LATTE: Learning Aligned Transactions and Textual Embeddings for Bank Clients
Fuente:
arXiv
Enregistré dans:
| Auteurs principaux: | , , , , , , , , |
|---|---|
| Format: | Preprint |
| Publié: |
2025
|
| Sujets: | |
| Accès en ligne: | |
| Tags: |
Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
|
| _version_ | 1866908716458049536 |
|---|---|
| author | Fadeev, Egor Mollaev, Dzhambulat Shestov, Aleksei Zoloev, Omar Sakhno, Artem Korolev, Dmitry Kireev, Ivan Savchenko, Andrey Makarenko, Maksim |
| author_facet | Fadeev, Egor Mollaev, Dzhambulat Shestov, Aleksei Zoloev, Omar Sakhno, Artem Korolev, Dmitry Kireev, Ivan Savchenko, Andrey Makarenko, Maksim |
| contents | Learning clients embeddings from sequences of their historic communications is central to financial applications. While large language models (LLMs) offer general world knowledge, their direct use on long event sequences is computationally expensive and impractical in real-world pipelines. In this paper, we propose LATTE, a contrastive learning framework that aligns raw event embeddings with semantic embeddings from frozen LLMs. Behavioral features are summarized into short prompts, embedded by the LLM, and used as supervision via contrastive loss. The proposed approach significantly reduces inference cost and input size compared to conventional processing of complete sequence by LLM. We experimentally show that our method outperforms state-of-the-art techniques for learning event sequence representations on real-world financial datasets while remaining deployable in latency-sensitive environments. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2508_10021 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | LATTE: Learning Aligned Transactions and Textual Embeddings for Bank Clients Fadeev, Egor Mollaev, Dzhambulat Shestov, Aleksei Zoloev, Omar Sakhno, Artem Korolev, Dmitry Kireev, Ivan Savchenko, Andrey Makarenko, Maksim Computation and Language Artificial Intelligence Learning clients embeddings from sequences of their historic communications is central to financial applications. While large language models (LLMs) offer general world knowledge, their direct use on long event sequences is computationally expensive and impractical in real-world pipelines. In this paper, we propose LATTE, a contrastive learning framework that aligns raw event embeddings with semantic embeddings from frozen LLMs. Behavioral features are summarized into short prompts, embedded by the LLM, and used as supervision via contrastive loss. The proposed approach significantly reduces inference cost and input size compared to conventional processing of complete sequence by LLM. We experimentally show that our method outperforms state-of-the-art techniques for learning event sequence representations on real-world financial datasets while remaining deployable in latency-sensitive environments. |
| title | LATTE: Learning Aligned Transactions and Textual Embeddings for Bank Clients |
| topic | Computation and Language Artificial Intelligence |
| url | https://arxiv.org/abs/2508.10021 |