arXiv：cs.LG· Joonghui Cho, Minchan Kang, Daeshik Kim·· 4 小时前AI 评分44
果蝇全连接组网络可学习人类设计的认知任务
A Drosophila Whole-Connectome Network Can Learn Human-Designed Cognitive Tasks
AI 导读
研究用公开的 MaleCNS v1.0 果蝇连接组作为人工网络的固定循环拓扑，每条解剖边仅学习一个标量，在留出加法任务上达到 92.77% 平均准确率，高于保持度分布的随机重连的 67.93%。
正文
Abstract:Can a biological wiring diagram serve as a useful computational substrate beyond the behaviors for which it evolved? We use the publicly released MaleCNS v1.0 connectome, reconstructed from a single adult male Drosophila specimen, as the fixed recurrent topology of an artificial network. We train separate models for bounded addition and for a controlled grounded relational language task built from a fixed 100-word lexicon. In both models, one scalar is learned per anatomical edge. The anatomical graph reaches 92.77% mean accuracy on held-out addition, compared with 67.93% for directed degree-preserving rewires. On the strict paired language endpoint, which matches original and order-reversed scenes to their corresponding descriptions, it reaches 61.59% across four fixed interfaces, compared with 44.17% for matched rewires. At the canonical interface, it ranks first in a fixed 21-graph comparison. On the matched 48-group intervention subset, shuffling task-defined sensory features reduces its score from 60.94% to 19.27%. Together, these results show that higher-order MaleCNS wiring provides a reusable inductive bias for bounded addition and grounded relational language.
| Comments: | 14 pages, 4 figures. Code: this https URL |
| Subjects: | Machine Learning (cs.LG) |
| Cite as: | arXiv:2610.10014 [cs.LG] |
| (or arXiv:2610.10014v1 [cs.LG] for this version) | |
| https://doi.org/10.48550/arXiv.2610.10014 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Joonghui Cho [view email]
[v1]
Wed, 7 Oct 2026 13:04:17 UTC (2,408 KB)
来源：arXiv：cs.LG · arxiv.org