Why Better Cross-Lingual Alignment Fails for Better Cross-Lingual Transfer: Case of Encoders

ArXi:2603.18863v1 Announce Type: new Better cross-lingual alignment is often assumed to yield better cross-lingual transfer. However, explicit alignment techniques -- despite increasing embedding similarity -- frequently fail to improve token-level downstream performance. In this work, we show that this mismatch arises because alignment and downstream task objectives are largely orthogonal, and because the downstream benefits from alignment vary substantially across languages and task types.