Search code, repositories, users, issues, pull requests...
This is the official repository for our work "AlphaGo Moment for Model Architecture Discovery". We present a highly autonomous, multi-agent framework that empowers a Large Language Model (LLM) to conduct end-to-end scientific research in the challenging domain of linear attention mechanisms.
Autonomous Architecture Discovery Pipeline: Complete multi-agent system that autonomously hypothesizes novel architectural concepts, implements them as code, and empirically validates their performance through systematic experimentation.
Architecture Database: MongoDB-based cloud service storing all historical experimental data and enabling agent information retrieval, supporting multi-pipeline parallel execution.
Cognition Base: MongoDB-powered knowledge repository providing relevant paper cognitions and research insights to guide the autonomous research process.
ASI-Arch autonomous research framework demonstrating AI's capability to conduct end-to-end scientific discovery, from hypothesis generation to empirical validation.
