Researchers at Microsoft introduce Aurora, a large-scale foundation model of the atmosphere trained on diverse weather and climate data. Plus, a new AI paper explores the self-improvement of LLMs in complex tasks. Also, aligning large language models with user preferences using multifaceted system messages. And, a deep learning method adds 301 planets to Kepler's total count.

Sources:https://www.marktechpost.com/2024/06/03/researchers-at-microsoft-introduce-aurora-a-large-scale-foundation-model-of-the-atmosphere-trained-on-over-a-million-hours-of-diverse-weather-and-climate-data/https://www.marktechpost.com/2024/06/02/this-ai-paper-explores-the-extent-to-which-llms-can-self-improve-their-performance-as-agents-in-long-horizon-tasks-in-a-complex-environment-using-the-webarena-benchmark/https://www.marktechpost.com/2024/06/02/aligning-large-language-models-with-diverse-user-preferences-using-multifaceted-system-messages-the-janus-approach/https://science.nasa.gov/universe/exoplanets/new-deep-learning-method-adds-301-planets-to-keplers-total-count/

Outline:(00:00:00) Introduction(00:00:52) Researchers at Microsoft Introduce Aurora: A Large-Scale Foundation Model of the Atmosphere Trained on Over a Million Hours of Diverse Weather and Climate Data(00:04:08) This AI Paper Explores the Extent to which LLMs can Self-Improve their Performance as Agents in Long-Horizon Tasks in a Complex Environment Using the WebArena Benchmark(00:07:01) Aligning Large Language Models with Diverse User Preferences Using Multifaceted System Messages: The JANUS Approach(00:09:51) New Deep Learning Method Adds 301 Planets to Kepler's Total Count