Michael Ginn

Nirajan Paudel, Michael Ginn, Luc De Nardi et al.

Speculative decoding has become a crucial component of large language model (LLM) inference, enabling faster generation by drafting multiple tokens and verifying them in parallel. However, small draft models tend to suffer from disproportionately poor multilingual capabilities. Thus, when generating text in a non-English language, speculative decoding is far less effective. We compare three strategies to improve speculative decoding efficiency for eleven languages: finetuning the draft model on task-specific data (translation); finetuning the draft model on unlabeled monolingual corpora; and training simple n-gram draft models on the same monolingual corpora. We evaluate efficiency on translation (from English into the target language) and the held-out task of story generation. We find that while task-specific distillation can significantly improve efficiency, distilled models generalize poorly to a new task. Meanwhile, n-gram draft models, despite lower acceptance rates, consistently provide large speed-ups due to much faster draft generation.

Michael Ginn, Lindia Tjuatja, Enora Rice et al.

Automated interlinear gloss prediction with neural networks is a promising approach to accelerate language documentation efforts. However, while state-of-the-art models like GlossLM achieve high scores on glossing benchmarks, user studies with linguists have found critical barriers to the usefulness of such models in real-world scenarios. In particular, existing models typically generate morpheme-level glosses but assign them to whole words without predicting the actual morpheme boundaries, making the predictions less interpretable and thus untrustworthy to human annotators. We conduct the first study on neural models that jointly predict interlinear glosses and the corresponding morphological segmentation from raw text. We run experiments to determine the optimal way to train models that balance segmentation and glossing accuracy, as well as the alignment between the two tasks. We extend the training corpus of GlossLM and pretrain PolyGloss, a family of seq2seq multilingual models for joint segmentation and glossing that outperforms GlossLM on glossing and beats various open-source LLMs on segmentation, glossing, and alignment. In addition, we demonstrate that PolyGloss can be quickly adapted to a new dataset via low-rank adaptation.

Michael Ginn, Alexis Palmer

Generalization is of particular importance in resource-constrained settings, where the available training data may represent only a small fraction of the distribution of possible texts. We investigate the ability of morpheme labeling models to generalize by evaluating their performance on unseen genres of text, and we experiment with strategies for closing the gap between performance on in-distribution and out-of-distribution data. Specifically, we use weight decay optimization, output denoising, and iterative pseudo-labeling, and achieve a 2% improvement on a test set containing texts from unseen genres. All experiments are performed using texts written in the Mayan language Uspanteko.

Michael Ginn, Alexis Palmer, Mans Hulden

Finite-State Transducers (FSTs) are effective models for string-to-string rewriting tasks, often providing the efficiency necessary for high-performance applications, but constructing transducers by hand is difficult. In this work, we propose a novel method for automatically constructing unweighted FSTs following the hidden state geometry learned by a recurrent neural network. We evaluate our methods on real-world datasets for morphological inflection, grapheme-to-phoneme prediction, and historical normalization, showing that the constructed FSTs are highly accurate and robust for many datasets, substantially outperforming classical transducer learning algorithms by up to 87% accuracy on held-out test sets.

Michael Ginn, Alexis Palmer

Morpheme glossing is a critical task in automated language documentation and can benefit other downstream applications greatly. While state-of-the-art glossing systems perform very well for languages with large amounts of existing data, it is more difficult to create useful models for low-resource languages. In this paper, we propose the use of a taxonomic loss function that exploits morphological information to make morphological glossing more performant when data is scarce. We find that while the use of this loss function does not outperform a standard loss function with regards to single-label prediction accuracy, it produces better predictions when considering the top-n predicted labels. We suggest this property makes the taxonomic loss function useful in a human-in-the-loop annotation setting.

Michael Ginn

Language documentation is a critical aspect of language preservation, often including the creation of Interlinear Glossed Text (IGT). Creating IGT is time-consuming and tedious, and automating the process can save valuable annotator effort. This paper describes the baseline system for the SIGMORPHON 2023 Shared Task of Interlinear Glossing. In our system, we utilize a transformer architecture and treat gloss generation as a sequence labelling task.

Michael Ginn, Lindia Tjuatja, Taiqi He et al.

