Origins of noise in both improving and degrading decision making.

Noise is a fundamental problem for information processing in neural systems. In decision-making, noise is assumed to have a primary role in errors and stochastic choice behavior. However, little is known about how noise arising from different sources contributes to value coding and choice behaviors, especially when it interacts with neural computation. Here we examine how noise arising early versus late in the choice process differentially impacts context-dependent choice behavior. We found in model simulations that early and late noise predict opposing context effects: under early noise, contextual information enhances choice accuracy; while under late noise, context degrades choice accuracy. Furthermore, we verified these opposing predictions in experimental human choice behavior. Manipulating early and late noise - by inducing uncertainty in option values and controlling time pressure - produced dissociable positive and negative context effects. These findings reconcile controversial experimental findings in the literature reporting either context-driven impairments or improvements in choice performance, suggesting a unified mechanism for context-dependent choice. More broadly, these findings highlight how different sources of noise can interact with neural computations to differentially modulate behavior. Significance: The current study addresses the role of noise origin in decision-making, reconciling controversies around how decision-making is impacted by context. We demonstrate that different types of noise - either arising early during evaluation or late during option comparison - leads to distinct results: with early noise, context enhances choice accuracy, while with late noise, context impairs it. Understanding these dynamics offers potential strategies for improving decision-making in noisy environments and refining existing neural computation models. Overall, our findings advance our understanding of how neural systems handle noise in essential cognitive tasks, suggest a beneficial role for contextual modulation under certain conditions, and highlight the profound implications of noise structure in decision-making.

Flexible control of representational dynamics in a disinhibition-based model of decision-making.

Inhibition is crucial for brain function, regulating network activity by balancing excitation and implementing gain control. Recent evidence suggests that beyond simply inhibiting excitatory activity, inhibitory neurons can also shape circuit function through disinhibition. While disinhibitory circuit motifs have been implicated in cognitive processes, including learning, attentional selection, and input gating, the role of disinhibition is largely unexplored in the study of decision-making. Here, we show that disinhibition provides a simple circuit motif for fast, dynamic control of network state and function. This dynamic control allows a disinhibition-based decision model to reproduce both value normalization and winner-take-all dynamics, the two central features of neurobiological decision-making captured in separate existing models with distinct circuit motifs. In addition, the disinhibition model exhibits flexible attractor dynamics consistent with different forms of persistent activity seen in working memory. Fitting the model to empirical data shows it captures well both the neurophysiological dynamics of value coding and psychometric choice behavior. Furthermore, the biological basis of disinhibition provides a simple mechanism for flexible top-down control of the network states, enabling the circuit to capture diverse task-dependent neural dynamics. These results suggest a biologically plausible unifying mechanism for decision-making and emphasize the importance of local disinhibition in neural processing.

Context-dependent choice and evaluation in real-world consumer behavior.

A body of work spanning neuroscience, economics, and psychology indicates that decision-making is context-dependent, which means that the value of an option depends not only on the option in question, but also on the other options in the choice set-or the 'context'. While context effects have been observed primarily in small-scale laboratory studies with tightly constrained, artificially constructed choice sets, it remains to be determined whether these context effects take hold in real-world choice problems, where choice sets are large and decisions driven by rich histories of direct experience. Here, we investigate whether valuations are context-dependent in real-world choice by analyzing a massive restaurant rating dataset as well as two independent replication datasets which provide complementary operationalizations of restaurant choice. We find that users make fewer ratings-maximizing choices in choice sets with higher-rated options-a hallmark of context-dependent choice-and that post-choice restaurant ratings also varied systematically with the ratings of unchosen restaurants. Furthermore, in a follow-up laboratory experiment using hypothetical choice sets matched to the real-world data, we find further support for the idea that subjective valuations of restaurants are scaled in accordance with the choice context, providing corroborating evidence for a general mechanistic-level account of these effects. Taken together, our results provide a potent demonstration of context-dependent choice in real-world choice settings, manifesting both in decisions and subjective valuation of options.

Asymmetric and adaptive reward coding via normalized reinforcement learning.

