Research Group Hinrich Schütze

Transliterating related languages that use different scripts into a common script shows effectiveness in improving crosslingual transfer in downstream tasks. However, this methodology often makes pretraining a model from scratch unavoidable, as transliteration brings about new subwords not covered in existing multilingual pretrained language models (mPLMs). This is not desired because it takes a lot of computation budget for pretraining. A more promising way is to make full use of available mPLMs. To this end, this paper proposes a simple but effective framework: Transliterate-Merge-Initialize (TransMI), which can create a strong baseline well-suited for data that is transliterated into a common script by exploiting an mPLM and its accompanied tokenizer. TransMI has three stages: (a) transliterate the vocabulary of an mPLM into a common script; (b) merge the new vocabulary with the original vocabulary; and (c) initialize the embeddings of the new subwords. We applied TransMI to three recent strong mPLMs, and our experiments demonstrate that TransMI not only preserves their ability to handle non-transliterated data, but also enables the models to effectively process transliterated data: the results show a consistent improvement of 3% to 34%, varying across different models and tasks.

Recent studies have shown that post-aligning multilingual pretrained language models (mPLMs) using alignment objectives on both original and transliterated data can improve crosslingual alignment. This improvement further leads to better crosslingual transfer performance. However, it remains unclear how and why a better crosslingual alignment is achieved, as this technique only involves transliterations, and does not use any parallel data. This paper attempts to explicitly evaluate the crosslingual alignment and identify the key elements in transliteration-based approaches that contribute to better performance. For this, we train multiple models under varying setups for two pairs of related languages: (1) Polish and Ukrainian and (2) Hindi and Urdu. To assess alignment, we define four types of similarities based on sentence representations. Our experiments show that adding transliterations alone improves the overall similarities, even for random sentence pairs. With the help of auxiliary alignment objectives, especially the contrastive objective, the model learns to distinguish matched from random pairs, leading to better alignments. However, we also show that better alignment does not always yield better downstream performance, suggesting that further research is needed to clarify the connection between alignment and performance.

The need for large text corpora has increased with the advent of pretrained language models and, in particular, the discovery of scaling laws for these models. Most available corpora have sufficient data only for languages with large dominant communities. However, there is no corpus available that (i) covers a wide range of minority languages; (ii) is generated by an open-source reproducible pipeline; and (iii) is rigorously cleaned from noise, making it trustworthy to use. We present GlotCC, a clean, document-level, 2TB general domain corpus derived from CommonCrawl, covering more than 1000 languages. We make GlotCC and the system used to generate it - including the pipeline, language identification model, and filters - available to the research community.

Hate speech online remains an understudied issue for marginalized communities, and has seen rising relevance, especially in the Global South, which includes developing societies with increasing internet penetration. In this paper, we aim to provide marginalized communities living in societies where the dominant language is low-resource with a privacy-preserving tool to protect themselves from hate speech on the internet by filtering offensive content in their native languages. Our contribution in this paper is twofold: 1) we release REACT (REsponsive hate speech datasets Across ConTexts), a collection of high-quality, culture-specific hate speech detection datasets comprising seven distinct target groups in eight low-resource languages, curated by experienced data collectors; 2) we propose a solution to few-shot hate speech detection utilizing federated learning (FL), a privacy-preserving and collaborative learning approach, to continuously improve a central model that exhibits robustness when tackling different target groups and languages. By keeping the training local to the users’ devices, we ensure the privacy of the users’ data while benefitting from the efficiency of federated learning. Furthermore, we personalize client models to target-specific training data and evaluate their performance. Our results indicate the effectiveness of FL across different target groups, whereas the benefits of personalization on few-shot learning are not clear.

Full-parameter fine-tuning has become the go-to choice for adapting language models (LMs) to downstream tasks due to its excellent performance. As LMs grow in size, fine-tuning the full parameters of LMs requires a prohibitively large amount of GPU memory. Existing approaches utilize zeroth-order optimizer to conserve GPU memory, which can potentially compromise the performance of LMs as non-zero order optimizers tend to converge more readily on most downstream tasks. In this paper, we propose a novel optimizer-independent end-to-end hierarchical fine-tuning strategy, HiFT, which only updates a subset of parameters at each training step. HiFT can significantly reduce the amount of gradients and optimizer state parameters residing in GPU memory at the same time, thereby reducing GPU memory usage. Our results demonstrate that: (1) HiFT achieves comparable performance to parameter-efficient fine-tuning and standard full parameter fine-tuning. (2) HiFT supports various optimizers including AdamW, AdaGrad, SGD, etc. (3) HiFT can save more than 60% GPU memory compared with standard full-parameter fine-tuning for 7B model. (4) HiFT enables full-parameter fine-tuning of a 7B model on single 48G A6000 with a precision of 32 using the AdamW optimizer, without using any memory saving techniques.

Instruction tuning enables language models to more effectively generalize and better follow user intent. However, obtaining instruction data is costly and challenging. Prior work employs methods such as expensive human annotation, crowd-sourced datasets with alignment issues, and generating noisy examples via LLMs. We introduce the LongForm-C dataset, which is created by reverse instructions. We generate instructions via LLMs for human-written corpus examples using reverse instructions. First we select a diverse set of human-written documents from corpora such as C4 and Wikipedia; then we generate instructions for these documents via LLMs. This approach provides a cheaper and cleaner instruction-tuning dataset with natural output and one suitable for long text generation. Our models outperform 10x larger language models without instruction tuning on tasks such as story/recipe generation and long-form question answering. Moreover, LongForm models outperform prior instruction-tuned models such as FLAN-T5 and Alpaca by a large margin, and improve language understanding capabilities further.

Many datasets have been developed to train and evaluate document-level relation extraction (RE) models. Most of these are constructed using real-world data. It has been shown that RE models trained on real-world data suffer from factual biases. To evaluate and address this issue, we present CovEReD, a counterfactual data generation approach for document-level relation extraction datasets using entity replacement. We first demonstrate that models trained on factual data exhibit inconsistent behavior: while they accurately extract triples from factual data, they fail to extract the same triples after counterfactual modification. This inconsistency suggests that models trained on factual data rely on spurious signals such as specific entities and external knowledge – rather than on the input context – to extract triples. We show that by generating document-level counterfactual data with CovEReD and training models on them, consistency is maintained with minimal impact on RE performance. We release our CovEReD pipeline as well as Re-DocRED-CF, a dataset of counterfactual RE documents, to assist in evaluating and addressing inconsistency in document-level RE.

To ensure large language models contain up-to-date knowledge, they need to be updated regularly. However, model editing is challenging as it might also affect knowledge that is unrelated to the new data. State-of-the-art methods identify parameters associated with specific knowledge and then modify them via direct weight updates. However, these locate-and-edit methods suffer from heavy computational overhead and lack theoretical validation. In contrast, directly fine-tuning the model on requested edits affects the model’s behavior on unrelated knowledge, and significantly damages the model’s generation fluency and consistency. To address these challenges, we propose SAUL, a streamlined model editing method that uses sentence concatenation with augmented random facts for generation regularization. Evaluations on three model editing benchmarks show that SAUL is a practical and reliable solution for model editing outperforming state-of-the-art methods while maintaining generation quality and reducing computational overhead.

Multilingual pre-trained models (mPLMs) have shown impressive performance on cross-lingual transfer tasks. However, the transfer performance is often hindered when a low-resource target language is written in a different script than the high-resource source language, even though the two languages may be related or share parts of their vocabularies. Inspired by recent work that uses transliteration to address this problem, our paper proposes a transliteration-based post-pretraining alignment (PPA) method aiming to improve the cross-lingual alignment between languages using diverse scripts. We select two areal language groups, Mediterranean-Amharic-Farsi and South+East Asian Languages, wherein the languages are mutually influenced but use different scripts. We apply our method to these language groups and conduct extensive experiments on a spectrum of downstream tasks. The results show that after PPA, models consistently outperform the original model (up to 50% for some tasks) in English-centric transfer. In addition, when we use languages other than English as sources in transfer, our method obtains even larger improvements.

Multiple choice question answering tasks evaluate the reasoning, comprehension, and mathematical abilities of Large Language Models (LLMs). While existing benchmarks employ automatic translation for multilingual evaluation, this approach is error-prone and potentially introduces culturally biased questions, especially in social sciences. We introduce the first multitask, multiple-choice Turkish QA benchmark, TurkishMMLU, to evaluate LLMs’ understanding of the Turkish language. TurkishMMLU includes over 10,000 questions, covering 9 different subjects from Turkish high-school education curricula. These questions are written by curriculum experts, suitable for the high-school curricula in Turkey, covering subjects ranging from natural sciences and math questions to more culturally representative topics such as Turkish Literature and the history of the Turkish Republic. We evaluate over 20 LLMs, including multilingual open-source (e.g., Gemma, Llama, MT5), closed-source (GPT 4o, Claude, Gemini), and Turkish-adapted (e.g., Trendyol) models. We provide an extensive evaluation, including zero-shot and few-shot evaluation of LLMs, chain-of-thought reasoning, and question difficulty analysis along with model performance. We provide an in-depth analysis of the Turkish capabilities and limitations of current LLMs to provide insights for future LLMs for the Turkish language.

Traditional benchmarking in NLP typically involves using static held-out test sets. However, this approach often results in an overestimation of performance and lacks the ability to offer comprehensive, interpretable, and dynamic assessments of NLP models. Recently, works like DynaBench (Kiela et al., 2021) and CheckList (Ribeiro et al., 2020) have addressed these limitations through behavioral testing of NLP models with test types generated by a multistep human-annotated pipeline. Unfortunately, manually creating a variety of test types requires much human labor, often at prohibitive cost. In this work, we propose SYNTHEVAL, a hybrid behavioral testing framework that leverages large language models (LLMs) to generate a wide range of test types for a comprehensive evaluation of NLP models. SYNTHEVAL first generates sentences via LLMs using controlled generation, and then identifies challenging examples by comparing the predictions made by LLMs with task-specific NLP models. In the last stage, human experts investigate the challenging examples, manually design templates, and identify the types of failures the taskspecific models consistently exhibit. We apply SYNTHEVAL to two classification tasks, sentiment analysis and toxic language detection, and show that our framework is effective in identifying weaknesses of strong models on these tasks.

This study investigates the linguistic understanding of Large Language Models (LLMs) regarding signifier (form) and signified (meaning) by distinguishing two LLM evaluation paradigms: psycholinguistic and neurolinguistic. Traditional psycholinguistic evaluations often reflect statistical biases that may misrepresent LLMs’ true linguistic capabilities. We introduce a neurolinguistic approach, utilizing a novel method that combines minimal pair and diagnostic probing to analyze activation patterns across model layers. This method allows for a detailed examination of how LLMs represent form and meaning, and whether these representations are consistent across languages. Our contributions are three-fold: (1) We compare neurolinguistic and psycholinguistic methods, revealing distinct patterns in LLM assessment; (2) We demonstrate that LLMs exhibit higher competence in form compared to meaning, with the latter largely correlated to the former; (3) We present new conceptual minimal pair datasets for Chinese (COMPS-ZH) and German (COMPS-DE), complementing existing English datasets.

