Saturday, September 12, 2009

Should experimentalists have to "explain" their data?

Sometimes when I talk to experimentalists about their latest results they say something like, "We have this really interesting data. But we can't explain it and so we are not going to publish it until we can explain it." No doubt this is sometimes what referees say.
But, I disagree. The most interesting experimental results have no clear explaination!
Furthermore, many of the "explainations" I read in experimental papers seem to be either naive, wrong, or highly speculative.

2 comments:

  1. Further, and worse, these "explanations" are often then cited as if they are "facts" "proven" by the data!

    ReplyDelete
  2. We theorists do need something to do, after all!

    ReplyDelete

Reviewing emergent computational abilities in Large Language Models

Two years ago, I wrote a post about a paper by Wei et al,  Emergent Abilities of Large Language Models Then last year, I posted about a pap...