Neuro-computational mechanisms and individual biases in action-outcome learning under moral conflict

Learning to predict action outcomes in morally conflicting situations is essential for social decision-making but poorly understood. Here we tested which forms of Reinforcement Learning Theory capture how participants learn to choose between self-money and other-shocks, and how they adapt to changes in contingencies. We find choices were better described by a reinforcement learning model based on the current value of separately expected outcomes than by one based on the combined historical values of past outcomes. Participants track expected values of self-money and other-shocks separately, with the substantial individual difference in preference reflected in a valuation parameter balancing their relative weight. This valuation parameter also predicted choices in an independent costly helping task. The expectations of self-money and other-shocks were biased toward the favored outcome but fMRI revealed this bias to be reflected in the ventromedial prefrontal cortex while the pain-observation network represented pain prediction errors independently of individual preferences.

Humans depart from optimal computational models of interactive decision-making during competition under partial information

Decision making under uncertainty in multiagent settings is of increasing interest in decision science. The degree to which human agents depart from computationally optimal solutions in socially interactive settings is generally unknown. Such understanding provides insight into how social contexts affect human interaction and the underlying contributions of Theory of Mind. In this paper, we adapt the well-known ‘Tiger Problem’ from artificial-agent research to human participants in solo and interactive settings. Compared to computationally optimal solutions, participants gathered less information before outcome-related decisions when competing than cooperating with others. These departures from optimality were not haphazard but showed evidence of improved performance through learning. Costly errors emerged under conditions of competition, yielding both lower rates of rewarding actions and accuracy in predicting others. Taken together, this work provides a novel approach and insights into studying human social interaction when shared information is partial.

Automating incarceration Christopher Slobogin Cambridge University Press, 2021. 182 pp

Better algorithms are key to reducing bias in criminal sentencing, argues a legal scholar

AI empires Kate Crawford Yale University Press, 2021. 336 pp

A Microsoft researcher unpacks the power and perils of today's artificial intelligence

Theory of mind and decision science: Towards a typology of tasks and computational models

The ability to form a Theory of Mind (ToM), i.e., to theorize about others' mental states to explain and predict behavior in relation to attributed intentional states, constitutes a hallmark of human cognition. These abilities are multi-faceted and include a variety of different cognitive sub-functions. Here, we focus on decision processes in social contexts and review a number of experimental and computational modeling approaches in this field. We provide an overview of experimental accounts and formal computational models with respect to two dimensions: interactivity and uncertainty. Thereby, we aim at capturing the nuances of ToM functions in the context of social decision processes. We suggest there to be an increase in ToM engagement and multiplexing as social cognitive decision-making tasks become more interactive and uncertain. We propose that representing others as intentional and goal directed agents who perform consequential actions is elicited only at the edges of these two dimensions. Further, we argue that computational models of valuation and beliefs follow these dimensions to best allow researchers to effectively model sophisticated ToM-processes. Finally, we relate this typology to neuroimaging findings in neurotypical (NT) humans, studies of persons with autism spectrum (AS), and studies of nonhuman primates.

Semantic Incongruency Interferes With Endogenous Attention in Cross-Modal Integration of Semantically Congruent Objects

Efficient multisensory integration is often influenced by other cognitive processes including, but not limited to, semantic congruency and focused endogenous attention. Semantic congruency can re-allocate processing resources to the location of a congruent stimulus, while attention can prioritize the integration of multi-sensory stimuli under focus. Here, we explore the robustness of this phenomenon in the context of three stimuli, two of which are in the focus of endogenous attention. Participants completed an endogenous attention task with a stimulus compound consisting of 3 different objects: (1) a visual object (V) in the foreground, (2) an auditory object (A), and (3) a visual background scene object (B). Three groups of participants focused their attention on either the visual object and auditory sound (Group VA, n = 30), the visual object and the background (VB, n = 27), or the auditory sound and the background (AB, n = 30), and judged the semantic congruency of the objects under focus. Congruency varied systematically across all 3 stimuli: All stimuli could be semantically incongruent (e.g., V, ambulance; A, church bell; and B, swimming-pool) or all could be congruent (e.g., V, lion; A, roar; and B, savannah), or two objects could be congruent with the remaining one incongruent to the other two (e.g., V, duck; A, quack; and B, phone booth). Participants exhibited a distinct pattern of errors: when participants attended two congruent objects (e.g., group VA: V, lion; A, roar), in the presence of an unattended, incongruent third object (e.g., B, bath room) they tended to make more errors than in any other stimulus combination. Drift diffusion modeling of the behavioral data revealed a significantly smaller drift rate in two-congruent-attended condition, indicating slower evidence accumulation, which was likely due to interference from the unattended, incongruent object. Interference with evidence accumulation occurred independently of which pair of objects was in the focus of attention, which suggests that the vulnerability of congruency judgments to incongruent unattended distractors is not affected by sensory modalities. A control analysis ruled out the simple explanation of a negative response bias. These findings implicate that our perceptual system is highly sensitive to semantic incongruencies even when they are not endogenously attended.

