Ariel Zylberberg

Perceptual decisions are widely modeled as the accumulation of evidence to a bound. In the lateral intraparietal area (LIP), this computation is thought to be implemented in a low-dimensional population representation organized around the single action used to report the choice, consistent with an intentional framework. The intentional framework, however, implies that changing the behavioral demands on the report should change the representation itself, raising a question about the generality of the low-dimensional decision representation described in LIP: is it a special case of decisions reported through a single action, or does it reflect a more general computational architecture that can support multiple behavioral outputs? We tested this by training monkeys to report the termination of a motion-discrimination decision with a saccade to a choice-neutral target, and its content only later, with a saccade to one of two choice targets. Even though the two reports were behaviorally separable, the timing of termination remained systematically linked to the accumulation of sensory evidence supporting the eventual choice in both monkeys, indicating that both reports continued to draw on a common underlying computation. Using high-density Neuropixels recordings from LIP, however, we found that decision termination and decision content were represented along orthogonal population coding directions supported by largely non-overlapping groups of neurons. Yet the two representations were not independent: trial-by-trial fluctuations in the population encoding content predicted subsequent fluctuations in the population encoding termination, with their coupling strengthening as the decision evolved. These results suggest that a single decision computation can be flexibly reformatted into distinct, action-specific representations, coordinated by selective transfer of information between neural populations.

Attention plays a key role in decision-making by directing limited cognitive resources to relevant information. It has been proposed that attention also biases the decision process, due to a multiplicative interaction between attention and subjective value (e.g., Krajbich et al., 2010). We tested two predictions of models that posit a causal multiplicative effect of attention on decision formation: (i) the last fixation should be more informative about the choice when the overall value of the alternatives is high, and (ii) more attention should be directed to the chosen option when choices conflict with stated preferences than when they do not. Reanalyzing several datasets from a food-choice task, we found no evidence supporting these predictions. An alternative model where attention reflects choices after the decision has completed explains key observations, including the last-fixation bias, the gaze-cascade effect, and the effect of the overall value of the alternatives on response times. However, this model does not fully account for the association between dwell time and choice. We conclude that gaze behavior prior to the choice report likely reflects both decisional and post-decisional processes.

Attention plays a key role in decision-making by directing limited cognitive resources to relevant information. It has been proposed that attention also biases the decision process, due to a multiplicative interaction between attention and subjective value (e.g., Krajbich et al., 2010). We tested two predictions of models that posit a causal multiplicative effect of attention on decision formation: (i) the last fixation should be more informative about the choice when the overall value of the alternatives is high, and (ii) more attention should be directed to the chosen option when choices conflict with stated preferences than when they do not. Reanalyzing several datasets from a food-choice task, we found no evidence supporting these predictions. An alternative model where attention reflects choices after the decision has completed, explains key observations, including the last-fixation bias, the gaze-cascade effect and the effect of the overall value of the alternatives on response times. However, this model does not fully account for the association between dwell time and choice. We conclude that gaze behavior prior to the choice report likely reflects both decisional and post-decisional processes.

We study how confidence in perceptual decisions depends on whether it is communicated verbally (e.g., “very likely”) or numerically (e.g., “80% certainty”). We find that verbal expressions more reliably distinguish correct from incorrect choices than numerical reports, challenging the common assumption that numerical probabilities provide more precise representations of uncertainty. Additionally, in a dyadic decision-making task in which participants can revise their initial reports based on a partner’s choice and expressed confidence, verbal and numerical reports are equally effective in supporting accurate revisions of initial judgments. Together, these results underscore the effectiveness of verbal expressions as a means of conveying decision confidence.

The ability to evaluate one’s own knowledge states is often studied using paradigms in which participants make a decision and subsequently report their confidence. This structure has motivated hierarchical models in which confidence arises from a metacognitive process, distinct from the decision process itself, that estimates the probability that the choice is correct (Meyniel et al., 2015; Pouget et al., 2016; Fleming and Daw, 2017). Here, we contrast this framework with an alternative based on an intentional architecture (Shadlen et al., 2008). In this account, choice and confidence are determined simultaneously through a multidimensional drift–diffusion process, where each dimension represents one choice–confidence combination (Ratcliff and Starns, 2009, 2013). Choice, response time, and confidence jointly emerge when one of these accumulators reaches a decision bound. To adjudicate between these accounts, we fit both models to behavioral data from two perceptual tasks: a random-dots motion discrimination task with incentivized confidence reports, and a luminance discrimination task without feedback or incentives. The integrated model provided a superior fit for the incentivized motion task, whereas the hierarchical model more accurately captured behavior in the un-incentivized luminance task. These results suggest that confidence does not rely on a single computational mechanism, but rather its implementation may adapt to the specific demands and structure of the task.

