Debora Nozza

Federico Bianchi, Pratyusha Kalluri, Esin Durmus et al. · stanford

Machine learning models that convert user-written text descriptions into images are now widely available online and used by millions of users to generate millions of images a day. We investigate the potential for these models to amplify dangerous and complex stereotypes. We find a broad range of ordinary prompts produce stereotypes, including prompts simply mentioning traits, descriptors, occupations, or objects. For example, we find cases of prompting for basic traits or social roles resulting in images reinforcing whiteness as ideal, prompting for occupations resulting in amplification of racial and gender disparities, and prompting for objects resulting in reification of American norms. Stereotypes are present regardless of whether prompts explicitly mention identity and demographic language or avoid such language. Moreover, stereotypes persist despite mitigation strategies; neither user attempts to counter stereotypes by requesting images with specific counter-stereotypes nor institutional attempts to add system ``guardrails'' have prevented the perpetuation of stereotypes. Our analysis justifies concerns regarding the impacts of today's models, presenting striking exemplars, and connecting these findings with deep insights into harms drawn from social scientific and humanist disciplines. This work contributes to the effort to shed light on the uniquely complex biases in language-vision models and demonstrates the ways that the mass deployment of text-to-image generation models results in mass dissemination of stereotypes and resulting harms.

Paul Röttger, Debora Nozza, Federico Bianchi et al. · stanford

Hate speech is a global phenomenon, but most hate speech datasets so far focus on English-language content. This hinders the development of more effective hate speech detection models in hundreds of languages spoken by billions across the world. More data is needed, but annotating hateful content is expensive, time-consuming and potentially harmful to annotators. To mitigate these issues, we explore data-efficient strategies for expanding hate speech detection into under-resourced languages. In a series of experiments with mono- and multilingual models across five non-English languages, we find that 1) a small amount of target-language fine-tuning data is needed to achieve strong performance, 2) the benefits of using more such data decrease exponentially, and 3) initial fine-tuning on readily-available English data can partially substitute target-language data and improve model generalisability. Based on these findings, we formulate actionable recommendations for hate speech detection in low-resource language settings.

Giuseppe Attanasio, Debora Nozza, Federico Bianchi et al. · stanford

Language is constantly changing and evolving, leaving language models to become quickly outdated. Consequently, we should continuously update our models with new data to expose them to new events and facts. However, that requires additional computing, which means new carbon emissions. Do any measurable benefits justify this cost? This paper looks for empirical evidence to support continuous training. We reproduce existing benchmarks and extend them to include additional time periods, models, and tasks. Our results show that the downstream task performance of temporally adapted English models for social media data do not improve over time. Pretrained models without temporal adaptation are actually significantly more effective and efficient. However, we also note a lack of suitable temporal benchmarks. Our findings invite a critical reflection on when and how to temporally adapt language models, accounting for sustainability.

Giuseppe Attanasio, Debora Nozza, Dirk Hovy et al.

Natural Language Processing (NLP) models risk overfitting to specific terms in the training data, thereby reducing their performance, fairness, and generalizability. E.g., neural hate speech detection models are strongly influenced by identity terms like gay, or women, resulting in false positives, severe unintended bias, and lower performance. Most mitigation techniques use lists of identity terms or samples from the target domain during training. However, this approach requires a-priori knowledge and introduces further bias if important terms are neglected. Instead, we propose a knowledge-free Entropy-based Attention Regularization (EAR) to discourage overfitting to training-specific terms. An additional objective function penalizes tokens with low self-attention entropy. We fine-tune BERT via EAR: the resulting model matches or exceeds state-of-the-art performance for hate speech classification and bias metrics on three benchmark corpora in English and Italian. EAR also reveals overfitting terms, i.e., terms most likely to induce bias, to help identify their effect on the model, task, and predictions.

Paul Röttger, Haitham Seelawi, Debora Nozza et al.

Hate speech detection models are typically evaluated on held-out test sets. However, this risks painting an incomplete and potentially misleading picture of model performance because of increasingly well-documented systematic gaps and biases in hate speech datasets. To enable more targeted diagnostic insights, recent research has thus introduced functional tests for hate speech detection models. However, these tests currently only exist for English-language content, which means that they cannot support the development of more effective models in other languages spoken by billions across the world. To help address this issue, we introduce Multilingual HateCheck (MHC), a suite of functional tests for multilingual hate speech detection models. MHC covers 34 functionalities across ten languages, which is more languages than any other hate speech dataset. To illustrate MHC's utility, we train and test a high-performing multilingual hate speech detection model, and reveal critical model weaknesses for monolingual and cross-lingual applications.

