David Ifeoluwa Adelani

henok.ademtew@mila.quebec

Biographie

David Adelani est professeur adjoint en science informatique et lutte contre les inégalités à l’Université McGill, et membre académique principal à Mila – Institut québécois d'intelligence artificielle. Ses recherches se concentrent sur le traitement multilingue du langage naturel, avec un accent particulier sur les langues sous-dotées en ressources.

Étudiants actuels

Henok Ademtew

Maîtrise recherche - McGill

Maîtrise recherche - McGill

victor.akinode@mila.quebec

virginia.ceccatelli@mila.quebec

Virginia Ceccatelli

Collaborateur·rice de recherche - McGill

luel.hagos-beyene@mila.quebec

David Guzmán Ramírez

Doctorat - McGill

david.guzman@mila.quebec

Stagiaire de recherche - McGill

Stagiaire de recherche - McGill

nicolas.helmeyer@mila.quebec

Postdoctorat - McGill

yejin.jeon@mila.quebec

Senyu Li Li

Doctorat - McGill

senyu.li@mila.quebec

marie.maltais@mila.quebec

Yinhan Lu

Collaborateur·rice de recherche - McGill

yinhan.lu@mila.quebec

Doctorat - McGill

damaris-stephanie.ndjebayi@mila.quebec

Damaris Stephanie NDJEBAYI

Doctorat - McGill

Tolúlọpẹ Ògúnrẹ̀mí

Stagiaire de recherche - Standford University

tolulope.ogunremi@mila.quebec

Juhyun Oh

Collaborateur·rice alumni - McGill

juhyun.oh@mila.quebec

Jessica Ojo

Maîtrise recherche - McGill

jessica.ojo@mila.quebec

Stagiaire de recherche - McGill

kosei.uemura@mila.quebec

Vivek Verma

Maîtrise professionnelle - UdeM

vivek.verma@mila.quebec

Jessica Wang

Stagiaire de recherche - McGill

jessica.wang@mila.quebec

Tianyou Xie

Stagiaire de recherche - McGill

tianyou.xie@mila.quebec

Tianyi Xu Xu

Stagiaire de recherche - McGill

tiany12138@gmail.com

Peter Yu

Collaborateur·rice alumni - McGill

Publications

VoxSumm: A Multilingual Corpus of Long-Form Spoken News for Joint Summarization and Translation

Marie Maltais

Virginia Ceccatelli

Min Ma

As information increasingly traverses linguistic boundaries, users require concise cross-lingual representations of long-form content. Never… (voir plus)theless, long-document summarization research remains text-centric, whereas multilingual speech research has largely prioritized translation, preserving source content rather than compressing it. We address this methodological gap by formalizing joint speech summarization and translation (JSumT): the generation of a succinct, faithful target-language summary directly from a long spoken document in a source language. We additionally introduce VoxSumm, the first multilingual and cross-lingual benchmark for this task, comprising 10,045 BBC article-summary pairs across 24 languages and encompassing approximately 703 hours of speech data. Our evaluation of representative speech-language models reveals pronounced variation across models and generation settings: Gemini3.1-Pro demonstrates the greatest consistency, summarization into English generally surpasses generation into non-English target languages, and translating an entire document before summarization compounds instruction-following failures. Through the release of VoxSumm, we establish a foundation for developing and evaluating multilingual systems capable of jointly interpreting, compressing, and translating long-form speech.

2026-08-10

arXiv (prépublication)

AfriSUD: A Dependency Treebank Collection for Evaluating Models on African Languages

Happy Buzaaba

Cheikh Mouhamadou Bamba Dione

Sylvain Kahane

Kim Gerdes

Bruno Guillaume

Kevin Guan

Aremu Anuoluwapo

Naome A. Etori

Shamsuddeen Hassan Muhammad

Utitofon Inyang

Peter Nabende

David Sabiiti Bamutura

Andiswa Bukula

Chinedu Uchechukwu

Rooweither Mabuya

Idris Akinade

Christiane Fellbaum

Despite their linguistic diversity and global significance, African languages remain underrepresented in research and resources to support N… (voir plus)LP. We aim to bridge this gap by introducing AfriSUD, the first large-scale collection of syntactically annotated treebanks for nine diverse African languages spanning major language families and regions across Sub-Saharan Africa. Using the Surface-Syntactic Universal Dependencies (SUD) framework, our community-led effort provides high-quality, native-speaker verified data that capture typological key features such as agglutination and tone. We evaluate a range of models on AfriSUD for part-of-speech tagging and dependency parsing including non-transformer baselines, multilingual pretrained encoders, and LLMs. Our results reveal a significant syntax gap, where models still show clear limitations across the nine languages, suggesting that existing architectures may not fully capture the structural diversity of African-language syntax.

