Here, we report statistics about the uniqueness of neuron concepts, as we increase the maximum formula length of our explanations. Figure S1: Number of repeated concepts across probed vision and NLI models, by maximum formula length. Table S1: For probed Image Classification and NLI models, average number of occurrences of each detected concept and percentage of detected concepts that are unique (i.e. A.1 Image Classification Figure S1 (left) plots the number of times each unique concept appears across the 512 units of ResNet-18 as the maximum formula length increases. Table S1 displays the mean number of occurrences per concept, and percentage of concepts occurring that are unique (i.e.

When you see an image of a cat, how do you (your brain, that is) recognize that there is a cat in the image? And how do you tell the difference between a cat and a dog even though you haven't seen the specific animals before? These are tough questions to answer. How would you program a computer to recognize a cat in an image? A random cat placed randomly in an image that is.

Two hackers managed to trick a vein scanning system just by using a simple wax hand. As the name suggests, vein authentication systems use a computer to scan and verify a user's veins by studying their shape, size and position underneath a hand. But it seems the system isn't totally foolproof, as security researchers Jan Krissler and Julian Albrecht revealed the sensors could be duped by creating a fake hand out of wax. Security researchers Jan Krissler and Julian Albrecht discovered a vein authentication system could be fooled. Vein authentication is a increasingly popular form of biometric security.

The animals of neuroscience research are an eclectic bunch, and for good reason. Different model organisms--like zebra fish larvae, C. elegans worms, fruit flies, and mice--give researchers the opportunity to answer specific questions. The first two, for example, have transparent bodies, which let scientists easily peer into their brains; the last two have eminently tweakable genomes, which allow scientists to isolate the effects of specific genes. For cognition studies, researchers have relied largely on primates and, more recently, rats, which I use in my own work. But the time is ripe for this exclusive club of research animals to accept a new, avian member: the corvid family.

The animals of neuroscience research are an eclectic bunch, and for good reason. Different model organisms--like zebra fish larvae, C. elegans worms, fruit flies, and mice--give researchers the opportunity to answer specific questions. The first two, for example, have transparent bodies, which let scientists easily peer into their brains; the last two have eminently tweakable genomes, which allow scientists to isolate the effects of specific genes. For cognition studies, researchers have relied largely on primates and, more recently, rats, which I use in my own work. But the time is ripe for this exclusive club of research animals to accept a new, avian member: the corvid family.