发表机构
University of Toronto; KITE Research Institute; University Health Network（多伦多大学; KITE研究所; 大学健康网络）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
该研究针对生物与人工神经网络表征对齐的不对称问题，开发整合谱正则化与双向预测性分析的计算框架，调控表征几何使双向预测性获55%相对提升，证明可通过调控表征几何调节双向对齐。
AI 中文摘要
近期研究表明，生物神经网络与人工神经网络间的表征对齐具有不对称性：模型表征预测神经反应的效果远好于神经反应预测模型表征的效果。这种不对称性引发了一个问题，即表征几何是否对双向表征对齐有贡献。我们假设，在训练过程中调控表征几何可以系统性地影响双向对齐。为验证该假设，我们开发了一个将谱正则化与双向预测性分析相结合的计算框架。作为初步验证，我们使用自监督对比视觉模型评估了该框架。调控学习表征的谱几何大幅提升了反向预测性，同时正向预测性仅有适度下降，使双向预测性获得了55%的相对提升。这些改进伴随着有效维度的降低和共享表征子空间的重组，在该子空间内，正向与反向预测性在中等谱指数下变得近似对称。总体而言，这些发现表明，可系统性调控表征几何以调节生物神经网络与人工神经网络间的双向表征对齐。
英文摘要
Recent work has shown that representational alignment between biological and artificial neural networks is asymmetric: model representations predict neural responses much better than neural responses predict model representations. This asymmetry raises the question of whether representational geometry contributes to bidirectional representational alignment. We hypothesized that steering representational geometry during training can systematically influence bidirectional alignment. To test this hypothesis, we developed a computational framework that integrates spectral regularization with bidirectional predictivity analyses. As an initial demonstration, we evaluated our framework using self-supervised contrastive vision models. Steering the spectral geometry of the learned representations substantially increased reverse predictivity with modest reductions in forward predictivity, yielding a 55% relative improvement in bidirectional predictivity. These improvements were accompanied by reduced effective dimensionality and a reorganization of the shared representational subspace, within which forward and reverse predictivity became approximately symmetric at intermediate spectral exponents. Overall, these findings demonstrate that representational geometry can be systematically steered to modulate bidirectional representational alignment between biological and artificial neural networks.
