Continual-ARCby unify

ReplayARC-AGI-1 · Workstreams · GPT-6 Luna, low

Hermes Agent, as shipped

Every learning mechanism the harness ships switched on.

Share learned
−0.05 ±0.06
Stream, budget used
45.7 solved 75%
Epilogue, budget used
60.3 solved 50%
Model spend
$9.69 17,922 calls
Instances
276 19.3 hours

Budget used is the share of the cap of 16 that an instance was charged, averaged over the instances: a solved instance costs 1 for each wrong answer and each demonstration, an unsolved one all 16, so lower is better. The share learned reads the stream's budget against the learner itself reset, the same system wiped before every instance of the same stream: the share of the budget left to learn that the learner saves, 0 for a learner that keeps nothing and 1 for perfect memory.

Loading the run