arXiv：cs.LG· Haowei Xu, Wanyi Fu, Hongbin Han, Zhaoheng Xie·· 10 小时前
NeuronDiscover：用世界动作模型在神经元微环境中进行机制发现
NeuronDiscover: Agent-in-Twin for Mechanistic Discovery in Neuronal Microenvironments with World Action Models
AI 导读
arXiv 论文 NeuronDiscover 已被作者 Haowei Xu 撤回，原因是投稿未获得所有共同作者授权。该框架提出共享的、以机制为基础的世界动作模型（WAM），将预测、干预提议与观测设计耦合，并据独立裁决结果修订 MIOY 图；在 16 次实验预算、32 个源单元下每世界解决 4.0 条关系，强基线为 3.4，无图修订为 3.2。
正文
This paper has been withdrawn by Haowei Xu
No PDF available, click to view other formats
Abstract:Mechanistic discovery in neuronal microenvironments requires interventions and measurements that separate competing explanations of solute transport and neuronal response. Predictive accuracy cannot settle the question: a real mechanistic change and an error in the computational twin leave the same signature in sparse observations. We formalize this twin confounding and reason over a joint mechanism--discrepancy belief, designing experiments that separate the two. NeuronDiscover is an Agent-in-Twin framework whose shared, mechanism-grounded World Action Model (WAM) couples prediction, intervention proposals, and observation design; independently adjudicated outcomes revise a scoped Mechanism--Intervention--Observation--Outcome (MIOY) graph, whose supported relations compile into executable programs carrying discrepancy-adjusted acceptance bounds. We evaluate on simulated brain-fluid tracer-transport worlds adjudicated by an independently frozen finer-mesh reference solver, and on donor-disjoint public current-clamp recordings of cortical neurons. Counting only relations that reach a certified terminal status, and scoring abstentions as unresolved for every method, at a matched budget of 16 experiments over 32 source units NeuronDiscover resolves 4.0 relations per assigned world against 3.4 for the strongest baseline and 3.2 without graph revision, at 5% false support and 82% scope accuracy. Joint mechanism--discrepancy acquisition resolves 3.8 relations versus 2.9 for plug-in expected information gain; discrepancy-adjusted verification lowers accepted-program failure from 15% to 9% at 60% acceptance coverage; and transfer to the recordings yields 1.94 versus 1.53 relations per assigned world. Correctness is adjudicated within declared model worlds and archival recordings.
| Comments: | The submission was made without obtaining authorization from all co-authors |
| Subjects: | Machine Learning (cs.LG); Neurons and Cognition (q-bio.NC) |
| Cite as: | arXiv:2609.35338 [cs.LG] |
| (or arXiv:2609.35338v2 [cs.LG] for this version) | |
| https://doi.org/10.48550/arXiv.2609.35338 arXiv-issued DOI via DataCite |
Submission history
From: Haowei Xu [view email]
[v1]
Mon, 28 Sep 2026 14:58:12 UTC (5,142 KB)
[v2]
Wed, 14 Oct 2026 14:15:37 UTC (1 KB) (withdrawn)
来源：arXiv：cs.LG · arxiv.org