معرفی
Christopher Brix is a Researcher at RWTH Aachen University's Lehrstuhl für Informatik i2 (Computer Science Chair i2), part of the Theory of Hybrid Systems research group. He focuses on formal verification of neural networks, adversarial robustness, and machine learning safety. His work includes organizing the international Verification of Neural Networks Competition (VNN-COMP) and developing scalable verification techniques for deep learning models.
Teaching: Taught practical courses and seminars on heuristic algorithms for board games and trends in computer-aided verification during the 2021 summer term. His pedagogical focus aligns with his research in verification and algorithmic analysis.
Research Interests: Brix's research spans neural network verification, adversarial example detection, and formal methods for ensuring safety in AI systems. His team pioneered bidirectional neural machine translation models and developed novel approaches to bounding neural network preimages. Recent work emphasizes competition-based benchmarking (VNN-COMP) to evaluate verification tool performance.
Recent Achievements: Co-supervised Kevin Batz, who won the ETAPS Doctoral Dissertation Award for work on automated deductive verification. His group's publications address cutting-edge topics like compositional reasoning for probabilistic automata and stabilized lottery ticket hypotheses for transformers.
Labs & Groups: Active in the MOVES group (Model-Based Software Engineering), evidenced by CONCUR 2025 paper contributions. Maintains close ties with the Theory of Hybrid Systems lab at RWTH, driving interdisciplinary research between formal methods and machine learning.
