On the Use of Anchoring for Training Vision Models | Read Paper on Bytez