The paper introduces Representational Similarity Analysis (RSA), a framework designed to quantitatively connect the three major branches of systems neuroscience: brain-activity measurement, behavioral measurement, and computational modeling. RSA addresses the challenge of relating these branches by computing representational dissimilarity matrices (RDMs), which characterize the information carried by a given representation in a brain or model. The authors propose using RDMs to compare multi-channel measures of neural activity, computational models, and behavioral data, thereby bridging the gap between different modalities and experimental designs. The framework is demonstrated through an example where fMRI data from early visual cortex and the fusiform face area are compared to computational models of visual object perception. RSA is shown to provide a rich and informationally dense interface for relating different representations, allowing for a more integrated and quantitative analysis of systems neuroscience data. The paper also discusses the broader potential of RSA, including its application to experimental design and its ability to address a wide range of neuroscientific questions.The paper introduces Representational Similarity Analysis (RSA), a framework designed to quantitatively connect the three major branches of systems neuroscience: brain-activity measurement, behavioral measurement, and computational modeling. RSA addresses the challenge of relating these branches by computing representational dissimilarity matrices (RDMs), which characterize the information carried by a given representation in a brain or model. The authors propose using RDMs to compare multi-channel measures of neural activity, computational models, and behavioral data, thereby bridging the gap between different modalities and experimental designs. The framework is demonstrated through an example where fMRI data from early visual cortex and the fusiform face area are compared to computational models of visual object perception. RSA is shown to provide a rich and informationally dense interface for relating different representations, allowing for a more integrated and quantitative analysis of systems neuroscience data. The paper also discusses the broader potential of RSA, including its application to experimental design and its ability to address a wide range of neuroscientific questions.