The APECS-SRN: towards a model of SRT sequence learning