Appendix for Self-Weighted Contrastive Learning among Multiple Views for Mitigating Representation Degeneration