Join the discussion

Write your take first — we'll ask for email only when you're ready to publish.

  • Hacker News
  • Excited to play with this.
  • Cool idea! How do you guarantee privacy?
  • Cool project
  • The title is misleading. This is model routing, not distillation.
  • Interesting approach. What does the cold-start phase look like for a new agent? How many traces or runs do you typically need before the router has enough signal to safely offload tasks from the frontier model??
  • Probably a silly question, but ...: if it's optimizing itself purely inside a simulator, what keeps it from just exploiting the simulator's glitches to look good on paper, while not being so good when it hits the real world?
  • Not sure I get it. The model you're improving is local? If so how do you even calculate cost compared to an API
  • Local models need to be tuned to work well so this looks useful. Seems to be for general purpose model serving. I’ve been using https://github.com/adrianco/retort to run experiments for coding models across 13 different programming languages to see which frontier and local models work.

Explore Birbla archives