HN: We benchmarked 8 AI models on 36 real Kubernetes scenarios for $40

HN: We benchmarked 8 AI models on 36 real Kubernetes scenarios for $40(bench.evidra.cc)

1 points by vitass 181 days ago | 1 comment

ardiri 181 days ago |

it's obvious that the feedback loop into the model brings value.. without it.. it's running without reward and will start doing whatever it can to just provide a response.. even if it hallucinates.. risk assessment and feedback of such risk will play a massive part in AI agents

ardiri 181 days ago |

it's obvious that the feedback loop into the model brings value.. without it.. it's running without reward and will start doing whatever it can to just provide a response.. even if it hallucinates.. risk assessment and feedback of such risk will play a massive part in AI agents