The exact nature of autistic cognition and intelligence has been a subject of extensive controversy. Individuals with autism spectrum condition (ASC) sometimes perform better than neurotypical individuals on certain tasks, including visual search and the Block Design and Object Assembly subtests of the Wechsler Intelligence Scale (WAIS) 
. However, these strengths of autistic ability are often thought to be limited to domain-specific low-level perceptual functions and the ASC′s enhanced sensation of the local information to arise at the sacrifice of high-level integrative and hierarchical processes involving diverse types of computation and information 
. Yet, such characterization of ASC′s intelligence is not directly compatible with a recent finding showing an unexpectedly high performance of autistic individuals on the Raven's Progressive Matrices (RPM) test, one of the most widely-used neuropsychological tests for general fluid intelligence. In the RPM, adults and children with ASC have been reported to score 30 percentile points higher on average than scores expected based on their performance on the Wechsler scales of intelligence 
. Moreover, children with Asperger's syndrome have been shown to outperform typically developing children matched for full-scale IQ on the WISC in the Raven's Standard Progressive Matrices (RSPM) test 
. Because the Raven's Matrices tests critically involve various types of high-level processes including abstraction, rule inference, and hierarchical goal management 
, these findings suggest that perceptual and cognitive strengths of ASC are not limited to simple low-level tasks but are utilized for solving novel problems.
The RSPM test consists of figures of geometric matrices that can be categorized into two major classes of “figural” and “analytic” items 
. “Figural” items are often characterized as “Gestalt reasoning,” which requires mostly visuospatial analysis with minimal analytic/analogical reasoning. In contrast, “analytic” problems require abstract “analogical reasoning” in addition to figural processing 
. Given that analytic items occupy the majority of the RSPM set (41 analytic items out of 60) and that figural items are relatively simple for high-functioning individuals, the analytic items constitute a core component of the task that critically contributes to the RPM advantage of high-functioning ASC.
Neural substrates for general intelligence have been examined in healthy adult individuals by using functional imaging techniques 
. Although exact locations of brain activation may vary depending on task parameters, a recent meta-analysis of the literature emphasized that a set of the frontoparietal regions, including the inferior parietal cortex (IPC), the ventrolateral prefrontal cortex (VLPFC) and the dorsolateral prefrontal cortex, is consistently involved. This observation led to the proposal of the Parietal-Frontal Integration Theory (P-FIT) of intelligence which states that intelligence is subserved by the large-scale network of the prefrontal and parietal regions 
. Although the functional role of each individual region of the network remains unclear, a recent functional magnetic resonance imaging (fMRI) study reported a dichotomy of the fluid intelligence network such that activity in the right lateralized network was modulated by rule complexity, whereas activity in the left anterior lateral occipital cortex (LOC) and VLPFC was sensitive to analogical reasoning demand 
In the present study, brain activity of high-functioning ASC participants during the RSPM task was measured using fMRI to examine functional properties of their brain network of geometric analogical reasoning. A previous fMRI study reported that adult individuals with ASC showed enhanced RSPM task-related activity in the extrastriate visual cortex and reduced activity in the lateral frontal regions including the precentral gyrus 
. Whereas that study combined all the figural and analytic items for analysis, the first purpose of the present experiment was to identify neural bases of ASC for high-level cognitive processes critical for analogical reasoning, including abstraction, rule inference, and hierarchical goal management. We aimed to isolate activation for these processes by directly comparing between the analytical and figural items based on the view that cognitive components for the analytical items subsume those for the figural items 
. From previous evidence of enhanced visual perception as well as increased activation in the occipital and lateral occipitotemporal cortex (LOTC) during intelligence tasks in ASC 
, we hypothesized that ASC would show enhanced activity in the localized posterior regions for geometric analogical reasoning. In addition to localized brain activity, we also examined activation coupling between regions in the large-scale network of fluid intelligence 
. Because analysis of functional connectivity does not technically require comparison between task conditions, we aimed to examine possible alterations of functional connectivity during solving the figural and analytical items. From the previous findings that the ASC brain can be characterized by scant long-range connections in the presence of excessive local connections 
, we hypothesized that ASC would show reduced long-range inter-regional activity coupling either between the anterior and posterior regions or inter-hemispheric regions.