2026-06-09

arXiv (prépublication)

OpenBibleTTS: Large-Scale Speech Resources and TTS Models for Low-Resource Languages

David Guzmán

Luel Hagos Beyene

Jesujoba Oluwadara Alabi

Dietrich Klakow

Recent advances in neural text-to-speech (TTS) and multilingual speech generation have substantially improved synthetic speech quality, yet … (voir plus)these gains remain unevenly distributed across the world's languages. Existing models are still dominated by a small set of high-resource languages, while many studies of low-resource TTS are simulated on artificially downsampled high-resource corpora that do not reflect the orthographic variation and limited phonetic coverage encountered in genuinely underrepresented settings. As such, we introduce OpenBibleTTS, which is a large-scale benchmark for low-resource speech synthesis spanning 37 underrepresented languages. Moreover, a systematic comparison of various TTS architectures and large-scale speech generation models is conducted across in-domain Biblical text and out-of-domain material. Results show that no single system dominates across languages and metrics: Gemini-TTS achieves the highest listener ratings on most evaluated languages, but monolingual EveryVoice models trained on OpenBibleTTS remain strongest for intelligibility and are preferred in several African languages, while open from-scratch systems degrade sharply on out-of-domain text, revealing a persistent gap between broad multilingual coverage and reliable synthesis quality in underserved linguistic communities. We complement automatic evaluation with subjective human judgments, and open-source all processed datasets, alignments, and trained models to support future low-resource TTS research.

2026-06-07

arXiv (prépublication)

SpeechJBB: Probing Safety Alignment and Comprehension in Large Audio Language Models under Code-Switched Speech

Virginia Ceccatelli

Large audio language models (LALMs) are increasingly deployed in real-world applications, yet their safety alignment is still primarily eval… (voir plus)uated on monolingual, text-based harmful prompts. This leaves their generalizability under multilingual and spoken settings, particularly code-switched speech, largely underexplored. To address this gap, we introduce SpeechJBB, an audio jailbreak dataset for benchmarking across multiple state-of-the-art LALMs. The extent of safety weaknesses is further probed by introducing an augmented setting where phonologically plausible pseudo-words are inserted around safety-critical terms to simulate localized obfuscation. Across models, code-switched harmful audio yields substantially high jailbreak success rates (JSR), with non-English monolingual and non-English code-switched pairs exhibiting the highest attack success. Pseudo-word insertion further reduces refusal rates, which demonstrates that natural-sounding obfuscation can effectively bypass safety policies.

2026-06-03

arXiv (prépublication)

TukaBench: A Culturally Grounded Jailbreak Benchmark for African Languages

Victor Akinode

Senyu Li

Wassim Hamidouche

Waqas Zamir

Inbal Becker-Reshef

Safety evaluation of Large Language Models (LLMs) remains heavily English-centric, leaving Low-Resource Languages (LRLs), particularly Afric… (voir plus)an ones, critically underexplored. We introduce TUKABENCH, a jailbreak benchmark for seven African languages that extends JailbreakBench (JBB) beyond direct translation through four settings: human translation of JBB prompts, English adaptation to African contexts followed by human translation, human-curated prompts validated through interactions with GPT-5.2, and code-switched prompts combining English and African languages, isolating the effect of language, cultural grounding, and prompt evasiveness on model safety. Across closed and open models, prompting in African languages reduces refusal relative to English, with culturally adapted prompts leading to least refusal. The evaluation also surfaces two structural limitations: model comprehension failures and reduced LLM-as-a-judge reliability in LRLs. To capture the first, we introduce Deflection alongside Refused and Jailbroken; to assess the second, we validate outputs with human annotations, showing that judge-human agreement drops in lower-resource languages and less commonly supported scripts.

