FEGIS is a schema-driven memory engine that gives LLMs cognitive tools and structured persistent memory. Developed by Perry Golden, it uses Qdrant vector database with FastEmbed for efficient storage and retrieval of information based on predefined archetypes. The system allows models to create, store, and search through structured cognitive artifacts like thoughts, reflections, and decisions using a facet-based organization system. This implementation enables LLMs to maintain context across conversations, build knowledge bases with qualitative dimensions, and create meaningful connections between related ideas - making it particularly valuable for applications requiring persistent memory and structured thinking.
Aucun avis encore. Soyez le premier à donner votre avis !
Connectez-vous pour rejoindre la conversation
Highlight AIIdentify reasoning blind spots, cognitive biases, and systematic errors in AI thinking patterns through structured self-examination. Parameters: BiasScope (range of bias detection), IntrospectionDepth (depth of examination).
Search through stored memories and analyses using semantic similarity. Examples: Find analyses of privacy concerns, retrieve creative ideas generated last week, show UncertaintyNavigator results, or search for decision-making memories.
