Authors: Jonathan Pei, Kevin Yang, Dan Klein
We propose Prefix-Adaptive Decoding (PREADD), a flexible method for controlled text generation. Unlike existing methods that use auxiliary expert models to control for attributes, PREADD does not require an external model, instead relying on linearly combining output logits from multiple prompts. Specifically, PREADD contrasts the output logits generated using a raw prompt against those generated using a prefix-prepended prompt, enabling both positive and negative control with respect to any attribute encapsulated by the prefix. We evaluate PREADD on three tasks -- toxic output mitigation, gender bias reduction, and sentiment control -- and find that PREADD outperforms not only prompting baselines, but also an auxiliary-expert control method, by 12% or more in relative gain on our main metrics for each task.
Authors: Clayton Fields, Casey Kennington
Vision language tasks, such as answering questions about or generating captions that describe an image, are difficult tasks for computers to perform. A relatively recent body of research has adapted the pretrained transformer architecture introduced in \citet{vaswani2017attention} to vision language modeling. Transformer models have greatly improved performance and versatility over previous vision language models. They do so by pretraining models on a large generic datasets and transferring their learning to new tasks with minor changes in architecture and parameter values. This type of transfer learning has become the standard modeling practice in both natural language processing and computer vision. Vision language transformers offer the promise of producing similar advancements in tasks which require both vision and language. In this paper, we provide a broad synthesis of the currently available research on vision language transformer models and offer some analysis of their strengths, limitations and some open questions that remain.
Authors: Enfa George, Mihai Surdeanu
We introduce SexTok, a multi-modal dataset composed of TikTok videos labeled as sexually suggestive (from the annotator's point of view), sex-educational content, or neither. Such a dataset is necessary to address the challenge of distinguishing between sexually suggestive content and virtual sex education videos on TikTok. Children's exposure to sexually suggestive videos has been shown to have adversarial effects on their development. Meanwhile, virtual sex education, especially on subjects that are more relevant to the LGBTQIA+ community, is very valuable. The platform's current system removes or penalizes some of both types of videos, even though they serve different purposes. Our dataset contains video URLs, and it is also audio transcribed. To validate its importance, we explore two transformer-based models for classifying the videos. Our preliminary results suggest that the task of distinguishing between these types of videos is learnable but challenging. These experiments suggest that this dataset is meaningful and invites further study on the subject.
Authors: Aref Farhadipour, Hadi Veisi
Dysarthria is a disability that causes a disturbance in the human speech system and reduces the quality and intelligibility of a person's speech. Because of this effect, the normal speech processing systems can not work properly on impaired speech. This disability is usually associated with physical disabilities. Therefore, designing a system that can perform some tasks by receiving voice commands in the smart home can be a significant achievement. In this work, we introduce gammatonegram as an effective method to represent audio files with discriminative details, which is used as input for the convolutional neural network. On the other word, we convert each speech file into an image and propose image recognition system to classify speech in different scenarios. Proposed CNN is based on the transfer learning method on the pre-trained Alexnet. In this research, the efficiency of the proposed system for speech recognition, speaker identification, and intelligibility assessment is evaluated. According to the results on the UA dataset, the proposed speech recognition system achieved 91.29% accuracy in speaker-dependent mode, the speaker identification system acquired 87.74% accuracy in text-dependent mode, and the intelligibility assessment system achieved 96.47% accuracy in two-class mode. Finally, we propose a multi-network speech recognition system that works fully automatically. This system is located in a cascade arrangement with the two-class intelligibility assessment system, and the output of this system activates each one of the speech recognition networks. This architecture achieves an accuracy of 92.3% WRR. The source code of this paper is available.
Authors: Siddharth Khincha, Chelsi Jain, Vivek Gupta, Tushar Kataria, Shuo Zhang
Information Synchronization of semi-structured data across languages is challenging. For instance, Wikipedia tables in one language should be synchronized across languages. To address this problem, we introduce a new dataset InfoSyncC and a two-step method for tabular synchronization. InfoSync contains 100K entity-centric tables (Wikipedia Infoboxes) across 14 languages, of which a subset (3.5K pairs) are manually annotated. The proposed method includes 1) Information Alignment to map rows and 2) Information Update for updating missing/outdated information for aligned tables across multilingual tables. When evaluated on InfoSync, information alignment achieves an F1 score of 87.91 (en <-> non-en). To evaluate information updation, we perform human-assisted Wikipedia edits on Infoboxes for 603 table pairs. Our approach obtains an acceptance rate of 77.28% on Wikipedia, showing the effectiveness of the proposed method.
Authors: Roni Rabin, Alexandre Djerbetian, Roee Engelberg, Lidan Hackmon, Gal Elidan, Reut Tsarfaty, Amir Globerson
Human communication often involves information gaps between the interlocutors. For example, in an educational dialogue, a student often provides an answer that is incomplete, and there is a gap between this answer and the perfect one expected by the teacher. Successful dialogue then hinges on the teacher asking about this gap in an effective manner, thus creating a rich and interactive educational experience. We focus on the problem of generating such gap-focused questions (GFQs) automatically. We define the task, highlight key desired aspects of a good GFQ, and propose a model that satisfies these. Finally, we provide an evaluation by human annotators of our generated questions compared against human generated ones, demonstrating competitive performance.
Authors: Abhirut Gupta, Ananya B. Sai, Richard Sproat, Yuri Vasilevski, James S. Ren, Ambarish Jash, Sukhdeep S. Sodhi, Aravindan Raghuveer
A large number of people are forced to use the Web in a language they have low literacy in due to technology asymmetries. Written text in the second language (L2) from such users often contains a large number of errors that are influenced by their native language (L1). We propose a method to mine phoneme confusions (sounds in L2 that an L1 speaker is likely to conflate) for pairs of L1 and L2. These confusions are then plugged into a generative model (Bi-Phone) for synthetically producing corrupted L2 text. Through human evaluations, we show that Bi-Phone generates plausible corruptions that differ across L1s and also have widespread coverage on the Web. We also corrupt the popular language understanding benchmark SuperGLUE with our technique (FunGLUE for Phonetically Noised GLUE) and show that SoTA language understating models perform poorly. We also introduce a new phoneme prediction pre-training task which helps byte models to recover performance close to SuperGLUE. Finally, we also release the FunGLUE benchmark to promote further research in phonetically robust language models. To the best of our knowledge, FunGLUE is the first benchmark to introduce L1-L2 interactions in text.