Debora Nozza, Dirk Hovy

Work on hate speech has made the consideration of rude and harmful examples in scientific publications inevitable. This raises various problems, such as whether or not to obscure profanities. While science must accurately disclose what it does, the unwarranted spread of hate speech is harmful to readers, and increases its internet frequency. While maintaining publications' professional appearance, obfuscating profanities makes it challenging to evaluate the content, especially for non-native speakers. Surveying 150 ACL papers, we discovered that obfuscation is usually employed for English but not other languages, and even so quite uneven. We discuss the problems with obfuscation and suggest a multilingual community resource called PrOf that has a Python module to standardize profanity obfuscation processes. We believe PrOf can help scientific publication policies to make hate speech work accessible and comparable, irrespective of language.

Giuseppe Attanasio, Flor Miriam Plaza-del-Arco, Debora Nozza et al.

Recent instruction fine-tuned models can solve multiple NLP tasks when prompted to do so, with machine translation (MT) being a prominent use case. However, current research often focuses on standard performance benchmarks, leaving compelling fairness and ethical considerations behind. In MT, this might lead to misgendered translations, resulting, among other harms, in the perpetuation of stereotypes and prejudices. In this work, we address this gap by investigating whether and to what extent such models exhibit gender bias in machine translation and how we can mitigate it. Concretely, we compute established gender bias metrics on the WinoMT corpus from English to German and Spanish. We discover that IFT models default to male-inflected translations, even disregarding female occupational stereotypes. Next, using interpretability methods, we unveil that models systematically overlook the pronoun indicating the gender of a target occupation in misgendered translations. Finally, based on this finding, we propose an easy-to-implement and effective bias mitigation solution based on few-shot learning that leads to significantly fairer translations.

Giuseppe Attanasio, Eliana Pastor, Chiara Di Bonaventura et al.

As Transformers are increasingly relied upon to solve complex NLP problems, there is an increased need for their decisions to be humanly interpretable. While several explainable AI (XAI) techniques for interpreting the outputs of transformer-based models have been proposed, there is still a lack of easy access to using and comparing them. We introduce ferret, a Python library to simplify the use and comparisons of XAI methods on transformer-based classifiers. With ferret, users can visualize and compare transformers-based models output explanations using state-of-the-art XAI methods on any free-text or existing XAI corpora. Moreover, users can also evaluate ad-hoc XAI metrics to select the most faithful and plausible explanations. To align with the recently consolidated process of sharing and using transformers-based models from Hugging Face, ferret interfaces directly with its Python library. In this paper, we showcase ferret to benchmark XAI methods used on transformers for sentiment analysis and hate speech detection. We show how specific methods provide consistently better explanations and are preferable in the context of transformer models.

Helena Bonaldi, Giuseppe Attanasio, Debora Nozza et al.

Recent computational approaches for combating online hate speech involve the automatic generation of counter narratives by adapting Pretrained Transformer-based Language Models (PLMs) with human-curated data. This process, however, can produce in-domain overfitting, resulting in models generating acceptable narratives only for hatred similar to training data, with little portability to other targets or to real-world toxic language. This paper introduces novel attention regularization methodologies to improve the generalization capabilities of PLMs for counter narratives generation. Overfitting to training-specific terms is then discouraged, resulting in more diverse and richer narratives. We experiment with two attention-based regularization techniques on a benchmark English dataset. Regularized models produce better counter narratives than state-of-the-art approaches in most cases, both in terms of automatic metrics and human evaluation, especially when hateful targets are not present in the training data. This work paves the way for better and more flexible counter-speech generation models, a task for which datasets are highly challenging to produce.

Donya Rooein, Sankalan Pal Chowdhury, Mariia Eremeeva et al.