What mechanisms underlie linguistic generalization in large language models (LLMs)? This question has attracted considerable attention, with most studies analyzing the extent to which the language skills of LLMs resemble rules. As of yet, it is not known whether linguistic generalization in LLMs could equally well be explained as the result of analogical processes, which can be formalized as similarity operations on stored exemplars. A key shortcoming of prior research is its focus on linguistic phenomena with a high degree of regularity, for which rule-based and analogical approaches make the same predictions. Here, we instead examine derivational morphology, specifically English adjective nominalization, which displays notable variability. We introduce a new method for investigating linguistic generalization in LLMs: focusing on GPT-J, we fit cognitive models that instantiate rule-based and analogical learning to the LLM training data and compare their predictions on a set of nonce adjectives with those of the LLM, allowing us to draw direct conclusions regarding underlying mechanisms. As expected, rule-based and analogical models explain the predictions of GPT-J equally well for adjectives with regular nominalization patterns. However, for adjectives with variable nominalization patterns, the analogical model provides a much better match. Furthermore, GPT-J’s behavior is sensitive to the individual word frequencies, even for regular forms, a behavior that is consistent with an analogical account of regular forms but not a rule-based one. These findings refute the hypothesis that GPT-J’s linguistic generalization on adjective nominalization involves rules, suggesting similarity operations on stored exemplars as the underlying mechanism. Overall, our study suggests that analogical processes play a bigger role in the linguistic generalization of LLMs than previously thought.

As LLMs are increasingly integrated into user-facing applications, addressing biases that perpetuate societal inequalities is crucial. While much work has gone into measuring or mitigating biases in these models, fewer studies have investigated their origins. Therefore, this study examines the correlation between gender-occupation bias in pre-training data and their manifestation in LLMs, focusing on the Dolma dataset and the OLMo model. Using zero-shot prompting and token co-occurrence analyses, we explore how biases in training data influence model outputs. Our findings reveal that biases present in pre-training data are amplified in model outputs. The study also examines the effects of prompt types, hyperparameters, and instruction-tuning on bias expression, finding instruction-tuning partially alleviating representational bias while still maintaining overall stereotypical gender associations, whereas hyperparameters and prompting variation have a lesser effect on bias expression. Our research traces bias throughout the LLM development pipeline and underscores the importance of mitigating bias at the pretraining stage.

Although large language models(LLMs) show amazing capabilities, among various exciting applications discovered for LLMs fall short in other low-resource languages. Besides, most existing methods depend on large-scale dialogue corpora and thus building systems for dialogue generation in a zero-shot scenario remains a considerable challenge. To address this challenge, we propose a novel end-to-end zero-shot dialogue generation model ChatZero based on cross-lingual code-switching method. First, we construct code-switching language and pseudo-target language with placeholders. Then for cross-lingual semantic transfer, we employ unsupervised contrastive learning to minimize the semantics gap of the source language, code-switching language, and pseudo-target language that are mutually positive examples in the high dimensional semantic space. Experiments on the multilingual DailyDialog and DSTC7-AVSD datasets demonstrate that ChatZero can achieve more than 90% of the original performance under the zero-shot case compared to supervised learning, and achieve state-of-the-art performance compared with other baselines.

English-centric large language models (LLMs) often show strong multilingual capabilities. However, the multilingual performance of these models remains unclear and is not thoroughly evaluated for many languages. Most benchmarks for multilinguality focus on classic NLP tasks, or cover a minimal number of languages. We introduce MEXA, a method for assessing the multilingual capabilities of pre-trained English-centric LLMs using parallel sentences, which are available for more languages than existing downstream tasks. MEXA leverages the fact that English-centric LLMs use English as a kind of pivot language in their intermediate layers. It computes the alignment between English and non-English languages using parallel sentences to evaluate the transfer of language understanding from English to other languages. This alignment can be used to estimate model performance in other languages. We conduct studies using various parallel datasets (FLORES-200 and Bible), models (Llama family, Gemma family, Mistral, and OLMo), and established downstream tasks (Belebele, m-MMLU, and m-ARC). We explore different methods to compute embeddings in decoder-only models. Our results show that MEXA, in its default settings, achieves a statistically significant average Pearson correlation of 0.90 with three established downstream tasks across nine models and two parallel datasets. This suggests that MEXA is a reliable method for estimating the multilingual capabilities of English-centric LLMs, providing a clearer understanding of their multilingual potential and the inner workings of LLMs.

Current state-of-the-art dialogue systems heavily rely on extensive training datasets. However, challenges arise in domains where domain-specific training datasets are insufficient or entirely absent. To tackle this challenge, we propose a novel data Augmentation framework for Multi-Domain Dialogue Generation, referred to as AMDG. The AMDG framework consists of a data augmentation process and a two-stage training approach: domain-agnostic training and domain adaptation training. We posit that domain corpora are a blend of domain-agnostic and domain-specific features, with certain representation patterns shared among diverse domains. Domain-agnostic training aims to enable models to learn these common expressive patterns. To construct domain-agnostic dialogue corpora, we employ a de-domaining data processing technique used to remove domain-specific features. By mitigating the effects of domain-specific features, the model trained on the de-domained corpora can effectively learn common expression patterns in different domains. Subsequently, we adapt the learned domain-agnostic features to the target domain through domain adaptation training. We conduct experiments on Chinese dialogue datasets from five different domains and show that AMDG achieves superior performance compared to both direct training on the target domain corpus and collective training on all five domain corpora. Our work underscores AMDG as a viable alternative solution for low-resource multi-domain dialogue generation.

Instruction tuning enhances large language models (LLMs) by aligning them with human preferences across diverse tasks. Traditional approaches to create instruction tuning datasets face serious challenges for low-resource languages due to their dependence on data annotation. This work introduces a novel method, Multilingual Reverse Instructions (MURI), which generates high-quality instruction tuning datasets for low-resource languages without requiring human annotators or pre-existing multilingual models. Utilizing reverse instructions and a translation pipeline, MURI produces instruction-output pairs from existing human-written texts in low-resource languages. This method ensures cultural relevance and diversity by sourcing texts from different native domains and applying filters to eliminate inappropriate content. Our dataset, MURI-IT, includes more than 2 million instruction-output pairs across 200 languages. Evaluation by native speakers and fine-tuning experiments with mT5 models demonstrate the approach’s effectiveness for both NLU and open-ended generation.

Recent multilingual pretrained language models (mPLMs) often avoid using language embeddings – learnable vectors assigned to different languages. These embeddings are discarded for two main reasons: (1) mPLMs are expected to have a single, unified parameter set across all languages, and (2) they need to function seamlessly as universal text encoders without requiring language IDs as input. However, this removal increases the burden on token embeddings to encode all language-specific information, which may hinder the model’s ability to produce more language-neutral representations. To address this challenge, we propose Language-Script Aware Multilingual Pretraining (LangSAMP), a method that incorporates both language and script embeddings to enhance representation learning while maintaining a simple architecture. Specifically, we integrate these embeddings into the output of the transformer blocks before passing the final representations to the language modeling head for prediction. We apply LangSAMP to the continual pretraining of XLM-R on a highly multilingual corpus covering more than 500 languages. The resulting model consistently outperforms the baseline. Extensive analysis further shows that language/script embeddings encode language/script-specific information, which improves the selection of source languages for crosslingual transfer.

Building high-quality datasets for specialized tasks is a time-consuming and resource-intensive process that often requires specialized domain knowledge. We propose Corpus Retrieval and Augmentation for Fine-Tuning (CRAFT), a method for generating synthetic datasets, given a small number of user-written few-shots that demonstrate the task to be performed. Given the few-shot examples, we use large-scale public web-crawled corpora and similarity-based document retrieval to find other relevant human-written documents. Lastly, instruction-tuned large language models (LLMs) augment the retrieved documents into custom-formatted task samples, which then can be used for fine-tuning. We demonstrate that CRAFT can efficiently generate large-scale task-specific training datasets for four diverse tasks: biology question-answering (QA), medicine QA and commonsense QA as well as summarization. Our experiments show that CRAFT-based models outperform or achieve comparable performance to general LLMs for QA tasks, while CRAFT-based summarization models outperform models trained on human-curated data by 46 preference points.

Natural language processing (NLP) has largely focused on modelling standardized languages. More recently, attention has increasingly shifted to local, non-standardized languages and dialects. However, the relevant speaker populations’ needs and wishes with respect to NLP tools are largely unknown. In this paper, we focus on dialects and regional languages related to German – a group of varieties that is heterogeneous in terms of prestige and standardization. We survey speakers of these varieties (N=327) and present their opinions on hypothetical language technologies for their dialects. Although attitudes vary among subgroups of our respondents, we find that respondents are especially in favour of potential NLP tools that work with dialectal input (especially audio input) such as virtual assistants, and less so for applications that produce dialectal output such as machine translation or spellcheckers.

We present MaskLID, a simple, yet effective, code-switching (CS) language identification (LID) method. MaskLID does not require any training and is designed to complement current high-performance sentence-level LIDs. Sentence-level LIDs are classifiers trained on monolingual texts to provide single labels, typically using a softmax layer to turn scores into probabilities. However, in cases where a sentence is composed in both L1 and L2 languages, the LID classifier often only returns the dominant label L1. To address this limitation, MaskLID employs a strategy to mask text features associated with L1, allowing the LID to classify the text as L2 in the next round. This method uses the LID itself to identify the features that require masking and does not rely on any external resource. In this work, we explore the use of MaskLID for two open-source LIDs (GlotLID and OpenLID), that are both based on the FastText architecture.

The world’s more than 7000 languages are written in at least 293 scripts. Due to various reasons, many closely related languages use different scripts, which poses a difficulty for multilingual pretrained language models (mPLMs) in learning crosslingual knowledge through lexical overlap. As a consequence, mPLMs are faced with a script barrier: representations from different scripts are located in different subspaces, which can result in crosslingual transfer involving languages of different scripts performing suboptimally. To address this problem, we propose TransliCo, a framework that optimizes the Transliteration Contrastive Modeling (TCM) objective to fine-tune an mPLM by contrasting sentences in its training data and their transliterations in a unified script (in our case Latin), which enhances uniformity in the representation space for different scripts. Using Glot500-m, an mPLM pretrained on over 500 languages, as our source model, we fine-tune it on a small portion (5%) of its training data, and refer to the resulting model as Furina. We show that Furina not only better aligns representations from distinct scripts but also outperforms the original Glot500-m on various zero-shot crosslingual transfer tasks. Additionally, we achieve consistent improvement in a case study on the Indic group where the languages exhibit areal features but use different scripts. We make our code and models publicly available.

Recommender systems are widely used to suggest engaging content, and Large Language Models (LLMs) have given rise to generative recommenders. Such systems can directly generate items, including for open-set tasks like question suggestion. While the world knowledge of LLMs enable good recommendations, improving the generated content through user feedback is challenging as continuously fine-tuning LLMs is prohibitively expensive. We present a training-free approach for optimizing generative recommenders by connecting user feedback loops to LLM-based optimizers. We propose a generative explore-exploit method that can not only exploit generated items with known high engagement, but also actively explore and discover hidden population preferences to improve recommendation quality. We evaluate our approach on question generation in two domains (e-commerce and general knowledge), and model user feedback with Click Through Rate (CTR). Experiments show our LLM-based explore-exploit approach can iteratively improve recommendations, and consistently increase CTR. Ablation analysis shows that generative exploration is key to learning user preferences, avoiding the pitfalls of greedy exploit-only approaches. A human evaluation strongly supports our quantitative findings.

Despite the ubiquity of large language models (LLMs) in AI research, the question of embodiment in LLMs remains underexplored, distinguishing them from embodied systems in robotics where sensory perception directly informs physical action.Our investigation navigates the intriguing terrain of whether LLMs, despite their non-embodied nature, effectively capture implicit human intuitions about fundamental, spatial building blocks of language. We employ insights from spatial cognitive foundations developed through early sensorimotor experiences, guiding our exploration through the reproduction of three psycholinguistic experiments. Surprisingly, correlations between model outputs and human responses emerge, revealing adaptability without a tangible connection to embodied experiences. Notable distinctions include polarized language model responses and reduced correlations in vision language models. This research contributes to a nuanced understanding of the interplay between language, spatial experiences, and the computations made by large language models.