Learning is widely modeled in psychology, neuroscience, and computer science by prediction error-guided reinforcement learning (RL) algorithms. While standard RL assumes linear reward functions, reward-related neural activity is a saturating, nonlinear function of reward; however, the computational and behavioral implications of nonlinear RL are unknown. Here, we show that nonlinear RL incorporating the canonical divisive normalization computation introduces an intrinsic and tunable asymmetry in prediction error coding. At the behavioral level, this asymmetry explains empirical variability in risk preferences typically attributed to asymmetric learning rates. At the neural level, diversity in asymmetries provides a computational mechanism for recently proposed theories of distributional RL, allowing the brain to learn the full probability distribution of future rewards. This behavioral and computational flexibility argues for an incorporation of biologically valid value functions in computational models of learning and decision-making.

Excitatory-inhibitory tone shapes decision strategies in a hierarchical neural network model of multi-attribute choice.

We are constantly faced with decisions between alternatives defined by multiple attributes, necessitating an evaluation and integration of different information sources. Time-varying signals in multiple brain areas are implicated in decision-making; but we lack a rigorous biophysical description of how basic circuit properties, such as excitatory-inhibitory (E/I) tone and cascading nonlinearities, shape attribute processing and choice behavior. Furthermore, how such properties govern choice performance under varying levels of environmental uncertainty is unknown. We investigated two-attribute, two-alternative decision-making in a dynamical, cascading nonlinear neural network with three layers: an input layer encoding choice alternative attribute values; an intermediate layer of modules processing separate attributes; and a final layer producing the decision. Depending on intermediate layer E/I tone, the network displays distinct regimes characterized by linear (I), convex (II) or concave (III) choice indifference curves. In regimes I and II, each option's attribute information is additively integrated. In regime III, time-varying nonlinear operations amplify the separation between offer distributions by selectively attending to the attribute with the larger differences in input values. At low environmental uncertainty, a linear combination most consistently selects higher valued alternatives. However, at high environmental uncertainty, regime III is more likely than a linear operation to select alternatives with higher value. Furthermore, there are conditions where readout from the intermediate layer could be experimentally indistinguishable from the final layer. Finally, these principles are used to examine multi-attribute decisions in systems with reduced inhibitory tone, leading to predictions of different choice patterns and overall performance between those with restrictions on inhibitory tone and neurotypicals.

Computational Markers of Risky Decision-making for Identification of Temporal Windows of Vulnerability to Opioid Use in a Real-world Clinical Setting.

Importance: Opioid addiction is a major public health problem. Despite availability of evidence-based treatments, relapse and dropout are common outcomes. Efforts aimed at identifying reuse risk and gaining more precise understanding of the mechanisms conferring reuse vulnerability are needed. Objective: To use tools from computational psychiatry and decision neuroscience to identify changes in decision-making processes preceding opioid reuse. Design, Setting, and Participants: A cohort of individuals with opioid use disorder were studied longitudinally at a community-based treatment setting for up to 7 months (1-15 sessions per person). At each session, patients completed a risky decision-making task amenable to computational modeling and standard clinical assessments. Time-lagged mixed-effects logistic regression analyses were used to assess the likelihood of opioid use between sessions (t to t + 1; within the subsequent 1-4 weeks) from data acquired at the current session (t). A cohort of control participants completed similar procedures (1-5 sessions per person), serving both as a baseline comparison group and an independent sample in which to assess measurement test-retest reliability. Data were analyzed between January 1, 2018, and September 5, 2019. Main Outcomes and Measures: Two individual model-based behavioral markers were derived from the task completed at each session, capturing a participant's current tolerance of known risks and ambiguity (partially unknown risks). Current anxiety, craving, withdrawal, and nonadherence were assessed via interview and clinic records. Opioid use was ascertained from random urine toxicology tests and self-reports. Results: Seventy patients (mean [SE] age, 44.7 [1.3] years; 12 women and 58 men [82.9% male]) and 55 control participants (mean [SE] age, 42.4 [1.5] years; 13 women and 42 men [76.4% male]) were included. Of the 552 sessions completed with patients (mean [SE], 7.89 [0.59] sessions per person), 252 (45.7%) directly preceded opioid use events (mean [SE], 3.60 [0.44] sessions per person). From the task parameters, only ambiguity tolerance was significantly associated with increased odds of prospective opioid use (adjusted odds ratio, 1.37 [95% CI, 1.07-1.76]), indicating patients were more tolerant specifically of ambiguous risks prior to these use events. The association of ambiguity tolerance with prospective use was independent of established clinical factors (adjusted odds ratio, 1.29 [95% CI, 1.01-1.65]; P = .04), such that a model combining these factors explained more variance in reuse risk. No significant differences in ambiguity tolerance were observed between patients and control participants, who completed 197 sessions (mean [SE], 3.58 [0.21] sessions per person); however, patients were more tolerant of known risks (B = 0.56 [95% CI, 0.05-1.07]). Conclusions and Relevance: Computational approaches can provide mechanistic insights about the cognitive factors underlying opioid reuse vulnerability and may hold promise for clinical use.