Rapid face orienting in infants and school-age children with and without autism: Exploring measurement invariance in eye-tracking

Questions concerning the ontogenetic stability of autism have recently received increased attention as long-term longitudinal studies have appeared in the literature. Most experimental measures are designed for specific ages and functioning levels, yet developing experimental tasks appropriate for a wide range of ages and functioning levels is critical for future long-term longitudinal studies, and treatment studies implemented at different ages. Accordingly, we designed an eye-tracking task to measure preferential orienting to facial features and implemented it with groups of participants with varying levels of functioning: infants, and school-age children with and without autism. All groups fixated eyes first, revealing an early and stable orienting bias. This indicates common bias towards the eyes across participants regardless of age or diagnosis. We also demonstrate that this eye-tracking task can be used with diverse populations who range in age and cognitive functioning. Our developmental approach has conceptual implications for future work focused on task development and particularly new experimental measures that offer measurement equivalence across broad age ranges, intellectual functioning and verbal abilities.

Emotional arousal in agenesis of the corpus callosum

While the processing of verbal and psychophysiological indices of emotional arousal have been investigated extensively in relation to the left and right cerebral hemispheres, it remains poorly understood how both hemispheres normally function together to generate emotional responses to stimuli. Drawing on a unique sample of nine high-functioning subjects with complete agenesis of the corpus callosum (AgCC), we investigated this issue using standardized emotional visual stimuli. Compared to healthy controls, subjects with AgCC showed a larger variance in their cognitive ratings of valence and arousal, and an insensitivity to the emotion category of the stimuli, especially for negatively-valenced stimuli, and especially for their arousal. Despite their impaired cognitive ratings of arousal, some subjects with AgCC showed large skin-conductance responses, and in general skin-conductance responses discriminated emotion categories and correlated with stimulus arousal ratings. We suggest that largely intact right hemisphere mechanisms can support psychophysiological emotional responses, but that the lack of interhemispheric communication between the hemispheres, perhaps together with dysfunction of the anterior cingulate cortex, interferes with normal verbal ratings of arousal, a mechanism in line with some models of alexithymia.

Role of the amygdala in processing visual social stimuli

We review the evidence implicating the amygdala as a critical component of a neural network of social cognition, drawing especially on research involving the processing of faces and other visual social stimuli. We argue that, although it is clear that social behavioral representations are not stored in the amygdala, the most parsimonious interpretation of the data is that the amygdala plays a role in guiding social behaviors on the basis of socioenvironmental context. Thus, it appears to be required for normal social cognition. We propose that the amygdala plays this role by attentionally modulating several areas of visual and somatosensory cortex that have been implicated in social cognition, and in helping to direct overt visuospatial attention in face gaze. We also hypothesize that the amygdala exerts attentional modulation of simulation in somatosensory cortices such as supramarginal gyrus and insula. Finally, we argue that the term emotion be broadened to include increased attention to bodily responses and their representation in cortex.

Crystal structure of the catalytic domain of a thermophilic endocellulase

One way to improve the economic feasibility of biomass conversion is to enhance the catalytic efficiency of cellulases through protein engineering. This requires that high-resolution structures of cellulases be available. Here we present the structure of E2cd, the catalytic domain of the thermophilic endocellulase E2 from Thermomonospora fusca, as determined by X-ray crystallography. The structure was solved by multiple isomorphous replacement at 2.6-A resolution and has been refined at 1.8-A resolution to an R-value of 18.4% for all reflections between 10- and 1.8-A resolution. The fold of E2cd is based on an unusual parallel beta-barrel and is equivalent to the fold determined for the catalytic domain of cellobiohydrolase II, an exocellulase from Trichoderma reesei [Rouvinen et al. (1990) Science 249, 380-385]. The active site cleft of the enzyme, approximately 11 A deep and running the entire length of the molecule, is seen to be completely free for ligand binding in the crystal. A 2.2-A resolution analysis of crystals of E2cd complexed with cellobiose, an inhibitor, shows how cellobiose binds in the active site and interacts with several residues which line the cleft. Catalytic roles are suggested for three aspartic acid residues at the active site. A comparison of the E2cd and CBHIIcd structures reveals a large difference in their active site accessibilities and supports the hypothesis that the main difference between endo- and exocellulases is the degree to which their active sites are accessible to substrate.