The apparent advantage is exaggerated by them running Astra at six different effort levels, and almost everything else at just the maximum available effort.

I don't really understand why they keep doing this. Either run and report everyone at multiple effort levels, or run everyone at only one.

But alsi, token efficiency seems pretty artificial? For example tokenizers are different from model to model. The cost/perf Pareto frontier seems a lot more meaningful (and Astra does very well at that too, just to be clear. It seems to be a great model.)

Duh, that's just a benchmarking artifact. If you run one model at 4x recurrence compared to another, you get 4x thinking without increasing the tokens. So of course it's going to dominate the perf at given output token level. The truly sensible comparison is perf -vs- thinking flops (because each model might be a different unknown size, but labs are very secretive about what they're actually running under the hood) or perhaps cost (which can be misleading because of subsidies, but is at least practically relevant in the moment).

GPT-6 has a looped transformer / recurrent depth architecture, so it gets some internal CoT reasoning "for free" with no output tokens.

This also makes it much harder to monitor its reasoning.