Phase 1 ships the project knowledge system and explorer. Phase 2 will expand each project into a dedicated research case study with architecture diagrams and implementation notes.
A structured knowledge base for core AI agent concepts, research papers, and engineering patterns.
A deep search agent direction for decomposing questions, retrieving evidence, reasoning over sources, and generating research-grade answers.
An AI agent system exploration project focused on emergent workflows, coordination, and agent environment design.
A simulator direction for studying emergent multi-agent behavior, world state, and autonomous interaction loops.
A personal agent research project exploring agent roles, interaction patterns, and structured workflows.
A fine-tuning pipeline study covering dataset preparation, supervised fine-tuning, LoRA-style adaptation, and evaluation.
A vision-oriented project used to study perception pipelines, model architecture, and applied AI experimentation.
A computer vision project direction connected to visual understanding, restoration, or multimodal learning practice.
A Qwen model adaptation project focused on dataset formatting, parameter-efficient tuning, and model behavior inspection.
A foundation track for Bayesian networks, optimization algorithms, deep reinforcement learning, and ML framework implementation.