Igor Ivanov comments on Concept Poisoning: Probing LLMs without probes