Marion-Cécile Martin

Mélanie Ducoffe, Guillaume Povéda, Audrey Galametz et al.

Surrogate Neural Networks are nowadays routinely used in industry as substitutes for computationally demanding engineering simulations (e.g., in structural analysis). They allow to generate faster predictions and thus analyses in industrial applications e.g., during a product design, testing or monitoring phases. Due to their performance and time-efficiency, these surrogate models are now being developed for use in safety-critical applications. Neural network verification and in particular the assessment of their robustness (e.g., to perturbations) is the next critical step to allow their inclusion in real-life applications and certification. We assess the applicability and scalability of empirical and formal methods in the context of aircraft predictive maintenance for surrogate neural networks designed to predict the stress sustained by an aircraft part from external loads. The case study covers a high-dimensional input and output space and the verification process thus accommodates multi-objective constraints. We explore the complementarity of verification methods in assessing the local stability property of such surrogate models to input noise. We showcase the effectiveness of sequentially combining methods in one verification 'pipeline' and demonstrate the subsequent gain in runtime required to assess the targeted property.

Catherine Kobus, François Lancelot, Marion-Cécile Martin et al.

This paper presents the contributions of the ATLANTIS team to SemEval-2025 Task 3, focusing on detecting hallucinated text spans in question answering systems. Large Language Models (LLMs) have significantly advanced Natural Language Generation (NLG) but remain susceptible to hallucinations, generating incorrect or misleading content. To address this, we explored methods both with and without external context, utilizing few-shot prompting with a LLM, token-level classification or LLM fine-tuned on synthetic data. Notably, our approaches achieved top rankings in Spanish and competitive placements in English and German. This work highlights the importance of integrating relevant context to mitigate hallucinations and demonstrate the potential of fine-tuned models and prompt engineering.

Jarod Duret, Salima Mdhaffar, Gaëlle Laperrière et al.

In this study, we investigate the benefits of domain-specific self-supervised pre-training for both offline and streaming ASR in Air Traffic Control (ATC) environments. We train BEST-RQ models on 4.5k hours of unlabeled ATC data, then fine-tune on a smaller supervised ATC set. To enable real-time processing, we propose using chunked attention and dynamic convolutions, ensuring low-latency inference. We compare these in-domain SSL models against state-of-the-art, general-purpose speech encoders such as w2v-BERT 2.0 and HuBERT. Results show that domain-adapted pre-training substantially improves performance on standard ATC benchmarks, significantly reducing word error rates when compared to models trained on broad speech corpora. Furthermore, the proposed streaming approach further improves word error rate under tighter latency constraints, making it particularly suitable for safety-critical aviation applications. These findings highlight that specializing SSL representations for ATC data is a practical path toward more accurate and efficient ASR systems in real-world operational settings.