I'm not sure this is true. Mutation happens with approximately fixed probability with each copy made.

"peak viral load did not differ by vaccination status or variant type"[1]

[1]: https://www.thelancet.com/journals/laninf/article/PIIS1473-3...