, Shuo Han, Yunyang Zeng, Ankit Shah, Bhiksha Raj | 2023-11-21T07:54:34Z | cs.SD, cs.CL, eess.AS | Within the ambit of VoIP (Voice over Internet Protocol) telecommunications,
the complexities introduced by acoustic transformations merit rigorous
analysis. This research, rooted in the exploration of proprietary sender-side
denoising effects, meticulously evaluates platforms such as Google Meets and
Zoom. The study draws upon the Deep Noise Suppression (DNS) 2020 dataset,
ensuring a structured examination tailored to various denoising settings and
receiver interfaces. A methodological novelty is introduced via the Oaxaca
decomposition, traditionally an econometric tool, repurposed herein to analyze
acoustic-phonetic perturbations within VoIP systems. To further ground the
implications of these transformations, psychoacoustic metrics, specifically
PESQ and STOI, were harnessed to furnish a comprehensive understanding of
speech alterations. Cumulatively, the insights garnered underscore the
intricate landscape of VoIP-influenced acoustic dynamics. In addition to the
primary findings, a multitude of metrics are reported, extending the research
purview. Moreover, out-of-domain benchmarking for both time and time-frequency
domain speech enhancement models is included, thereby enhancing the depth and
applicability of this inquiry. Repository:
github.com/deepology/VoIP-DNS-Challenge | cs.SD | Psychoacoustic Challenges Of Speech Enhancement On VoIP Platforms |
2023-10-11T03:01:42Z | http://arxiv.org/pdf/2310.07155v2 | http://arxiv.org/abs/2310.07155v2 | Shamik Roy, Dan Goldwasser | 2023-10-21T16:28:07Z | cs.CL | Social media has become a major driver of social change, by facilitating the
formation of online social movements. Automatically understanding the
perspectives driving the movement and the voices opposing it, is a challenging
task as annotated data is difficult to obtain. We propose a weakly supervised
graph-based approach that explicitly models perspectives in
#BackLivesMatter-related tweets. Our proposed approach utilizes a
social-linguistic representation of the data. We convert the text to a graph by
breaking it into structured elements and connect it with the social network of
authors, then structured prediction is done over the elements for identifying
perspectives. Our approach uses a small seed set of labeled examples. We
experiment with large language models for generating artificial training
examples, compare them to manual annotation, and find that it achieves
comparable performance. We perform quantitative and qualitative analyses using
a human-annotated test set. Our model outperforms multitask baselines by a
large margin, successfully characterizing the perspectives supporting and
opposing #BLM. | cs.CL | "A Tale of Two Movements": Identifying and Comparing Perspectives in #BlackLivesMatter and #BlueLivesMatter Movements-related Tweets using Weakly Supervised Graph-based Structured Prediction |
2023-10-11T02:47:40Z | http://arxiv.org/pdf/2310.07147v1 | http://arxiv.org/abs/2310.07147v1 | Zhikai Li, Xiaoxuan Liu, Banghua Zhu, Zhen Dong, Qingyi Gu, Kurt Keutzer | 2023-10-11T02:47:40Z | cs.CL, cs.LG | Large Language Models (LLMs) have showcased remarkable impacts across a wide
spectrum of natural language processing tasks. Fine-tuning these pre-trained
models on downstream datasets provides further significant performance gains,
but this process has been challenging due to its extraordinary resource
requirements. To this end, existing efforts focus on parameter-efficient
fine-tuning, which, unfortunately, fail to capitalize on the powerful potential
of full-parameter fine-tuning. In this work, we propose QFT, a novel Quantized
Full-parameter Tuning framework for LLMs that enables memory-efficient
fine-tuning without harming performance. Our framework incorporates two novel
ideas: (i) we adopt the efficient Lion optimizer, which only keeps track of the
momentum and has consistent update magnitudes for each parameter, an inherent
advantage for robust quantization; and (ii) we quantize all model states and
store them as integer values, and present a gradient flow and parameter update
scheme for the quantized weights. As a result, QFT reduces the model state
memory to 21% of the standard solution while achieving comparable performance,
e.g., tuning a LLaMA-7B model requires only <30GB of memory, satisfied by a
single A6000 GPU. | cs.CL | QFT: Quantized Full-parameter Tuning of LLMs with Affordable Resources |
2023-10-11T02:47:21Z | http://arxiv.org/pdf/2310.07146v1 | http://arxiv.org/abs/2310.07146v1 | Zhiyu Chen, Yujie Lu, William Yang Wang | 2023-10-11T02:47:21Z | cs.CL | Mental illness remains one of the most critical public health issues of our
time, due to the severe scarcity and accessibility limit of professionals.
Psychotherapy requires high-level expertise to conduct deep, complex reasoning
and analysis on the cognition modeling of the patients. In the era of Large
Language Models, we believe it is the right time to develop AI assistance for
computational psychotherapy. We study the task of cognitive distortion
detection and propose the Diagnosis of Thought (DoT) prompting. DoT performs
diagnosis on the patient's speech via three stages: subjectivity assessment to
separate the facts and the thoughts; contrastive reasoning to elicit the
reasoning processes supporting and contradicting the thoughts; and schema
analysis to summarize the cognition schemas. The generated diagnosis rationales
through the three stages are essential for assisting the professionals.
