Yeah, but it's likely you can RLVR on this capability. Extrapolate from there.
But why would you train on this capability versus, say, optimizing RL training runs?
1. It's more direct
2. Could lead to interpretability gains
3. Might generalize
Probably more reasons I haven't thought of. I am not an expert, so take this as layman's opinion.
Edit: Catastrophic forgetting might also be able to be mitigated.
Catastrophic forgetting is just lossy compression.