First, model each row as a bitmask where each bit represents whether a seat is occupied (1) or not (0). Ensure no two adjacent bits are set in the same row to prevent cheating.

Precompute all valid bitmasks for each row where no two adjacent seats are occupied (e.g., for a row of 3 seats, valid masks are `000`, `001`, `010`, `100`, `101`).

Use dynamic programming where `dp[i][mask]` represents the maximum students seated up to row `i` with `mask` representing occupied seats in row `i`. Transition by checking compatibility between `mask` of the current row and the previous row's `mask` (no overlapping or adjacent conflicts).