Can we say anything about "base b valuated primes"?

51 Views Asked by At

"Base $b$ valuated prime" is an odd prime, whose representations in all natural number bases $2,3,\dots,b$ are also prime when evaluated as base $b$ numbers.

That is, write the (prime) number in bases $2,3,\dots,b$. Then read those expressions as base $b$ expressions, to obtain new numbers, which all must be prime to satisfy the definition.

Here is a table of first ten examples, for base $b=2,3,\dots,10$ valuated primes:

 2:        3,      5,       7,       11,       13,       17,       19,       23,       29,       31, ...
 3:        7,     11,      13,       41,       47,       67,       73,       79,      109,      127, ...
 4:       43,     61,     139,      151,      229,      277,      379,      439,      787,      823, ...
 5:       41,    151,     181,      271,     1759,     2011,     3061,     5869,     9601,    14591, ...
 6:    45481, 102181,  202621,   229261,   275881,   498301,   570421,   894541,   906541,  1103101, ...
 7:      701, 805583, 3550733, 29725363, 39296861, 44098673, 50976553, 55402397, 60343249, 62351203, ...
 8: 65484343,      ?,       ?,        ?,        ?,        ?,        ?,        ?,        ?,        ?, ...
 9:        ?,      ?,       ?,        ?,        ?,        ?,        ?,        ?,        ?,        ?, ...
10:        ?,      ?,       ?,        ?,        ?,        ?,        ?,        ?,        ?,        ?, ...
  • For $b=2$, we have the sequences of all primes themselves.

  • For $b=3$, they are in the OEIS sequence $A235266$, which links to similar sequences.

  • For $b\ge 4$, I haven't found similar sequences or relevant information.

The ? stands in places where I didn't find all ten examples.

For $b=8,9,10$, I checked first $10^8$ primes. - But, the examples are probably much larger?

Can these primes exits in all $b$? (Can we find second $b=8$ or first $b=9,10$ examples?)

Can we estimate the number of such $b$ valuated primes below some upper bound?

Roughly speaking, we are taking some prime $p_n$, and evaluating numbers of $\approx$ sizes:

$$b\land\{\log_2p_n\},b\land\{\log_3p_n\},\dots,b\land\{\log_bp_n\}$$

Which need to be all simultaneously prime. (Where $\land$ is exponent, for readability).

Can we test for these numbers faster (Can we establish any useful characterizations)?


This question was motivated by a recent similar question.