I have a table with column columna with data in the same row:

columna
-----------------
a,b,a,b,a,b,a,b

I tried this, but it’s not good

duplicated_rows AS 
(
    SELECT 
        *,
        ROW_NUMBER() OVER (PARTITION BY columna) AS row_num
    FROM 
        tablea
)
SELECT * 
FROM duplicated_rows
WHERE row_num > 1;

Not sure how to remove duplicate so I will get result

columna
-------
a,b