Publications

Brain-like neural dynamics for behavioral control develop through reinforcement learning

Olivier Codol

Nanda H. Krishna

Guillaume Lajoie

Matthew G. Perich

During development, neural circuits are shaped continuously as we learn to control our bodies. The ultimate goal of this process is to produ… (see more)ce neural dynamics that enable the rich repertoire of behaviors we perform. What begins as a series of “babbles” coalesces into skilled motor output as the brain rapidly learns to control the body. However, the nature of the teaching signal underlying this normative learning process remains elusive. Here, we test two well-established and biologically plausible theories—supervised learning (SL) and reinforcement learning (RL)—that could explain how neural circuits develop the capacity for skilled movements. We trained recurrent neural networks to control a biomechanical model of a primate arm using either SL or RL and compared the resulting neural dynamics to populations of neurons recorded from the motor cortex of monkeys performing the same movements. Intriguingly, only RL-trained networks produced neural activity that matched their biological counterparts in terms of both the geometry and dynamics of population activity. We show that this similarity with biological brains depends critically on matching biomechanical properties of the limb. Dynamical analysis on network activity revealed that our RL-trained networks operate at the “edge of chaos”, a dynamical regime known for its computational richness, greater memory capacity, and robust plasticity properties. We then demonstrated that monkeys and RL-trained networks, but not SL-trained networks, show a strikingly similar capacity for robust short-term behavioral adaptation to a movement perturbation, indicating a fundamental and general commonality in the neural control policy. Together, our results support the hypothesis that neural dynamics for behavioral control emerge through a process akin to reinforcement learning. The resulting neural circuits offer numerous advantages for adaptable behavioral control over simpler and more efficient learning rules and expand our understanding of how developmental processes shape neural dynamics.

2025-11-01

bioRxiv (preprint)

Survey on <scp>AI</scp> Ethics: A Socio‐Technical Perspective

Dave Mbiazi

Meghana Bhange

Maryam Babaei

Ivaxi Sheth

Patrik Joslin Kenfack

Samira Ebrahimi Kahou

Abstract The past decade has observed a significant advancement in AI, with deep learning‐based models being deployed in diverse scenarios… (see more), including safety‐critical applications. As these AI systems become deeply embedded in our societal infrastructure, the repercussions of their decisions and actions have significant consequences, making the ethical implications of AI deployment highly relevant and essential. The ethical concerns associated with AI are multifaceted, including challenging issues of fairness, privacy and data protection, responsibility and accountability, safety and robustness, transparency and explainability, and environmental impact. These principles together form the foundations of ethical AI considerations that concern every stakeholder in the AI system lifecycle. In light of the present ethical and future x‐risk concerns, governments have shown increasing interest in establishing guidelines for the ethical deployment of AI. This work unifies the current and future ethical concerns of deploying AI into society. While we acknowledge and appreciate the technical surveys for each of the ethical principles concerned, in this paper, we aim to provide a comprehensive overview that not only addresses each principle from a technical point of view but also discusses them from a social perspective.

2025-11-01

Computational Intelligence (published)

AFRIDOC-MT: Document-level MT Corpus for African Languages

Jesujoba Oluwadara Alabi

Israel Abebe Azime

Miaoran Zhang

Cristina España-Bonet

Rachel Bawden

Dawei Zhu

Clement Odoje

Idris Akinade

Iffat Maab

Davis David

Shamsuddeen Hassan Muhammad

Neo Putini

David O. Ademuyiwa

Andrew Caines

Dietrich Klakow

This paper introduces AFRIDOC-MT, a document-level multi-parallel translation dataset covering English and five African languages: Amharic, … (see more)Hausa, Swahili, Yor\`ub\'a, and Zulu. The dataset comprises 334 health and 271 information technology news documents, all human-translated from English to these languages. We conduct document-level translation benchmark experiments by evaluating neural machine translation (NMT) models and large language models (LLMs) for translations between English and these languages, at both the sentence and pseudo-document levels. These outputs are realigned to form complete documents for evaluation. Our results indicate that NLLB-200 achieved the best average performance among the standard NMT models, while GPT-4o outperformed general-purpose LLMs. Fine-tuning selected models led to substantial performance gains, but models trained on sentences struggled to generalize effectively to longer documents. Furthermore, our analysis reveals that some LLMs exhibit issues such as under-generation, repetition of words or phrases, and off-target translations, especially for African languages.

2025-10-31

Proceedings of the 2025 Conference on Empirical Methods in Natural Language Processing (published)

CAVE: Detecting and Explaining Commonsense Anomalies in Visual Environments

Rishika Bhagwatkar

Syrielle Montariol

Angelika Romanou

Beatriz Borges

Irina Rish

Antoine Bosselut

2025-10-31

Proceedings of the 2025 Conference on Empirical Methods in Natural Language Processing (published)

Characterization of astrocytic primary cilia in the adult mouse cortex and hippocampus

Sylvie C. Lahaie

Albert HK. Fok

Jessica M. Nicholls

Hannah Lee

Tabish A. Syed

Sabrina Chierzi

Sayuri Hatada

Naomi Egawa

Alex L. Schober

Tak Yi Mayumi Wong

Robert Royston

Kaleem Siddiqi

Yoshiyuki Kubota

Keith K. Murai

2025-10-31

Heliyon (published)

Charting the Landscape of African NLP: Mapping Progress and Shaping the Road Ahead

Jesujoba Oluwadara Alabi

Michael A. Hedderich

Dietrich Klakow

2025-10-31

Proceedings of the 2025 Conference on Empirical Methods in Natural Language Processing (published)

