MeMo's memory model lets teams upgrade their LLM without retraining it — and performance jumps 26%

Researchers' MeMo keeps AI memory separate from reasoning, so teams can upgrade their LLM without retraining it and see a 26% ...

AI’s Memory Crisis Is Here: Don’t Hoard, Optimize

The GPUs powering today's models carry limited high-bandwidth memory (HBM) before external memory is required—that's the memory wall, and at inference scale, every model hits it. As the industry ...
Scientific American

Bumblebees use tools to solve complex problems—despite not being trained to do so

Bumblebees appear to be capable of coming up with creative solutions to new problems to get a sugary reward—and their ...

Tether Brings AI Memory Compression To Consumer Devices

Tether’s TurboQuant enables useful and powerful local AI applications on consumer devices at much lower costs and without ...

A 0.12% parameter add-on gives AI agents the working memory RAG can't

Researchers built delta-mem to give AI agents working memory at 0.12% parameter overhead, outperforming RAG and context expansion on memory-heavy agent tasks.
IEEE

Identifying Mild Cognitive Impairment and Healthy Aging Using EEG Visual Working Memory Task with XGBoost

Abstract: Mild Cognitive Impairment (MCI) is a stage that marks the transition from healthy aging to severe cognitive decline, often associated with impaired working memory. This study investigates ...