Liquid neural networks, comprehensive health checks for model deployments, Bayesian inference framework for in-context learning in LLMs, a paper on Hard Prompts Made Easy, and more.