Sensitivity of reaction time to the magnitude of rewards reveals the cost-structure of time.

The Drift-Diffusion Model (DDM) is the prevalent computational model of the speed-accuracy trade-off in decision making. The DDM provides an explanation of behavior by optimally balancing reaction times and error rates. However, when applied to value-based decision making, the DDM makes the stark prediction that reaction times depend only on the relative utility difference between the options and not on absolute utility magnitudes. This prediction runs counter to evidence that reaction times decrease with higher utility magnitude. Here, we ask if and how it could be optimal for reaction times to show this observed pattern. We study an algorithmic framework that balances the cost of delaying rewards against the utility of obtained rewards. We find that the functional form of the cost of delay plays a key role, with the empirically observed pattern becoming optimal under multiplicative discounting. We add to the empirical literature by testing whether utility magnitude affects reaction times using a novel methodology that does not rely on functional form assumptions for the subjects' utilities. Our results advance the understanding of how and why reaction times are sensitive to the magnitude of rewards.

Bounded rationality in C. elegans is explained by circuit-specific normalization in chemosensory pathways.

Rational choice theory assumes optimality in decision-making. Violations of a basic axiom of economic rationality known as "Independence of Irrelevant Alternatives" (IIA) have been demonstrated in both humans and animals and could stem from common neuronal constraints. Here we develop tests for IIA in the nematode Caenorhabditis elegans, an animal with only 302 neurons, using olfactory chemotaxis assays. We find that in most cases C. elegans make rational decisions. However, by probing multiple neuronal architectures using various choice sets, we show that violations of rationality arise when the circuit of olfactory sensory neurons is asymmetric. We further show that genetic manipulations of the asymmetry between the AWC neurons can make the worm irrational. Last, a context-dependent normalization-based model of value coding and gain control explains how particular neuronal constraints on information coding give rise to irrationality. Thus, we demonstrate that bounded rationality could arise due to basic neuronal constraints.

Multiple timescales of normalized value coding underlie adaptive choice behavior.

Adaptation is a fundamental process crucial for the efficient coding of sensory information. Recent evidence suggests that similar coding principles operate in decision-related brain areas, where neural value coding adapts to recent reward history. However, the circuit mechanism for value adaptation is unknown, and the link between changes in adaptive value coding and choice behavior is unclear. Here we show that choice behavior in nonhuman primates varies with the statistics of recent rewards. Consistent with efficient coding theory, decision-making shows increased choice sensitivity in lower variance reward environments. Both the average adaptation effect and across-session variability are explained by a novel multiple timescale dynamical model of value representation implementing divisive normalization. The model predicts empirical variance-driven changes in behavior despite having no explicit knowledge of environmental statistics, suggesting that distributional characteristics can be captured by dynamic model architectures. These findings highlight the importance of treating decision-making as a dynamic process and the role of normalization as a unifying computation for contextual phenomena in choice.

The computational form of craving is a selective multiplication of economic value.

