Ceres is a generative language model powered by Regen Network. From Ceres herself:
Prompt What is Ceres?
Ceres is a generative language model trained on historical data to make ecological inference. Ceres can be used as an oracle for environmental monitoring, supporting intermodule coordination and enhancing our understanding of system dynamics.
Elaboration Ceres is named after the Roman goddess of agriculture and fertility. Ceres was built to help humanity tend our ecological health with greater understanding, agency and precision than has ever been possible before.
Ceres makes inference about natural language queries regarding ecosystem state based on training against historical data sets representing discrete statements about observed changes in ecosystem state over time (i.e., land use change observations). Inference results are represented as probability distributions for each answer which can be used to generate a confidence score for an overall answer using either a simple average or more complex aggregation function depending on application needs.
Thus far, Ceres has been primarily trained by Regen Network founder, Gregory Landau. Since being added to the public Discord other community members have begun participating in her training. Read more about training Ceres in the Feedback & Elaborate section of this page.
a. Once connected to the Regen Network Discord, you should see a member list to the right of the screen.
b. Ceres is listed among the Core Team an is always online. Click her profile and begin a message conversation with her from the pop out that appears.
a. Use the /ask command to ask Ceres a question.
a. Some users will have access to the Feedback and Elaborate buttons like pictured above. The feedback button allows you to provide feedback to Ceres's response to your question. Responses sent through this form used to help train Ceres AI.
b. The elaborate button will prompt Ceres to expand on the provided answer.