Experiments demonstrate that DoT obtains significant improvements over ChatGPT
for cognitive distortion detection, while generating high-quality rationales
approved by human experts. | cs.CL | Empowering Psychotherapy with Large Language Models: Cognitive Distortion Detection through Diagnosis of Thought Prompting |
2023-10-11T02:22:28Z | http://arxiv.org/pdf/2310.07137v1 | http://arxiv.org/abs/2310.07137v1 | Zhongfen Deng, Wei-Te Chen, Lei Chen, Philip S. Yu | 2023-10-11T02:22:28Z | cs.IR, cs.CL, cs.LG | Product attribute value extraction plays an important role for many
real-world applications in e-Commerce such as product search and
recommendation. Previous methods treat it as a sequence labeling task that
needs more annotation for position of values in the product text. This limits
their application to real-world scenario in which only attribute values are
weakly-annotated for each product without their position. Moreover, these
methods only use product text (i.e., product title and description) and do not
consider the semantic connection between the multiple attribute values of a
given product and its text, which can help attribute value extraction. In this
paper, we reformulate this task as a multi-label classification task that can
be applied for real-world scenario in which only annotation of attribute values
is available to train models (i.e., annotation of positional information of
attribute values is not available). We propose a classification model with
semantic matching and negative label sampling for attribute value extraction.
Semantic matching aims to capture semantic interactions between attribute
values of a given product and its text. Negative label sampling aims to enhance
the model's ability of distinguishing similar values belonging to the same
attribute. Experimental results on three subsets of a large real-world
e-Commerce dataset demonstrate the effectiveness and superiority of our
proposed model. | cs.IR | AE-smnsMLC: Multi-Label Classification with Semantic Matching and Negative Label Sampling for Product Attribute Value Extraction |
2023-10-11T02:16:12Z | http://arxiv.org/pdf/2310.07135v2 | http://arxiv.org/abs/2310.07135v2 | Shreya Havaldar, Matthew Pressimone, Eric Wong, Lyle Ungar | 2023-12-05T02:18:40Z | cs.CL | Understanding how styles differ across languages is advantageous for training
both humans and computers to generate culturally appropriate text. We introduce
an explanation framework to extract stylistic differences from multilingual LMs
and compare styles across languages. Our framework (1) generates comprehensive
style lexica in any language and (2) consolidates feature importances from LMs
into comparable lexical categories. We apply this framework to compare
politeness, creating the first holistic multilingual politeness dataset and
exploring how politeness varies across four languages. Our approach enables an
effective evaluation of how distinct linguistic categories contribute to
stylistic variations and provides interpretable insights into how people
communicate differently around the world. | cs.CL | Comparing Styles across Languages |
2023-10-11T01:03:42Z | http://arxiv.org/pdf/2310.07106v1 | http://arxiv.org/abs/2310.07106v1 | Ariel Goldstein, Eric Ham, Mariano Schain, Samuel Nastase, Zaid Zada, Avigail Dabush, Bobbi Aubrey, Harshvardhan Gazula, Amir Feder, Werner K Doyle, Sasha Devore, Patricia Dugan, Daniel Friedman, Roi Reichart, Michael Brenner, Avinatan Hassidim, Orrin Devinsky, Adeen Flinker, Omer Levy, Uri Hasson | 2023-10-11T01:03:42Z | cs.CL, cs.AI, cs.LG, q-bio.NC | Deep Language Models (DLMs) provide a novel computational paradigm for
understanding the mechanisms of natural language processing in the human brain.
Unlike traditional psycholinguistic models, DLMs use layered sequences of
continuous numerical vectors to represent words and context, allowing a
plethora of emerging applications such as human-like text generation. In this
paper we show evidence that the layered hierarchy of DLMs may be used to model
the temporal dynamics of language comprehension in the brain by demonstrating a
strong correlation between DLM layer depth and the time at which layers are
most predictive of the human brain. Our ability to temporally resolve
individual layers benefits from our use of electrocorticography (ECoG) data,
which has a much higher temporal resolution than noninvasive methods like fMRI.
Using ECoG, we record neural activity from participants listening to a
30-minute narrative while also feeding the same narrative to a high-performing
DLM (GPT2-XL). We then extract contextual embeddings from the different layers
of the DLM and use linear encoding models to predict neural activity. We first
focus on the Inferior Frontal Gyrus (IFG, or Broca's area) and then extend our
model to track the increasing temporal receptive window along the linguistic
processing hierarchy from auditory to syntactic and semantic areas. Our results
reveal a connection between human language processing and DLMs, with the DLM's
layer-by-layer accumulation of contextual information mirroring the timing of
neural activity in high-order language areas. | cs.CL | The Temporal Structure of Language Processing in the Human Brain Corresponds to The Layered Hierarchy of Deep Language Models |
2023-10-11T00:38:57Z | http://arxiv.org/pdf/2310.07096v1 | http://arxiv.org/abs/2310.07096v1 | Shawn Tan, Yikang Shen, Zhenfang Chen, Aaron Courville, Chuang Gan | 2023-10-11T00:38:57Z | cs.CL, cs.AI | The Universal Transformer (UT) is a variant of the Transformer that shares
parameters across its layers. Empirical evidence shows that UTs have better
compositional generalization than Vanilla Transformers (VTs) in formal language
tasks. The parameter-sharing also affords it better parameter efficiency than
VTs. Despite its many advantages, scaling UT parameters is much more compute
and memory intensive than scaling up a VT. This paper proposes the Sparse
Universal Transformer (SUT), which leverages Sparse Mixture of Experts (SMoE)
and a new stick-breaking-based dynamic halting mechanism to reduce UT's
computation complexity while retaining its parameter efficiency and
generalization ability. Experiments show that SUT achieves the same performance
as strong baseline models while only using half computation and parameters on
WMT'14 and strong generalization results on formal language tasks (Logical
inference and CFQ). The new halting mechanism also enables around 50\%
reduction in computation during inference with very little performance decrease
on formal language tasks. | cs.CL | Sparse Universal Transformer |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.