Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: He, Zirui, Zhang, Huopu, Liu, Yanguang, Wu, Sirui, Du, Mengnan
Format: Preprint
Veröffentlicht: 2026
Schlagworte:
Online-Zugang:https://arxiv.org/abs/2602.20859
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866918353647435776
author He, Zirui
Zhang, Huopu
Liu, Yanguang
Wu, Sirui
Du, Mengnan
author_facet He, Zirui
Zhang, Huopu
Liu, Yanguang
Wu, Sirui
Du, Mengnan
contents Financial prediction from long documents involves significant challenges, as actionable signals are often sparse and obscured by noise, and the optimal LLM for generating embeddings varies across tasks and time periods. In this paper, we propose FinAnchor(Financial Anchored Representations), a lightweight framework that integrates embeddings from multiple LLMs without fine-tuning the underlying models. FinAnchor addresses the incompatibility of feature spaces by selecting an anchor embedding space and learning linear mappings to align representations from other models into this anchor. These aligned features are then aggregated to form a unified representation for downstream prediction. Across multiple financial NLP tasks, FinAnchor consistently outperforms strong single-model baselines and standard ensemble methods, demonstrating the effectiveness of anchoring heterogeneous representations for robust financial prediction.
format Preprint
id arxiv_https___arxiv_org_abs_2602_20859
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle FinAnchor: Aligned Multi-Model Representations for Financial Prediction
He, Zirui
Zhang, Huopu
Liu, Yanguang
Wu, Sirui
Du, Mengnan
Computation and Language
Financial prediction from long documents involves significant challenges, as actionable signals are often sparse and obscured by noise, and the optimal LLM for generating embeddings varies across tasks and time periods. In this paper, we propose FinAnchor(Financial Anchored Representations), a lightweight framework that integrates embeddings from multiple LLMs without fine-tuning the underlying models. FinAnchor addresses the incompatibility of feature spaces by selecting an anchor embedding space and learning linear mappings to align representations from other models into this anchor. These aligned features are then aggregated to form a unified representation for downstream prediction. Across multiple financial NLP tasks, FinAnchor consistently outperforms strong single-model baselines and standard ensemble methods, demonstrating the effectiveness of anchoring heterogeneous representations for robust financial prediction.
title FinAnchor: Aligned Multi-Model Representations for Financial Prediction
topic Computation and Language
url https://arxiv.org/abs/2602.20859