New technique helps LLMs improve reasoning by ignoring irrelevant information

AWS adds Guardrails for Amazon Bedrock to help safeguard LLMs

New technique can accelerate language models by 300x

Martian’s tool automatically switches between LLMs to reduce costs