Authors: Sara Papi, Peidong Wan, Junkun Chen, Jian Xue, Jinyu Li, Yashesh Gaur
In real-world applications, users often require both translations and transcriptions of speech to enhance their comprehension, particularly in streaming scenarios where incremental generation is necessary. This paper introduces a streaming Transformer-Transducer that jointly generates automatic speech recognition (ASR) and speech translation (ST) outputs using a single decoder. To produce ASR and ST content effectively with minimal latency, we propose a joint token-level serialized output training method that interleaves source and target words by leveraging an off-the-shelf textual aligner. Experiments in monolingual (it-en) and multilingual (\{de,es,it\}-en) settings demonstrate that our approach achieves the best quality-latency balance. With an average ASR latency of 1s and ST latency of 1.3s, our model shows no degradation or even improves output quality compared to separate ASR and ST models, yielding an average improvement of 1.1 WER and 0.4 BLEU in the multilingual case.
Authors: Shiva Omrani Sabbaghi, Robert Wolfe, Aylin Caliskan
Language models are trained on large-scale corpora that embed implicit biases documented in psychology. Valence associations (pleasantness/unpleasantness) of social groups determine the biased attitudes towards groups and concepts in social cognition. Building on this established literature, we quantify how social groups are valenced in English language models using a sentence template that provides an intersectional context. We study biases related to age, education, gender, height, intelligence, literacy, race, religion, sex, sexual orientation, social class, and weight. We present a concept projection approach to capture the valence subspace through contextualized word embeddings of language models. Adapting the projection-based approach to embedding association tests that quantify bias, we find that language models exhibit the most biased attitudes against gender identity, social class, and sexual orientation signals in language. We find that the largest and better-performing model that we study is also more biased as it effectively captures bias embedded in sociocultural data. We validate the bias evaluation method by overperforming on an intrinsic valence evaluation task. The approach enables us to measure complex intersectional biases as they are known to manifest in the outputs and applications of language models that perpetuate historical biases. Moreover, our approach contributes to design justice as it studies the associations of groups underrepresented in language such as transgender and homosexual individuals.
Authors: Angel Felipe Magnossão de Paula, Paolo Rosso, Damiano Spina
This paper proposes a novelty approach to mitigate the negative transfer problem. In the field of machine learning, the common strategy is to apply the Single-Task Learning approach in order to train a supervised model to solve a specific task. Training a robust model requires a lot of data and a significant amount of computational resources, making this solution unfeasible in cases where data are unavailable or expensive to gather. Therefore another solution, based on the sharing of information between tasks, has been developed: Multi-Task Learning (MTL). Despite the recent developments regarding MTL, the problem of negative transfer has still to be solved. Negative transfer is a phenomenon that occurs when noisy information is shared between tasks, resulting in a drop in performance. This paper proposes a new approach to mitigate the negative transfer problem based on the task awareness concept. The proposed approach results in diminishing the negative transfer together with an improvement of performance over classic MTL solution. Moreover, the proposed approach has been implemented in two unified architectures to detect Sexism, Hate Speech, and Toxic Language in text comments. The proposed architectures set a new state-of-the-art both in EXIST-2021 and HatEval-2019 benchmarks.
Authors: Bruce W. Lee, BongSeok Yang, Jason Hyung-Jong Lee
Though discourse parsing can help multiple NLP fields, there has been no wide language model search done on implicit discourse relation classification. This hinders researchers from fully utilizing public-available models in discourse analysis. This work is a straightforward, fine-tuned discourse performance comparison of seven pre-trained language models. We use PDTB-3, a popular discourse relation annotated dataset. Through our model search, we raise SOTA to 0.671 ACC and obtain novel observations. Some are contrary to what has been reported before (Shi and Demberg, 2019b), that sentence-level pre-training objectives (NSP, SBO, SOP) generally fail to produce the best performing model for implicit discourse relation classification. Counterintuitively, similar-sized PLMs with MLM and full attention led to better performance.
Authors: Angel Felipe Magnossão de Paula, Giulia Rizzi, Elisabetta Fersini, Damiano Spina
With the increasing influence of social media platforms, it has become crucial to develop automated systems capable of detecting instances of sexism and other disrespectful and hateful behaviors to promote a more inclusive and respectful online environment. Nevertheless, these tasks are considerably challenging considering different hate categories and the author's intentions, especially under the learning with disagreements regime. This paper describes AI-UPV team's participation in the EXIST (sEXism Identification in Social neTworks) Lab at CLEF 2023. The proposed approach aims at addressing the task of sexism identification and characterization under the learning with disagreements paradigm by training directly from the data with disagreements, without using any aggregated label. Yet, performances considering both soft and hard evaluations are reported. The proposed system uses large language models (i.e., mBERT and XLM-RoBERTa) and ensemble strategies for sexism identification and classification in English and Spanish. In particular, our system is articulated in three different pipelines. The ensemble approach outperformed the individual large language models obtaining the best performances both adopting a soft and a hard label evaluation. This work describes the participation in all the three EXIST tasks, considering a soft evaluation, it obtained fourth place in Task 2 at EXIST and first place in Task 3, with the highest ICM-Soft of -2.32 and a normalized ICM-Soft of 0.79. The source code of our approaches is publicly available at https://github.com/AngelFelipeMP/Sexism-LLM-Learning-With-Disagreement.
Authors: Dean Ninalga
Recent work suggests that interpolating between the weights of two specialized language models can transfer knowledge between tasks in a way that multi-task learning cannot. However, very few have explored interpolation between more than two models, where each has a distinct knowledge base. In this paper, we introduce Derivative Free Weight-space Ensembling (DFWE), a new few-sample task transfer approach for open-domain dialogue. Our framework creates a set of diverse expert language models trained using a predefined set of source tasks. Next, we finetune each of the expert models on the target task, approaching the target task from several distinct knowledge bases. Finally, we linearly interpolate between the model weights using a gradient-free-optimization algorithm, to efficiently find a good interpolation weighting. We demonstrate the effectiveness of the method on FETA-Friends outperforming the standard pretrain-finetune approach.
Authors: Sarenne Wallbridge, Peter Bell, Catherine Lai
Speech is a fundamental means of communication that can be seen to provide two channels for transmitting information: the lexical channel of which words are said, and the non-lexical channel of how they are spoken. Both channels shape listener expectations of upcoming communication; however, directly quantifying their relative effect on expectations is challenging. Previous attempts require spoken variations of lexically-equivalent dialogue turns or conspicuous acoustic manipulations. This paper introduces a generalised paradigm to study the value of non-lexical information in dialogue across unconstrained lexical content. By quantifying the perceptual value of the non-lexical channel with both accuracy and entropy reduction, we show that non-lexical information produces a consistent effect on expectations of upcoming dialogue: even when it leads to poorer discriminative turn judgements than lexical content alone, it yields higher consensus among participants.