Attention plays a key role in decision-making by directing limited cognitive resources to relevant information. It has been proposed that attention also biases the decision process, due to a multiplicative interaction between attention and subjective value (e.g., Krajbich et al., 2010). We tested two predictions of models that posit a causal multiplicative effect of attention on decision formation: (i) the last fixation should be more informative about the choice when the overall value of the alternatives is high, and (ii) more attention should be directed to the chosen option when choices conflict with stated preferences than when they do not. Reanalyzing several datasets from a food-choice task, we found no evidence supporting these predictions. An alternative model where attention reflects choices after the decision has completed, explains key observations, including the last-fixation bias, the gaze-cascade effect and the effect of the overall value of the alternatives on response times. However, this model does not fully account for the association between dwell time and choice. We conclude that gaze behavior prior to the choice report likely reflects both decisional and post-decisional processes.

Many decisions benefit from the accumulation of evidence obtained sequentially over time. In such circumstances, the decision-maker must balance speed against accuracy, while considering the cost associated with the passage of time. A neural mechanism to achieve this balance is to accumulate evidence and to terminate the deliberation when enough evidence has accrued. To accommodate time costs, it has been hypothesized that the criterion to terminate a decision may become lax as a function of time. Here we tested this hypothesis by manipulating the cost of time in a perceptual choice–reaction time (RT) task. Human participants (both sexes) discriminated the direction of motion in a dynamic random-dot display, which varied in difficulty across trials. Unbeknownst to the participants, halfway through the experiment, we increased the time pressure by canceling a small fraction of trials, mimicking a broken fixation, if they had not made a decision by a provisional deadline. This subtle manipulation led participants to make faster but less accurate decisions. Choice and RT were well explained by a bounded evidence-accumulation process. We developed a novel computational method to estimate the time-dependent changes in the stopping bounds directly from the participants’ RT and choice data. Our analysis revealed that the bounds decline as a function of time, and that this decline is steeper following the time–cost manipulation. The time-varying decision bounds approximate an optimal stopping policy, although the specific bound shape is idiosyncratic across individuals.

Deciding between a pair of familiar items is thought to rely on a comparison of their subjective values. When the values are similar, decisions take longer, and the choice may be inconsistent with stated value. These regularities are thought to be explained by the same mechanism of noisy evidence accumulation that leads to perceptual errors under conditions of low signal to noise. However, unlike perceptual decisions, subjective values may vary with internal states (e.g. desires, priorities) that change over time. This raises the possibility that the apparent stochasticity of choice reflects changes in value rather than mere noise. We hypothesized that these changes would manifest in serial dependencies across decision sequences. We analyzed data from a task in which participants chose between snack items. We developed an algorithm, Reval, that revealed significant fluctuations of the subjective values of items within an experimental session. The dynamic values predicted choices and response times more accurately than stated values. The dynamic values also furnished a superior account of the BOLD signal in ventromedial prefrontal cortex. A novel bounded-evidence accumulation model with temporally correlated evidence samples supports the idea that revaluation reflects the dynamic construction of subjective value during deliberation, which in turn influences subsequent decisions.

Many decisions are expressed as a preference for one item over another. When these items are familiar, it is often assumed that the decision maker assigns a value to each of the items and chooses the item with the highest value. These values may be imperfectly recalled, but are assumed to be stable over the course of an interview or psychological experiment. Choices that are inconsistent with a stated valuation are thought to occur because of unspecified noise that corrupts the neural representation of value. Assuming that the noise is uncorrelated over time, the pattern of choices and response times in value-based decisions are modeled within the framework of Bounded Evidence Accumulation (BEA), similar to that used in perceptual decision-making. In BEA, noisy evidence samples accumulate over time until the accumulated evidence for one of the options reaches a threshold. Here, we argue that the assumption of temporally uncorrelated noise, while reasonable for perceptual decisions, is not reasonable for value-based decisions. Subjective values depend on the internal state of the decision maker, including their desires, needs, priorities, attentional state, and goals. These internal states may change over time, or undergo revaluation, as will the subjective values. We reasoned that these hypothetical value changes should be detectable in the pattern of choices made over a sequence of decisions. We reanalyzed data from a well-studied task in which participants were presented with pairs of snacks and asked to choose the one they preferred. Using a novel algorithm (Reval), we show that the subjective value of the items changes significantly during a short experimental session (about 1 hour). Values derived with Reval explain choice and response time better than explicitly stated values. They also better explain the BOLD signal in the ventromedial prefrontal cortex, known to represent the value of decision alternatives. Revaluation is also observed in a BEA model in which successive evidence samples are not assumed to be independent. We argue that revaluation is a consequence of the process by which values are constructed during deliberation to resolve preference choices.