Large Language Models (LLMs) exhibit strong In-Context Learning (ICL) capabilities when prompts with demonstrations are applied to them. However, fine-tuning still remains crucial to further enhance their adaptability. Prompt-based fine-tuning proves to be an effective fine-tuning method in low-data scenarios, but high demands on computing resources limit its practicality. We address this issue by introducing a prompt-based parameter-efficient fine-tuning (PEFT) approach. GNNavi leverages insights into ICL’s information flow dynamics, which indicates that label words act in prompts as anchors for information propagation. GNNavi employs a Graph Neural Network (GNN) layer to precisely guide the aggregation and distribution of information flow during the processing of prompts by hardwiring the desired information flow into the GNN. Our experiments on text classification tasks with GPT-2 and Llama2 shows GNNavi surpasses standard prompt-based fine-tuning methods in few-shot settings by updating just 0.2% to 0.5% of parameters. We compare GNNavi with prevalent PEFT approaches, such as prefix tuning, LoRA and Adapter in terms of performance and efficiency. Our analysis reveals that GNNavi enhances information flow and ensures a clear aggregation process.

In-context learning is a popular inference strategy where large language models solve a task using only a few labeled demonstrations without needing any parameter updates. Although there have been extensive studies on English in-context learning, multilingual in-context learning remains under-explored, and we lack an in-depth understanding of the role of demonstrations in this context. To address this gap, we conduct a multidimensional analysis of multilingual in-context learning, experimenting with 5 models from different model families, 9 datasets covering classification and generation tasks, and 56 typologically diverse languages. Our results reveal that the effectiveness of demonstrations varies significantly across models, tasks, and languages. We also find that strong instruction-following models including Llama 2-Chat, GPT-3.5, and GPT-4 are largely insensitive to the quality of demonstrations. Instead, a carefully crafted template often eliminates the benefits of demonstrations for some tasks and languages altogether. These findings show that the importance of demonstrations might be overestimated. Our work highlights the need for granular evaluation across multiple axes towards a better understanding of in-context learning.

Multiple choice question answering tasks evaluate the reasoning, comprehension, and mathematical abilities of Large Language Models (LLMs). While existing benchmarks employ automatic translation for multilingual evaluation, this approach is error-prone and potentially introduces culturally biased questions, especially in social sciences. We introduce the first multitask, multiple-choice Turkish QA benchmark, TurkishMMLU, to evaluate LLMs’ understanding of the Turkish language. TurkishMMLU includes over 10,000 questions, covering 9 different subjects from Turkish high-school education curricula. These questions are written by curriculum experts, suitable for the high-school curricula in Turkey, covering subjects ranging from natural sciences and math questions to more culturally representative topics such as Turkish Literature and the history of the Turkish Republic. We evaluate over 20 LLMs, including multilingual open-source (e.g., Gemma, Llama, MT5), closed-source (GPT 4o, Claude, Gemini), and Turkish-adapted (e.g., Trendyol) models. We provide an extensive evaluation, including zero-shot and few-shot evaluation of LLMs, chain-of-thought reasoning, and question difficulty analysis along with model performance. We provide an in-depth analysis of the Turkish capabilities and limitations of current LLMs to provide insights for future LLMs for the Turkish language.

In this work, we present a collaboratively and continuously developed open-source educational resource (OSER) for teaching natural language processing at two different universities. We shed light on the principles we followed for the initial design of the course and the rationale for ongoing developments, followed by a reflection on the inter-university collaboration for designing and maintaining teaching material. When reflecting on the latter, we explicitly emphasize the considerations that need to be made when facing heterogeneous groups and when having to accommodate multiple examination regulations within one single course framework. Relying on the fundamental principles of OSER developments as defined by Bothmann et al. (2023) proved to be an important guideline during this process. The final part pertains to open-sourcing our teaching material, coping with the increasing speed of developments in the field, and integrating the course digitally, also addressing conflicting priorities and challenges we are currently facing.

Abstract notions are often comprehended through analogies, wherein there exists correspondence or partial similarity with more concrete concepts. A fundamental aspect of human cognition involves synthesising embodied experiences into spatial schemas, which profoundly influence conceptualisation and underlie language acquisition. Recent studies have demonstrated that Large Language Models (LLMs) exhibit certain spatial intuitions akin to human language. For instance, both humans and LLMs tend to associate ↑ with hope more readily than with warn. However, the nuanced partial similarities between concrete (e.g., ↑) and abstract (e.g., hope) concepts, remain insufficiently explored. Therefore, we propose a novel methodology utilising analogical reasoning to elucidate these associations and examine whether LLMs adjust their associations in response to analogy-prompts. We find that analogy-prompting is slightly increasing agreement with human choices and the answers given by models include valid explanations supported by analogies, even when in disagreement with human results.

Recent studies have highlighted the potential of exploiting parallel corpora to enhance multilingual large language models, improving performance in both bilingual tasks, e.g., machine translation, and general-purpose tasks, e.g., text classification. Building upon these findings, our comprehensive study aims to identify the most effective strategies for leveraging parallel corpora. We investigate the impact of parallel corpora quality and quantity, training objectives, and model size on the performance of multilingual large language models enhanced with parallel corpora across diverse languages and tasks. Our analysis reveals several key insights: (i) filtering noisy translations is essential for effectively exploiting parallel corpora, while language identification and short sentence filtering have little effect; (ii) even a corpus containing just 10K parallel sentences can yield results comparable to those obtained from much larger datasets; (iii) employing only the machine translation objective yields the best results among various training objectives and their combinations; (iv) larger multilingual language models benefit more from parallel corpora than smaller models due to their stronger capacity for cross-task transfer. Our study offers valuable insights into the optimal utilization of parallel corpora to enhance multilingual large language models, extending the generalizability of previous findings from limited languages and tasks to a broader range of scenarios.

Decoder-only large language models (LLMs) excel in high-resource languages across various tasks through few-shot or even zero-shot in-context learning (ICL). However, their performance often does not transfer well to low-resource languages, especially those written in non-Latin scripts. Inspired by recent work that leverages transliteration in encoder-only models, we investigate whether transliteration is also effective in improving LLMs’ performance for low-resource languages written in non-Latin scripts. To this end, we propose three prompt templates, where the target-language text is represented in (1) its original script, (2) Latin script, or (3) both. We apply these methods to several representative LLMs of different sizes on various tasks including text classification and sequential labeling. Our findings show that the effectiveness of transliteration varies by task type and model size. For instance, all models benefit from transliterations for sequential labeling (with increases of up to 25%).

Transformer-based pre-trained language models (PLMs) have achieved remarkable performance in various natural language processing (NLP) tasks. However, pre-training such models can take considerable resources that are almost only available to high-resource languages. On the contrary, static word embeddings are easier to train in terms of computing resources and the amount of data required. In this paper, we introduce MoSECroT Model Stitching with Static Word Embeddings for Crosslingual Zero-shot Transfer), a novel and challenging task that is especially relevant to low-resource languages for which static word embeddings are available. To tackle the task, we present the first framework that leverages relative representations to construct a common space for the embeddings of a source language PLM and the static word embeddings of a target language. In this way, we can train the PLM on source-language training data and perform zero-shot transfer to the target language by simply swapping the embedding layer. However, through extensive experiments on two classification datasets, we show that although our proposed framework is competitive with weak baselines when addressing MoSECroT, it fails to achieve competitive results compared with some strong baselines. In this paper, we attempt to explain this negative result and provide several thoughts on possible improvement.

Continual learning aims at incrementally acquiring new knowledge while not forgetting existing knowledge. To overcome catastrophic forgetting, methods are either rehearsal-based, i.e., store data examples from previous tasks for data replay, or isolate parameters dedicated to each task. However, rehearsal-based methods raise privacy and memory issues, and parameter-isolation continual learning does not consider interaction between tasks, thus hindering knowledge transfer. In this work, we propose MoCL, a rehearsal-free Modular and Compositional Continual Learning framework which continually adds new modules to language models and composes them with existing modules. Experiments on various benchmarks show that MoCL outperforms state of the art and effectively facilitates knowledge transfer.

Instead of pretraining multilingual language models from scratch, a more efficient method is to adapt existing pretrained language models (PLMs) to new languages via vocabulary extension and continued pretraining. However, this method usually randomly initializes the embeddings of new subwords and introduces substantially more embedding parameters to the model, thus weakening the efficiency. To address these issues, we propose a novel framework: One For All (OFA), which wisely initializes the embeddings of unseen subwords and thus can adapt a PLM to multiple languages efficiently and effectively. OFA takes advantage of external well-aligned multilingual static word vectors and injects the alignment knowledge into the subword embeddings. In addition, OFA applies matrix factorization and replaces the cumbersome embeddings with two lower-dimensional matrices, which largely reduces the number of parameters. We show OFA accelerates the convergence of continued pretraining, which is environmentally friendly as much fewer carbon footprints are generated. Through extensive experiments, we demonstrate OFA can achieve competitive or better performance than default continued pretraining baselines on a wide range of crosslingual downstream tasks. We make our code and models publicly available.

The development of assistive robotic agents to support household tasks is advancing, yet the underlying models often operate in virtual settings that do not reflect real-world complexity. For assistive care robots to be effective in diverse environments, their models must be robust and integrate multiple modalities. Consider a caretaker needing assistance in a dimly lit room or navigating around a newly installed glass door. Models relying solely on visual input might fail in low light, while those using depth information could avoid the door. This demonstrates the necessity for models that can process various sensory inputs. Our ongoing study evaluates state-of-the-art robotic models in the AI2Thor virtual environment. We introduce disturbances, such as dimmed lighting and mirrored walls, to assess their impact on modalities like movement or vision, and object recognition. Our goal is to gather input from the Geriatronics community to understand and model the challenges faced by practitioners.

Recent studies indicate that leveraging off-the-shelf or fine-tuned retrievers, capable of retrieving relevant in-context examples tailored to the input query, enhances few-shot in-context learning of English. However, adapting these methods to other languages, especially low-resource ones, poses challenges due to the scarcity of cross-lingual retrievers and annotated data. Thus, we introduce XAMPLER: Cross-Lingual Example Retrieval, a method tailored to tackle the challenge of cross-lingual in-context learning using only annotated English data. XAMPLER first trains a retriever based on Glot500, a multilingual small language model, using positive and negative English examples constructed from the predictions of a multilingual large language model, i.e., MaLA500. Leveraging the cross-lingual capacity of the retriever, it can directly retrieve English examples as few-shot examples for in-context learning of target languages. Experiments on the multilingual text classification benchmark SIB200 with 176 languages show that XAMPLER substantially improves the in-context learning performance across languages.

While natural language processing tools have been developed extensively for some of the world’s languages, a significant portion of the world’s over 7000 languages are still neglected. One reason for this is that evaluation datasets do not yet cover a wide range of languages, including low-resource and endangered ones. We aim to address this issue by creating a text classification dataset encompassing a large number of languages, many of which currently have little to no annotated data available. We leverage parallel translations of the Bible to construct such a dataset by first developing applicable topics and employing a crowdsourcing tool to collect annotated data. By annotating the English side of the data and projecting the labels onto other languages through aligned verses, we generate text classification datasets for more than 1500 languages. We extensively benchmark several existing multilingual language models using our dataset. To facilitate the advancement of research in this area, we will release our dataset and code.

Large language models (LLMs) possess extensive parametric knowledge, but this knowledge is difficult to update with new information because retraining is very expensive and infeasible for closed-source models. Knowledge editing (KE) has emerged as a viable solution for updating the knowledge of LLMs without compromising their overall performance. On-the-fly KE methods, inspired by in-context learning (ICL), have shown great promise and allow LLMs to be treated as black boxes. In the past, KE was primarily employed in English contexts, whereas the potential for cross-lingual KE in current English-centric LLMs has not been fully explored. To foster more research in this direction, we introduce the BMIKE-53 benchmark for evaluating cross-lingual KE on 53 diverse languages across three KE task types. We also propose a gradient-free KE method called Multilingual In-context Knowledge Editing (MIKE) and evaluate it on BMIKE-53. Our evaluation focuses on cross-lingual knowledge transfer in terms of reliability, generality, locality, and portability, offering valuable insights and a framework for future research in cross-lingual KE.

