Large multilingual models have inspired a new class of word alignment methods, which work well for the model's pretraining languages. However, the languages most in need of automatic alignment are low-resource and, thus, not typically included in the pretraining data. In this work, we ask: How do modern aligners perform on unseen languages, and are they better than traditional methods? We contribute gold-standard alignments for Bribri--Spanish, Guarani--Spanish, Quechua--Spanish, and Shipibo-Konibo--Spanish. With these, we evaluate state-of-the-art aligners with and without model adaptation to the target language. Finally, we also evaluate the resulting alignments extrinsically through two downstream tasks: named entity recognition and part-of-speech tagging. We find that although transformer-based methods generally outperform traditional models, the two classes of approach remain competitive with each other.

By focusing on one such two "how to share" categories into task, the model ignores knowledge from the training five categories, including feature learning approach, signals of related tasks (Ruder, 2017). There low-rank approach, task clustering approach, task are a great number of tasks in NLP, from syntax relation learning approach, and decomposition approach; parsing to information extraction, from machine Crawshaw (2020) presented more recent translation to question answering: each requires models in both single-domain and multi-modal architectures, a model dedicated to learning from data. Biologically, as well as an overview of optimization humans learn natural languages, from basic methods in MTL. Nevertheless, it is still not clearly grammar to complex semantics in a single brain understood how to design and train a single model (Hashimoto et al., 2017). In the field of machine to handle a variety of NLP tasks according to task learning, multi-task learning (MTL) aims to leverage relatedness. Especially when faced with a set of useful information shared across multiple related tasks that are seldom simultaneously trained previously, tasks to improve the generalization performance it is of crucial importance that researchers on all tasks (Caruana, 1997). In deep neural find proper auxiliary tasks and assess the feasibility networks, it is generally achieved by sharing part of of such multi-task learning attempt.

Significant developments in techniques such as encoder-decoder models have enabled us to represent information comprising multiple modalities. This information can further enhance many downstream tasks in the field of information retrieval and natural language processing; however, improvements in multi-modal techniques and their performance evaluation require large-scale multi-modal data which offers sufficient diversity. Multi-lingual modeling for a variety of tasks like multi-modal summarization, text generation, and translation leverages information derived from high-quality multi-lingual annotated data. In this work, we present the current largest multi-lingual multi-modal summarization dataset (M3LS), and it consists of over a million instances of document-image pairs along with a professionally annotated multi-modal summary for each pair. It is derived from news articles published by British Broadcasting Corporation(BBC) over a decade and spans 20 languages, targeting diversity across five language roots, it is also the largest summarization dataset for 13 languages and consists of cross-lingual summarization data for 2 languages. We formally define the multi-lingual multi-modal summarization task utilizing our dataset and report baseline scores from various state-of-the-art summarization techniques in a multi-lingual setting. We also compare it with many similar datasets to analyze the uniqueness and difficulty of M3LS.

The field of visual question answering (VQA) has recently seen a surge in research focused on providing explanations for predicted answers. However, current systems mostly rely on separate models to predict answers and generate explanations, leading to less grounded and frequently inconsistent results. To address this, we propose a multitask learning approach towards a Unified Model for Answer and Explanation generation (UMAE). Our approach involves the addition of artificial prompt tokens to training data and fine-tuning a multimodal encoder-decoder model on a variety of VQA-related tasks. In our experiments, UMAE models surpass the prior state-of-the-art answer accuracy on A-OKVQA by 10~15%, show competitive results on OK-VQA, achieve new state-of-the-art explanation scores on A-OKVQA and VCR, and demonstrate promising out-of-domain performance on VQA-X.

In this work, therefore, we study three types of plagiarism (i.e., verbatim, paraphrase, and idea) among GPT-2 generated texts, Language Models (LMs) have become core elements of Natural in comparison to its training data, and further analyze the plagiarism Language Processing (NLP) solutions, excelling in a wide range of patterns of fine-tuned LMs with domain-specific corpora which are tasks such as natural language generation (NLG), speech recognition, extensively used in practice. Our results suggest that (1) three types machine translation, and question answering. The development of plagiarism widely exist in LMs beyond memorization, (2) both of large-scale text corpora (generally scraped from the Web) has size and decoding methods of LMs are strongly associated with the enabled researchers to train increasingly large-scale LMs. Especially, degrees of plagiarism they exhibit, and (3) fine-tuned LMs' plagiarism large-scale LMs have demonstrated unprecedented performance on patterns vary based on their corpus similarity and homogeneity. NLG such that LM-generated texts routinely show more novel and Given that a majority of LMs' training data is scraped from the Web interesting stories than human writings do [35], and the distinction without informing content owners, their reiteration of words, phrases, between machine-authored and human-written texts has become and even core ideas from training sets into generated texts has ethical non-trivial [52, 53]. As a result, there has been a significant increase implications. Their patterns are likely to exacerbate as both in the use of LMs in user-facing products and critical applications.

