Document-level relation extraction (DocRE) aims to identify relations for an entity pair within a document. Existing methods can be broadly classified into two categories: direct encoding of the entire document or enhancement using extracted evidence sentences. However, the former often introduces noise unrelated to relations, while the latter is heavily dependent on the quality of evidence extraction. Moreover, these DocRE models typically use an adaptive threshold to predict all potential relations for an entity pair. As a result, class imbalance in DocRE often leads the model to learn a high threshold for an entity pair, which in turn causes the model to frequently predict that the entity pair has no relation.

To address these issues, we propose a Multi-Discrimination framework (MD-RE) that does not rely on evidence sentences. MD-RE employs three discriminators with dynamically adjusted thresholds to independently predict relations, and aggregates their outputs via a weighted fusion strategy. Furthermore, we propose an Adaptive Threshold Shifted Loss (ATSL), which encourages a lower threshold to alleviate the high false negative rate resulting from class imbalance. Experiments on three datasets demonstrate that MD-RE achieves new state-of-the-art results. In addition, ATSL significantly improves the performance of various existing DocRE models. Moreover, combining other losses with MD-RE also yields competitive results. Code is available at https://github.com/xhm-code/MD-RE.

Document-level relation extraction (DocRE) aims to determine which relations hold between a given entity pair within a document. As a multi-label classification task, the most commonly adopted paradigm introduces a learnable threshold to distinguish positive and negative classes for an entity pair. Under this paradigm, existing losses decouple the optimization into independent positive and negative losses, which interact solely with a shared threshold. This leads to two inherent limitations: (i) threshold instability caused by conflicting gradient updates from the decoupled losses; and (ii) optimization bias exacerbated by the severe imbalance between limited positive samples and abundant negative samples inherent in DocRE, which makes the model more likely to predict that no relation exists.To address these issues, we propose the Adaptive-Threshold Global Loss (ATGL). Unlike prior work, ATGL integrates positive, negative, and threshold optimization into a unified logit space and explicitly enforces ranking constraints on their contributions to the objective. Furthermore, ATGL incorporates an imbalance-aware optimization mechanism, thereby effectively addressing the severe class imbalance in DocRE. Our ATGL serves as a general optimization objective that can be readily applied to different DocRE models. Experiments on four datasets show that ATGL outperforms other DocRE losses and achieves state-of-the-art results, while consistently improving the performance of existing DocRE models. Code is available at https://github.com/xhm-code/ATGL.

Cross-document Relation Extraction (CodRE) requires reasoning over scattered evidence to identify relations between target entities across multiple documents. Existing methods indiscriminately fuse target entities and the intermediate bridge entities that link them into a unified representation. This leads to intermediate evidence that often aligns with only one side of the entity pair, resulting in one-sided relation transfer contextual bias and incomplete reasoning chains. Moreover, these methods typically employ a global threshold to determine relation existence for all entity pairs, limiting the model’s reasoning performance.To address these issues, we propose DEBAR (Dual-stream Entity Bias Reduction), a framework designed to explicitly decouple and preserve bidirectional bridge evidence, combined with a novel dynamic loss optimization objective. Specifically, DEBAR employs a bridge-aware input construction strategy and a dual-stream graph reasoning network to separately encode head and tail contexts, preventing semantic interference while capturing global dependencies through iterative message passing. Furthermore, we introduce a curriculum-aware ranking optimization objective that progressively tightens classification constraints to stabilize training and enforce discriminative decision boundaries. Experiments on the CodRE benchmarks show that DEBAR achieves state-of-the-art performance while effectively mitigating cross-document contextual bias. Moreover, extensive experiments on our proposed loss across backbones confirm its generalization, suggesting it as a reliable replacement for existing CodRE losses. Code is available at https://github.com/newyuyou/DEBAR.