In real-world environments, continual learning is essential for machine learning models, as they need to acquire new knowledge incrementally without forgetting what they have already learned. While pretrained language models have shown impressive capabilities on various static tasks, applying them to continual learning poses significant challenges, including avoiding catastrophic forgetting, facilitating knowledge transfer, and maintaining parameter efficiency. In this paper, we introduce MoCL-P, a novel lightweight continual learning method that addresses these challenges simultaneously. Unlike traditional approaches that continuously expand parameters for newly arriving tasks, MoCL-P integrates task representation-guided module composition with adaptive pruning, effectively balancing knowledge integration and computational overhead. Our evaluation across three continual learning benchmarks with up to 176 tasks shows that MoCL-P achieves state-of-the-art performance and improves parameter efficiency by up to three times, demonstrating its potential for practical applications where resource requirements are constrained.

Despite the success of the Universal Dependencies (UD) project exemplified by its impressive language breadth, there is still a lack in `within-language breadth’: most treebanks focus on standard languages. Even for German, the language with the most annotations in UD, so far no treebank exists for one of its language varieties spoken by over 10M people: Bavarian. To contribute to closing this gap, we present the first multi-dialect Bavarian treebank (MaiBaam) manually annotated with part-of-speech and syntactic dependency information in UD, covering multiple text genres (wiki, fiction, grammar examples, social, non-fiction). We highlight the morphosyntactic differences between the closely-related Bavarian and German and showcase the rich variability of speakers’ orthographies. Our corpus includes 15k tokens, covering dialects from all Bavarian-speaking areas spanning three countries. We provide baseline parsing and POS tagging results, which are lower than results obtained on German and vary substantially between different graph-based parsers. To support further research on Bavarian syntax, we make our dataset, language-specific guidelines and code publicly available.

We present GlotScript, an open resource and tool for low resource writing system identification. GlotScript-R is a resource that provides the attested writing systems for more than 7,000 languages. It is compiled by aggregating information from existing writing system resources. GlotScript-T is a writing system identification tool that covers all 161 Unicode 15.0 scripts. For an input text, it returns its script distribution where scripts are identified by ISO 15924 codes. We also present two use cases for GlotScript. First, we demonstrate that GlotScript can help cleaning multilingual corpora such as mC4 and OSCAR. Second, we analyze the tokenization of a number of language models such as GPT-4 using GlotScript and provide insights on the coverage of low resource scripts and languages by each language model. We hope that GlotScript will become a useful resource for work on low resource languages in the NLP community.

Word alignments are essential for a variety of NLP tasks. Therefore, choosing the best approaches for their creation is crucial. However, the scarce availability of gold evaluation data makes the choice difficult. We propose SilverAlign, a new method to automatically create silver data for the evaluation of word aligners by exploiting machine translation and minimal pairs. We show that performance on our silver data correlates well with gold benchmarks for 9 language pairs, making our approach a valid resource for evaluation of different domains and languages when gold data are not available. This addresses the important scenario of missing gold data alignments for low-resource languages.

Lexical-syntactic flexibility, in the form of conversion (or zero-derivation) is a hallmark of English morphology. In conversion, a word with one part of speech is placed in a non-prototypical context, where it is coerced to behave as if it had a different part of speech. However, while this process affects a large part of the English lexicon, little work has been done to establish the degree to which language models capture this type of generalization. This paper reports the first study on the behavior of large language models with reference to conversion. We design a task for testing lexical-syntactic flexibility—the degree to which models can generalize over words in a construction with a non-prototypical part of speech. This task is situated within a natural language inference paradigm. We test the abilities of five language models—two proprietary models (GPT-3.5 and GPT-4), three open source model (Mistral 7B, Falcon 40B, and Llama 2 70B). We find that GPT-4 performs best on the task, followed by GPT-3.5, but that the open source language models are also able to perform it and that the 7-billion parameter Mistral displays as little difference between its baseline performance on the natural language inference task and the non-prototypical syntactic category task, as the massive GPT-4.

The Universal Dependencies (UD) project has created an invaluable collection of treebanks with contributions in over 140 languages. However, the UD annotations do not tell the full story. Grammatical constructions that convey meaning through a particular combination of several morphosyntactic elements – for example, interrogative sentences with special markers and/or word orders – are not labeled holistically. We argue for (i) augmenting UD annotations with a ‘UCxn’ annotation layer for such meaning-bearing grammatical constructions, and (ii) approaching this in a typologically informed way so that morphosyntactic strategies can be compared across languages. As a case study, we consider five construction families in ten languages, identifying instances of each construction in UD treebanks through the use of morphosyntactic patterns. In addition to findings regarding these particular constructions, our study yields important insights on methodology for describing and identifying constructions in language-general and language-particular ways, and lays the foundation for future constructional enrichment of UD treebanks.

In this paper, we make a contribution that can be understood from two perspectives: from an NLP perspective, we introduce a small challenge dataset for NLI with large lexical overlap, which minimises the possibility of models discerning entailment solely based on token distinctions, and show that GPT-4 and Llama 2 fail it with strong bias. We then create further challenging sub-tasks in an effort to explain this failure. From a Computational Linguistics perspective, we identify a group of constructions with three classes of adjectives which cannot be distinguished by surface features. This enables us to probe for LLM’s understanding of these constructions in various ways, and we find that they fail in a variety of ways to distinguish between them, suggesting that they don’t adequately represent their meaning or capture the lexical properties of phrasal heads.

Large language models (LLMs) have advanced the state of the art in natural language processing. However, their predominant design for English or a limited set of languages creates a substantial gap in their effectiveness for low-resource languages. To bridge this gap, we introduce MaLA-500, a novel large language model designed to cover an extensive range of 534 languages. To train MaLA-500, we employ vocabulary extension and continued pretraining on LLaMA 2 with Glot500-c. Our intrinsic evaluation demonstrates that MaLA-500 is better at predicting the given texts of low-resource languages than existing multilingual LLMs. Moreover, the extrinsic evaluation of in-context learning shows that MaLA-500 outperforms previous LLMs on SIB200 and Taxi1500 by a significant margin, i.e., 11.68% and 4.82% marco-average accuracy across languages.

While current large language models (LLMs) demonstrate some capabilities in knowledge-intensive tasks, they are limited by relying on their parameters as an implicit storage mechanism. As a result, they struggle with infrequent knowledge and temporal degradation. In addition, the uninterpretable nature of parametric memorization makes it challenging to understand and prevent hallucination. Parametric memory pools and model editing are only partial solutions. Retrieval Augmented Generation (RAG) – though non-parametric – has its own limitations: it lacks structure, complicates interpretability and makes it hard to effectively manage stored knowledge. In this paper, we introduce MemLLM, a novel method of enhancing LLMs by integrating a structured and explicit read-and-write memory module. MemLLM tackles the aforementioned challenges by enabling dynamic interaction with the memory and improving the LLM’s capabilities in using stored knowledge. Our experiments indicate that MemLLM enhances the LLM’s performance and interpretability, in language modeling in general and knowledge-intensive tasks in particular. We see MemLLM as an important step towards making LLMs more grounded and factual through memory augmentation.

We introduce HATELEXICON, a lexicon of slurs and targets of hate speech for Brazil, Germany, India and Kenya, to aid model development and interpretability. First, we demonstrate how HATELEXICON can be used to interpret model predictions, showing that models developed to classify extreme speech rely heavily on target group names. Further, we propose a culturally-informed method to aid shot selection for training in low-resource settings. In few-shot learning, shot selection is of paramount importance to model performance and we need to ensure we make the most of available data. We work with HASOC German and Hindi data for training and the Multilingual HateCheck (MHC) benchmark for evaluation. We show that selecting shots based on our lexicon leads to models performing better than models trained on shots sampled randomly. Thus, when given only a few training examples, using HATELEXICON to select shots containing more sociocultural information leads to better few-shot performance. With these two use-cases we show how our HATELEXICON can be used for more effective hate speech detection.

Prompt-based methods have been successfully applied to multilingual pretrained language models for zero-shot cross-lingual understanding. However, most previous studies primarily focused on sentence-level classification tasks, and only a few considered token-level labeling tasks such as Named Entity Recognition (NER) and Part-of-Speech (POS) tagging. In this paper, we propose Token-Level Prompt Decomposition (ToPro), which facilitates the prompt-based method for token-level sequence labeling tasks. The ToPro method decomposes an input sentence into single tokens and applies one prompt template to each token. Our experiments on multilingual NER and POS tagging datasets demonstrate that ToPro-based fine-tuning outperforms Vanilla fine-tuning and Prompt-Tuning in zero-shot cross-lingual transfer, especially for languages that are typologically different from the source language English. Our method also attains state-of-the-art performance when employed with the mT5 model. Besides, our exploratory study in multilingual large language models shows that ToPro performs much better than the current in-context learning method. Overall, the performance improvements show that ToPro could potentially serve as a novel and simple benchmarking method for sequence labeling tasks.

Cross-lingual transfer (XLT) driven by massively multilingual language models (mmLMs) has been shown largely ineffective for low-resource (LR) target languages with little (or no) representation in mmLM’s pretraining, especially if they are linguistically distant from the high-resource (HR) source language. Much of the recent focus in XLT research has been dedicated to LR language families, i.e., families without any HR languages (e.g., families of African languages or indigenous languages of the Americas). In this work, in contrast, we investigate a configuration that is arguably of practical relevance for more of the world’s languages: XLT to LR languages that do have a close HR relative. To explore the extent to which a HR language can facilitate transfer to its LR relatives, we (1) introduce Kardeş-NLU, an evaluation benchmark with language understanding datasets in five LR Turkic languages: Azerbaijani, Kazakh, Kyrgyz, Uzbek, and Uyghur; and (2) investigate (a) intermediate training and (b) fine-tuning strategies that leverage Turkish in XLT to these target languages. Our experimental results show that both - integrating Turkish in intermediate training and in downstream fine-tuning - yield substantial improvements in XLT to LR Turkic languages. Finally, we benchmark cutting-edge instruction-tuned large language models on Kardeş-NLU, showing that their performance is highly task- and language-dependent.

Recent multilingual pretrained language models (mPLMs) have been shown to encode strong language-specific signals, which are not explicitly provided during pretraining. It remains an open question whether it is feasible to employ mPLMs to measure language similarity, and subsequently use the similarity results to select source languages for boosting cross-lingual transfer. To investigate this, we propose mPLM-Sim, a language similarity measure that induces the similarities across languages from mPLMs using multi-parallel corpora. Our study shows that mPLM-Sim exhibits moderately high correlations with linguistic similarity measures, such as lexicostatistics, genealogical language family, and geographical sprachbund. We also conduct a case study on languages with low correlation and observe that mPLM-Sim yields more accurate similarity results. Additionally, we find that similarity results vary across different mPLMs and different layers within an mPLM. We further investigate whether mPLM-Sim is effective for zero-shot cross-lingual transfer by conducting experiments on both low-level syntactic tasks and high-level semantic tasks. The experimental results demonstrate that mPLM-Sim is capable of selecting better source languages than linguistic measures, resulting in a 1%-2% improvement in zero-shot cross-lingual transfer performance.