Recent advances in large language models (LLMs) demonstrate their potential as educational tutors. However, different tutoring strategies benefit different student personalities, and mismatches can be counterproductive to student outcomes. Despite this, current LLM tutoring systems do not take into account student personality traits. To address this problem, we first construct a taxonomy that links pedagogical methods to personality profiles, based on pedagogical literature. We simulate student-teacher conversations and use our framework to let the LLM tutor adjust its strategy to the simulated student personality. We evaluate the scenario with human teachers and find that they consistently prefer our approach over two baselines. Our method also increases the use of less common, high-impact strategies such as role-playing, which human and LLM annotators prefer significantly. Our findings pave the way for developing more personalized and effective LLM use in educational applications.

Samia Touileb, Debora Nozza

Scandinavian countries are perceived as role-models when it comes to gender equality. With the advent of pre-trained language models and their widespread usage, we investigate to what extent gender-based harmful and toxic content exist in selected Scandinavian language models. We examine nine models, covering Danish, Swedish, and Norwegian, by manually creating template-based sentences and probing the models for completion. We evaluate the completions using two methods for measuring harmful and toxic completions and provide a thorough analysis of the results. We show that Scandinavian pre-trained language models contain harmful and gender-based stereotypes with similar values across all languages. This finding goes against the general expectations related to gender equality in Scandinavian countries and shows the possible problematic outcomes of using such models in real-world settings.

Flor Miriam Plaza-del-Arco, Debora Nozza, Dirk Hovy

Large Language Models (LLMs) exhibit remarkable text classification capabilities, excelling in zero- and few-shot learning (ZSL and FSL) scenarios. However, since they are trained on different datasets, performance varies widely across tasks between those models. Recent studies emphasize the importance of considering human label variation in data annotation. However, how this human label variation also applies to LLMs remains unexplored. Given this likely model specialization, we ask: Do aggregate LLM labels improve over individual models (as for human annotators)? We evaluate four recent instruction-tuned LLMs as annotators on five subjective tasks across four languages. We use ZSL and FSL setups and label aggregation from human annotation. Aggregations are indeed substantially better than any individual model, benefiting from specialization in diverse tasks or languages. Surprisingly, FSL does not surpass ZSL, as it depends on the quality of the selected examples. However, there seems to be no good information-theoretical strategy to select those. We find that no LLM method rivals even simple supervised models. We also discuss the tradeoffs in accuracy, cost, and moral/ethical considerations between LLM and human annotation.

Beatrice Savoldi, Giuseppe Attanasio, Olga Gorodetskaya et al.

The rise of Artificial Intelligence (AI) language technologies, particularly generative AI (GenAI) chatbots accessible via conversational interfaces, is transforming digital interactions. While these tools hold societal promise, they also risk widening digital divides due to uneven adoption and low awareness of their limitations. This study presents the first comprehensive empirical mapping of GenAI adoption, usage patterns, and literacy in Italy, based on newly collected survey data from 1,906 Italian-speaking adults. Our findings reveal widespread adoption for both work and personal use, including sensitive tasks like emotional support and medical advice. Crucially, GenAI is supplanting other technologies to become a primary information source: this trend persists despite low user digital literacy, posing a risk as users struggle to recognize errors or misinformation. Moreover, we identify a significant gender divide -- particularly pronounced in older generations -- where women are half as likely to adopt GenAI and use it less frequently than men. While we find literacy to be a key predictor of adoption, it only partially explains this disparity, suggesting that other barriers are at play. Overall, our data provide granular insights into the multipurpose usage of GenAI, highlighting the dual need for targeted educational initiatives and further investigation into the underlying barriers to equitable participation that competence alone cannot explain.

Tanise Ceron, Dmitry Nikolaev, Dominik Stammbach et al.

Large language models (LLMs) are known to generate politically biased text, yet how such biases arise remains unclear. A crucial step toward answering this question is the analysis of training data, whose political content remains largely underexplored in current LLM research. To address this gap, we present in this paper an analysis of the pre- and post-training corpora of OLMO2, the largest fully open-source model released together with its complete dataset. From these corpora, we draw large random samples, automatically annotate documents for political orientation, and analyze their source domains and content. We then assess how political content in the training data correlates with models' stance on specific policy issues. Our analysis shows that left-leaning documents predominate across datasets, with pre-training corpora containing significantly more politically engaged content than post-training data. We also find that left- and right-leaning documents frame similar topics through distinct values and sources of legitimacy. Finally, the predominant stance in the training data strongly correlates with models' political biases when evaluated on policy issues. These findings underscore the need to integrate political content analysis into future data curation pipelines as well as in-depth documentation of filtering strategies for transparency.

