Because it seems that the ResearchGate score relies more on the number of citations and less on the quality of the work or the significance of the journal in which the article was published.
The "Research Interest Score" calculated by ResearchGate is based on citations, "Reads" by members and recommendations (see https://help.researchgate.net/hc/en-us/articles/14293473316753-Research-Interest-Score). It should not be taken too seriously. Generally, all metrics based on citations reflect the quality of publications only to a very limited extent. In my opinion, the actual "quality", which consists of many factors, cannot really be measured. See also these discussions: