You can find the potential file under the link below.
Almost all values match the original publication or the erratum with corrected parameters. The only differences are for parameters a1 and a3 for both Mo and S.
The epsilon value for S-S interactions is also different. I believe this is an error in the erratum though, as they claimed epsilon for Mo-S was calculated as a geometrical mean, but it isn’t true, unless you use the value for S-S epsilon from the file in the link.
The same potential file is also here: