This is a list of key contributors who have made key contributions which are "powering" this repository:
Patrick Hammer @ Stockholm University (Providing a view of NARS memory to GPT and initial prototype, integration of the following ideas of other contributors into a coherent system using GPT-3.5 and ONA)
Peter Isaev @ Temple University (New improvements, migration to new API, and connection to MeTTa-NARS)
Julius Pfrommer @ Fraunhofer Institute (Extensive experiments regarding sub-symbolic NARS long-term memory maintainence and reasoning with vector embeddings, and related encodings to translate Narsese to vectors)
Hugo Latapie @ Cisco (Neuro-symbolic hybrid representation for relational representation which combines the benefits of Narsese and sentence embeddings, which is now the key memory structure of gptONA! And GPT prompting for symbolic knowledge extraction from text)
Robert Johannson @ Stockholm University (Relational reasoning experiments regarding symmetric and transitivity usage, including comparisons between NarsGPT and gptONA which revealed NarsGPT limitations in reasoning compared to gptONA)
Peter Voss @ Aigo (comprehensive incremental Q&A benchmark which covers question answering in a natural personal assistant setting, this is the first benchmark this system has been tried on)
This is a list of key contributors who have made key contributions which are "powering" this repository: