I think you might have an XY Problem here.

Correct me if I'm wrong, but I think that what you really want to do is to munge this input:

1,1,0,0,0,0,0,1,1,1,1,0,0,0,1,1,1,1,0,0
0,0,1,1,0,0,1,1,0,0,1,1,0,1,0,1,0,1,1,0
[download]

using your %translation hash, to produce this output:

Odp-
kIr+
[download]

This implies that each 20-character-long line in your input file may be divided into four segments, of respectively 6 + 6 + 6 + 2 bytes (counting from left to right).

In which case, why not simplify matter by stringifying the 'segments' in question? Your %reverse_translation could then look like this:

(
  110011 => 'I',
  111101 => 'u',
  11 => '*',
  01 => '/',
  ... etc ...,
)
[download]

Here's a possible implementation, which works at least with the (rather sparse) data you provide:

my %translation = (
  # as in OP
);

my %reverse_translation;
for ( keys %translation ) {
  my $bin = join '', @{ $translation{$_}} ;
  $reverse_translation{$bin} = $_;
}

while ( <DATA> ) {
  my @segments = unpack 'A6A6A6A2', join '', split /,/;
  print $reverse_translation{$_} for @segments;
  print "\n";
}

__DATA__
1,1,0,0,0,0,0,1,1,1,1,0,0,0,1,1,1,1,0,0
0,0,1,1,0,0,1,1,0,0,1,1,0,1,0,1,0,1,1,0
[download]