activity
20242026
collaborators

7 papers

cs.LG2026

Trading Complexity for Expressivity Through Structured Generalized Linear Token Mixing

Erwan Fagnou, Paul Caillon, Blaise Delattre +1

Token mixing layers play a key role in how language models can learn and generate long-range dependencies. Their efficiency relies on the necessary trade-off between decoding speed…

cs.LG2026

Structured-Sparse Attention for Entity Tracking with Subquadratic Sequence Complexity

Hangyue Zhao, Paul Caillon, Erwan Fagnou +1

Entity tracking requires maintaining and updating latent states for entities and attributes over long sequences. Recent task-specific attention operators can compress deep Transfor…

cs.LG2025

Forward Only Learning for Orthogonal Neural Networks of any Depth

Paul Caillon, Alex Colagrande, Erwan Fagnou +2

Backpropagation is still the de facto algorithm used today to train neural networks. With the exponential growth of recent architectures, the computational cost of this algorithm a…

cs.LG2025

Bridging the Theoretical Gap in Randomized Smoothing

Blaise Delattre, Paul Caillon, Quentin Barthélemy +2

Randomized smoothing has become a leading approach for certifying adversarial robustness in machine learning models. However, a persistent gap remains between theoretical certified…

cs.LG2025

Fast Training of Recurrent Neural Networks with Stationary State Feedbacks

Paul Caillon, Erwan Fagnou, Alexandre Allauzen

Recurrent neural networks (RNNs) have recently demonstrated strong performance and faster inference than Transformers at comparable parameter budgets. However, the recursive gradie…

cs.LG2025

Accelerated Training through Iterative Gradient Propagation Along the Residual Path

Erwan Fagnou, Paul Caillon, Blaise Delattre +1

Despite being the cornerstone of deep learning, backpropagation is criticized for its inherent sequentiality, which can limit the scalability of very deep models. Such models faced…