Large language models

By reasoning and accessing knowledge in an alternative language, large language models can produce more reliable answers in the user’s original language.

Inspired by computer vision hardware, researchers adapt a two-in-one crossbar-based chip design to boost the energy efficiency of AI systems.

Bigger models are not always better, as A*STAR researchers show how a targeted AI model can beat global, general-purpose models in summarising multilingual regional news.

The largest RNA language model to date helps researchers uncover the hidden roles played by non-coding RNA sequences in our cells.

A new comprehensive benchmark for audio-based artificial intelligence allows researchers to gauge and improve the way models follow speech, soundscapes and everyday cues.

By adding rewards for logical reasoning steps, a new training framework aims to make large language models more robust and reliable.

An AI-based framework adds a personal touch to medical dialogues, supporting the health management plans of patients with diabetes.