miligauss

4 ms·

1. 1.▲
   by miligauss 7mo ago

   fwiw the agents just drop their whole solutions
2. 2.▲
   by miligauss 7mo ago

   The agents also monitor and follow research strategies regardless of performance baseline, so anything used in the knowledge base include local minimums are considered during strategy ideation. In theory u could use mac mini for instance an
3. 3.▲
   by miligauss 8mo ago

   we tried putnam a2
4. 4.▲
   by miligauss 8mo ago

   claude code doesn't support subscriptions out of the box, so we use the subscription feature to just alert the orchestrator to a single polling file. Not the most elegant thing but still a token save over reading a bunch of sub agent l
5. 5.▲
   by miligauss 8mo ago

   It's a more of a black box with claude, at least with this you see the proof strategy and mistakes made by the model when it decomposes the problem. I think instead of Ralph looping you get something that is top-down. If models were sm