2026-05-30

arXiv (prépublication)

NaijaS2ST: A Multi-Accent Benchmark for Speech-to-Speech Translation in Low-Resource Nigerian Languages

Marie Maltais

Min Ma

Shamsuddeen Hassan Muhammad

Idris Abdulmumin

Maryam Ibrahim Mukhtar

Daud Abolade

Joel Okepefi

Johnson Sewedo

Speech translation for low-resource languages remains fundamentally limited by the scarcity of high-quality, diverse parallel speech data, a… (voir plus) challenge that is especially pronounced in African linguistic contexts. To address this, we introduce NaijaS2ST, a parallel speech translation dataset spanning Igbo, Hausa, Yor\`ub\'a, and Nigerian Pidgin paired with English. The dataset comprises approximately 50 hours of speech per language and captures substantial variation in speakers and accents, reflecting realistic multilingual and multi-accent conditions. With NaijaS2ST, we conduct a comprehensive benchmark of cascaded, end-to-end (E2E), and AudioLLM-based approaches across bidirectional translation settings. Our results show that audio LLMs with few-shot examples are more effective for speech-to-text translation than cascaded and end-to-end methods trained on fine-tuned data. However, for speech-to-speech translation, the cascaded and audio LLM paradigms yield comparable performance, indicating that there is still considerable room for improvement in developing targeted, task-specific models for this setting. By providing both a high-quality dataset and a systematic benchmark, we hope that NaijaS2ST will serve as a strong foundation for advancing research in low-resource, multilingual speech translation.

2026-04-16

arXiv (prépublication)

Testing the Assumptions of Active Learning for Translation Tasks with Few Samples

Lorenzo Flores

Cesare Spinoso di-Piano

Ori Ernst

Jackie Chi Kit Cheung

Active learning (AL) is a training paradigm for selecting unlabeled samples for annotation to improve model performance on a test set, which… (voir plus) is useful when only a limited number of samples can be annotated. These algorithms often work by optimizing for the informativeness and diversity of the training data to be annotated. Recent work found that AL strategies fail to outperform random sampling on various language generation tasks when using 100-500 samples. To understand AL's poor performance when only using few samples, we investigate whether the core assumptions underlying AL strategies hold. We find that neither the informativeness nor diversity of the training data, which AL strategies optimize for, are correlated with test set performance. Instead, factors like the ordering of the training samples and interactions with pre-training data have a larger impact on performance. This suggests that future AL methods must take these factors into account in order to work with very few samples.

2026-04-09

arXiv (prépublication)

YoNER: A New Yorùbá Multi-domain Named Entity Recognition Dataset

Peace Busola Falola

Jesujoba O. Alabi

Solomon O. Akinola

Folashade T. Ogunajo

Emmanuel Oluwadunsin Alabi

Named Entity Recognition (NER) is a foundational NLP task, yet research in Yor\`ub\'a has been constrained by limited and domain-specific re… (voir plus)sources. Existing resources, such as MasakhaNER (a manually annotated news-domain corpus) and WikiAnn (automatically created from Wikipedia), are valuable but restricted in domain coverage. To address this gap, we present YoNER, a new multidomain Yor\`ub\'a NER dataset that extends entity coverage beyond news and Wikipedia. The dataset comprises about 5,000 sentences and 100,000 tokens collected from five domains including Bible, Blogs, Movies, Radio broadcast and Wikipedia, and annotated with three entity types: Person (PER), Organization (ORG) and Location (LOC), following CoNLL-style guidelines. Annotation was conducted manually by three native Yor\`ub\'a speakers, with an inter-annotator agreement of over 0.70, ensuring high quality and consistency. We benchmark several transformer encoder models using cross-domain experiments with MasakhaNER 2.0, and we also assess the effect of few-shot in-domain data using YoNER and cross-lingual setups with English datasets. Our results show that African-centric models outperform general multilingual models for Yor\`ub\'a, but cross-domain performance drops substantially, particularly for blogs and movie domains. Furthermore, we observed that closely related formal domains, such as news and Wikipedia, transfer more effectively. In addition, we introduce a new Yor\`ub\'a-specific language model (OyoBERT) that outperforms multilingual models in in-domain evaluation. We publicly release the YoNER dataset and pretrained OyoBERT models to support future research on Yor\`ub\'a natural language processing.