Authors: Benjamin Clavié, Guillaume Soulié
Understanding labour market dynamics requires accurately identifying the skills required for and possessed by the workforce. Automation techniques are increasingly being developed to support this effort. However, automatically extracting skills from job postings is challenging due to the vast number of existing skills. The ESCO (European Skills, Competences, Qualifications and Occupations) framework provides a useful reference, listing over 13,000 individual skills. However, skills extraction remains difficult and accurately matching job posts to the ESCO taxonomy is an open problem. In this work, we propose an end-to-end zero-shot system for skills extraction from job descriptions based on large language models (LLMs). We generate synthetic training data for the entirety of ESCO skills and train a classifier to extract skill mentions from job posts. We also employ a similarity retriever to generate skill candidates which are then re-ranked using a second LLM. Using synthetic data achieves an RP@10 score 10 points higher than previous distant supervision approaches. Adding GPT-4 re-ranking improves RP@10 by over 22 points over previous methods. We also show that Framing the task as mock programming when prompting the LLM can lead to better performance than natural language prompts, especially with weaker LLMs. We demonstrate the potential of integrating large language models at both ends of skills matching pipelines. Our approach requires no human annotations and achieve extremely promising results on skills extraction against ESCO.
Authors: Ipek Baris Schlicht, Lynn Khellaf, Defne Altiok
This paper describes our submission for the subjectivity detection task at the CheckThat! Lab. To tackle class imbalances in the task, we have generated additional training materials with GPT-3 models using prompts of different styles from a subjectivity checklist based on journalistic perspective. We used the extended training set to fine-tune language-specific transformer models. Our experiments in English, German and Turkish demonstrate that different subjective styles are effective across all languages. In addition, we observe that the style-based oversampling is better than paraphrasing in Turkish and English. Lastly, the GPT-3 models sometimes produce lacklustre results when generating style-based texts in non-English languages.
Authors: Björn Engelmann, Fabian Haak, Christin Katharina Kreutz, Narjes Nikzad Khasmakhi, Philipp Schaer
Reading levels are highly individual and can depend on a text's language, a person's cognitive abilities, or knowledge on a topic. Text simplification is the task of rephrasing a text to better cater to the abilities of a specific target reader group. Simplification of scientific abstracts helps non-experts to access the core information by bypassing formulations that require domain or expert knowledge. This is especially relevant for, e.g., cancer patients reading about novel treatment options. The SimpleText lab hosts the simplification of scientific abstracts for non-experts (Task 3) to advance this field. We contribute three runs employing out-of-the-box summarization models (two based on T5, one based on PEGASUS) and one run using ChatGPT with complex phrase identification.
Authors: S. Magalí López Cortez, Cassandra L. Jacobs
Time pressure and topic negotiation may impose constraints on how people leverage discourse relations (DRs) in spontaneous conversational contexts. In this work, we adapt a system of DRs for written language to spontaneous dialogue using crowdsourced annotations from novice annotators. We then test whether discourse relations are used differently across several types of multi-utterance contexts. We compare the patterns of DR annotation within and across speakers and within and across turns. Ultimately, we find that different discourse contexts produce distinct distributions of discourse relations, with single-turn annotations creating the most uncertainty for annotators. Additionally, we find that the discourse relation annotations are of sufficient quality to predict from embeddings of discourse units.
Authors: Ethan Gotlieb Wilcox, Tiago Pimentel, Clara Meister, Ryan Cotterell, Roger P. Levy
A fundamental result in psycholinguistics is that less predictable words take a longer time to process. One theoretical explanation for this finding is Surprisal Theory (Hale, 2001; Levy, 2008), which quantifies a word's predictability as its surprisal, i.e. its negative log-probability given a context. While evidence supporting the predictions of Surprisal Theory have been replicated widely, most have focused on a very narrow slice of data: native English speakers reading English texts. Indeed, no comprehensive multilingual analysis exists. We address this gap in the current literature by investigating the relationship between surprisal and reading times in eleven different languages, distributed across five language families. Deriving estimates from language models trained on monolingual and multilingual corpora, we test three predictions associated with surprisal theory: (i) whether surprisal is predictive of reading times; (ii) whether expected surprisal, i.e. contextual entropy, is predictive of reading times; (iii) and whether the linking function between surprisal and reading times is linear. We find that all three predictions are borne out crosslinguistically. By focusing on a more diverse set of languages, we argue that these results offer the most robust link to-date between information theory and incremental language processing across languages.
This research focuses on assessing the ability of large language models (LLMs) in representing geometries and their spatial relations. We utilize LLMs including GPT-2 and BERT to encode the well-known text (WKT) format of geometries and then feed their embeddings into classifiers and regressors to evaluate the effectiveness of the LLMs-generated embeddings for geometric attributes. The experiments demonstrate that while the LLMs-generated embeddings can preserve geometry types and capture some spatial relations (up to 73% accuracy), challenges remain in estimating numeric values and retrieving spatially related objects. This research highlights the need for improvement in terms of capturing the nuances and complexities of the underlying geospatial data and integrating domain knowledge to support various GeoAI applications using foundation models.
Authors: Shreyanth S
By combining the undecimated wavelet transform within a Word Embedded Semantic Marginal Autoencoder (WESMA), this research study provides a novel strategy for improving security measures and denoising multiple languages. The incorporation of these strategies is intended to address the issues of robustness, privacy, and multilingualism in data processing applications. The undecimated wavelet transform is used as a feature extraction tool to identify prominent language patterns and structural qualities in the input data. The proposed system may successfully capture significant information while preserving the temporal and geographical links within the data by employing this transform. This improves security measures by increasing the system's ability to detect abnormalities, discover hidden patterns, and distinguish between legitimate content and dangerous threats. The Word Embedded Semantic Marginal Autoencoder also functions as an intelligent framework for dimensionality and noise reduction. The autoencoder effectively learns the underlying semantics of the data and reduces noise components by exploiting word embeddings and semantic context. As a result, data quality and accuracy are increased in following processing stages. The suggested methodology is tested using a diversified dataset that includes several languages and security scenarios. The experimental results show that the proposed approach is effective in attaining security enhancement and denoising capabilities across multiple languages. The system is strong in dealing with linguistic variances, producing consistent outcomes regardless of the language used. Furthermore, incorporating the undecimated wavelet transform considerably improves the system's ability to efficiently address complex security concerns
Authors: Reagan Mozer, Aaron R. Kaufman, Leo A. Celi, Luke Miratrix
Text is a ubiquitous component of medical data, containing valuable information about patient characteristics and care that are often missing from structured chart data. Despite this richness, it is rarely used in clinical research, owing partly to its complexity. Using a large database of patient records and treatment histories accompanied by extensive notes by attendant physicians and nurses, we show how text data can be used to support causal inference with electronic health data in all stages, from conception and design to analysis and interpretation, with minimal additional effort. We focus on studies using matching for causal inference. We augment a classic matching analysis by incorporating text in three ways: by using text to supplement a multiple imputation procedure, we improve the fidelity of imputed values to handle missing data; by incorporating text in the matching stage, we strengthen the plausibility of the matching procedure; and by conditioning on text, we can estimate easily interpretable text-based heterogeneous treatment effects that may be stronger than those found across categories of structured covariates. Using these techniques, we hope to expand the scope of secondary analysis of clinical data to domains where quantitative data is of poor quality or nonexistent, but where text is available, such as in developing countries.
