Comment by iLoveOncall

An experiment is not a proof.

If this is the level of one of the contributors to the OpenTSLM paper (which you very obviously are), no wonder due diligence wasn't done properly.