无需训练的 DNA Socket 接口：将冻结的 Evo 2 特征接入大语言模型
A Training-Free Socket for DNA in Frozen Language Models
AI 导读
arXiv 论文提出一种免训练的“socket”方法，用岭回归拟合的 plug 将冻结的 Evo 2 DNA 特征映射到宿主 LLM 早期层的隐藏状态，全程无梯度更新。
正文 · 原文
Abstract:Connecting a DNA foundation model to a large language model (LLM) usually requires training a connector, and often the LLM too. We show that a closed-form map is enough for gene-level questions. Our method, the socket, fits one ridge regression (the plug) from frozen Evo 2 features to the host LLM's own representation of the gene's symbol, name, and description at an early layer. The plug's output replaces the hidden state at placeholder positions, and no gradient passes through the host.
We test it on GeneHop, a five-way multiple-choice benchmark split by homology. In a single run on the held-out test set, under a frozen analysis plan, real DNA beats shuffled DNA on all four hosts: by +22.2 points [19.8, 25.0] on Qwen3.5-4B, +26.1 [24.3, 27.9] on Qwen3.6-27B, +16.7 [14.2, 19.7] on Mistral-7B, and +18.4 [16.2, 21.0] on OLMo-2-7B. All four survive Holm correction, every null control stays within two points of the sham level, and the development results agree within 2.3 points.
A layer rule that uses no task items was selected on three hosts; it predicted the working layer of the fourth before any DNA-plug scoring on it. A text-only linear map lets one plug feed any other host, at a cost of at most 0.9 points on test. The socket conveys gene identity: nearest-neighbour retrieval recovers 72-84% of the effect, but the plug beats it on every host by 3.0-5.2 points and keeps most of its effect without training homologs.
| Subjects: | Genomics (q-bio.GN); Artificial Intelligence (cs.AI) |
| Cite as: | arXiv:2611.05401 [q-bio.GN] |
| (or arXiv:2611.05401v1 [q-bio.GN] for this version) | |
| https://doi.org/10.48550/arXiv.2611.05401 arXiv-issued DOI via DataCite |
Submission history
From: Arash Fatehi [view email]
[v1]
Fri, 16 Oct 2026 19:55:12 UTC (418 KB)
来源：arXiv cs.AI · arxiv.org