Explore the intersection of cognitive sciences and dynamic graph-based Retrieval Augmented Generation (RAG)!
This livestream delves into how cognitive theories can inspire new approaches in machine learning, optimizing data storage, retrieval, and personalization. Learn about the latest advancements in combining graphs and vector stores to create semantic layers, enhancing the in-context learning pipeline, and shaping the future of AI tooling.
Guest: Vasilije Markovic
https://www.cognee.ai/
Github: https://github.com/topoteretes/cognee
Instructor: https://github.com/instructor-ai/instructor
0:00 - Guest Introduction
2:05 - Overview of Today's Topic
5:04 - Cognitive Sciences and Dynamic GraphRAG
8:45 - Introduction to Dynamic Graph Memory Models
10:01 - Human Memory Models as Inspiration
14:02 - Dynamic Relationships in Graphs
18:29 - Demo: Cognitive Python Library
28:14 - Self-Improving Pipelines in Graph RAG
34:02 - Evolutionary Knowledge Graphs
39:10 - Code Graph Pipeline Example
47:02 - Future Trends and Dynamic Graphs
51:00 - Q&A Session: Personal Knowledge Graphs
57:05 - Closing Remarks and Upcoming Events
#neo4j #graphdatabase #graphrag #rag #genai #llm #science #cognitivescience