On the singular arithmetic module L_sing, choose the cyclic generator w from the projected standard basis. Then
[ w , T_3 w , T_11 w ]
is a Z_3-basis, so every Heegner operator acts by a concrete 3 x 3 matrix in this basis.
This is the regular representation of the singular Hecke order on its arithmetic module.
3T_3:[0, 0, 0]
[1, 1, 0]
[0, 0, 0]
T_7:[2, 0, 0]
[0, 2, 0]
[1, 0, 2]
T_11:[0, 0, 0]
[0, 0, 0]
[1, 0, 0]
T_19:[0, 0, 0]
[2, 2, 0]
[0, 0, 0]
T_43:[1, 0, 0]
[1, 2, 0]
[2, 0, 1]
T_67:[1, 0, 0]
[1, 2, 0]
[2, 0, 1]
T_163:[2, 0, 0]
[2, 1, 0]
[0, 0, 2]
27T_3 modulo 27:[0, 18, 0]
[1, 4, 12]
[0, 3, 18]
T_7 modulo 27:[8, 0, 9]
[24, 8, 18]
[10, 9, 8]
T_11 modulo 27:[0, 0, 9]
[0, 12, 0]
[1, 18, 0]
T_19 modulo 27:[12, 9, 0]
[20, 20, 24]
[3, 6, 21]
T_43 modulo 27:[10, 18, 18]
[7, 17, 3]
[5, 3, 1]
T_67 modulo 27:[19, 18, 18]
[22, 14, 21]
[26, 21, 10]
T_163 modulo 27:[8, 9, 0]
[14, 1, 6]
[24, 15, 17]
Modulo 3, these matrices recover the syndrome algebra in a cyclic module basis. Modulo 9 and 27, the matrices split the 43/67 collision and thicken the node without destroying cyclicity. The residue module already shows a nontrivial flag:
T_3 w is the split semisimple branch;T_11 w is the nilpotent socle;T_3 w and T_11 w is an invariant two-step memory layer.So the singular Hecke factor is not just a ring defect: it acts as an explicit finite-state transition system on its own arithmetic module.