Argument Structure Constructions (ASCs) are one of the most well-studied construction groups, providing a unique opportunity to demonstrate the usefulness of Construction Grammar (CxG). For example, the caused-motion construction (CMC, She sneezed the foam off her cappuccino'') demonstrates that constructions must carry meaning, otherwise the fact that sneeze’’ in this context causes movement cannot be explained. We form the hypothesis that this remains challenging even for state-of-the-art Large Language Models (LLMs), for which we devise a test based on substituting the verb with a prototypical motion verb. To be able to perform this test at statistically significant scale, in the absence of adequate CxG corpora, we develop a novel pipeline of NLP-assisted collection of linguistically annotated text. We show how dependency parsing and GPT-3.5 can be used to significantly reduce annotation cost and thus enable the annotation of rare phenomena at scale. We then evaluate GPT, Gemini, Llama2 and Mistral models for their understanding of the CMC using the newly collected corpus. We find that all models struggle with understanding the motion component that the CMC adds to a sentence.

Despite the predominance of English in their training data, English-centric Large Language Models (LLMs) like GPT-3 and LLaMA display a remarkable ability to perform multilingual tasks, raising questions about the depth and nature of their cross-lingual capabilities. This paper introduces the decomposed prompting approach to probe the linguistic structure understanding of these LLMs in sequence labeling tasks. Diverging from the single text-to-text prompt, our method generates for each token of the input sentence an individual prompt which asks for its linguistic label. We assess our method on the Universal Dependencies part-of-speech tagging dataset for 38 languages, utilizing both English-centric and multilingual LLMs. Our findings show that decomposed prompting surpasses the iterative prompting baseline in efficacy and efficiency under zero- and few-shot settings. Further analysis reveals the influence of evaluation methods and the use of instructions in prompts. Our multilingual investigation shows that English-centric language models perform better on average than multilingual models. Our study offers insights into the multilingual transferability of English-centric LLMs, contributing to the understanding of their multilingual linguistic knowledge.

The convergence of embodied agents and large language models (LLMs) has brought significant advancements to embodied instruction following.Particularly, the strong reasoning capabilities of LLMs make it possible for robots to perform long-horizon tasks without expensive annotated demonstrations.However, public benchmarks for testing the long-horizon reasoning capabilities of language-conditioned robots in various scenarios are still missing. To fill this gap, this work focuses on the tabletopmanipulation task and releases a simulation benchmark,textit{LoHoRavens}, which covers various long-horizonreasoning aspects spanning color, size, space, arithmeticsand reference.Furthermore, there is a key modality bridging problem forlong-horizon manipulation tasks with LLMs: how toincorporate the observation feedback during robot executionfor the LLM’s closed-loop planning, which is however less studied by prior work. We investigate two methods of bridging the modality gap: caption generation and learnable interface for incorporating explicit and implicit observation feedback to the LLM, respectively.These methods serve as the two baselines for our proposed benchmark. Experiments show that both methods struggle to solve most tasks, indicating long-horizon manipulation tasks are still challenging for current popular models.We expect the proposed public benchmark and baselines can help the community develop better models for long-horizon tabletop manipulation tasks.

The remarkable ability of Large Language Models (LLMs) to understand and follow instructions has sometimes been limited by their in-context learning (ICL) performance in low-resource languages. To address this, we introduce a novel approach that leverages cross-lingual retrieval-augmented in-context learning (CREA-ICL). By extracting semantically similar prompts from high-resource languages, we aim to bolster the zero-shot performance of multilingual pretrained language models (MPLMs) across diverse tasks. Though our approach yields steady improvements in classification tasks, it faces challenges in generation tasks, with Bangla serving as a key case study. Our evaluation offers insights into the performance dynamics of retrieval-augmented in-context learning across both classification and generation domains.

The promise of Large Language Models (LLMs) in Natural Language Processing has often been overshadowed by their limited performance in low-resource languages such as Bangla. To address this, our paper presents a pioneering approach that utilizes cross-lingual retrieval augmented in-context learning. By strategically sourcing semantically similar prompts from high-resource language, we enable multilingual pretrained language models (MPLMs), especially the generative model BLOOMZ, to successfully boost performance on Bangla tasks. Our extensive evaluation highlights that the cross-lingual retrieval augmented prompts bring steady improvements to MPLMs over the zero-shot performance.

While large language models (LLMs) are proficient at question-answering (QA), it is not always clear how (or even if) an answer follows from their latent ‘beliefs’. This lack of interpretability is a growing impediment to widespread use of LLMs. To address this, our goals are to make model beliefs and their inferential relationships explicit, and to resolve inconsistencies that may exist, so that answers are supported by interpretable chains of reasoning drawn from a consistent network of beliefs. Our approach, which we call REFLEX, is to add a rational, self-reflecting layer on top of the LLM. First, given a question, we construct a belief graph using a backward-chaining process to materialize relevant model beliefs (including beliefs about answer candidates) and their inferential relationships. Second, we identify and minimize contradictions in that graph using a formal constraint reasoner. We find that REFLEX significantly improves consistency (by 8%-11% absolute) without harming overall answer accuracy, resulting in answers supported by faithful chains of reasoning drawn from a more consistent belief system. This suggests a new style of system architecture in which an LLM extended with a rational layer can provide an interpretable window into system beliefs, add a systematic reasoning capability, and repair latent inconsistencies present in the LLM.

Most languages of the world pose low-resource challenges to natural language processing models. With multilingual training, knowledge can be shared among languages. However, not all languages positively influence each other and it is an open research question how to select the most suitable set of languages for multilingual training and avoid negative interference among languages whose characteristics or data distributions are not compatible. In this paper, we propose GradSim, a language grouping method based on gradient similarity. Our experiments on three diverse multilingual benchmark datasets show that it leads to the largest performance gains compared to other similarity measures and it is better correlated with cross-lingual model performance. As a result, we set the new state of the art on AfriSenti, a benchmark dataset for sentiment analysis on low-resource African languages. In our extensive analysis, we further reveal that besides linguistic features, the topics of the datasets play an important role for language grouping and that lower layers of transformer models encode language-specific features while higher layers capture task-specific information.

Large language models (LLMs) have recently reached an impressive level of linguistic capability, prompting comparisons with human language skills. However, there have been relatively few systematic inquiries into the linguistic capabilities of the latest generation of LLMs, and those studies that do exist (i) ignore the remarkable ability of humans to generalize, (ii) focus only on English, and (iii) investigate syntax or semantics and overlook other capabilities that lie at the heart of human language, like morphology. Here, we close these gaps by conducting the first rigorous analysis of the morphological capabilities of ChatGPT in four typologically varied languages (specifically, English, German, Tamil, and Turkish). We apply a version of Berko’s (1958) wug test to ChatGPT, using novel, uncontaminated datasets for the four examined languages. We find that ChatGPT massively underperforms purpose-built systems, particularly in English. Overall, our results—through the lens of morphology—cast a new light on the linguistic capabilities of ChatGPT, suggesting that claims of human-like language skills are premature and misleading.

Several recent papers have published good solutions for language identification (LID) for about 300 high-resource and medium-resource languages. However, there is no LID available that (i) covers a wide range of low-resource languages, (ii) is rigorously evaluated and reliable and (iii) efficient and easy to use. Here, we publish GlotLID-M, an LID model that satisfies the desiderata of wide coverage, reliability and efficiency. It identifies 1665 languages, a large increase in coverage compared to prior work. In our experiments, GlotLID-M outperforms four baselines (CLD3, FT176, OpenLID and NLLB) when balancing F1 and false positive rate (FPR). We analyze the unique challenges that low-resource LID poses: incorrect corpus metadata, leakage from high-resource languages, difficulty separating closely related languages, handling of macrolanguage vs varieties and in general noisy data. We hope that integrating GlotLID-M into dataset creation pipelines will improve quality and enhance accessibility of NLP technology for low-resource languages and cultures.

Few-shot classification has made great strides due to foundation models that, through priming and prompting, are highly effective few-shot learners. However, this approach has high variance both across different sets of few shots (data selection) and across different finetuning runs (run variability). This is problematic not only because it impedes the fair comparison of different approaches, but especially because it makes few-shot learning too unreliable for many real-world applications. To alleviate these issues, we make two contributions for more stable and effective few-shot learning: First, we propose novel ensembling methods and show that they substantially reduce run variability. Second, we introduce a new active learning (AL) criterion for data selection and present the first AL-based approach specifically tailored towards prompt-based learning. In our experiments, we show that our combined method, MEAL (Multiprompt finetuning and prediction Ensembling with Active Learning), improves overall performance of prompt-based finetuning by 2.3 points on five diverse tasks.

Pretrained language models (PLMs) are key components in NLP, but they contain strong social biases. Quantifying these biases is challenging because current methods focusing on fill-the-mask objectives are sensitive to slight changes in input. To address this, we propose a bias probing technique called LABDet, for evaluating social bias in PLMs with a robust and language-agnostic method. For nationality as a case study, we show that LABDet “surfaces” nationality bias by training a classifier on top of a frozen PLM on non-nationality sentiment detection. We find consistent patterns of nationality bias across monolingual PLMs in six languages that align with historical and political context. We also show for English BERT that bias surfaced by LABDet correlates well with bias in the pretraining data; thus, our work is one of the few studies that directly links pretraining data to PLM behavior. Finally, we verify LABDet’s reliability and applicability to different templates and languages through an extensive set of robustness checks.

In comparative linguistics, colexification refers to the phenomenon of a lexical form conveying two or more distinct meanings. Existing work on colexification patterns relies on annotated word lists, limiting scalability and usefulness in NLP. In contrast, we identify colexification patterns of more than 2,000 concepts across 1,335 languages directly from an unannotated parallel corpus. We then propose simple and effective methods to build multilingual graphs from the colexification patterns: ColexNet and ColexNet+. ColexNet’s nodes are concepts and its edges are colexifications. In ColexNet+, concept nodes are additionally linked through intermediate nodes, each representing an ngram in one of 1,334 languages. We use ColexNet+ to train ColexNet+, high-quality multilingual embeddings that are well-suited for transfer learning. In our experiments, we first show that ColexNet achieves high recall on CLICS, a dataset of crosslingual colexifications. We then evaluate ColexNet+ on roundtrip translation, sentence retrieval and sentence classification and show that our embeddings surpass several transfer learning baselines. This demonstrates the benefits of using colexification as a source of information in multilingual NLP.

Pretrained multilingual encoder models can directly perform zero-shot multilingual tasks or linguistic probing by reformulating the input examples into cloze-style prompts. This is accomplished by predicting the probabilities of the label words at the masked token position, without requiring any updates to the model parameters. However, the performance of this method is limited by the model’s bias toward predicting label words which frequently occurred during the pretraining. These words typically receive high probabilities. To address this issue, we combine the models with calibration techniques which modify the probabilities of label words predicted by the models. We first validate the effectiveness of a proposed simple calibration method together with other existing techniques on monolingual encoders in both zero- and few-shot scenarios. We subsequently employ these calibration techniques on multilingual encoders, resulting in substantial performance improvements across a wide range of tasks.

Very low-resource languages, having only a few million tokens worth of data, are not well-supported by multilingual NLP approaches due to poor quality cross-lingual word representations. Recent work showed that good crosslingual performance can be achieved if a source language is related to the low-resource target language. However, not all language pairs are related. In this paper, we propose to build multilingual word embeddings (MWEs) via a novel language chain-based approach, that incorporates intermediate related languages to bridge the gap between the distant source and target. We build MWEs one language at a time by starting from the resource rich source and sequentially adding each language in the chain till we reach the target. We extend a semi-joint bilingual approach to multiple languages in order to eliminate the main weakness of previous works, i.e., independently trained monolingual embeddings, by anchoring the target language around the multilingual space. We evaluate our method on bilingual lexicon induction for 4 language families, involving 4 very low-resource (≤ 5M tokens) and 4 moderately low-resource (≤ 50M) target languages, showing improved performance in both categories. Additionally, our analysis reveals the importance of good quality embeddings for intermediate languages as well as the importance of leveraging anchor points from all languages in the multilingual space.

