It turns out there’s actually a term for that: semantic satiation.
The explanation for the phenomenon was that verbal repetition repeatedly aroused a specific neural pattern in the cortex which corresponds to the meaning of the word. Rapid repetition causes both the peripheral sensorimotor activity and the central neural activation to fire repeatedly, which is known to cause reactive inhibition, hence a reduction in the intensity of the activity with each repetition. Jakobovits James (1962) calls this theory the beginning of “experimental neurosemantics.”
There is also a reverse but related phenomenon, semantic generation, which describes an increase in the intensity of meaning of a word when it is repeated:
It would seem at first that semantic generation and semantic satiation are contradictory phenomena, since at one time repeated exposure leads to an increase in the intensity of meaning and at another time to a decrease. This seeming paradox can be resolved by presenting a frequency law (see Jakobovits & Lambert, 1963, and Jakobovits & Hogenraad, 1967) which states that the relation between the intensity of a response and the frequency of exposure of the stimulus approximates an inverted U-shaped distribution. The rising part of the curve represents the semantic generation phase, indicating an increase in intensity of meaning during the initial stages of repetition. The falling part of the curve represents the semantic satiation phase, indicating a decrease in intensity of meaning as exposure continues. The point where the curve changes inflection is referred to as the “critical point” and marks the stage after which exposure results in an inhibitory rather than a facilitative effect.