Collaborative Rational Speech Act: Pragmatic Reasoning for Multi-Turn Dialog

Lautaro Estienne

Gabriel Ben Zenou

Nona Naderi

Jackie Chi Kit Cheung

Pablo Piantanida

As AI systems take on collaborative roles, they must reason about shared goals and beliefs-not just generate fluent language. The Rational S… (see more)peech Act (RSA) framework offers a principled approach to pragmatic reasoning, but existing extensions face challenges in scaling to multi-turn, collaborative scenarios. In this paper, we introduce Collaborative Rational Speech Act (CRSA), an information-theoretic (IT) extension of RSA that models multi-turn dialog by optimizing a gain function adapted from rate-distortion theory. This gain is an extension of the gain model that is maximized in the original RSA model but takes into account the scenario in which both agents in a conversation have private information and produce utterances conditioned on the dialog. We demonstrate the effectiveness of CRSA on referential games and template-based doctor-patient dialogs in the medical domain. Empirical results show that CRSA yields more consistent, interpretable, and collaborative behavior than existing baselines-paving the way for more pragmatic and socially aware language agents.

2025-10-31

Proceedings of the 2025 Conference on Empirical Methods in Natural Language Processing (published)

Combining Virtual Reality and Hypnosis? A User Experience Study in Patients with Multiple Myeloma Following Stem Cell Transplantation

Jade Véronneau

Alexandra Chevestrier-Lefeuvre

Valentyn Fournier

Audrey Laurin

Rémi Caron-Trahan

Mathieu Landry

Joséphine Guiné

Odile Dubey-Harispe

Nadia Godin

Idrissi Moulay

Danny Wade

Sandie Oberoi

Caroline Arbour

Philippe Richebé

Karim Jerbi

Pierre Rainville

Richard LeBlanc

Floriane Rousseaux

David Ogez

Multiple myeloma (MM) and stem cell transplantation (SCT) significantly impact patients’ quality of life. Virtual reality with hypnosis (V… (see more)RH) has emerged as a promising nonpharmacological intervention to address these challenges, yet data on its acceptability and user experience remain scarce. This study assessed the acceptability and user experience of a VRH intervention among adult patients with MM who had undergone allogeneic SCT. Participants used a VRH application and rated their experience through standardized questionnaires and semistructured interviews. Quantitative data were analyzed descriptively, and qualitative data underwent descriptive content analysis. Findings indicated high patients’ satisfaction, strong perceived relevance, and low cybersickness. Qualitative analysis revealed perceived emotional and psychological benefits. VRH was deemed particularly suitable during hospitalization and treatment periods. This study shows the potential of combining virtual reality and hypnosis for MM patients following SCT. Indeed, they showed high satisfaction levels, paving the way for further studies evaluating the clinical efficacy of such interventions.

2025-10-31

Journal of Patient Experience (published)

Efficiency vs. Alignment: Investigating Safety and Fairness Risks in Parameter-Efficient Fine-Tuning of LLMs

Mina Taraghi

Yann Batiste Pequignot

Amin Nikanjam

Mohamed Amine Merzouk

Foutse Khomh

Organizations are increasingly adopting and adapting Large Language Models (LLMs) hosted on public repositories such as HuggingFace. Althoug… (see more)h these adaptations often improve performance on specialized downstream tasks, recent evidence indicates that they can also degrade a model's safety or fairness. Since different fine-tuning techniques may exert distinct effects on these critical dimensions, this study undertakes a systematic assessment of their trade-offs. Four widely used Parameter-Efficient Fine-Tuning methods, LoRA, IA3, Prompt-Tuning, and P-Tuning, are applied to four instruction-tuned model families (Meta-Llama-3-8B, Qwen2.5-7B, Mistral-7B, and Gemma-7B). In total, 235 fine-tuned variants are evaluated across eleven safety hazard categories and nine demographic fairness dimensions. The results show that adapter-based approaches (LoRA, IA3) tend to improve safety scores and are the least disruptive to fairness, retaining higher accuracy and lower bias scores. In contrast, prompt-based methods (Prompt-Tuning and P-Tuning) generally reduce safety and cause larger fairness regressions, with decreased accuracy and increased bias. Alignment shifts are strongly moderated by base model type: LLaMA remains stable, Qwen records modest gains, Gemma experiences the steepest safety decline, and Mistral, which is released without an internal moderation layer, displays the greatest variance. Improvements in safety do not necessarily translate into improvements in fairness, and no single configuration optimizes all fairness metrics simultaneously, indicating an inherent trade-off between these objectives. These findings suggest a practical guideline for safety-critical deployments: begin with a well-aligned base model, favour adapter-based PEFT, and conduct category-specific audits of both safety and fairness.

2025-10-31

ArXiv (preprint)

Evaluating WMT 2025 Metrics Shared Task Submissions on the SSA-MTE African Challenge Set

Senyu Li

Felermino Dario Mario Ali

Jiayi Wang

Rui Sousa-Silva

Henrique Lopes Cardoso

Pontus Stenetorp

Colin Cherry

2025-10-31

Proceedings of the Tenth Conference on Machine Translation (published)

Findings of the WMT25 Shared Task on Automated Translation Evaluation Systems: Linguistic Diversity is Challenging and References Still Help

Alon Lavie

Greg Hanneman

Sweta Agrawal

Diptesh Kanojia

Chi-kiu Lo

Vilém Zouhar

Frédéric Blain

Chrysoula Zerva

Eleftherios Avramidis

Sourabh Dattatray Deoghare

Archchana Sindhujan

Jiayi Wang

Brian Thompson

Tom Kocmi

Markus Freitag

Daniel Deutsch

2025-10-31

Proceedings of the Tenth Conference on Machine Translation (published)