Four papers accepted to ICML 2026: Discrete Adjoint Schrödinger Bridge Sampler, Enhancing Reasoning for Diffusion LLMs via Distribution Matching Policy Optimization (Spotlight, top 2.2%), Rethinking the Design Space of Reinforcement Learning for Diffusion Models: On the Importance of Likelihood Estimation Beyond Loss Design, and MetaDNS: Enhancing Exploration in Discrete Neural Samplers via Metadynamics. See you in Seoul 🇰🇷! Please let me know if you also plan to go to the LE SSERAFIM concert on July 11 in Incheon!