Brief

Researchers have introduced AdaNAGED, a novel parameter-free optimization method designed for efficient fine-tuning of large language models (LLMs). This approach unifies gradient-free training, adaptive parameter tuning, and geometry-aware updates, addressing the memory overhead associated with traditional backpropagation methods. The method has demonstrated convergence guarantees and has been validated on the OPT-1.3B model for large-scale LLM fine-tuning tasks. AI

This article details how to create an "LLM Wiki" using Zo, a platform designed to build persistent, queryable knowledge bases from personal documents. It addresses the limitations of traditional search and RAG workflows, which often require manual synthesis or re-processing of information for each query. The Zo approach focuses on continuously building structured understanding by treating documents as part of an evolving knowledge layer, rather than isolated chunks, leveraging its persistent workspace and always-on compute environment. AI

MiniMax AI showcased its Zo chatbot in a livestream event held in Singapore. The demonstration highlighted the capabilities of MiniMax models integrated within the Zo platform. AI