Medial Orbitofrontal Neurons Preferentially Signal Cues Predicting Changes in Reward during Unblocking.

UNLABELLED: The orbitofrontal cortex (OFC) has been broadly implicated in the ability to use the current value of expected outcomes to guide behavior. Although value correlates have been prominently reported in lateral OFC, they are more often associated with more medial areas. Further, recent studies in primates have suggested a dissociation in which the lateral OFC is involved in credit assignment and representation of reward identity and more medial areas are critical to representing value. Previously, we used unblocking to test more specifically what information about outcomes is represented by OFC neurons in rats; consistent with the proposed dichotomy between the lateral and medial OFC, we found relatively little linear value coding in the lateral OFC (Lopatina et al., 2015). Here we have repeated this experiment, recording in the medial OFC, to test whether such value signals might be found there. Neurons were recorded in an unblocking task as rats learned about cues that signaled either more, less, or the same amount of reward. We found that medial OFC neurons acquired responses to these cues; however, these responses did not signal different reward values across cues. Surprisingly, we found that cells developed responses to cues predicting a change, particularly a decrease, in reward value. This is consistent with a special role for medial OFC in representing current value to support devaluation/revaluation sensitive changes in behavior. SIGNIFICANCE STATEMENT: This study uniquely examines encoding in rodent mOFC at the single-unit level in response to cues that predict more, less, or no change in reward in rats during training in a Pavlovian unblocking task, finding more cells responding to change-predictive cues and stronger activity in response to cues predictive of less reward.

Model-based learning and the contribution of the orbitofrontal cortex to the model-free world.

Learning is proposed to occur when there is a discrepancy between reward prediction and reward receipt. At least two separate systems are thought to exist: one in which predictions are proposed to be based on model-free or cached values; and another in which predictions are model-based. A basic neural circuit for model-free reinforcement learning has already been described. In the model-free circuit the ventral striatum (VS) is thought to supply a common-currency reward prediction to midbrain dopamine neurons that compute prediction errors and drive learning. In a model-based system, predictions can include more information about an expected reward, such as its sensory attributes or current, unique value. This detailed prediction allows for both behavioral flexibility and learning driven by changes in sensory features of rewards alone. Recent evidence from animal learning and human imaging suggests that, in addition to model-free information, the VS also signals model-based information. Further, there is evidence that the orbitofrontal cortex (OFC) signals model-based information. Here we review these data and suggest that the OFC provides model-based information to this traditional model-free circuitry and offer possibilities as to how this interaction might occur.

Orbitofrontal cortex and learning predictions of state transitions.

The orbitofrontal cortex (OFC) has been implicated in goal-directed planning and model-based decision-making. One key prerequisite for model-based decision-making is learning the transition structure of the environment-the probabilities of transitioning from one environmental state to another. In this work, we investigated how the OFC might be involved in learning this transition structure, by using fMRI to assess OFC activity while humans experienced probabilistic cue-outcome transitions. We found that OFC activity was indeed correlated with behavioral measures of learning about transition structure. On a trial-by-trial basis, OFC activity was associated with subsequently increased expectation of the more probable outcome; that is, with subsequently more optimal cue-outcome predictions. Interestingly, this relationship was observed no matter what outcome occurred at the time of the OFC activity, and thus is inconsistent with an interpretation of the OFC activity as representing a "state prediction error" that would facilitate learning transitions via error-correcting mechanisms. Finally, OFC activity was related to more optimal predictions only for subsequent trials involving the same cue that was observed at the time of OFC activity-this relationship was not observed for subsequent trials involving a different cue. All together, these results indicate that the OFC is involved in updating or reinforcing a learned transition model on a trial-by-trial basis, specifically for the currently observed cue-outcome associations. (PsycInfo Database Record (c) 2021 APA, all rights reserved).

Orbitofrontal neurons signal sensory associations underlying model-based inference in a sensory preconditioning task.

Using knowledge of the structure of the world to infer value is at the heart of model-based reasoning and relies on a circuit that includes the orbitofrontal cortex (OFC). Some accounts link this to the representation of biological significance or value by neurons in OFC, while other models focus on the representation of associative structure or cognitive maps. Here we tested between these accounts by recording OFC neurons in rats during an OFC-dependent sensory preconditioning task. We found that while OFC neurons were strongly driven by biological significance or reward predictions at the end of training, they also showed clear evidence of acquiring the incidental stimulus-stimulus pairings in the preconditioning phase, prior to reward training. These results support a role for OFC in representing associative structure, independent of value.

Ensembles in medial and lateral orbitofrontal cortex construct cognitive maps emphasizing different features of the behavioral landscape.

The orbitofrontal cortex (OFC) has long been implicated in the ability to use the current value of expected outcomes to guide behavior. More recently, this specific role has been conceptualized as a special case of a more general function that OFC plays in constructing a "cognitive map" of the behavioral task space by labeling the current task state and learning relationships among task states. Here, we have used single unit recording data from 2 prior studies to examine whether and how information relating different states within and across trials is represented in medial versus lateral OFC in rats. Using a hierarchical clustering analysis, we examined how neurons from each area represented information about differently valued trial types, defined by the cue-outcome pairings, versus how those same neurons represented information about similar epochs between these different trial types, such as the stimulus sample, delay, and reward consumption epochs. This analysis revealed that ensembles in the lateral OFC (lOFC) group states according to trial epoch, whereas those in the medial OFC (mOFC) organize the same states by trial type. These results suggest that the lOFC and mOFC construct cognitive maps that emphasize different features of the behavioral landscape, with lOFC tracking events based on local similarities, irrespective of their values and mOFC tracking more distal or higher order relationships relevant to value. (PsycINFO Database Record

Lateral orbitofrontal neurons acquire responses to upshifted, downshifted, or blocked cues during unblocking.

The lateral orbitofrontal cortex (lOFC) has been described as signaling either outcome expectancies or value. Previously, we used unblocking to show that lOFC neurons respond to a predictive cue signaling a 'valueless' change in outcome features (McDannald, 2014). However, many lOFC neurons also fired to a cue that simply signaled more reward. Here, we recorded lOFC neurons in a variant of this task in which rats learned about cues that signaled either more (upshift), less (downshift) or the same (blocked) amount of reward. We found that neurons acquired responses specifically to one of the three cues and did not fire to the other two. These results show that, at least early in learning, lOFC neurons fire to valued cues in a way that is more consistent with signaling of the predicted outcome's features than with signaling of a general, abstract or cached value that is independent of the outcome.