Institution profile

IRT Saint Exupéry

Academic institutioneurope · fr
Official website
Research library38linked papers
Opportunities0open roles
Selected work

Representative Papers

Efficient Robust Conformal Prediction via Lipschitz-Bounded Networks

Jun 05, 2025

Traditional conformal prediction (CP) fails under adversarial attacks, while existing robust CP methods suffer from excessively large prediction sets or high computational overhead on large-scale tasks. To address this, we propose Lip-RCP—the first efficient robust prediction framework that deeply integrates 1-Lipschitz robust neural networks with CP. Methodologically, we impose Lipschitz constraints to ensure output stability and derive, for the first time, a theoretical worst-case coverage bound for standard CP under arbitrary attack magnitudes. Experiments on medium- and large-scale benchmarks (e.g., ImageNet) show that Lip-RCP reduces robust prediction set size by up to 42% over state-of-the-art methods while accelerating inference by 3.8×. Crucially, it strictly guarantees both nominal coverage ≥90% and finite-sample robust coverage—without compromising statistical validity.

1 citationsRead paper

When Do Concepts Become Functionally Sufficient During Language-Model Training?

Aug 15, 2026

This study addresses the unclear evolution of functional sufficiency regarding internal concepts in language models by proposing an interventional verification framework centered on this criterion. Through activation decomposition, sparse soft masking, and cross-checkpoint alignment, interpretability hypotheses are translated into functional tests involving reconstruction, decoding, and downstream tasks. Experiments across seven models demonstrate that downstream mask soft quality is significantly lower than reconstruction mask quality, accompanied by minimal predictive distribution shift. These findings elucidate the functional evolutionary dynamics of conceptual structures during training, providing empirical evidence and a novel paradigm for assessing the functional completeness of internal model representations.

0 citationsRead paper
Recent publications

Latest Papers

When Do Concepts Become Functionally Sufficient During Language-Model Training?

Aug 15, 2026

This study addresses the unclear evolution of functional sufficiency regarding internal concepts in language models by proposing an interventional verification framework centered on this criterion. Through activation decomposition, sparse soft masking, and cross-checkpoint alignment, interpretability hypotheses are translated into functional tests involving reconstruction, decoding, and downstream tasks. Experiments across seven models demonstrate that downstream mask soft quality is significantly lower than reconstruction mask quality, accompanied by minimal predictive distribution shift. These findings elucidate the functional evolutionary dynamics of conceptual structures during training, providing empirical evidence and a novel paradigm for assessing the functional completeness of internal model representations.

0 citationsRead paper

Industrial Application of a Multi-Disciplinary Design Optimization with Uncertainties to a Pair of Telecommunication Satellites

Jul 21, 2026

Traditional satellite design relies on conservative safety margins to account for uncertainties, often resulting in substantial performance penalties. This work proposes an integrated framework that combines uncertainty quantification, sensitivity analysis, and reliability-based multidisciplinary design optimization, applied for the first time to an industrial-scale dual-communication rideshare satellite system. By eliminating conventional safety margins and instead ensuring that all design constraints are satisfied with high probability, the proposed approach reduces performance loss by 66%. The framework demonstrates strong scalability and is readily extendable to the collaborative optimization of satellite constellations comprising any number of spacecraft.

0 citationsRead paper