Language documentation projects often involve the creation of annotated text in a format such as interlinear glossed text (IGT), which captures fine-grained morphosyntactic analyses in a morpheme-by-morpheme format. However, there are few existing resources providing large amounts of standardized, easily accessible IGT data, limiting their applicability to linguistic research, and making it difficult to use such data in NLP modeling. We compile the largest existing corpus of IGT data from a variety of sources, covering over 450k examples across 1.8k languages, to enable research on crosslingual transfer and IGT generation. We normalize much of our data to follow a standard set of labels across languages. Furthermore, we explore the task of automatically generating IGT in order to aid documentation projects. As many languages lack sufficient monolingual data, we pretrain a large multilingual model on our corpus. We demonstrate the utility of this model by finetuning it on monolingual corpora, outperforming SOTA models by up to 6.6\%. Our pretrained model and dataset are available on Hugging Face.

Michael Ginn

Linguists have long held that a key aspect of natural language syntax is the recursive organization of language units into constituent structures, and research has suggested that current state-of-the-art language models lack an inherent bias towards this feature. A number of alternative models have been proposed to provide inductive biases towards constituency, including the Tree Transformer, which utilizes a modified attention mechanism to organize tokens into constituents. We investigate Tree Transformers to study whether they utilize meaningful and/or useful constituent structures. We pretrain a large Tree Transformer on language modeling in order to investigate the learned constituent tree representations of sentences, finding little evidence for meaningful structures. Next, we evaluate Tree Transformers with similar transformer models on error detection tasks requiring constituent structure. We find that while the Tree Transformer models may slightly outperform at these tasks, there is little evidence to suggest a meaningful improvement. In general, we conclude that there is little evidence to support Tree Transformer as an effective model of syntactic constituency.

Ray Groshan, Michael Ginn, Alexis Palmer

Data augmentation, a widely-employed technique for addressing data scarcity, involves generating synthetic data examples which are then used to augment available training data. Researchers have seen surprising success from simple methods, such as random perturbations from natural examples, where models seem to benefit even from data with nonsense words, or data that doesn't conform to the rules of the language. A second line of research produces synthetic data that does in fact follow all linguistic constraints; these methods require some linguistic expertise and are generally more challenging to implement. No previous work has done a systematic, empirical comparison of both linguistically-naive and linguistically-motivated data augmentation strategies, leaving uncertainty about whether the additional time and effort of linguistically-motivated data augmentation work in fact yields better downstream performance. In this work, we conduct a careful and comprehensive comparison of augmentation strategies (both linguistically-naive and linguistically-motivated) for two low-resource languages with different morphological properties, Uspanteko and Arapaho. We evaluate the effectiveness of many different strategies and their combinations across two important sequence-to-sequence tasks for low-resource languages: machine translation and interlinear glossing. We find that linguistically-motivated strategies can have benefits over naive approaches, but only when the new examples they produce are not significantly unlike the training data distribution.

Michael Ginn, Mans Hulden

Dynamic topic models have been proposed as a tool for historical analysis, but traditional approaches have had limited usefulness, being difficult to configure, interpret, and evaluate. In this work, we experiment with a recent approach for dynamic topic modeling using BERT embeddings. We compare topic models built using traditional statistical models (LDA and NMF) and the BERT-based model, modeling topics over the entire surviving corpus of Roman literature. We find that while quantitative metrics prefer statistical models, qualitative evaluation finds better insights from the neural model. Furthermore, the neural topic model is less sensitive to hyperparameter configuration and thus may make dynamic topic modeling more viable for historical researchers.

Michael Ginn, Mans Hulden, Alexis Palmer

Interlinear glossed text (IGT) is a popular format in language documentation projects, where each morpheme is labeled with a descriptive annotation. Automating the creation of interlinear glossed text would be desirable to reduce annotator effort and maintain consistency across annotated corpora. Prior research has explored a number of statistical and neural methods for automatically producing IGT. As large language models (LLMs) have showed promising results across multilingual tasks, even for rare, endangered languages, it is natural to wonder whether they can be utilized for the task of generating IGT. We explore whether LLMs can be effective at the task of interlinear glossing with in-context learning, without any traditional training. We propose new approaches for selecting examples to provide in-context, observing that targeted selection can significantly improve performance. We find that LLM-based methods beat standard transformer baselines, despite requiring no training at all. These approaches still underperform state-of-the-art supervised systems for the task, but are highly practical for researchers outside of the NLP community, requiring minimal effort to use.