Attribution is a key concept in large language models (LLMs) as it enables control over information sources and enhances the factuality of LLMs. While existing approaches utilize open book question answering to improve attribution, factual datasets may reward language models to recall facts that they already know from their pretraining data, not attribution. In contrast, counterfactual open book QA datasets would further improve attribution because the answer could only be grounded in the given text. We propose Hallucination Augmented Recitations (HAR) for creating counterfactual datasets by utilizing hallucination in LLMs to improve attribution. For open book QA as a case study, we demonstrate that models finetuned with our counterfactual datasets improve text grounding, leading to better open book QA performance, with up to an 8.0% increase in F1 score. Our counterfactual dataset leads to significantly better performance than using humanannotated factual datasets, even with 4x smaller datasets and 4x smaller models. We observe that improvements are consistent across various model sizes and datasets, including multi-hop, biomedical, and adversarial QA datasets.

Construction Grammar (CxG) is a paradigm from cognitive linguistics emphasizing the connection between syntax and semantics. Rather than rules that operate on lexical items, it posits constructions as the central building blocks of language, i.e., linguistic units of different granularity that combine syntax and semantics. As a first step toward assessing the compatibility of CxG with the syntactic and semantic knowledge demonstrated by state-of-the-art pretrained language models (PLMs), we present an investigation of their capability to classify and understand one of the most commonly studied constructions, the English comparative correlative (CC). We conduct experiments examining the classification accuracy of a syntactic probe on the one hand and the models’ behavior in a semantic application task on the other, with BERT, RoBERTa, and DeBERTa as the example PLMs. Our results show that all three investigated PLMs, as well as OPT, are able to recognize the structure of the CC but fail to use its meaning. While human-like performance of PLMs on many NLP tasks has been alleged, this indicates that PLMs still suffer from substantial shortcomings in central domains of linguistic knowledge.

Multilingual pretrained language models (MPLMs) have demonstrated substantial performance improvements in zero-shot cross-lingual transfer across various natural language understanding tasks by finetuning MPLMs on task-specific labelled data of a source language (e.g. English) and evaluating on a wide range of target languages. Recent studies show that prompt-based finetuning surpasses regular finetuning in few-shot scenarios. However, the exploration of prompt-based learning in multilingual tasks remains limited. In this study, we propose the PROFIT pipeline to investigate the cross-lingual capabilities of Prompt-based Finetuning. We conduct comprehensive experiments on diverse cross-lingual language understanding tasks (sentiment classification, paraphrase identification, and natural language inference) and empirically analyze the variation trends of prompt-based finetuning performance in cross-lingual transfer across different few-shot and full-data settings. Our results reveal the effectiveness and versatility of prompt-based finetuning in cross-lingual language understanding. Our findings indicate that prompt-based finetuning outperforms vanilla finetuning in full-data scenarios and exhibits greater advantages in few-shot scenarios, with different performance patterns dependent on task types. Additionally, we analyze underlying factors such as language similarity and pretraining data size that impact the cross-lingual performance of prompt-based finetuning. Overall, our work provides valuable insights into the cross-lingual prowess of prompt-based finetuning.

In industry settings, machine learning is an attractive tool to automatize processes. Unfortunately, annotated and high-quality data is expensive to source. This problem is exacerbated in settings spanning multiple markets and languages. Thus, developing solutions for multilingual tasks with little available data is challenging. Few-shot learning is a compelling approach when building solutions in multilingual and low-resource settings, since the method not only requires just a few training examples to achieve high performance, but is also a technique agnostic to language. Even though the technique can be applied to multilingual settings, optimizing performance is an open question. In our work we show that leveraging higher-resource, task-specific language data can boost overall performance and we propose a method to select training examples per their average performance in a Monte Carlo simulation, resulting in a training set more conducive to learning. We demonstrate the effectiveness of our methods in fashion text reviews moderation, classifying reviews as related or unrelated to the given product. We show that our methodology boosts performance in multilingual (English, French, German) settings, increasing F1 score and significantly decreasing false positives.

Constituency parsing plays a fundamental role in advancing natural language processing (NLP) tasks. However, training an automatic syntactic analysis system for ancient languages solely relying on annotated parse data is a formidable task due to the inherent challenges in building treebanks for such languages. It demands extensive linguistic expertise, leading to a scarcity of available resources. To overcome this hurdle, cross-lingual transfer techniques which require minimal or even no annotated data for low-resource target languages offer a promising solution. In this study, we focus on building a constituency parser for Middle High German (MHG) under realistic conditions, where no annotated MHG treebank is available for training. In our approach, we leverage the linguistic continuity and structural similarity between MHG and Modern German (MG), along with the abundance of MG treebank resources. Specifically, by employing the delexicalization method, we train a constituency parser on MG parse datasets and perform cross-lingual transfer to MHG parsing. Our delexicalized constituency parser demonstrates remarkable performance on the MHG test set, achieving an F1-score of 67.3%. It outperforms the best zero-shot cross-lingual baseline by a margin of 28.6% points. The encouraging results underscore the practicality and potential for automatic syntactic analysis in other ancient languages that face similar challenges as MHG.

The NLP community has mainly focused on scaling Large Language Models (LLMs) vertically, i.e., making them better for about 100 languages. We instead scale LLMs horizontally: we create, through continued pretraining, Glot500-m, an LLM that covers 511 predominantly low-resource languages. An important part of this effort is to collect and clean Glot500-c, a corpus that covers these 511 languages and allows us to train Glot500-m. We evaluate Glot500-m on five diverse tasks across these languages. We observe large improvements for both high-resource and low-resource languages compared to an XLM-R baseline. Our analysis shows that no single factor explains the quality of multilingual LLM representations. Rather, a combination of factors determines quality including corpus size, script, ‘help’ from related languages and the total capacity of the model. Our work addresses an important goal of NLP research: we should notlimit NLP to a small fraction of the world’s languages and instead strive to support as many languages as possible to bring the benefits of NLP technology to all languages and cultures.

Languages differ in how they divide up the world into concepts and words; e.g., in contrast to English, Swahili has a single concept for ‘belly’ and ‘womb’. We investigate these differences in conceptualization across 1,335 languages by aligning concepts in a parallel corpus. To this end, we propose Conceptualizer, a method that creates a bipartite directed alignment graph between source language concepts and sets of target language strings. In a detailed linguistic analysis across all languages for one concept (‘bird’) and an evaluation on gold standard data for 32 Swadesh concepts, we show that Conceptualizer has good alignment accuracy. We demonstrate the potential of research on conceptualization in NLP with two experiments. (1) We define crosslingual stability of a concept as the degree to which it has 1-1 correspondences across languages, and show that concreteness predicts stability. (2) We represent each language by its conceptualization pattern for 83 concepts, and define a similarity measure on these representations. The resulting measure for the conceptual similarity between two languages is complementary to standard genealogical, typological, and surface similarity measures. For four out of six language families, we can assign languages to their correct family based on conceptual similarity with accuracies between 54% and 87%.

We investigate response generation for multi-turn dialogue in generative chatbots. Existing generative modelsbased on RNNs (Recurrent Neural Networks) usually employ the last hidden state to summarize the history, which makesmodels unable to capture the subtle variability observed in different dialogues and cannot distinguish the differencesbetween dialogues that are similar in composition. In this paper, we propose Pseudo-Variational Gated Recurrent Unit (PVGRU). The key novelty of PVGRU is a recurrent summarizing variable thataggregates the accumulated distribution variations of subsequences. We train PVGRU without relying on posterior knowledge, thus avoiding the training-inference inconsistency problem. PVGRU can perceive subtle semantic variability through summarizing variables that are optimized by two objectives we employ for training: distribution consistency and reconstruction. In addition, we build a Pseudo-Variational Hierarchical Dialogue(PVHD) model based on PVGRU. Experimental results demonstrate that PVGRU can broadly improve the diversity andrelevance of responses on two benchmark datasets.

An emerging solution for explaining Transformer-based models is to use vector-based analysis on how the representations are formed. However, providing a faithful vector-based explanation for a multi-layer model could be challenging in three aspects: (1) Incorporating all components into the analysis, (2) Aggregating the layer dynamics to determine the information flow and mixture throughout the entire model, and (3) Identifying the connection between the vector-based analysis and the model’s predictions. In this paper, we present DecompX to tackle these challenges. DecompX is based on the construction of decomposed token representations and their successive propagation throughout the model without mixing them in between layers. Additionally, our proposal provides multiple advantages over existing solutions for its inclusion of all encoder components (especially nonlinear feed-forward networks) and the classification head. The former allows acquiring precise vectors while the latter transforms the decomposition into meaningful prediction-based values, eliminating the need for norm- or summation-based vector aggregation. According to the standard faithfulness evaluations, DecompX consistently outperforms existing gradient-based and vector-based approaches on various datasets.

Since conventional knowledge embedding models cannot take full advantage of the abundant textual information, there have been extensive research efforts in enhancing knowledge embedding using texts. However, existing enhancement approaches cannot apply to temporal knowledge graphs (tKGs), which contain time-dependent event knowledge with complex temporal dynamics. Specifically, existing enhancement approaches often assume knowledge embedding is time-independent. In contrast, the entity embedding in tKG models usually evolves, which poses the challenge of aligning temporally relevant texts with entities. To this end, we propose to study enhancing temporal knowledge embedding with textual data in this paper. As an approach to this task, we propose Enhanced Temporal Knowledge Embeddings with Contextualized Language Representations (ECOLA), which takes the temporal aspect into account and injects textual information into temporal knowledge embedding. To evaluate ECOLA, we introduce three new datasets for training and evaluating ECOLA. Extensive experiments show that ECOLA significantly enhances temporal KG embedding models with up to 287% relative improvements regarding Hits@1 on the link prediction task.

Multilingual Pretrained Language Models (MPLMs) perform strongly in cross-lingual transfer. We propose Prompts Augmented by Retrieval Crosslingually (PARC) to improve zero-shot performance on low-resource languages (LRLs) by augmenting the context with prompts consisting of semantically similar sentences retrieved from a high-resource language (HRL). PARC improves zero-shot performance on three downstream tasks (sentiment classification, topic categorization, natural language inference) with multilingual parallel test sets across 10 LRLs covering 6 language families in unlabeled (+5.1%) and labeled settings (+16.3%). PARC also outperforms finetuning by 3.7%. We find a significant positive correlation between cross-lingual transfer performance on one side, and the similarity between high- and low-resource languages as well as the amount of low-resource pretraining data on the other side. A robustness analysis suggests that PARC has the potential to achieve even stronger performance with more powerful MPLMs.

Figurative language is a challenge for language models since its interpretation is based on the use of words in a way that deviates from their conventional order and meaning. Yet, humans can easily understand and interpret metaphors, similes or idioms as they can be derived from embodied metaphors. Language is a proxy for embodiment and if a metaphor is conventional and lexicalised, it becomes easier for a system without a body to make sense of embodied concepts. Yet, the intricate relation between embodiment and features such as concreteness or age of acquisition has not been studied in the context of figurative language interpretation concerning language models. Hence, the presented study shows how larger language models perform better at interpreting metaphoric sentences when the action of the metaphorical sentence is more embodied. The analysis rules out multicollinearity with other features (e.g. word length or concreteness) and provides initial evidence that larger language models conceptualise embodied concepts to a degree that facilitates figurative language understanding.

Although unsupervised neural machine translation (UNMT) has achieved success in many language pairs, the copying problem, i.e., directly copying some parts of the input sentence as the translation, is common among distant language pairs, especially when low-resource languages are involved. We find this issue is closely related to an unexpected copying behavior during online back-translation (BT). In this work, we propose a simple but effective training schedule that incorporates a language discriminator loss. The loss imposes constraints on the intermediate translation so that the translation is in the desired language. By conducting extensive experiments on different language pairs, including similar and distant, high and low-resource languages, we find that our method alleviates the copying problem, thus improving the translation performance on low-resource languages.

