Randall C. O’Reilly

A unified framework for inhibitory control

Inhibiting unwanted thoughts, actions and emotions figures centrally in daily life, and the prefrontal cortex (PFC) is widely viewed as a source of this inhibitory control. We argue that the function of the PFC is best understood in terms of representing and actively maintaining abstract information, such as goals, which produces two types of inhibitory effects on other brain regions. Inhibition of some subcortical regions takes a directed global form, with prefrontal regions providing contextual information relevant to when to inhibit all processing in a region. Inhibition within neocortical (and some subcortical) regions takes an indirect competitive form, with prefrontal regions providing excitation of goal-relevant options. These distinctions are crucial for understanding the mechanisms of inhibition and how they can be impaired or improved.

Regional specialization within the human striatum for diverse psychological functions

Significance The subcortical striatum is critical for the planning and execution of motor behavior, and its dysfunction is associated with disorders such as Parkinson’s disease. More recently, the human striatum has also been reported to be involved in heterogeneous nonmotor psychological functions. However, detailed functional mappings of human psychological processes to striatal regions have been bound by theoretical and methodological limitations, including a strong focus on experimental paradigms derived from animal research, and the tendency to infer function from anatomical connectivity, rather than task-related activation. To overcome these limitations, we used a large-scale, unbiased, data-driven approach, and generated a precise, comprehensive functional map, directly associating striatal zones with the broadest range of psychological processes to date. Decades of animal and human neuroimaging research have identified distinct, but overlapping, striatal zones, which are interconnected with separable corticostriatal circuits, and are crucial for the organization of functional systems. Despite continuous efforts to subdivide the human striatum based on anatomical and resting-state functional connectivity, characterizing the different psychological processes related to each zone remains a work in progress. Using an unbiased, data-driven approach, we analyzed large-scale coactivation data from 5,809 human imaging studies. We (i) identified five distinct striatal zones that exhibited discrete patterns of coactivation with cortical brain regions across distinct psychological processes and (ii) identified the different psychological processes associated with each zone. We found that the reported pattern of cortical activation reliably predicted which striatal zone was most strongly activated. Critically, activation in each functional zone could be associated with distinct psychological processes directly, rather than inferred indirectly from psychological functions attributed to associated cortices. Consistent with well-established findings, we found an association of the ventral striatum (VS) with reward processing. Confirming less well-established findings, the VS and adjacent anterior caudate were associated with evaluating the value of rewards and actions, respectively. Furthermore, our results confirmed a sometimes overlooked specialization of the posterior caudate nucleus for executive functions, often considered the exclusive domain of frontoparietal cortical circuits. Our findings provide a precise functional map of regional specialization within the human striatum, both in terms of the differential cortical regions and psychological functions associated with each striatal zone.

Recurrent Processing during Object Recognition.

How does the brain learn to recognize objects visually, and perform this difficult feat robustly in the face of many sources of ambiguity and variability? We present a computational model based on the biology of the relevant visual pathways that learns to reliably recognize 100 different object categories in the face of naturally occurring variability in location, rotation, size, and lighting. The model exhibits robustness to highly ambiguous, partially occluded inputs. Both the unified, biologically plausible learning mechanism and the robustness to occlusion derive from the role that recurrent connectivity and recurrent processing mechanisms play in the model. Furthermore, this interaction of recurrent connectivity and learning predicts that high-level visual representations should be shaped by error signals from nearby, associated brain areas over the course of visual learning. Consistent with this prediction, we show how semantic knowledge about object categories changes the nature of their learned visual representations, as well as how this representational shift supports the mapping between perceptual and conceptual knowledge. Altogether, these findings support the potential importance of ongoing recurrent processing throughout the brain’s visual system and suggest ways in which object recognition can be understood in terms of interactions within and between processes over time.

Thalamic pathways underlying prefrontal cortex–medial temporal lobe oscillatory interactions

As focus shifts to large-scale network interactions involved in memory, it is becoming increasingly clear that oscillatory dynamics are critically involved. A number of studies have shown a negative correlation between memory retrieval in alpha and beta power, and a positive correlation between retrieval and theta power. In this opinion article, we suggest three thalamic sub-regions responsible for the coordination of oscillatory activity and the facilitation of memory processes. Specifically, the medial dorsal nucleus is related to changes in beta synchrony, the pulvinar is responsible for alpha synchrony, and the anterior thalamus is related to theta synchrony. These pathways may be modulated via frontal control, and changes in oscillations could be used to track the engagement of underlying memory systems.