We extend a current sequence-tagging approach to Grammatical Error Correction (GEC) by introducing specialised tags for spelling correction and morphological inflection using the SymSpell and LemmInflect algorithms. Our approach improves generalisation: the proposed new tagset allows a smaller number of tags to correct a larger range of errors. Our results show a performance improvement both overall and in the targeted error categories. We further show that ensembles trained with our new tagset outperform those trained with the baseline tagset on the public BEA benchmark.

The exponential growth of social media platforms such as Twitter and Facebook has revolutionized textual communication and textual content publication in human society. However, they have been increasingly exploited to propagate toxic content, such as hate speech, malicious advertisement, and pornography, which can lead to highly negative impacts (e.g., harmful effects on teen mental health). Researchers and practitioners have been enthusiastically developing and extensively deploying textual content moderation software to address this problem. However, we find that malicious users can evade moderation by changing only a few words in the toxic content. Moreover, modern content moderation software performance against malicious inputs remains underexplored. To this end, we propose MTTM, a Metamorphic Testing framework for Textual content Moderation software. Specifically, we conduct a pilot study on 2,000 text messages collected from real users and summarize eleven metamorphic relations across three perturbation levels: character, word, and sentence. MTTM employs these metamorphic relations on toxic textual contents to generate test cases, which are still toxic yet likely to evade moderation. In our evaluation, we employ MTTM to test three commercial textual content moderation software and two state-of-the-art moderation algorithms against three kinds of toxic content. The results show that MTTM achieves up to 83.9%, 51%, and 82.5% error finding rates (EFR) when testing commercial moderation software provided by Google, Baidu, and Huawei, respectively, and it obtains up to 91.2% EFR when testing the state-of-the-art algorithms from the academy. In addition, we leverage the test cases generated by MTTM to retrain the model we explored, which largely improves model robustness (0% to 5.9% EFR) while maintaining the accuracy on the original test set.

Exploiting social media to spread hate has tremendously increased over the years. Lately, multi-modal hateful content such as memes has drawn relatively more traction than uni-modal content. Moreover, the availability of implicit content payloads makes them fairly challenging to be detected by existing hateful meme detection systems. In this paper, we present a use case study to analyze such systems' vulnerabilities against external adversarial attacks. We find that even very simple perturbations in uni-modal and multi-modal settings performed by humans with little knowledge about the model can make the existing detection models highly vulnerable. Empirically, we find a noticeable performance drop of as high as 10% in the macro-F1 score for certain attacks. As a remedy, we attempt to boost the model's robustness using contrastive learning as well as an adversarial training-based method - VILLA. Using an ensemble of the above two approaches, in two of our high resolution datasets, we are able to (re)gain back the performance to a large extent for certain attacks. We believe that ours is a first step toward addressing this crucial problem in an adversarial setting and would inspire more such investigations in the future.

We derive an alternative yet equivalent However, there are noticeably fewer success cases in employing formulation of the sampling from discrete diffusion models for large-scale text generation diffusion processes and leverage this insight to tasks. This is possibly due to the discrete nature of natural develop a family of reparameterized discrete diffusion languages, while most conventional diffusion models focus models. The derived generic framework is on continuous-valued contents. To bridge the discrepancy, highly flexible, offers a fresh perspective of the recent work aims at conducting the diffusion process over token generation process in discrete diffusion models, embeddings so that the continuous diffusion models can and features more effective training and decoding be applied to discrete texts (Li et al., 2022; Gong et al., 2022; techniques. We conduct extensive experiments Strudel et al., 2022; Dieleman et al., 2022) or logits (Han to evaluate the text generation capability of our et al., 2022; Richemond et al., 2022). Nevertheless, these model, demonstrating significant improvements approaches often require designing a well-crafted rounding over existing diffusion models.

Back-translation is widely known for its effectiveness in neural machine translation when there is little to no parallel data. In this approach, a source-to-target model is coupled with a target-to-source model trained in parallel. The target-to-source model generates noisy sources, while the source-to-target model is trained to reconstruct the targets and vice versa. Recent developments of multilingual pre-trained sequence-to-sequence models for programming languages have been very effective for a broad spectrum of downstream software engineering tasks. Hence, training them to build programming language translation systems via back-translation is compelling. However, these models cannot be further trained via back-translation since they learn to output sequences in the same language as the inputs during pre-training. As an alternative, we propose performing back-translation via code summarization and generation. In code summarization, a model learns to generate natural language (NL) summaries given code snippets. In code generation, the model learns to do the opposite. Therefore, target-to-source generation in back-translation can be viewed as a target-to-NL-to-source generation. We show that our proposed approach performs competitively with state-of-the-art methods. We have made the code publicly available.