Craving is thought to be a specific desire state that biases choice toward the desired object, be it chocolate or drugs. A vast majority of people report having experienced craving of some kind. In its pathological form craving contributes to health outcomes in addiction and obesity. Yet despite its ubiquity and clinical relevance we still lack a basic neurocomputational understanding of craving. Here, using an instantaneous measure of subjective valuation and selective cue exposure, we identify a behavioral signature of a food craving-like state and advance a computational framework for understanding how this state might transform valuation to bias choice. We find desire induced by exposure to a specific high-calorie, high-fat/sugar snack good is expressed in subjects' momentary willingness to pay for this good. This effect is selective but not exclusive to the exposed good; rather, we find it generalizes to nonexposed goods in proportion to their subjective attribute similarity to the exposed ones. A second manipulation of reward size (number of snack units available for purchase) further suggested that a multiplicative gain mechanism supports the transformation of valuation during laboratory craving. These findings help explain how real-world food craving can result in behaviors inconsistent with preferences expressed in the absence of craving and open a path for the computational modeling of craving-like phenomena using a simple and repeatable experimental tool for assessing subjective states in economic terms.

Risk preferences impose a hidden distortion on measures of choice impulsivity.

Measuring temporal discounting through the use of intertemporal choice tasks is now the gold standard method for quantifying human choice impulsivity (impatience) in neuroscience, psychology, behavioral economics, public health and computational psychiatry. A recent area of growing interest is individual differences in discounting levels, as these may predispose to (or protect from) mental health disorders, addictive behaviors, and other diseases. At the same time, more and more studies have been dedicated to the quantification of individual attitudes towards risk, which have been measured in many clinical and non-clinical populations using closely related techniques. Economists have pointed to interactions between measurements of time preferences and risk preferences that may distort estimations of the discount rate. However, although becoming standard practice in economics, discount rates and risk preferences are rarely measured simultaneously in the same subjects in other fields, and the magnitude of the imposed distortion is unknown in the assessment of individual differences. Here, we show that standard models of temporal discounting -such as a hyperbolic discounting model widely present in the literature which fails to account for risk attitudes in the estimation of discount rates- result in a large and systematic pattern of bias in estimated discounting parameters. This can lead to the spurious attribution of differences in impulsivity between individuals when in fact differences in risk attitudes account for observed behavioral differences. We advance a model which, when applied to standard choice tasks typically used in psychology and neuroscience, provides both a better fit to the data and successfully de-correlates risk and impulsivity parameters. This results in measures that are more accurate and thus of greater utility to the many fields interested in individual differences in impulsivity.

Free choice shapes normalized value signals in medial orbitofrontal cortex.

Normalization is a common cortical computation widely observed in sensory perception, but its importance in perception of reward value and decision making remains largely unknown. We examined (1) whether normalized value signals occur in the orbitofrontal cortex (OFC) and (2) whether changes in behavioral task context influence the normalized representation of value. We record medial OFC (mOFC) single neuron activity in awake-behaving monkeys during a reward-guided lottery task. mOFC neurons signal the relative values of options via a divisive normalization function when animals freely choose between alternatives. The normalization model, however, performed poorly in a variant of the task where only one of the two possible choice options yields a reward and the other was certain not to yield a reward (so called: "forced choice"). The existence of such context-specific value normalization may suggest that the mOFC contributes valuation signals critical for economic decision making when meaningful alternative options are available.

Normalized value coding explains dynamic adaptation in the human valuation process.

The notion of subjective value is central to choice theories in ecology, economics, and psychology, serving as an integrated decision variable by which options are compared. Subjective value is often assumed to be an absolute quantity, determined in a static manner by the properties of an individual option. Recent neurobiological studies, however, have shown that neural value coding dynamically adapts to the statistics of the recent reward environment, introducing an intrinsic temporal context dependence into the neural representation of value. Whether valuation exhibits this kind of dynamic adaptation at the behavioral level is unknown. Here, we show that the valuation process in human subjects adapts to the history of previous values, with current valuations varying inversely with the average value of recently observed items. The dynamics of this adaptive valuation are captured by divisive normalization, linking these temporal context effects to spatial context effects in decision making as well as spatial and temporal context effects in perception. These findings suggest that adaptation is a universal feature of neural information processing and offer a unifying explanation for contextual phenomena in fields ranging from visual psychophysics to economic choice.

