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

ACL ARR 2026 January Submission9122 Authors

Keywords: Embedding, Multi Task Learning, Pre-finetuning, Representation, Debiasing, Regularization

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.

Paper Type: Long

Research Area: Semantics: Lexical, Sentence-level Semantics, Textual Inference and Other areas

Research Area Keywords: "Language Modeling", "Efficient/Low-Resource Methods for NLP", "Semantics: Lexical and Sentence-Level"

Contribution Types: NLP engineering experiment, Approaches to low-resource settings

Languages Studied: English

Submission Number: 9122