Orthogonal similarity transformation refers to a linear transformation that preserves angles and distances, typically represented by orthogonal matrices. The discussion emphasizes that the property of asymmetry remains unchanged when subjected to such transformations. Participants explore the mathematical proof of this invariance, highlighting the role of eigenvalues and eigenvectors in maintaining asymmetry. The conversation also touches on the implications of this property in various mathematical and physical contexts. Understanding orthogonal similarity transformations is crucial for grasping the concept of asymmetry invariance.