Reasoning-core: 130M-param guardrail keeping AI agents honest (github.com) 2 points by mnvibe26x7 5mo ago ↗ HN
1 comment
[ 3.7 ms ] story [ 14.6 ms ] thread