What Is Intern-Discovery
Intern-Discovery is a scientific discovery platform launched by Shanghai Artificial Intelligence Laboratory. Through model sharing, data interoperability, and facility interconnection, the platform provides one-stop support for the entire scientific research process with human-machine collaboration, integrated data and computing, and the combination of dry and wet experiments. The platform is based on the "Shusheng" general-purpose and specialized integrated foundation large model, integrating multi-domain professional agents, massive scientific data, and experimental equipment resources. The platform supports the complete scientific research process from hypothesis generation to experimental validation, helping researchers efficiently complete interdisciplinary knowledge discovery and innovation. Currently, the platform has opened trial applications, which can promote the co-construction of the scientific intelligence ecosystem and accelerate the large-scale development of scientific research.
Main Features of Intern-Discovery
- Scientific Data Plaza: Gathers high-quality professional data resources in core scientific research fields, supporting data-driven scientific discovery and interdisciplinary research.
- Scientific Research Assistant: Provides one-stop end-to-end algorithmic tools for interdisciplinary researchers, enabling full-cycle automation of scientific research workflows in AI+X fields and helping accelerate scientific breakthroughs.
- Scientific Agents: Integrates agent tools for scientific research tasks, covering multiple scenarios such as dry-wet experiment collaboration, data analysis, and research assistance.
- Scientific Model Development: Provides high-performance computing power and modeling tools required for scientific computing, supporting model training and inference deployment throughout the scientific research process.
How to Use Intern-Discovery
- Submit a Trial Application: Fill out the trial application form:https://aicarrier.feishu.cn/share/base/form/shrcngFNOyh39UtEILODs7CJHfh?sessionid=994664144。
- Log In to the Platform: After receiving the approval notification, log in to Intern-Discovery using the account information provided by the platform.
- Become Familiar with Platform Features: Browse the platform interface, review the user manual and tutorials, and quickly master platform operations.
- Use Platform Features：
- Model Sharing: Select and combine general-purpose large models and professional agents, and create new agents with a low-code development environment.
- Data Interoperability: Search and use professional datasets on the platform for data preprocessing and modeling analysis.
- Facility Interconnection: Connect experimental equipment through the SCP protocol, and remotely invoke and automatically schedule experimental workflows.
Application Scenarios of Intern-Discovery
- Protein Structure Prediction and Design: Rapidly predict protein structures and design variants, significantly shortening experimental cycles.
- Drug Development: Integrate drug target data and virtual screening tools to accelerate drug discovery and optimization and improve R&D efficiency.
- New Material Design and Synthesis: Based on agents in physics, chemistry, and materials science, design new materials and predict performance, supporting the full process from theory to experiment.
- Climate Change Research: Integrate authoritative meteorological datasets, build and validate climate models, and analyze climate change trends.
- Geological Disaster Prediction: Combine AI models and GIS data to develop geological disaster prediction models and improve the accuracy and timeliness of early warnings.