Institution profile

National Research Center for Applied Cybersecurity

Academic institutioneurope · de
Official website
Research library12linked papers
Opportunities0open roles
Selected work

Representative Papers

Mechanizing Choreographic Programs and Hoare Logic with State Transformers

Aug 17, 2026

This study addresses the challenges of complex binding substitution and the inherent difficulty of focusing on distributed semantics in choreography mechanization. We formalize a choreography language in Lean based on a state transformer model, which abstracts local operational details to streamline binding substitution and support diverse communication patterns. The research establishes key metatheoretic properties, including the soundness and completeness of endpoint projection, deadlock freedom and confluence of processes, and the correctness of Hoare logic. By successfully integrating dependent types with state transformers, this work provides a rigorous and scalable theoretical foundation and tooling support for verifying distributed choreographic programs.

0 citationsRead paper

Parameter Exploration for RLVR via Variational Learning

Aug 10, 2026

This work addresses a key limitation in current reinforcement learning approaches for large language models, which predominantly rely on action-space exploration—such as temperature scaling—and struggle to effectively reorder tokens, often leading to training divergence or stagnation. To overcome this, the paper introduces Perturbed Parameter Policy Optimization (3PO), the first systematic framework leveraging parameter-space exploration. Built upon a variational formulation of the policy posterior, 3PO generates diverse trajectories through parameter perturbations and enhances exploration efficiency via a reward-based grouping mechanism. Evaluated on OLMo-3-1025-7B and Qwen2.5-Math-7B across mathematical reasoning and code generation tasks, 3PO consistently outperforms standard GRPO, yielding substantial gains in downstream performance with negligible computational overhead while significantly reducing zero-advantage groups and erroneous outputs.

0 citationsRead paper

Component-Level Ensemble Fusion for Speech and Environmental Sound Deepfake Detection

Jul 17, 2026

This work addresses the fine-grained detection of joint speech and background audio spoofing by proposing a component-level ensemble method capable of distinguishing among five scenarios: speech-only forgery, background-only forgery, both forged, both genuine, and authentic recordings. The approach integrates four pretrained anti-spoofing models—XLSR-Mamba, DF-Arena, SLS, and TCM-ADD—and incorporates RawBoost data augmentation, multi-head fine-tuning, margin-space score fusion, and a lightweight class-bias calibration strategy. Evaluated in the ICME 2026 ESDD2 Challenge, the method achieves macro F1 scores of 0.7715 and 0.7828 on the evaluation and test sets, respectively, ranking 5th out of 31 participating teams and significantly outperforming the official baseline.

0 citationsRead paper
Recent publications

Latest Papers

Mechanizing Choreographic Programs and Hoare Logic with State Transformers

Aug 17, 2026

This study addresses the challenges of complex binding substitution and the inherent difficulty of focusing on distributed semantics in choreography mechanization. We formalize a choreography language in Lean based on a state transformer model, which abstracts local operational details to streamline binding substitution and support diverse communication patterns. The research establishes key metatheoretic properties, including the soundness and completeness of endpoint projection, deadlock freedom and confluence of processes, and the correctness of Hoare logic. By successfully integrating dependent types with state transformers, this work provides a rigorous and scalable theoretical foundation and tooling support for verifying distributed choreographic programs.

0 citationsRead paper

Parameter Exploration for RLVR via Variational Learning

Aug 10, 2026

This work addresses a key limitation in current reinforcement learning approaches for large language models, which predominantly rely on action-space exploration—such as temperature scaling—and struggle to effectively reorder tokens, often leading to training divergence or stagnation. To overcome this, the paper introduces Perturbed Parameter Policy Optimization (3PO), the first systematic framework leveraging parameter-space exploration. Built upon a variational formulation of the policy posterior, 3PO generates diverse trajectories through parameter perturbations and enhances exploration efficiency via a reward-based grouping mechanism. Evaluated on OLMo-3-1025-7B and Qwen2.5-Math-7B across mathematical reasoning and code generation tasks, 3PO consistently outperforms standard GRPO, yielding substantial gains in downstream performance with negligible computational overhead while significantly reducing zero-advantage groups and erroneous outputs.

0 citationsRead paper

Component-Level Ensemble Fusion for Speech and Environmental Sound Deepfake Detection

Jul 17, 2026

This work addresses the fine-grained detection of joint speech and background audio spoofing by proposing a component-level ensemble method capable of distinguishing among five scenarios: speech-only forgery, background-only forgery, both forged, both genuine, and authentic recordings. The approach integrates four pretrained anti-spoofing models—XLSR-Mamba, DF-Arena, SLS, and TCM-ADD—and incorporates RawBoost data augmentation, multi-head fine-tuning, margin-space score fusion, and a lightweight class-bias calibration strategy. Evaluated in the ICME 2026 ESDD2 Challenge, the method achieves macro F1 scores of 0.7715 and 0.7828 on the evaluation and test sets, respectively, ranking 5th out of 31 participating teams and significantly outperforming the official baseline.

0 citationsRead paper