This sounds like the interaction between run-of-the-mill execution algorithms.
Bidding algo is likely a market maker putting out small quotes (100 shares is a standard lot) and adjusting down (up) the price each time the bid (offer) is hit (lifted). Selling algo trying to liquidate a larger block without moving the market. It is hitting the top bid from time to time. If those two are the only market participants talking for a few milliseconds, they'll walk down the price in 100-share increments. Given how volatile GameStop was, I suspect the predatory algorithms, who sniff out this sort of stuff, were offline.
One of the things I'm realizing from all this is there aren't many good, succinct sources on market microstructure. It's complicated. But it's not that complicated. (It's just usually boring.)
The best I can recommend is how I learned it. Start with a respected paper [1]. Trace through the references until you find something you understand. Then work your way forward.
[1] https://www.smallake.kr/wp-content/uploads/2016/03/optliq.pd...
It was not. Only retail that played in a couple of YOLO gamified market brokerages that was restricted.
The fact that retail was restricted to selling and hedge funds were not, further corroborates foul play.