Research · Hacker News ·
Δ-Mem: Efficient Online Memory for Large Language Models
Δ-Mem proposes an online memory mechanism for large language models that aims to improve efficiency during long-context use. The work is presented as an arXiv paper on memory handling for LLMs.