@inproceedings{xu-etal-2026-beyond-conditional,
title = "Beyond Conditional Computation: Retrieval-Augmented Genomic Foundation Models with Gengram",
author = "Xu, Huinan and
Feng, Xuyang and
Chen, Junhong and
Liu, Junchen and
Deng, Kaiwen and
Ding, Kai and
Long, Shengning and
Shuai, Jiaxue and
li, Zhaorong and
Liu, Shiping and
Xue, Guirong and
Xiao, Zhan",
editor = "Lee, Hung-Yi and
Dev, Sunipa and
Li, Jessy and
Elliott, Desmond",
booktitle = "Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing",
month = oct,
year = "2026",
address = "Budapest, Hungary",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2026.emnlp-main.1464/",
pages = "33480--33511",
ISBN = "979-8-89176-459-0",
abstract = "Current genomic foundation models (GFMs) rely largely on extensive neural computation to implicitly infer conserved biological motifs from single-nucleotide inputs. We introduce Gengram, a conditional memory module that provides an explicit and efficient lookup primitive for multi-base motifs through a genome-specific hashing scheme. Integrated into state-of-the-art (SOTA) GFM backbones, Gengram achieves substantial improvements of up to 22.6{\%} across functional genomics tasks and generalizes across diverse architectures. Further analysis of Gengram{'}s latent space reveals biologically meaningful representations aligned with established genomic principles. By introducing structured motif memory as a modeling primitive, Gengram improves empirical performance and mechanistic interpretability, offering a scalable and biology-aligned direction for next-generation GFMs. The code is available at https://github.com/zhejianglab/Gengram, and the model weights are available at https://huggingface.co/ZhejiangLab/Gengram."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="xu-etal-2026-beyond-conditional">
<titleInfo>
<title>Beyond Conditional Computation: Retrieval-Augmented Genomic Foundation Models with Gengram</title>
</titleInfo>
<name type="personal">
<namePart type="given">Huinan</namePart>
<namePart type="family">Xu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Xuyang</namePart>
<namePart type="family">Feng</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Junhong</namePart>
<namePart type="family">Chen</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Junchen</namePart>
<namePart type="family">Liu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Kaiwen</namePart>
<namePart type="family">Deng</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Kai</namePart>
<namePart type="family">Ding</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Shengning</namePart>
<namePart type="family">Long</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jiaxue</namePart>
<namePart type="family">Shuai</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Zhaorong</namePart>
<namePart type="family">li</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Shiping</namePart>
<namePart type="family">Liu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Guirong</namePart>
<namePart type="family">Xue</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Zhan</namePart>
<namePart type="family">Xiao</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2026-10</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing</title>
</titleInfo>
<name type="personal">
<namePart type="given">Hung-Yi</namePart>
<namePart type="family">Lee</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Sunipa</namePart>
<namePart type="family">Dev</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jessy</namePart>
<namePart type="family">Li</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Desmond</namePart>
<namePart type="family">Elliott</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Budapest, Hungary</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
<identifier type="isbn">979-8-89176-459-0</identifier>
</relatedItem>
<abstract>Current genomic foundation models (GFMs) rely largely on extensive neural computation to implicitly infer conserved biological motifs from single-nucleotide inputs. We introduce Gengram, a conditional memory module that provides an explicit and efficient lookup primitive for multi-base motifs through a genome-specific hashing scheme. Integrated into state-of-the-art (SOTA) GFM backbones, Gengram achieves substantial improvements of up to 22.6% across functional genomics tasks and generalizes across diverse architectures. Further analysis of Gengram’s latent space reveals biologically meaningful representations aligned with established genomic principles. By introducing structured motif memory as a modeling primitive, Gengram improves empirical performance and mechanistic interpretability, offering a scalable and biology-aligned direction for next-generation GFMs. The code is available at https://github.com/zhejianglab/Gengram, and the model weights are available at https://huggingface.co/ZhejiangLab/Gengram.</abstract>
<identifier type="citekey">xu-etal-2026-beyond-conditional</identifier>
<location>
<url>https://aclanthology.org/2026.emnlp-main.1464/</url>
</location>
<part>
<date>2026-10</date>
<extent unit="page">
<start>33480</start>
<end>33511</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Beyond Conditional Computation: Retrieval-Augmented Genomic Foundation Models with Gengram
%A Xu, Huinan
%A Feng, Xuyang
%A Chen, Junhong
%A Liu, Junchen
%A Deng, Kaiwen
%A Ding, Kai
%A Long, Shengning
%A Shuai, Jiaxue
%A li, Zhaorong
%A Liu, Shiping
%A Xue, Guirong
%A Xiao, Zhan
%Y Lee, Hung-Yi
%Y Dev, Sunipa
%Y Li, Jessy
%Y Elliott, Desmond
%S Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing
%D 2026
%8 October
%I Association for Computational Linguistics
%C Budapest, Hungary
%@ 979-8-89176-459-0
%F xu-etal-2026-beyond-conditional
%X Current genomic foundation models (GFMs) rely largely on extensive neural computation to implicitly infer conserved biological motifs from single-nucleotide inputs. We introduce Gengram, a conditional memory module that provides an explicit and efficient lookup primitive for multi-base motifs through a genome-specific hashing scheme. Integrated into state-of-the-art (SOTA) GFM backbones, Gengram achieves substantial improvements of up to 22.6% across functional genomics tasks and generalizes across diverse architectures. Further analysis of Gengram’s latent space reveals biologically meaningful representations aligned with established genomic principles. By introducing structured motif memory as a modeling primitive, Gengram improves empirical performance and mechanistic interpretability, offering a scalable and biology-aligned direction for next-generation GFMs. The code is available at https://github.com/zhejianglab/Gengram, and the model weights are available at https://huggingface.co/ZhejiangLab/Gengram.
%U https://aclanthology.org/2026.emnlp-main.1464/
%P 33480-33511
Markdown (Informal)
[Beyond Conditional Computation: Retrieval-Augmented Genomic Foundation Models with Gengram](https://aclanthology.org/2026.emnlp-main.1464/) (Xu et al., EMNLP 2026)
ACL
- Huinan Xu, Xuyang Feng, Junhong Chen, Junchen Liu, Kaiwen Deng, Kai Ding, Shengning Long, Jiaxue Shuai, Zhaorong li, Shiping Liu, Guirong Xue, and Zhan Xiao. 2026. Beyond Conditional Computation: Retrieval-Augmented Genomic Foundation Models with Gengram. In Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing, pages 33480–33511, Budapest, Hungary. Association for Computational Linguistics.