Authors: Jessica Echterhoff, An Yan, Julian McAuley
It is time-consuming to find the best product among many similar alternatives. Comparative sentences can help to contrast one item from others in a way that highlights important features of an item that stand out. Given reviews of one or multiple items and relevant item features, we generate comparative review sentences to aid users to find the best fit. Specifically, our model consists of three successive components in a transformer: (i) an item encoding module to encode an item for comparison, (ii) a comparison generation module that generates comparative sentences in an autoregressive manner, (iii) a novel decoding method for user personalization. We show that our pipeline generates fluent and diverse comparative sentences. We run experiments on the relevance and fidelity of our generated sentences in a human evaluation study and find that our algorithm creates comparative review sentences that are relevant and truthful.
Authors: Waseem AlShikh, Manhal Daaboul, Kirk Goddard, Brock Imel, Kiran Kamble, Parikshith Kulkarni, Melisa Russak
In this paper, we introduce the Instruction Following Score (IFS), a metric that detects language models' ability to follow instructions. The metric has a dual purpose. First, IFS can be used to distinguish between base and instruct models. We benchmark publicly available base and instruct models, and show that the ratio of well formatted responses to partial and full sentences can be an effective measure between those two model classes. Secondly, the metric can be used as an early stopping criteria for instruct tuning. We compute IFS for Supervised Fine-Tuning (SFT) of 7B and 13B LLaMA models, showing that models learn to follow instructions relatively early in the training process, and the further finetuning can result in changes in the underlying base model semantics. As an example of semantics change we show the objectivity of model predictions, as defined by an auxiliary metric ObjecQA. We show that in this particular case, semantic changes are the steepest when the IFS tends to plateau. We hope that decomposing instruct tuning into IFS and semantic factors starts a new trend in better controllable instruct tuning and opens possibilities for designing minimal instruct interfaces querying foundation models.
Authors: Chuanbo Hu, Bin Liu, Xin Li, Yanfang Ye
Social media platforms such as Instagram and Twitter have emerged as critical channels for drug marketing and illegal sale. Detecting and labeling online illicit drug trafficking activities becomes important in addressing this issue. However, the effectiveness of conventional supervised learning methods in detecting drug trafficking heavily relies on having access to substantial amounts of labeled data, while data annotation is time-consuming and resource-intensive. Furthermore, these models often face challenges in accurately identifying trafficking activities when drug dealers use deceptive language and euphemisms to avoid detection. To overcome this limitation, we conduct the first systematic study on leveraging large language models (LLMs), such as ChatGPT, to detect illicit drug trafficking activities on social media. We propose an analytical framework to compose \emph{knowledge-informed prompts}, which serve as the interface that humans can interact with and use LLMs to perform the detection task. Additionally, we design a Monte Carlo dropout based prompt optimization method to further to improve performance and interpretability. Our experimental findings demonstrate that the proposed framework outperforms other baseline language models in terms of drug trafficking detection accuracy, showing a remarkable improvement of nearly 12\%. By integrating prior knowledge and the proposed prompts, ChatGPT can effectively identify and label drug trafficking activities on social networks, even in the presence of deceptive language and euphemisms used by drug dealers to evade detection. The implications of our research extend to social networks, emphasizing the importance of incorporating prior knowledge and scenario-based prompts into analytical tools to improve online security and public safety.
Authors: Lakshmi Nair, Mikhail Bernadskiy, Arulselvan Madhavan, Craig Chan, Ayon Basumallik, Darius Bunandar
The recent rise of large language models (LLMs) has resulted in increased efforts towards running LLMs at reduced precision. Running LLMs at lower precision supports resource constraints and furthers their democratization, enabling users to run billion-parameter LLMs on their personal devices. To supplement this ongoing effort, we propose INT-FP-QSim: an open-source simulator that enables flexible evaluation of LLMs and vision transformers at various numerical precisions and formats. INT-FP-QSim leverages existing open-source repositories such as TensorRT, QPytorch and AIMET for a combined simulator that supports various floating point and integer formats. With the help of our simulator, we survey the impact of different numerical formats on the performance of LLMs and vision transformers at 4-bit weights and 4-bit or 8-bit activations. We also compare recently proposed methods like Adaptive Block Floating Point, SmoothQuant, GPTQ and RPTQ on the model performances. We hope INT-FP-QSim will enable researchers to flexibly simulate models at various precisions to support further research in quantization of LLMs and vision transformers.
Authors: Krishnapriya Vishnubhotla, Frank Rudzicz, Graeme Hirst, Adam Hammond
Current models for quotation attribution in literary novels assume varying levels of available information in their training and test data, which poses a challenge for in-the-wild inference. Here, we approach quotation attribution as a set of four interconnected sub-tasks: character identification, coreference resolution, quotation identification, and speaker attribution. We benchmark state-of-the-art models on each of these sub-tasks independently, using a large dataset of annotated coreferences and quotations in literary novels (the Project Dialogism Novel Corpus). We also train and evaluate models for the speaker attribution task in particular, showing that a simple sequential prediction model achieves accuracy scores on par with state-of-the-art models.
Authors: Tommaso Pegolotti, Elias Frantar, Dan Alistarh, Markus Püschel
We present ongoing work on a new automatic code generation approach for supporting quantized generative inference on LLMs such as LLaMA or OPT on off-the-shelf CPUs. Our approach is informed by the target architecture and a performance model, including both hardware characteristics and method-specific accuracy constraints. Results on CPU-based inference for LLaMA models show that our approach can lead to high performance and high accuracy, comparing favorably to the best existing open-source solution. A preliminary implementation is available at https://github.com/IST-DASLab/QIGen.
Authors: Clara Meister, Tiago Pimentel, Luca Malagutti, Ethan G. Wilcox, Ryan Cotterell
Sampling is a common strategy for generating text from probabilistic models, yet standard ancestral sampling often results in text that is incoherent or ungrammatical. To alleviate this issue, various modifications to a model's sampling distribution, such as nucleus or top-k sampling, have been introduced and are now ubiquitously used in language generation systems. We propose a unified framework for understanding these techniques, which we term sampling adapters. Sampling adapters often lead to qualitatively better text, which raises the question: From a formal perspective, how are they changing the (sub)word-level distributions of language generation models? And why do these local changes lead to higher-quality text? We argue that the shift they enforce can be viewed as a trade-off between precision and recall: while the model loses its ability to produce certain strings, its precision rate on desirable text increases. While this trade-off is not reflected in standard metrics of distribution quality (such as perplexity), we find that several precision-emphasizing measures indeed indicate that sampling adapters can lead to probability distributions more aligned with the true distribution. Further, these measures correlate with higher sequence-level quality scores, specifically, Mauve.