Zehui Yu, Indira Sen, Dennis Assenmacher et al.

Machine learning (ML)-based content moderation tools are essential to keep online spaces free from hateful communication. Yet, ML tools can only be as capable as the quality of the data they are trained on allows them. While there is increasing evidence that they underperform in detecting hateful communications directed towards specific identities and may discriminate against them, we know surprisingly little about the provenance of such bias. To fill this gap, we present a systematic review of the datasets for the automated detection of hateful communication introduced over the past decade, and unpack the quality of the datasets in terms of the identities that they embody: those of the targets of hateful communication that the data curators focused on, as well as those unintentionally included in the datasets. We find, overall, a skewed representation of selected target identities and mismatches between the targets that research conceptualizes and ultimately includes in datasets. Yet, by contextualizing these findings in the language and location of origin of the datasets, we highlight a positive trend towards the broadening and diversification of this research space.

Giuseppe Russo, Debora Nozza, Paul Röttger et al.

People increasingly rely on Large Language Models (LLMs) for moral advice, which may influence humans' decisions. Yet, little is known about how closely LLMs align with human moral judgments. To address this, we introduce the Moral Dilemma Dataset, a benchmark of 1,618 real-world moral dilemmas paired with a distribution of human moral judgments consisting of a binary evaluation and a free-text rationale. We treat this problem as a pluralistic distributional alignment task, comparing the distributions of LLM and human judgments across dilemmas. We find that models reproduce human judgments only under high consensus; alignment deteriorates sharply when human disagreement increases. In parallel, using a 60-value taxonomy built from 3,783 value expressions extracted from rationales, we show that LLMs rely on a narrower set of moral values than humans. These findings reveal a pluralistic moral gap: a mismatch in both the distribution and diversity of values expressed. To close this gap, we introduce Dynamic Moral Profiling (DMP), a Dirichlet-based sampling method that conditions model outputs on human-derived value profiles. DMP improves alignment by 64.3% and enhances value diversity, offering a step toward more pluralistic and human-aligned moral guidance from LLMs.

Donya Rooein, Vilém Zouhar, Debora Nozza et al.

Stories play a pivotal role in human communication, shaping beliefs and morals, particularly in children. As parents increasingly rely on large language models (LLMs) to craft bedtime stories, the presence of cultural and gender stereotypes in these narratives raises significant concerns. To address this issue, we present Biased Tales, a comprehensive dataset designed to analyze how biases influence protagonists' attributes and story elements in LLM-generated stories. Our analysis uncovers striking disparities. When the protagonist is described as a girl (as compared to a boy), appearance-related attributes increase by 55.26%. Stories featuring non-Western children disproportionately emphasize cultural heritage, tradition, and family themes far more than those for Western children. Our findings highlight the role of sociocultural bias in making creative AI use more equitable and diverse.

Donya Rooein, Flor Miriam Plaza-del-Arco, Debora Nozza et al.

Given Farsi's speaker base of over 127 million people and the growing availability of digital text, including more than 1.3 million articles on Wikipedia, it is considered a middle-resource language. However, this label quickly crumbles when the situation is examined more closely. We focus on three subjective tasks (Sentiment Analysis, Emotion Analysis, and Toxicity Detection) and find significant challenges in data availability and quality, despite the overall increase in data availability. We review 110 publications on subjective tasks in Farsi and observe a lack of publicly available datasets. Furthermore, existing datasets often lack essential demographic factors, such as age and gender, that are crucial for accurately modeling subjectivity in language. When evaluating prediction models using the few available datasets, the results are highly unstable across both datasets and models. Our findings indicate that the volume of data is insufficient to significantly improve a language's prospects in NLP.

Mattia Setzu, Marta Marchiori Manerba, Pasquale Minervini et al.