Distinct contributions of the caudate nucleus, rostral prefrontal cortex, and parietal cortex to the execution of instructed tasks

When we behave according to rules and instructions, our brains interpret abstract representations of what to do and transform them into actual behavior. In order to investigate the neural mechanisms behind this process, we devised an fMRI experiment that explicitly isolated rule interpretation from rule encoding and execution. Our results showed that a specific network of regions (including the left rostral prefrontal cortex, the caudate nucleus, and the bilateral posterior parietal cortices) is responsible for translating rules into executable form. An analysis of activation patterns across conditions revealed that the posterior parietal cortices represent a mental template for the task to perform, that the inferior parietal gyrus and the caudate nucleus are responsible for instantiating the template in the proper context, and that the left rostral prefrontal cortex integrates information across complex relationships.

Individual Differences in Cognitive Flexibility

I s it better to be flexible, or persistent? A colleague once said that the secret to success in science is perseveration (no, not mere perseverence—that wouldn’t get you very far at all). Yet we all believe that the highest levels of cognitive function are associated with extreme flexibility—the ability to juggle many things at once and not get hung up on just one. Indeed, this ability to juggle the many demands of being a scientist are seemingly even more important these days than perseverating on one important deep problem. Clearly, these two poles of flexibility and perseverence (aka stability) both have benefits and costs (even in the world beyond the ivory tower), and it would make sense that, at a population level, individuals might be differentially distributed across this spectrum to cover our collective bases (1). Samanez-Larkin et al. (2) provide an important advance in understanding the biological basis for these individual differences, leveraging the powerful technique of dopamine (DA) receptor availability measurements under radioligand positron emission tomography (PET) imaging. This technique is uniquely important for understanding the function of DA (and other neuromodulators) in humans, by virtue of being able to noninvasively assess both DA receptor availability and differences in DA levels, across different brain areas. Interestingly, they found that individual differences in cognitive flexibility were predicted by baseline DA D2/D3 receptor availability in the prefrontal cortex (PFC), parietal cortex, and thalamus, whereas amphetamineinduced DA release in the anterior striatum additionally predicted individual differences (and partially mediated the cortical and thalamic effects). Interpreting these results requires wading into a complex sea of countervailing effects of dopamine and dopamine receptors across the striatum and PFC, as nicely reviewed by Cools and D’Esposito (3). Broadly speaking, consistent with a variety of data, D1 receptors in the PFC are associated with greater stability, whereas D2 receptors promote flexibility. However, the opposite pattern seems to hold in the striatum: D1 receptors promote flexibility, and D2 promotes stability. Furthermore, in the striatum, DA activation of D2 receptors has an overall inhibitory effect, whereas it is excitatory on D1. Putting this all together, we see that Samanez-Larkin et al. (2) bowled a perfect strike: they found D2 baseline availability effects on flexibility in PFC but not in the striatum, whereas they found increased DA levels in the striatum associated with greater flexibility, which would presumably lead to greater D1 receptor activation. If someone ever figures out a way to image D1 receptor availability using PET, then one would predict the opposite pattern, in which flexibility is associated with greater D1 baseline availability effects in striatum and lower levels of D1 availability in the PFC. As emphasized by Cools and D’Esposito (3), these opposing dynamics in PFC versus striatum provide a nice opportunity for the commonly seen U-shaped

Anthropomorphic reasoning about neuromorphic AGI safety

Abstract One candidate approach to creating artificial general intelligence (AGI) is to imitate the essential computations of human cognition. This process is sometimes called ‘reverse-engineering the brain’ and the end product called ‘neuromorphic.’ We argue that, unlike with other approaches to AGI, anthropomorphic reasoning about behaviour and safety concerns is appropriate and crucial in a neuromorphic context. Using such reasoning, we offer some initial ideas to make neuromorphic AGI safer. In particular, we explore how basic drives that promote social interaction may be essential to the development of cognitive capabilities as well as serving as a focal point for human-friendly outcomes.

Reply to Aksentijevic: It is a matter of what is countable and how neurons learn.

Aksentijevic (1) raises issues with our recent report (2) that bear further clarification because they reflect some fundamental confusion about the nature of random sequences. First, it is essential to distinguish statistics for individual elements (i.e., patterns of length one) vs. statistics for patterns consisting of more than one element (i.e., higher order patterns). Aksentijevic’s comments reflect some confusion between these two. The independent Bernoulli process only guarantees that the statistics for patterns of length one are fully random in the usual sense (e.g., independent and identically distributed), where the mean time and waiting time are indeed equal. Once one starts looking at higher order patterns, different statistical structures can, and do, emerge. For example, in a sequence of length three, pattern HH can happen twice but pattern HT cannot. No “assumption of self-correction” is required here. Different waiting times, clustering, or spreading of pattern occurrences are simply consequences intrinsic to the patterns’ composition (3, 4). Although somewhat difficult to comprehend precisely, we hope it is recognized that we did not fudge our random sequences to create structures where none actually exist. These differences are mathematical facts that are readily observed and replicable by anyone.