Authors: Barry Menglong Yao (1), Aditya Shah (1), Lichao Sun (2), Jin-Hee Cho (1), Lifu Huang (1) ((1) Virginia Tech, (2) Lehigh University)
We propose end-to-end multimodal fact-checking and explanation generation, where the input is a claim and a large collection of web sources, including articles, images, videos, and tweets, and the goal is to assess the truthfulness of the claim by retrieving relevant evidence and predicting a truthfulness label (e.g., support, refute or not enough information), and to generate a statement to summarize and explain the reasoning and ruling process. To support this research, we construct Mocheg, a large-scale dataset consisting of 15,601 claims where each claim is annotated with a truthfulness label and a ruling statement, and 33,880 textual paragraphs and 12,112 images in total as evidence. To establish baseline performances on Mocheg, we experiment with several state-of-the-art neural architectures on the three pipelined subtasks: multimodal evidence retrieval, claim verification, and explanation generation, and demonstrate that the performance of the state-of-the-art end-to-end multimodal fact-checking does not provide satisfactory outcomes. To the best of our knowledge, we are the first to build the benchmark dataset and solutions for end-to-end multimodal fact-checking and explanation generation. The dataset, source code and model checkpoints are available at https://github.com/VT-NLP/Mocheg.
Authors: Shengyao Zhuang, Houxing Ren, Linjun Shou, Jian Pei, Ming Gong, Guido Zuccon, Daxin Jiang
The Differentiable Search Index (DSI) is an emerging paradigm for information retrieval. Unlike traditional retrieval architectures where index and retrieval are two different and separate components, DSI uses a single transformer model to perform both indexing and retrieval.
In this paper, we identify and tackle an important issue of current DSI models: the data distribution mismatch that occurs between the DSI indexing and retrieval processes. Specifically, we argue that, at indexing, current DSI methods learn to build connections between the text of long documents and the identifier of the documents, but then retrieval of document identifiers is based on queries that are commonly much shorter than the indexed documents. This problem is further exacerbated when using DSI for cross-lingual retrieval, where document text and query text are in different languages.
To address this fundamental problem of current DSI models, we propose a simple yet effective indexing framework for DSI, called DSI-QG. When indexing, DSI-QG represents documents with a number of potentially relevant queries generated by a query generation model and re-ranked and filtered by a cross-encoder ranker. The presence of these queries at indexing allows the DSI models to connect a document identifier to a set of queries, hence mitigating data distribution mismatches present between the indexing and the retrieval phases. Empirical results on popular mono-lingual and cross-lingual passage retrieval datasets show that DSI-QG significantly outperforms the original DSI model.
Authors: Ahmed Sabir
This paper focuses on enhancing the captions generated by image-caption generation systems. We propose an approach for improving caption generation systems by choosing the most closely related output to the image rather than the most likely output produced by the model. Our model revises the language generation output beam search from a visual context perspective. We employ a visual semantic measure in a word and sentence level manner to match the proper caption to the related information in the image. The proposed approach can be applied to any caption system as a post-processing based method.
Authors: Joel Lamy-Poirier
We introduce Breadth-First Pipeline Parallelism, a novel training schedule which optimizes the combination of pipeline and data parallelism. Breadth-First Pipeline Parallelism lowers training time, cost and memory usage by combining a high GPU utilization with a small batch size per GPU, and by making use of fully sharded data parallelism. Experimentally, we observed an increase of up to 43% in training throughput for a 52 billion-parameter model using a small batch size per GPU compared to Megatron-LM, which would reduce the training time and cost by the same amount on a large GPU cluster.
Authors: Elias Stengel-Eskin, Benjamin Van Durme
Sequence generation models are increasingly being used to translate natural language into programs, i.e. to perform executable semantic parsing. The fact that semantic parsing aims to predict programs that can lead to executed actions in the real world motivates developing safe systems. This in turn makes measuring calibration -- a central component to safety -- particularly important. We investigate the calibration of popular generation models across four popular semantic parsing datasets, finding that it varies across models and datasets. We then analyze factors associated with calibration error and release new confidence-based challenge splits of two parsing datasets. To facilitate the inclusion of calibration in semantic parsing evaluations, we release a library for computing calibration metrics.
Authors: Ping Yu, Tianlu Wang, Olga Golovneva, Badr Alkhamissy, Gargi Ghosh, Mona Diab, Asli Celikyilmaz
Current large language models can perform reasonably well on complex tasks that require step-by-step reasoning with few-shot learning. Are these models applying reasoning skills they have learnt during pre-training and reason outside of their training context, or are they simply memorizing their training corpus at finer granularity and have learnt to better understand their context? To tease apart these possibilities, we introduce ALERT, a benchmark and suite of analyses for assessing language models' reasoning ability comparing pre-trained and finetuned models on complex tasks that require reasoning skills to solve. ALERT provides a test bed to asses any language model on fine-grained reasoning skills, which spans over 20 datasets and covers 10 different reasoning skills. We leverage ALERT to further investigate the role of finetuning. With extensive empirical analysis we find that language models learn more reasoning skills such as textual entailment, abductive reasoning, and analogical reasoning during finetuning stage compared to pretraining state. We also find that when language models are finetuned they tend to overfit to the prompt template, which hurts the robustness of models causing generalization problems.
Authors: Wenda Xu, Xian Qian, Mingxuan Wang, Lei Li, William Yang Wang
Is it possible to train a general metric for evaluating text generation quality without human annotated ratings? Existing learned metrics either perform unsatisfactorily across text generation tasks or require human ratings for training on specific tasks. In this paper, we propose SESCORE2, a self-supervised approach for training a model-based metric for text generation evaluation. The key concept is to synthesize realistic model mistakes by perturbing sentences retrieved from a corpus. The primary advantage of the SESCORE2 is its ease of extension to many other languages while providing reliable severity estimation. We evaluate SESCORE2 and previous methods on four text generation tasks across three languages. SESCORE2 outperforms unsupervised metric PRISM on four text generation evaluation benchmarks, with a Kendall improvement of 0.078. Surprisingly, SESCORE2 even outperforms the supervised BLEURT and COMET on multiple text generation tasks. The code and data are available at https://github.com/xu1998hz/SEScore2.