Language Models (LMs) have been shown to inherit undesired biases that might hurt minorities and underrepresented groups if such systems were integrated into real-world applications without careful fairness auditing. This paper proposes FairBelief, an analytical approach to capture and assess beliefs, i.e., propositions that an LM may embed with different degrees of confidence and that covertly influence its predictions. With FairBelief, we leverage prompting to study the behavior of several state-of-the-art LMs across different previously neglected axes, such as model scale and likelihood, assessing predictions on a fairness dataset specifically designed to quantify LMs' outputs' hurtfulness. Finally, we conclude with an in-depth qualitative assessment of the beliefs emitted by the models. We apply FairBelief to English LMs, revealing that, although these architectures enable high performances on diverse natural language processing tasks, they show hurtful beliefs about specific genders. Interestingly, training procedure and dataset, model scale, and architecture induce beliefs of different degrees of hurtfulness.

Anne Lauscher, Debora Nozza, Archie Crowley et al.

As 3rd-person pronoun usage shifts to include novel forms, e.g., neopronouns, we need more research on identity-inclusive NLP. Exclusion is particularly harmful in one of the most popular NLP applications, machine translation (MT). Wrong pronoun translations can discriminate against marginalized groups, e.g., non-binary individuals (Dev et al., 2021). In this ``reality check'', we study how three commercial MT systems translate 3rd-person pronouns. Concretely, we compare the translations of gendered vs. gender-neutral pronouns from English to five other languages (Danish, Farsi, French, German, Italian), and vice versa, from Danish to English. Our error analysis shows that the presence of a gender-neutral pronoun often leads to grammatical and semantic translation errors. Similarly, gender neutrality is often not preserved. By surveying the opinions of affected native speakers from diverse languages, we provide recommendations to address the issue in future MT research.

Federico Bianchi, Debora Nozza, Dirk Hovy

Meaning is context-dependent, but many properties of language (should) remain the same even if we transform the context. For example, sentiment, entailment, or speaker properties should be the same in a translation and original of a text. We introduce language invariant properties: i.e., properties that should not change when we transform text, and how they can be used to quantitatively evaluate the robustness of transformation algorithms. We use translation and paraphrasing as transformation examples, but our findings apply more broadly to any transformation. Our results indicate that many NLP transformations change properties like author characteristics, i.e., make them sound more male. We believe that studying these properties will allow NLP to address both social factors and pragmatic aspects of language. We also release an application suite that can be used to evaluate the invariance of transformation applications.

Federico Bianchi, Silvia Terragni, Dirk Hovy et al.

Many data sets (e.g., reviews, forums, news, etc.) exist parallelly in multiple languages. They all cover the same content, but the linguistic differences make it impossible to use traditional, bag-of-word-based topic models. Models have to be either single-language or suffer from a huge, but extremely sparse vocabulary. Both issues can be addressed by transfer learning. In this paper, we introduce a zero-shot cross-lingual topic model. Our model learns topics on one language (here, English), and predicts them for unseen documents in different languages (here, Italian, French, German, and Portuguese). We evaluate the quality of the topic predictions for the same document in different languages. Our results show that the transferred topics are coherent and stable across languages, which suggests exciting future research directions.

Debora Nozza, Federico Bianchi, Dirk Hovy

Recently, Natural Language Processing (NLP) has witnessed an impressive progress in many areas, due to the advent of novel, pretrained contextual representation models. In particular, Devlin et al. (2019) proposed a model, called BERT (Bidirectional Encoder Representations from Transformers), which enables researchers to obtain state-of-the art performance on numerous NLP tasks by fine-tuning the representations on their data set and task, without the need for developing and training highly-specific architectures. The authors also released multilingual BERT (mBERT), a model trained on a corpus of 104 languages, which can serve as a universal language model. This model obtained impressive results on a zero-shot cross-lingual natural inference task. Driven by the potential of BERT models, the NLP community has started to investigate and generate an abundant number of BERT models that are trained on a particular language, and tested on a specific data domain and task. This allows us to evaluate the true potential of mBERT as a universal language model, by comparing it to the performance of these more specific models. This paper presents the current state of the art in language-specific BERT models, providing an overall picture with respect to different dimensions (i.e. architectures, data domains, and tasks). Our aim is to provide an immediate and straightforward overview of the commonalities and differences between Language-Specific (language-specific) BERT models and mBERT. We also provide an interactive and constantly updated website that can be used to explore the information we have collected, at https://bertlang.unibocconi.it.