RAM: Towards an Ever-Improving Memory System by Learning from Communications
Li, Jiaqi, Wang, Xiaobo, Ding, Wentao, Wang, Zihao, Kang, Yipeng, Jia, Zixia, Zheng, Zilong
–arXiv.org Artificial Intelligence
More in the human direction until they were recently, retrieval-augmented generation (RAG; present." Lewis et al. (2020)) is proposed to enable accessing and precisely manipulating the memory with a --Tomasello (2010) disentangled knowledge storage system; refer to 4 Human learning, extended as a lifelong process, for details. However, conventional RAG augments typically operates in a communicative and cooperative LLMs with a static and exterior knowledge to address framework among people via different forms knowledge-intensive tasks. Fundamentally, of interactions within the physical and social world, the main challenge of building CLAI agents lies as evidenced by the above quotes of Tomasello in determining when and how to update dynamic (2010). From toddlers to academic graduates, the and internal knowledge given communicative feedback.
arXiv.org Artificial Intelligence
Jul-5-2024
- Country:
- Asia
- Europe
- Germany (0.05)
- Ireland > Leinster
- County Dublin > Dublin (0.04)
- Italy (0.14)
- United Kingdom (0.14)
- North America > United States (0.14)
- Genre:
- Research Report > New Finding (0.67)
- Industry:
- Government (0.94)
- Leisure & Entertainment > Sports (0.47)
- Technology: