Stjepan Picek

This paper introduces the first backdoor attack specifically targeting pipeline parallelism in decentralized post-training, demonstrating that a limited adversary controlling an intermediate stage can significantly degrade model alignment.

NeuroLip proposes an event-based spatiotemporal framework for visual speaker recognition that achieves robust cross-scene generalization by capturing fine-grained lip dynamics, outperforming existing methods by over 8%.

The paper investigates the ability of evolutionary computation to discover monotone Boolean functions with high nonlinearity, demonstrating that genetic programming is a highly effective encoding for this task, especially in higher dimensions.

MASCing is a novel framework that enables flexible, non-retraining reconfiguration of Mixture-of-Experts (MoE) models for specific safety objectives by applying activation steering masks to control expert selection.

The paper introduces NeWTral, a framework that restores safety alignment to specialized LLM adapters without sacrificing their domain-specific knowledge, achieving a significant reduction in attack success rates while maintaining high fidelity.