evhub comments on Relaxed adversarial training for inner alignment

[evhub]

The idea is that we’re thinking of pseudo-inputs as “predicates that constrain X” here, so, for $\alpha \in X_{\text{pseudo}}$, we have $\alpha :X\to B$.

[Anthony DiGiovanni]

Ah right, thanks! (My background is more stats than comp sci, so I’m used to “indicator” instead of “predicate.”)