Authors: Siqi Ouyang, Rong Ye, Lei Li
End-to-end Speech Translation (E2E ST) aims to directly translate source speech into target text. Existing ST methods perform poorly when only extremely small speech-text data are available for training. We observe that an ST model's performance closely correlates with its embedding similarity between speech and source transcript. In this paper, we propose Word-Aligned COntrastive learning (WACO), a simple and effective method for extremely low-resource speech-to-text translation. Our key idea is bridging word-level representations for both speech and text modalities via contrastive learning. We evaluate WACO and other methods on the MuST-C dataset, a widely used ST benchmark, and on a low-resource direction Maltese-English from IWSLT 2023. Our experiments demonstrate that WACO outperforms the best baseline by 9+ BLEU points with only 1-hour parallel ST data. Code is available at https://github.com/owaski/WACO.
Authors: Yeskendir Koishekenov, Alexandre Berard, Vassilina Nikoulina
The recently released NLLB-200 is a set of multilingual Neural Machine Translation models that cover 202 languages. The largest model is based on a Mixture of Experts architecture and achieves SoTA results across many language pairs. It contains 54.5B parameters and requires at least four 32GB GPUs just for inference. In this work, we propose a pruning method that enables the removal of up to 80% of experts without further finetuning and with a negligible loss in translation quality, which makes it feasible to run the model on a single 32GB GPU. Further analysis suggests that our pruning metrics can identify language-specific experts.
Authors: Zekun Li, Baolin Peng, Pengcheng He, Michel Galley, Jianfeng Gao, Xifeng Yan
We introduce a novel prompting framework called Directional Stimulus Prompting for guiding black-box large language models (LLMs) toward desired outputs. The framework introduces a new component called directional stimulus into the prompt, providing more fine-grained guidance and control over LLMs. The directional stimulus serves as hints or cues for each input query to guide LLMs toward the desired output, such as keywords that the desired summary should include for summarization. We utilize a small tunable model (e.g., T5) to generate such directional stimulus for each query, allowing us to optimize black-box LLMs by optimizing a small policy model. This policy model can be trained through 1) supervised fine-tuning using labeled data and 2) reinforcement learning from offline or online rewards to explore directional stimulus that better aligns LLMs with desired behaviors. We evaluate our framework on summarization and dialogue response generation tasks. Experimental results show that our framework consistently improves ChatGPT's performance over standard prompting with a small collection of training data, and reinforcement learning further improves the performance. Notably, on the MultWOZ dataset, our framework enables ChatGPT to achieve a remarkable 41.4% improvement in its combined score with only 80 dialogues, matching or even surpassing the performance of some fully trained state-of-the-art models. We have made our code publicly available.
Authors: Karim Gamal, Ahmed Gaber, Hossam Amer
Federated learning is a growing field in the machine learning community due to its decentralized and private design. Model training in federated learning is distributed over multiple clients giving access to lots of client data while maintaining privacy. Then, a server aggregates the training done on these multiple clients without access to their data, which could be emojis widely used in any social media service and instant messaging platforms to express users' sentiments. This paper proposes federated learning-based multilingual emoji prediction in both clean and attack scenarios. Emoji prediction data have been crawled from both Twitter and SemEval emoji datasets. This data is used to train and evaluate different transformer model sizes including a sparsely activated transformer with either the assumption of clean data in all clients or poisoned data via label flipping attack in some clients. Experimental results on these models show that federated learning in either clean or attacked scenarios performs similarly to centralized training in multilingual emoji prediction on seen and unseen languages under different data sources and distributions. Our trained transformers perform better than other techniques on the SemEval emoji dataset in addition to the privacy as well as distributed benefits of federated learning.
Authors: Brian Yan, Jiatong Shi, Yun Tang, Hirofumi Inaguma, Yifan Peng, Siddharth Dalmia, Peter Polák, Patrick Fernandes, Dan Berrebbi, Tomoki Hayashi, Xiaohui Zhang, Zhaoheng Ni, Moto Hira, Soumi Maiti, Juan Pino, Shinji Watanabe
ESPnet-ST-v2 is a revamp of the open-source ESPnet-ST toolkit necessitated by the broadening interests of the spoken language translation community. ESPnet-ST-v2 supports 1) offline speech-to-text translation (ST), 2) simultaneous speech-to-text translation (SST), and 3) offline speech-to-speech translation (S2ST) -- each task is supported with a wide variety of approaches, differentiating ESPnet-ST-v2 from other open source spoken language translation toolkits. This toolkit offers state-of-the-art architectures such as transducers, hybrid CTC/attention, multi-decoders with searchable intermediates, time-synchronous blockwise CTC/attention, Translatotron models, and direct discrete unit models. In this paper, we describe the overall design, example models for each task, and performance benchmarking behind ESPnet-ST-v2, which is publicly available at https://github.com/espnet/espnet.
Authors: Ehsan Kamalloo, Xinyu Zhang, Odunayo Ogundepo, Nandan Thakur, David Alfonso-Hermelo, Mehdi Rezagholizadeh, Jimmy Lin
The ever-increasing size of language models curtails their widespread availability to the community, thereby galvanizing many companies into offering access to large language models through APIs. One particular type, suitable for dense retrieval, is a semantic embedding service that builds vector representations of input text. With a growing number of publicly available APIs, our goal in this paper is to analyze existing offerings in realistic retrieval scenarios, to assist practitioners and researchers in finding suitable services according to their needs. Specifically, we investigate the capabilities of existing semantic embedding APIs on domain generalization and multilingual retrieval. For this purpose, we evaluate these services on two standard benchmarks, BEIR and MIRACL. We find that re-ranking BM25 results using the APIs is a budget-friendly approach and is most effective in English, in contrast to the standard practice of employing them as first-stage retrievers. For non-English retrieval, re-ranking still improves the results, but a hybrid model with BM25 works best, albeit at a higher cost. We hope our work lays the groundwork for evaluating semantic embedding APIs that are critical in search and more broadly, for information access.
Authors: Ehsan Kamalloo, Nouha Dziri, Charles L. A. Clarke, Davood Rafiei
Lexical matching remains the de facto evaluation method for open-domain question answering (QA). Unfortunately, lexical matching fails completely when a plausible candidate answer does not appear in the list of gold answers, which is increasingly the case as we shift from extractive to generative models. The recent success of large language models (LLMs) for QA aggravates lexical matching failures since candidate answers become longer, thereby making matching with the gold answers even more challenging. Without accurate evaluation, the true progress in open-domain QA remains unknown. In this paper, we conduct a thorough analysis of various open-domain QA models, including LLMs, by manually evaluating their answers on a subset of NQ-open, a popular benchmark. Our assessments reveal that while the true performance of all models is significantly underestimated, the performance of the InstructGPT (zero-shot) LLM increases by nearly +60%, making it on par with existing top models, and the InstructGPT (few-shot) model actually achieves a new state-of-the-art on NQ-open. We also find that more than 50% of lexical matching failures are attributed to semantically equivalent answers. We further demonstrate that regex matching ranks QA models consistent with human judgments, although still suffering from unnecessary strictness. Finally, we demonstrate that automated evaluation models are a reasonable surrogate for lexical matching in some circumstances, but not for long-form answers generated by LLMs. The automated models struggle in detecting hallucinations in LLM answers and are thus unable to evaluate LLMs. At this time, there appears to be no substitute for human evaluation.
