Reasoning-core: 130M-param guardrail keeping AI agents honest

(github.com)

2 points | by mnvibe26x7 9 hours ago ago

1 comments