For these release notes, we used an LLM (OpenAI o3) to review and summarise the code changes, along with the associated issue and PR descriptions. It may contain typical errors and inaccuracies. You can read the prompt here.