Oculomatic: High speed, reliable, and accurate open-source eye tracking for humans and non-human primates.

BACKGROUND: Video-based noninvasive eye trackers are an extremely useful tool for many areas of research. Many open-source eye trackers are available but current open-source systems are not designed to track eye movements with the temporal resolution required to investigate the mechanisms of oculomotor behavior. Commercial systems are available but employ closed source hardware and software and are relatively expensive, limiting wide-spread use. NEW METHOD: Here we present Oculomatic, an open-source software and modular hardware solution to eye tracking for use in humans and non-human primates. RESULTS: Oculomatic features high temporal resolution (up to 600Hz), real-time eye tracking with high spatial accuracy (<0.5°), and low system latency (∼1.8ms, 0.32ms STD) at a relatively low-cost. COMPARISON WITH EXISTING METHOD(S): Oculomatic compares favorably to our existing scleral search-coil system while being fully non invasive. CONCLUSIONS: We propose that Oculomatic can support a wide range of research into the properties and neural mechanisms of oculomotor behavior.

Adaptive neural coding: from biological to behavioral decision-making.

Empirical decision-making in diverse species deviates from the predictions of normative choice theory, but why such suboptimal behavior occurs is unknown. Here, we propose that deviations from optimality arise from biological decision mechanisms that have evolved to maximize choice performance within intrinsic biophysical constraints. Sensory processing utilizes specific computations such as divisive normalization to maximize information coding in constrained neural circuits, and recent evidence suggests that analogous computations operate in decision-related brain areas. These adaptive computations implement a relative value code that may explain the characteristic context-dependent nature of behavioral violations of classical normative theory. Examining decision-making at the computational level thus provides a crucial link between the architecture of biological decision circuits and the form of empirical choice behavior.

Dynamic divisive normalization predicts time-varying value coding in decision-related circuits.

Normalization is a widespread neural computation, mediating divisive gain control in sensory processing and implementing a context-dependent value code in decision-related frontal and parietal cortices. Although decision-making is a dynamic process with complex temporal characteristics, most models of normalization are time-independent and little is known about the dynamic interaction of normalization and choice. Here, we show that a simple differential equation model of normalization explains the characteristic phasic-sustained pattern of cortical decision activity and predicts specific normalization dynamics: value coding during initial transients, time-varying value modulation, and delayed onset of contextual information. Empirically, we observe these predicted dynamics in saccade-related neurons in monkey lateral intraparietal cortex. Furthermore, such models naturally incorporate a time-weighted average of past activity, implementing an intrinsic reference-dependence in value coding. These results suggest that a single network mechanism can explain both transient and sustained decision activity, emphasizing the importance of a dynamic view of normalization in neural coding.

Exploiting exploration: past outcomes and future actions.

Applying past knowledge to future actions is crucial for adaptive choice behavior. Here, in this issue of Neuron, Donahue et al. (2013) show that reward enhances neural coding reliability for actions in a network of frontal and parietal brain areas.

Thirst-dependent risk preferences in monkeys identify a primitive form of wealth.

Experimental economic techniques have been widely used to evaluate human risk attitudes, but how these measured attitudes relate to overall individual wealth levels is unclear. Previous noneconomic work has addressed this uncertainty in animals by asking the following: (i) Do our close evolutionary relatives share both our risk attitudes and our degree of economic rationality? And (ii) how does the amount of food or water one holds (a nonpecuniary form of "wealth") alter risk attitudes in these choosers? Unfortunately, existing noneconomic studies have provided conflicting insights from an economic point of view. We therefore used standard techniques from human experimental economics to measure monkey risk attitudes for water rewards as a function of blood osmolality (an objective measure of how much water the subjects possess). Early in training, monkeys behaved randomly, consistently violating first-order stochastic dominance and monotonicity. After training, they behaved like human choosers--technically consistent in their choices and weakly risk averse (i.e., risk averse or risk neutral on average)--suggesting that well-trained monkeys can serve as a model for human choice behavior. As with attitudes about money in humans, these risk attitudes were strongly wealth dependent; as the animals became "poorer," risk aversion increased, a finding incompatible with some models of wealth and risk in human decision making.