PhD Thesis — Intertextuality & Causality
A semantic analysis of discourse bridging NLP with SAT-based formal methods, from echoes between migration narratives to causal network discovery in biographical trajectories.
NLP for languages the industry forgot
Ngambay · spoken by ~1M people in Chad
Sakayo Toadoum Sari
PhD candidate in AI at CRIL-CNRS, Université d'Artois, working on discourse semantics and causal reasoning, and on multilingual evaluation for African languages: whether benchmark numbers actually measure the reasoning they claim to. If your language isn't in the dataset, build the dataset.
01 · About
I grew up in Chad speaking a language with almost no digital footprint. Today I build the technology that gives languages like mine one. My research at CRIL-CNRS (UMR 8188) connects two worlds that rarely meet: discourse-level NLP and formal methods, probing how texts echo one another and how causal structure can be extracted from narratives, with formal guarantees rather than heuristics.
A second thread runs alongside it: multilingual evaluation. Having built speech and translation systems for African languages, I keep meeting the same problem, which is that the automatic metrics say very little about whether a model understood anything. That question is what most of my current benchmark work is about.
I teach algorithmics at Université d'Artois. On YouTube, Sakayo Tech brings AI education in French to students across francophone Africa, and my mentorship program supports a first cohort of 20 young innovators in Chad. I've worked with UNICEF, EPFL, CDIAL, and Batazia, and I contribute to the Masakhane and AfricaNLP communities as a researcher and reviewer.
Lens, France
02 · Current work
A semantic analysis of discourse bridging NLP with SAT-based formal methods, from echoes between migration narratives to causal network discovery in biographical trajectories.
Detecting experiential intertextuality in migration narratives. Shows lexical methods beat semantic embeddings on this task, and explains why.
Solo end-to-end French RAG pipeline built under strict compute constraints, scoring above the field mean on all six official metrics.
A diagnostic probe asking whether MT quality metrics track causal reasoning in African languages. They correlate with accuracy far more loosely than their use in evaluation assumes.
Logical Analysis of Data built into random forest construction via Minimal Support Sets, producing abductive explanations roughly five times smaller than standard forests on microarray data.
Separability and discriminability come apart under LoRA. A set-cover complexity measure guides rank selection, and a contrastive coverage loss improves low-resource and out-of-distribution results.
Probing with inclusion-minimality guarantees: which dimensions a model actually needs to encode a property, certified rather than estimated.
AI education series in French for African students: concepts, careers, and hands-on tools, published twice a week.
03 · Publications
Under review
Published
2026 IntertextMigra: Detecting Experiential Intertextuality in Migration Narratives SIGDIAL 2026 · accepted 2026 French RAG under Strict Constraints — EvalLLM Challenge CORIA-TALN 2026 · accepted 2025 Pretraining Strategies using Monolingual and Parallel Data for Low-Resource MT AfricaNLP 2025 2024 AfriMTE & AfriCOMET: Empowering COMET to Embrace Under-resourced African Languages NAACL 2024 2023 MasakhaNEWS: News Topic Classification for African Languages IJCNLP-AACL 2023 · most-cited work 2023 Ngambay–French Neural Machine Translation NLP4TIA @ RANLP 2023 2022 Error-Correction DNNs for South African Wildlife-Vehicle Collision Data MDPI Mathematics04 · Journey
First papers accepted at SIGDIAL and CORIA-TALN, with submissions under review at AAAI and EACL. Teaching algorithmics, and working on benchmarks for causal reasoning in African languages.
Joined CRIL on the ANR HYCI project. RCCE consultancy for UNICEF in Chad. Worked on ASR and TTS for African languages at CDIAL, including a human evaluation pipeline with 20+ annotators.
MSc in Machine Intelligence as a Google & Meta Scholar. Research internship at EPFL on polarization detection. NLP engineering at Batazia. Trained 40+ students in Python and ML.
MSc in Industrial Mathematics as a MasterCard Foundation Scholar.
BSc in Applied Physics, then MSc in Physics. First steps in web development and ML at Limbe City Council.
05 · Skills
06 · Contact
Building something for underserved languages, or want to collaborate on discourse, causality, or African NLP? I'd love to hear about it.