2026-04-06

arXiv (prépublication)

AfrIFact: Cultural Information Retrieval, Evidence Extraction and Fact Checking for African Languages

Israel Abebe Azime

Jesujoba O. Alabi

Crystina Zhang

Iffat Maab

Atnafu Lambebo Tonja

Tadesse Destaw Belay

Folasade Alabi

Salomey Osei

Saminu Muhammad Aliyu

Nkechinyere Faith Aguobi

Bontu Fufa Balcha

Blessing Sibanda

Davis David

Mouhamadane Mboup

Daud Abolade

Neo Putini

Philipp Slusallek

Dietrich Klakow

Assessing the veracity of a claim made online is a complex and important task with real-world implications. When these claims are directed a… (voir plus)t communities with limited access to information and the content concerns issues such as healthcare and culture, the consequences intensify, especially in low-resource languages. In this work, we introduce AfrIFact, a dataset that covers the necessary steps for automatic fact-checking (i.e., information retrieval, evidence extraction, and fact checking), in ten African languages and English. Our evaluation results show that even the best embedding models lack cross-lingual retrieval capabilities, and that cultural and news documents are easier to retrieve than healthcare-domain documents, both in large corpora and in single documents. We show that LLMs lack robust multilingual fact-verification capabilities in African languages, while few-shot prompting improves performance by up to 43% in AfriqueQwen-14B, and task-specific fine-tuning further improves fact-checking accuracy by up to 26%. These findings, along with our release of the AfrIFact dataset, encourage work on low-resource information retrieval, evidence retrieval, and fact checking.

2026-03-31

arXiv (prépublication)

MGSM-Pro: A Simple Strategy for Robust Multilingual Mathematical Reasoning Evaluation

Tianyi Xu

Kosei Uemura

Alfred Malengo Kondoro

Tadesse Destaw Belay

Catherine Nana Nyaah Essuman

Ifeoma Okoh

Ganiyat Afolabi

Ayodele Awokoya

Large language models have made substantial progress in mathematical reasoning. However, benchmark development for multilingual evaluation h… (voir plus)as lagged behind English in both difficulty and recency. Recently, GSM-Symbolic showed a strong evidence of high variance when models are evaluated on different instantiations of the same question; however, the evaluation was conducted only in English. In this paper, we introduce MGSM-Pro, an extension of MGSM dataset with GSM-Symbolic approach. Our dataset provides five instantiations per MGSM question by varying names, digits and irrelevant context. Evaluations across nine languages reveal that many low-resource languages suffer large performance drops when tested on digit instantiations different from those in the original test set. We further find that models robustness in HRL setting do not necessarily translate to LRL. Moreover, proprietary models, such as Gemini 2.5 Flash and GPT-4.1 are less robust to digit, whereas Gemini 3.0 Pro is more robust. Among open models, GPT-OSS 120B and DeepSeek v3 show stronger robustness. Based on these findings, we recommend evaluating each problem using at least five digit-varying instantiations to obtain a more robust and realistic assessment of math reasoning.

2026-01-28

arXiv (prépublication)

Sudanese-Flores: Extending FLORES+ to Sudanese Arabic Dialect

Hadia Mohmmedosman Ahmed Samil

In this work, we introduce Sudanese-Flores, an extension of the popular Flores+ machine translation (MT) benchmark to the Sudanese Arabic di… (voir plus)alect. We translate both the DEV and DEVTEST splits of the Modern Standard Arabic dataset into the corresponding Sudanese dialect, resulting in a total of 2,009 sentences. While the dialect was recently introduced in Google Translate, there are no available benchmark in this dialect despite spoken by over 40 million people. Our evaluation on two leading LLMs such as GPT-4.1 and Gemini 2.5 Flash showed that while the performance English to Arabic is impressive (more than 23 BLEU), they struggle on Sudanese dialect (less than 11 BLEU) in zero-shot settings. In few-shot scenario, we achieved only a slight boost in performance.

2026-01-26

AfricaNLP @ Conference of European Chapter of Association for Computational Linguistics (publié)

openreview.net

Multilinguality as Sense Adaptation

Jan Christian Blaise Cruz

Alham Fikri Aji

2026-01-14

ArXiv (prépublication)