Authors: I-Hung Hsu, Avik Ray, Shubham Garg, Nanyun Peng, Jing Huang
Existing efforts on text synthesis for code-switching mostly require training on code-switched texts in the target language pairs, limiting the deployment of the models to cases lacking code-switched data. In this work, we study the problem of synthesizing code-switched texts for language pairs absent from the training data. We introduce GLOSS, a model built on top of a pre-trained multilingual machine translation model (PMMTM) with an additional code-switching module. This module, either an adapter or extra prefixes, learns code-switching patterns from code-switched data during training, while the primary component of GLOSS, i.e., the PMMTM, is frozen. The design of only adjusting the code-switching module prevents our model from overfitting to the constrained training data for code-switching. Hence, GLOSS exhibits the ability to generalize and synthesize code-switched texts across a broader spectrum of language pairs. Additionally, we develop a self-training algorithm on target language pairs further to enhance the reliability of GLOSS. Automatic evaluations on four language pairs show that GLOSS achieves at least 55% relative BLEU and METEOR scores improvements compared to strong baselines. Human evaluations on two language pairs further validate the success of GLOSS.
Authors: Dawei Zhu, Xiaoyu Shen, Marius Mosbach, Andreas Stephan, Dietrich Klakow
Weakly supervised learning is a popular approach for training machine learning models in low-resource settings. Instead of requesting high-quality yet costly human annotations, it allows training models with noisy annotations obtained from various weak sources. Recently, many sophisticated approaches have been proposed for robust training under label noise, reporting impressive results. In this paper, we revisit the setup of these approaches and find that the benefits brought by these approaches are significantly overestimated. Specifically, we find that the success of existing weakly supervised learning approaches heavily relies on the availability of clean validation samples which, as we show, can be leveraged much more efficiently by simply training on them. After using these clean labels in training, the advantages of using these sophisticated approaches are mostly wiped out. This remains true even when reducing the size of the available clean data to just five samples per class, making these approaches impractical. To understand the true value of weakly supervised learning, we thoroughly analyze diverse NLP datasets and tasks to ascertain when and why weakly supervised approaches work. Based on our findings, we provide recommendations for future research.
Authors: Khanh Nguyen
How do language models "think"? This paper formulates a probabilistic cognitive model called the bounded pragmatic speaker, which can characterize the operation of different variations of language models. Specifically, we demonstrate that large language models fine-tuned with reinforcement learning from human feedback (Ouyang et al., 2022) embody a model of thought that conceptually resembles a fast-and-slow model (Kahneman, 2011), which psychologists have attributed to humans. We discuss the limitations of reinforcement learning from human feedback as a fast-and-slow model of thought and propose avenues for expanding this framework. In essence, our research highlights the value of adopting a cognitive probabilistic modeling approach to gain insights into the comprehension, evaluation, and advancement of language models.
Authors: Wen Yang, Chong Li, Jiajun Zhang, Chengqing Zong
Large language models (LLMs) demonstrate promising translation performance among various natural languages. However, many LLMs especially the open-sourced ones, such as BLOOM and LLaMA, are English-dominant and support only dozens of natural languages, making the potential of LLMs on language translation less explored. In this work, we present BigTranslate which adapts LLaMA that covers only 20 languages and enhances it with multilingual translation capability on more than 100 languages. BigTranslate is built upon LLaMA-13B and it is optimized in three steps. First, we continue training LLaMA with massive Chinese monolingual data. Second, we continue training the model with a large-scale parallel dataset that covers 102 natural languages. Third, we instruct-tune the foundation model with multilingual translation instructions, leading to our BigTranslate model. The preliminary experiments on multilingual translation show that BigTranslate performs comparably with ChatGPT and Google Translate in many languages and even outperforms ChatGPT in 8 language pairs. We release the BigTranslate model and hope it can advance the research progress.
Authors: Tamanna Hossain, Sunipa Dev, Sameer Singh
Content Warning: This paper contains examples of misgendering and erasure that could be offensive and potentially triggering.
Gender bias in language technologies has been widely studied, but research has mostly been restricted to a binary paradigm of gender. It is essential also to consider non-binary gender identities, as excluding them can cause further harm to an already marginalized group. In this paper, we comprehensively evaluate popular language models for their ability to correctly use English gender-neutral pronouns (e.g., singular they, them) and neo-pronouns (e.g., ze, xe, thon) that are used by individuals whose gender identity is not represented by binary pronouns. We introduce MISGENDERED, a framework for evaluating large language models' ability to correctly use preferred pronouns, consisting of (i) instances declaring an individual's pronoun, followed by a sentence with a missing pronoun, and (ii) an experimental setup for evaluating masked and auto-regressive language models using a unified method. When prompted out-of-the-box, language models perform poorly at correctly predicting neo-pronouns (averaging 7.7% accuracy) and gender-neutral pronouns (averaging 34.2% accuracy). This inability to generalize results from a lack of representation of non-binary pronouns in training data and memorized associations. Few-shot adaptation with explicit examples in the prompt improves performance for neo-pronouns, but only to 64.7% even with 20 shots. We release the full dataset, code, and demo at https://tamannahossainkay.github.io/misgendered/
Authors: Alessandro Wollek, Philip Haitzer, Thomas Sedlmeyr, Sardi Hyska, Johannes Rueckel, Bastian Sabel, Michael Ingrisch, Tobias Lasser
Radiologists are in short supply globally, and deep learning models offer a promising solution to address this shortage as part of clinical decision-support systems. However, training such models often requires expensive and time-consuming manual labeling of large datasets. Automatic label extraction from radiology reports can reduce the time required to obtain labeled datasets, but this task is challenging due to semantically similar words and missing annotated data. In this work, we explore the potential of weak supervision of a deep learning-based label prediction model, using a rule-based labeler. We propose a deep learning-based CheXpert label prediction model, pre-trained on reports labeled by a rule-based German CheXpert model and fine-tuned on a small dataset of manually labeled reports. Our results demonstrate the effectiveness of our approach, which significantly outperformed the rule-based model on all three tasks. Our findings highlight the benefits of employing deep learning-based models even in scenarios with sparse data and the use of the rule-based labeler as a tool for weak supervision.