In the field of Geriatronics, enabling effective and transparent communication between humans and robots is crucial for enhancing the acceptance and performance of assistive robots. Our early-stage research project investigates the potential of language-based modulation as a means to improve human-robot interaction. We propose to explore real-time modulation during task execution, leveraging language cues, visual references, and multimodal inputs. By developing transparent and interpretable methods, we aim to enable robots to adapt and respond to language commands, enhancing their usability and flexibility. Through the exchange of insights and knowledge at the workshop, we seek to gather valuable feedback to advance our research and contribute to the development of interactive robotic systems for Geriatronics and beyond.

In efforts to keep up with the rapid progress and use of large language models, gender bias research is becoming more prevalent in NLP. Non-English bias research, however, is still in its infancy with most work focusing on English. In our work, we study how grammatical gender bias relating to politeness levels manifests in Japanese and Korean language models. Linguistic studies in these languages have identified a connection between gender bias and politeness levels, however it is not yet known if language models reproduce these biases. We analyze relative prediction probabilities of the male and female grammatical genders using templates and find that informal polite speech is most indicative of the female grammatical gender, while rude and formal speech is most indicative of the male grammatical gender. Further, we find politeness levels to be an attack vector for allocational gender bias in cyberbullying detection models. Cyberbullies can evade detection through simple techniques abusing politeness levels. We introduce an attack dataset to (i) identify representational gender bias across politeness levels, (ii) demonstrate how gender biases can be abused to bypass cyberbullying detection models and (iii) show that allocational biases can be mitigated via training on our proposed dataset. Through our findings we highlight the importance of bias research moving beyond its current English-centrism.

Despite much progress in recent years, the vast majority of work in natural language processing (NLP) is on standard languages with many speakers. In this work, we instead focus on low-resource languages and in particular non-standardized low-resource languages. Even within branches of major language families, often considered well-researched, little is known about the extent and type of available resources and what the major NLP challenges are for these language varieties. The first step to address this situation is a systematic survey of available corpora (most importantly, annotated corpora, which are particularly valuable for NLP research). Focusing on Germanic low-resource language varieties, we provide such a survey in this paper. Except for geolocation (origin of speaker or document), we find that manually annotated linguistic resources are sparse and, if they exist, mostly cover morphosyntax. Despite this lack of resources, we observe that interest in this area is increasing: there is active development and a growing research community. To facilitate research, we make our overview of over 80 corpora publicly available.

Recently, various intermediate layer distillation (ILD) objectives have been shown to improve compression of BERT models via Knowledge Distillation (KD). However, a comprehensive evaluation of the objectives in both task-specific and task-agnostic settings is lacking. To the best of our knowledge, this is the first work comprehensively evaluating distillation objectives in both settings. We show that attention transfer gives the best performance overall. We also study the impact of layer choice when initializing the student from the teacher layers, finding a significant impact on the performance in task-specific distillation. For vanilla KD and hidden states transfer, initialisation with lower layers of the teacher gives a considerable improvement over higher layers, especially on the task of QNLI (up to an absolute percentage change of 17.8 in accuracy). Attention transfer behaves consistently under different initialisation settings. We release our code as an efficient transformer-based model distillation framework for further studies.

One of the challenges with finetuning pretrained language models (PLMs) is that their tokenizer is optimized for the language(s) it was pretrained on, but brittle when it comes to previously unseen variations in the data. This can for instance be observed when finetuning PLMs on one language and evaluating them on data in a closely related language variety with no standardized orthography. Despite the high linguistic similarity, tokenization no longer corresponds to meaningful representations of the target data, leading to low performance in, e.g., part-of-speech tagging. In this work, we finetune PLMs on seven languages from three different families and analyze their zero-shot performance on closely related, non-standardized varieties. We consider different measures for the divergence in the tokenization of the source and target data, and the way they can be adjusted by manipulating the tokenization during the finetuning step. Overall, we find that the similarity between the percentage of words that get split into subwords in the source and target data (the split word ratio difference) is the strongest predictor for model performance on target data.

LLMs (large language models) such as ChatGPT have shown remarkable language understanding and generation capabilities. Although reference-free evaluators based on LLMs show better human alignment than traditional reference-based evaluators, there are many challenges in using reference-free evaluators based on LLMs. Reference-free evaluators are more suitable for open-ended examples with different semantics responses. But not all examples are open-ended. For closed-ended examples with unique correct semantic response, reference-free evaluators will still consider it high quality when giving a response that is inconsistent with the facts and the semantic of reference. In order to comprehensively evaluate the reliability of evaluators based on LLMs, we construct two adversarial meta-evaluation dialogue generation datasets KdConv-ADV and DSTC7-ADV based on KdConv and DSTC7-AVSD, respectively. Compared to previous meta-evaluation benchmarks, KdConv-ADV and DSTC7-ADV are much more challenging since they requires evaluators to be able to reasonably evaluate closed-ended examples with the help of external knowledge or even its own knowledge. Empirical results show that the ability of LLMs to identify unreasonable responses is insufficient. There are risks in using eference-free evaluators based on LLMs to evaluate the quality of dialogue responses.

Large language models (LLMs) have significantly advanced the field of natural language processing (NLP) through their extensive parameters and comprehensive data utilization. However, existing LLMs lack a dedicated memory unit, limiting their ability to explicitly store and retrieve knowledge for various tasks. In this paper, we propose RET-LLM a novel framework that equips LLMs with a general write-read memory unit, allowing them to extract, store, and recall knowledge from the text as needed for task performance. Inspired by Davidsonian semantics theory, we extract and save knowledge in the form of triplets. The memory unit is designed to be scalable, aggregatable, updatable, and interpretable. Through qualitative evaluations, we demonstrate the superiority of our proposed framework over baseline approaches in question answering tasks. Moreover, our framework exhibits robust performance in handling temporal-based question answering tasks, showcasing its ability to effectively manage time-dependent information.

An interesting line of research in natural language processing (NLP) aims to incorporate linguistic typology to bridge linguistic diversity and assist the research of low-resource languages. While most works construct linguistic similarity measures based on lexical or typological features, such as word order and verbal inflection, recent work has introduced a novel approach to defining language similarity based on how they represent basic concepts, which is complementary to existing similarity measures. In this work, we study the conceptual similarity in detail and evaluate it extensively on a binary classification task.

Construction Grammar (CxG) has recently been used as the basis for probing studies that have investigated the performance of large pre-trained language models (PLMs) with respect to the structure and meaning of constructions. In this position paper, we make suggestions for the continuation and augmentation of this line of research. We look at probing methodology that was not designed with CxG in mind, as well as probing methodology that was designed for specific constructions. We analyse selected previous work in detail, and provide our view of the most important challenges and research questions that this promising new field faces.

Humor is a magnetic component in everyday human interactions and communications. Computationally modeling humor enables NLP systems to entertain and engage with users. We investigate the effectiveness of prompting, a new transfer learning paradigm for NLP, for humor recognition. We show that prompting performs similarly to finetuning when numerous annotations are available, but gives stellar performance in low-resource humor recognition. The relationship between humor and offense is also inspected by applying influence functions to prompting; we show that models could rely on offense to determine humor during transfer.

Part-of-Speech (POS) tagging is an important component of the NLP pipeline, but many low-resource languages lack labeled data for training. An established method for training a POS tagger in such a scenario is to create a labeled training set by transferring from high-resource languages. In this paper, we propose a novel method for transferring labels from multiple high-resource source to low-resource target languages. We formalize POS tag projection as graph-based label propagation. Given translations of a sentence in multiple languages, we create a graph with words as nodes and alignment links as edges by aligning words for all language pairs. We then propagate node labels from source to target using a Graph Neural Network augmented with transformer layers. We show that our propagation creates training sets that allow us to train POS taggers for a diverse set of languages. When combined with enhanced contextualized embeddings, our method achieves a new state-of-the-art for unsupervised POS tagging of low-resource languages.

Construction Grammar (CxG) is a paradigm from cognitive linguistics emphasising the connection between syntax and semantics. Rather than rules that operate on lexical items, it posits constructions as the central building blocks of language, i.e., linguistic units of different granularity that combine syntax and semantics. As a first step towards assessing the compatibility of CxG with the syntactic and semantic knowledge demonstrated by state-of-the-art pretrained language models (PLMs), we present an investigation of their capability to classify and understand one of the most commonly studied constructions, the English comparative correlative (CC). We conduct experiments examining the classification accuracy of a syntactic probe on the one hand and the models’ behaviour in a semantic application task on the other, with BERT, RoBERTa, and DeBERTa as the example PLMs. Our results show that all three investigated PLMs are able to recognise the structure of the CC but fail to use its meaning. While human-like performance of PLMs on many NLP tasks has been alleged, this indicates that PLMs still suffer from substantial shortcomings in central domains of linguistic knowledge.

To tackle the rising phenomenon of hate speech, efforts have been made towards data curation and analysis. When it comes to analysis of bias, previous work has focused predominantly on race. In our work, we further investigate bias in hate speech datasets along racial, gender and intersectional axes. We identify strong bias against African American English (AAE), masculine and AAE+Masculine tweets, which are annotated as disproportionately more hateful and offensive than from other demographics. We provide evidence that BERT-based models propagate this bias and show that balancing the training data for these protected attributes can lead to fairer models with regards to gender, but not race.

Research to tackle hate speech plaguing online media has made strides in providing solutions, analyzing bias and curating data. A challenging problem is ambiguity between hate speech and offensive language, causing low performance both overall and specifically for the hate speech class. It can be argued that misclassifying actual hate speech content as merely offensive can lead to further harm against targeted groups. In our work, we mitigate this potentially harmful phenomenon by proposing an adversarial debiasing method to separate the two classes. We show that our method works for English, Arabic German and Hindi, plus in a multilingual setting, improving performance over baselines.

Bilingual Word Embeddings (BWEs) are one of the cornerstones of cross-lingual transfer of NLP models. They can be built using only monolingual corpora without supervision leading to numerous works focusing on unsupervised BWEs. However, most of the current approaches to build unsupervised BWEs do not compare their results with methods based on easy-to-access cross-lingual signals. In this paper, we argue that such signals should always be considered when developing unsupervised BWE methods. The two approaches we find most effective are: 1) using identical words as seed lexicons (which unsupervised approaches incorrectly assume are not available for orthographically distinct language pairs) and 2) combining such lexicons with pairs extracted by matching romanized versions of words with an edit distance threshold. We experiment on thirteen non-Latin languages (and English) and show that such cheap signals work well and that they outperform using more complex unsupervised methods on distant language pairs such as Chinese, Japanese, Kannada, Tamil, and Thai. In addition, they are even competitive with the use of high-quality lexicons in supervised approaches. Our results show that these training signals should not be neglected when building BWEs, even for distant languages.

Parallel corpora are ideal for extracting a multilingual named entity (MNE) resource, i.e., a dataset of names translated into multiple languages. Prior work on extracting MNE datasets from parallel corpora required resources such as large monolingual corpora or word aligners that are unavailable or perform poorly for underresourced languages. We present CLC-BN, a new method for creating an MNE resource, and apply it to the Parallel Bible Corpus, a corpus of more than 1000 languages. CLC-BN learns a neural transliteration model from parallel-corpus statistics, without requiring any other bilingual resources, word aligners, or seed data. Experimental results show that CLC-BN clearly outperforms prior work. We release an MNE resource for 1340 languages and demonstrate its effectiveness in two downstream tasks: knowledge graph augmentation and bilingual lexicon induction.

