ELF: A Family of Encoder-Free ECG-Language Models

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Han, William, Chen, Tony, Duan, Chaojing, Song, Xiaoyu, Yao, Yihang, Yang, Yuzhe, Rosenberg, Michael A., Liu, Emerson, Zhao, Ding
Natura: Preprint
Pubblicazione: 2026
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866913175416340480
author Han, William
Chen, Tony
Duan, Chaojing
Song, Xiaoyu
Yao, Yihang
Yang, Yuzhe
Rosenberg, Michael A.
Liu, Emerson
Zhao, Ding
author_facet Han, William
Chen, Tony
Duan, Chaojing
Song, Xiaoyu
Yao, Yihang
Yang, Yuzhe
Rosenberg, Michael A.
Liu, Emerson
Zhao, Ding
contents ECG-Language Models (ELMs) extend recent advances in Multimodal Large Language Models (MLLMs) to automated ECG interpretation. However, most existing ELMs inherit Vision-Language Model (VLM) design choices and rely on pretrained ECG encoders, introducing substantial architectural and training complexity. Inspired by encoder-free VLMs, we introduce ELF, a family of three encoder-free ELM architectures that remain competitive with, and often outperform, prior state-of-the-art ELMs across two datasets despite substantially simpler architectures and training pipelines. All code and data is available at github.com/ELM-Research/ECG-Language-Models.
format Preprint
id arxiv_https___arxiv_org_abs_2601_18798
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle ELF: A Family of Encoder-Free ECG-Language Models
Han, William
Chen, Tony
Duan, Chaojing
Song, Xiaoyu
Yao, Yihang
Yang, Yuzhe
Rosenberg, Michael A.
Liu, Emerson
Zhao, Ding
Multimedia
Artificial Intelligence
ECG-Language Models (ELMs) extend recent advances in Multimodal Large Language Models (MLLMs) to automated ECG interpretation. However, most existing ELMs inherit Vision-Language Model (VLM) design choices and rely on pretrained ECG encoders, introducing substantial architectural and training complexity. Inspired by encoder-free VLMs, we introduce ELF, a family of three encoder-free ELM architectures that remain competitive with, and often outperform, prior state-of-the-art ELMs across two datasets despite substantially simpler architectures and training pipelines. All code and data is available at github.com/ELM-Research/ECG-Language-Models.
title ELF: A Family of Encoder-Free ECG-Language Models
topic Multimedia
Artificial Intelligence
url https://arxiv.org/abs/2601.18798