REZE: Representation Regularization for Domain-adaptive Text Embedding Pre-finetuning

arXiv:2604.17257v1 Announce Type: new Abstract: Recent text embedding models are often adapted to specialized domains via contrastive pre-finetuning (PFT) on a naive collection of scattered, heterogeneous tasks. However, this approach often introduces task-induced bias alongside domain knowledge, leading to uncontrolled representation shifts that distort the pretrained embedding geometry and cause substantial performance degradation. To address this issue, we propose REZE}, a representation regularization framework that explicitly controls representation shift during embedding pre-finetuning. REZE operates on the relations of anchor-positive pairs and decomposes them in an eigenspace. It then measures task-wise dispersion along each eigencomponent to identify task-variant directions and applies adaptive soft-shrinkage to suppress task-induced noise while preserving task-invariant semantic structure, without inference-time overhead. Experiments across multiple embedding backbones and specialized benchmarks show that REZE outperforms standard pre-finetuning and isotropy-oriented post-hoc regularization in most settings, remaining stable where existing PFT variants collapse. Embedding space analyses further confirm that REZE induces controlled shifts aligned with the original embedding manifold, underscoring representation shift control as a key principle for robust embedding pre-finetuning under heterogeneous supervision.

Related

• A Representation-Level Assessment of Bias Mitigation in Foundation Models
• CLEAR: Cross-Lingual Enhancement in Alignment via Reverse-training
• End-to-end Contrastive Language-Speech Pretraining Model For Long-form Spoken Question Answering
• Large Language Model Post-Training: A Unified View of Off-Policy and On-Policy Learning

Source: arXiv cs.CL | 2026-04-21