Nikolaus Kriegeskorte

Computational neuroscientist who introduced representational similarity analysis (RSA) — the method of comparing two systems by the geometry of their internal representations — in Kriegeskorte, Mur & Bandettini (2008), and helped make it the standard tool for relating deep neural networks to brain activity. He is the ancestry figure behind the vault's RSA-underdetermination thread: the method he originated is the one Grujičić and others interrogate when they argue that a representational match does not fix a shared mechanism. Named here to close the recurring blind spot of citing a method's critics without a hub for the person whose method is under critique.

References

• claim-representational-similarity-underdetermines-mechanism — RSA underdetermines shared mechanism
• claim-similarity-measures-pick-out-different-representational-vehicles — how the choice of RSA similarity measure selects the mechanism-type
• claim-similarity-measure-choice-reverses-neural-representational-conclusions — RSA similarity orderings reverse with the metric