Authors: Jifan Yu, Xiaozhi Wang, Shangqing Tu, Shulin Cao, Daniel Zhang-Li, Xin Lv, Hao Peng, Zijun Yao, Xiaohan Zhang, Hanming Li, Chunyang Li, Zheyuan Zhang, Yushi Bai, Yantao Liu, Amy Xin, Nianyi Lin, Kaifeng Yun, Linlu Gong, Jianhui Chen, Zhili Wu, Yunjia Qi, Weikai Li, Yong Guan, Kaisheng Zeng, Ji Qi, Hailong Jin, Jinxin Liu, Yu Gu, Yuan Yao, Ning Ding, Lei Hou, Zhiyuan Liu, Bin Xu, Jie Tang, Juanzi Li
The unprecedented performance of large language models (LLMs) necessitates improvements in evaluations. Rather than merely exploring the breadth of LLM abilities, we believe meticulous and thoughtful designs are essential to thorough, unbiased, and applicable evaluations. Given the importance of world knowledge to LLMs, we construct a Knowledge-oriented LLM Assessment benchmark (KoLA), in which we carefully design three crucial factors: (1) For ability modeling, we mimic human cognition to form a four-level taxonomy of knowledge-related abilities, covering $19$ tasks. (2) For data, to ensure fair comparisons, we use both Wikipedia, a corpus prevalently pre-trained by LLMs, along with continuously collected emerging corpora, aiming to evaluate the capacity to handle unseen data and evolving knowledge. (3) For evaluation criteria, we adopt a contrastive system, including overall standard scores for better numerical comparability across tasks and models and a unique self-contrast metric for automatically evaluating knowledge hallucination. We evaluate $21$ open-source and commercial LLMs and obtain some intriguing findings. The KoLA dataset and open-participation leaderboard are publicly released at https://kola.xlore.cn and will be continuously updated to provide references for developing LLMs and knowledge-related systems.
Authors: Le Zhuo, Ruibin Yuan, Jiahao Pan, Yinghao Ma, Yizhi LI, Ge Zhang, Si Liu, Roger Dannenberg, Jie Fu, Chenghua Lin, Emmanouil Benetos, Wenhu Chen, Wei Xue, Yike Guo
We introduce LyricWhiz, a robust, multilingual, and zero-shot automatic lyrics transcription method achieving state-of-the-art performance on various lyrics transcription datasets, even in challenging genres such as rock and metal. Our novel, training-free approach utilizes Whisper, a weakly supervised robust speech recognition model, and GPT-4, today's most performant chat-based large language model. In the proposed method, Whisper functions as the "ear" by transcribing the audio, while GPT-4 serves as the "brain," acting as an annotator with a strong performance for contextualized output selection and correction. Our experiments show that LyricWhiz significantly reduces Word Error Rate compared to existing methods in English and can effectively transcribe lyrics across multiple languages. Furthermore, we use LyricWhiz to create the first publicly available, large-scale, multilingual lyrics transcription dataset with a CC-BY-NC-SA copyright license, based on MTG-Jamendo, and offer a human-annotated subset for noise level estimation and evaluation. We anticipate that our proposed method and dataset will advance the development of multilingual lyrics transcription, a challenging and emerging task.
Authors: Xuan-Quy Dao
This paper presents a performance comparison of three large language models (LLMs), namely OpenAI ChatGPT, Microsoft Bing Chat, and Google Bard, on the VNHSGE English dataset. The results show that BingChat is better than ChatGPT and Bard. Therefore, BingChat and Bard can replace ChatGPT while ChatGPT is not yet officially available in Vietnam. The results also indicate that ChatGPT, Bing Chat, and Bard outperform Vietnamese students in English language proficiency. The findings of this study contribute to the understanding of the potential of LLMs in English language education. The remarkable performance of ChatGPT, Bing Chat, and Bard demonstrates their potential as effective tools for teaching and learning English at the high school level.
Authors: Xinming Tu, James Zou, Weijie J. Su, Linjun Zhang
The rapid advances of large language models (LLMs), such as ChatGPT, are revolutionizing data science and statistics. These state-of-the-art tools can streamline complex processes. As a result, it reshapes the role of data scientists. We argue that LLMs are transforming the responsibilities of data scientists, shifting their focus from hands-on coding, data-wrangling and conducting standard analyses to assessing and managing analyses performed by these automated AIs. This evolution of roles is reminiscent of the transition from a software engineer to a product manager. We illustrate this transition with concrete data science case studies using LLMs in this paper. These developments necessitate a meaningful evolution in data science education. Pedagogy must now place greater emphasis on cultivating diverse skillsets among students, such as LLM-informed creativity, critical thinking, AI-guided programming. LLMs can also play a significant role in the classroom as interactive teaching and learning tools, contributing to personalized education. This paper discusses the opportunities, resources and open challenges for each of these directions. As with any transformative technology, integrating LLMs into education calls for careful consideration. While LLMs can perform repetitive tasks efficiently, it's crucial to remember that their role is to supplement human intelligence and creativity, not to replace it. Therefore, the new era of data science education should balance the benefits of LLMs while fostering complementary human expertise and innovations. In conclusion, the rise of LLMs heralds a transformative period for data science and its education. This paper seeks to shed light on the emerging trends, potential opportunities, and challenges accompanying this paradigm shift, hoping to spark further discourse and investigation into this exciting, uncharted territory.
Authors: Fuxiang Tao, Wei Ma, Xuri Ge, Anna Esposito, Alessandro Vinciarelli
This work shows that depression changes the correlation between features extracted from speech. Furthermore, it shows that using such an insight can improve the training speed and performance of depression detectors based on SVMs and LSTMs. The experiments were performed over the Androids Corpus, a publicly available dataset involving 112 speakers, including 58 people diagnosed with depression by professional psychiatrists. The results show that the models used in the experiments improve in terms of training speed and performance when fed with feature correlation matrices rather than with feature vectors. The relative reduction of the error rate ranges between 23.1% and 26.6% depending on the model. The probable explanation is that feature correlation matrices appear to be more variable in the case of depressed speakers. Correspondingly, such a phenomenon can be thought of as a depression marker.
Authors: Sneha Singhania, Simon Razniewski, Gerhard Weikum
The widespread usage of latent language representations via pre-trained language models (LMs) suggests that they are a promising source of structured knowledge. However, existing methods focus only on a single object per subject-relation pair, even though often multiple objects are correct. To overcome this limitation, we analyze these representations for their potential to yield materialized multi-object relational knowledge. We formulate the problem as a rank-then-select task. For ranking candidate objects, we evaluate existing prompting techniques and propose new ones incorporating domain knowledge. Among the selection methods, we find that choosing objects with a likelihood above a learned relation-specific threshold gives a 49.5% F1 score. Our results highlight the difficulty of employing LMs for the multi-valued slot-filling task and pave the way for further research on extracting relational knowledge from latent language representations.