if for some combination we get multiple zero rows the same combination of entries of b vector should give 0 for the corresponding rows

In case of Ax=b where matrix A has dependent columns and the vector b is in the column space of A can the infinite solutions of this be imagined as the null space being moved to the tip of one particular solution vector x of Ax=b and all possible solutions are described as addition between x particular and any vector of the null space. In a sense x particular gives access to dimensions that x vector in null space does not

there can be infinite xparticular chosen depending on what values the free variables are assigned to and the null space added to any of them describes the same solution set as xparticular obtained from setting free variables to zero

There can be infinite xparticular solutions when free variables are present. Here's why: