小柯机器人

美国麻省理工学院Josh H. McDermott、Jenelle Feather课题组在研究中取得进展。他们的研究利用模型同构体揭示了生物神经网络和人工神经网络之间的一致性。该研究于2023年10月16日发表于国际学术期刊《自然-神经科学》杂志。

为了揭示这些不变性，研究人员生成了"模型元变量"，即在模型阶段激活与自然刺激相匹配的刺激。视觉和听觉有监督和无监督神经网络模型的元变量在模型后期生成时，人类往往完全无法识别，这表明模型和人类不变性之间存在差异。有针对性地改变元变量可以提高元变量在人类中的可识别性，但并不能消除人类与元变量之间的整体差异。

一个模型元变量组的人类可识别性可以很好地通过其他模型的可识别性来预测，这表明除了任务要求的不变性之外，模型还包含特异的不变性。元变量组识别性从传统的基于大脑的基准和对抗脆弱性中分离出来，它揭示了现有感官模型的独特失效模式，并为模型评估提供了一种补充基准。

研究人员表示，人们经常利用感官系统深度神经网络模型来探究表征转换，这种转换与大脑中的表征转换具有一致性。

附：英文原文

Title: Model metamers reveal divergent invariances between biological and artificial neural networks

Author: Feather, Jenelle, Leclerc, Guillaume, Mdry, Aleksander, McDermott, Josh H.

Issue&Volume: 2023-10-16

Abstract: Deep neural network models of sensory systems are often proposed to learn representational transformations with invariances like those in the brain. To reveal these invariances, we generated ‘model metamers’, stimuli whose activations within a model stage are matched to those of a natural stimulus. Metamers for state-of-the-art supervised and unsupervised neural network models of vision and audition were often completely unrecognizable to humans when generated from late model stages, suggesting differences between model and human invariances. Targeted model changes improved human recognizability of model metamers but did not eliminate the overall human–model discrepancy. The human recognizability of a model’s metamers was well predicted by their recognizability by other models, suggesting that models contain idiosyncratic invariances in addition to those required by the task. Metamer recognizability dissociated from both traditional brain-based benchmarks and adversarial vulnerability, revealing a distinct failure mode of existing sensory models and providing a complementary benchmark for model assessment.

DOI: 10.1038/s41593-023-01442-0

Source: https://www.nature.com/articles/s41593-023-01442-0

Nature Neuroscience：《自然—神经科学》，创刊于1998年。隶属于施普林格·自然出版集团，最新IF：28.771
官方网址：https://www.nature.com/neuro/
投稿链接：https://mts-nn.nature.com/cgi-bin/main.plex