Bias research in NLP is a rapidly growing and developing field. Similar to CrowS-Pairs (Nangia et al., 2020), we assess gender bias in masked-language models (MLMs) by studying pairs of sentences with gender swapped person references.Most bias research focuses on and often is specific to English.Using a novel methodology for creating sentence pairs that is applicable across languages, we create, based on CrowS-Pairs, a multilingual dataset for English, Finnish, German, Indonesian and Thai.Additionally, we propose SJSD, a new bias measure based on Jensen–Shannon divergence, which we argue retains more information from the model output probabilities than other previously proposed bias measures for MLMs.Using multilingual MLMs, we find that SJSD diagnoses the same systematic biased behavior for non-English that previous studies have found for monolingual English pre-trained MLMs. SJSD outperforms the CrowS-Pairs measure, which struggles to find such biases for smaller non-English datasets.

Vast efforts have been devoted to creating high-performance few-shot learners, i.e., large-scale pretrained language models (PLMs) that perform well with little downstream task training data. Training PLMs has incurred significant cost, but utilizing the few-shot learners is still challenging due to their enormous size. This work focuses on a crucial question: How to make effective use of these few-shot learners? We propose LMTurk, a novel approach that treats few-shotlearners as crowdsourcing workers. The rationale is that crowdsourcing workers are in fact few-shot learners: They are shown a few illustrative examples to learn about a task and then start annotating. LMTurk employs few-shot learners built upon PLMs as workers. We show that the resulting annotations can be utilized to train models that solve the task well and are small enough to be deployable in practical scenarios. Active learning is integrated into LMTurk to reduce the amount of queries made to PLMs, minimizing the computational cost of running PLM inference passes. Altogether, LMTurk is an important step towards making effective use of current PLMs.

We introduce CaMEL (Case Marker Extraction without Labels), a novel and challenging task in computational morphology that is especially relevant for low-resource languages. We propose a first model for CaMEL that uses a massively multilingual corpus to extract case markers in 83 languages based only on a noun phrase chunker and an alignment system. To evaluate CaMEL, we automatically construct a silver standard from UniMorph. The case markers extracted by our model can be used to detect and visualise similarities and differences between the case systems of different languages as well as to annotate fine-grained deep cases in languages in which they are not overtly marked.

Training scene graph classification models requires a large amount of annotated image data. Meanwhile, scene graphs represent relational knowledge that can be modeled with symbolic data from texts or knowledge graphs. While image annotation demands extensive labor, collecting textual descriptions of natural scenes requires less effort. In this work, we investigate whether textual scene descriptions can substitute for annotated image data. To this end, we employ a scene graph classification framework that is trained not only from annotated images but also from symbolic data. In our architecture, the symbolic entities are first mapped to their correspondent image-grounded representations and then fed into the relational reasoning pipeline. Even though a structured form of knowledge, such as the form in knowledge graphs, is not always available, we can generate it from unstructured texts using a transformer-based language model. We show that by fine-tuning the classification pipeline with the extracted knowledge from texts, we can achieve ~8x more accurate results in scene graph classification, ~3x in object classification, and ~1.5x in predicate classification, compared to the supervised baselines with only 1% of the annotated images.

Consistency of a model—that is, the invariance of its behavior under meaning-preserving alternations in its input—is a highly desirable property in natural language processing. In this paper we study the question: Are Pretrained Language Models (PLMs) consistent with respect to factual knowledge? To this end, we create ParaRel, a high-quality resource of cloze-style query English paraphrases. It contains a total of 328 paraphrases for 38 relations. Using ParaRel, we show that the consistency of all PLMs we experiment with is poor— though with high variance between relations. Our analysis of the representational spaces of PLMs suggests that they have a poor structure and are currently not suitable for representing knowledge robustly. Finally, we propose a method for improving model consistency and experimentally demonstrate its effectiveness.

With the advent of end-to-end deep learning approaches in machine translation, interest in word alignments initially decreased; however, they have again become a focus of research more recently. Alignments are useful for typological research, transferring formatting like markup to translated texts, and can be used in the decoding of machine translation systems. At the same time, massively multilingual processing is becoming an important NLP scenario, and pretrained language and machine translation models that are truly multilingual are proposed. However, most alignment algorithms rely on bitexts only and do not leverage the fact that many parallel corpora are multiparallel. In this work, we exploit the multiparallelity of corpora by representing an initial set of bilingual alignments as a graph and then predicting additional edges in the graph. We present two graph algorithms for edge prediction: one inspired by recommender systems and one based on network link prediction. Our experimental results show absolute improvements in F1 of up to 28{%} over the baseline bilingual word aligner in different datasets.

Although pretrained language models (PTLMs) contain significant amounts of world knowledge, they can still produce inconsistent answers to questions when probed, even after specialized training. As a result, it can be hard to identify what the model actually “believes” about the world, making it susceptible to inconsistent behavior and simple errors. Our goal is to reduce these problems. Our approach is to embed a PTLM in a broader system that also includes an evolving, symbolic memory of beliefs – a BeliefBank – that records but then may modify the raw PTLM answers. We describe two mechanisms to improve belief consistency in the overall system. First, a reasoning component – a weighted MaxSAT solver – revises beliefs that significantly clash with others. Second, a feedback component issues future queries to the PTLM using known beliefs as context. We show that, in a controlled experimental setting, these two mechanisms result in more consistent beliefs in the overall system, improving both the accuracy and consistency of its answers over time. This is significant as it is a first step towards PTLM-based architectures with a systematic notion of belief, enabling them to construct a more coherent picture of the world, and improve over time without model retraining.

With more than 7000 languages worldwide, multilingual natural language processing (NLP) is essential both from an academic and commercial perspective. Researching typological properties of languages is fundamental for progress in multilingual NLP. Examples include assessing language similarity for effective transfer learning, injecting inductive biases into machine learning models or creating resources such as dictionaries and inflection tables. We provide ParCourE, an online tool that allows to browse a word-aligned parallel corpus, covering 1334 languages. We give evidence that this is useful for typological research. ParCourE can be set up for any parallel corpus and can thus be used for typological research on other corpora as well as for exploring their quality and properties.

Recent research investigates factual knowledge stored in large pretrained language models (PLMs). Instead of structural knowledge base (KB) queries, masked sentences such as ‘Paris is the capital of [MASK]’ are used as probes. The good performance on this analysis task has been interpreted as PLMs becoming potential repositories of factual knowledge. In experiments across ten linguistically diverse languages, we study knowledge contained in static embeddings. We show that, when restricting the output space to a candidate set, simple nearest neighbor matching using static embeddings performs better than PLMs. E.g., static embeddings perform 1.6% points better than BERT while just using 0.3% of energy for training. One important factor in their good comparative performance is that static embeddings are standardly learned for a large vocabulary. In contrast, BERT exploits its more sophisticated, but expensive ability to compose meaningful representations from a much smaller subword vocabulary.

Recently, it has been found that monolingual English language models can be used as knowledge bases. Instead of structural knowledge base queries, masked sentences such as “Paris is the capital of [MASK]” are used as probes. We translate the established benchmarks TREx and GoogleRE into 53 languages. Working with mBERT, we investigate three questions. (i) Can mBERT be used as a multilingual knowledge base? Most prior work only considers English. Extending research to multiple languages is important for diversity and accessibility. (ii) Is mBERT’s performance as knowledge base language-independent or does it vary from language to language? (iii) A multilingual model is trained on more text, e.g., mBERT is trained on 104 Wikipedias. Can mBERT leverage this for better performance? We find that using mBERT as a knowledge base yields varying performance across languages and pooling predictions across languages improves performance. Conversely, mBERT exhibits a language bias; e.g., when queried in Italian, it tends to predict Italy as the country of origin.

Annotation projection is an important area in NLP that can greatly contribute to creating language resources for low-resource languages. Word alignment plays a key role in this setting. However, most of the existing word alignment methods are designed for a high resource setting in machine translation where millions of parallel sentences are available. This amount reduces to a few thousands of sentences when dealing with low-resource languages failing the existing established IBM models. In this paper, we propose subword sampling-based alignment of text units. This method’s hypothesis is that the aggregation of different granularities of text for certain language pairs can help word-level alignment. For certain languages for which gold-standard alignments exist, we propose an iterative Bayesian optimization framework to optimize selecting possible subwords from the space of possible subword representations of the source and target sentences. We show that the subword sampling method consistently outperforms word-level alignment on six language pairs: English-German, English-French, English-Romanian, English-Persian, English-Hindi, and English-Inuktitut. In addition, we show that the hyperparameters learned for certain language pairs can be applied to other languages at no supervision and consistently improve the alignment results. We observe that using 5K parallel sentences together with our proposed subword sampling approach, we obtain similar F1 scores to the use of 100K’s of parallel sentences in existing word-level fast-align/eflomal alignment methods.

How can pretrained language models (PLMs) learn factual knowledge from the training set? We investigate the two most important mechanisms: reasoning and memorization. Prior work has attempted to quantify the number of facts PLMs learn, but we present, using synthetic data, the first study that investigates the causal relation between facts present in training and facts learned by the PLM. For reasoning, we show that PLMs seem to learn to apply some symbolic reasoning rules correctly but struggle with others, including two-hop reasoning. Further analysis suggests that even the application of learned reasoning rules is flawed. For memorization, we identify schema conformity (facts systematically supported by other facts) and frequency as key factors for its success.

Khandelwal et al. (2020) use a k-nearest-neighbor (kNN) component to improve language model performance. We show that this idea is beneficial for open-domain question answering (QA). To improve the recall of facts encountered during training, we combine BERT (Devlin et al., 2019) with a traditional information retrieval step (IR) and a kNN search over a large datastore of an embedded text collection. Our contributions are as follows: i) BERT-kNN outperforms BERT on cloze-style QA by large margins without any further training. ii) We show that BERT often identifies the correct response category (e.g., US city), but only kNN recovers the factually correct answer (e.g.,“Miami”). iii) Compared to BERT, BERT-kNN excels for rare facts. iv) BERT-kNN can easily handle facts not covered by BERT’s training set, e.g., recent events.

Building on Petroni et al. 2019, we propose two new probing tasks analyzing factual knowledge stored in Pretrained Language Models (PLMs). (1) Negation. We find that PLMs do not distinguish between negated (‘‘Birds cannot [MASK]”) and non-negated (‘‘Birds can [MASK]”) cloze questions. (2) Mispriming. Inspired by priming methods in human psychology, we add “misprimes” to cloze questions (‘‘Talk? Birds can [MASK]”). We find that PLMs are easily distracted by misprimes. These results suggest that PLMs still have a long way to go to adequately learn human-like factual knowledge.

Prior work has determined domain similarity using text-based features of a corpus. However, when using pre-trained word embeddings, the underlying text corpus might not be accessible anymore. Therefore, we propose the CCA measure, a new measure of domain similarity based directly on the dimension-wise correlations between corresponding embedding spaces. Our results suggest that an inherent notion of domain can be captured this way, as we are able to reproduce our findings for different domain comparisons for English, German, Spanish and Czech as well as in cross-lingual comparisons. We further find a threshold at which the CCA measure indicates that two corpora come from the same domain in a monolingual setting by applying permutation tests. By evaluating the usability of the CCA measure in a domain adaptation application, we also show that it can be used to determine which corpora are more similar to each other in a cross-domain sentiment detection task.

Nowadays, classical count-based word embeddings using positive pointwise mutual information (PPMI) weighted co-occurrence matrices have been widely superseded by machine-learning-based methods like word2vec and GloVe. But these methods are usually applied using very large amounts of text data. In many cases, however, there is not much text data available, for example for specific domains or low-resource languages. This paper revisits PPMI by adding Dirichlet smoothing to correct its bias towards rare words. We evaluate on standard word similarity data sets and compare to word2vec and the recent state of the art for low-resource settings: Positive and Unlabeled (PU) Learning for word embeddings. The proposed method outperforms PU-Learning for low-resource settings and obtains competitive results for Maltese and Luxembourgish.