Thank you Charbel-Raphael! The agenda builds on proof-certificate-based verification, and there is growing interest in the topic of learnable neural certificates for AI systems safety. I mention two relevant papers from my group and collaborators, but there is much more:

• Neustroev, Giacobbe, Lukina, Neural Continuous-time Supermartingale Certificates, AAAI25

• Giacobbe, Kroening, Pal, Tautschnig. Neural Model Checking. NeurIPS24