It remains unclear whether and, if so, how non-human animals make on-the-fly predictions during pursuit. Here, we used a novel laboratory pursuit task that incentivizes prediction of future prey positions. We trained three macaques to perform a joystick-controlled pursuit task in which prey follow intelligent escape algorithms. Subjects aimed towards the prey's likely future positions, indicating that they generate internal predictions and use them to guide behavior. We then developed a generative model that explains real-time pursuit trajectories and showed that our subjects use prey position, velocity, and acceleration to make predictions. We identified neurons in the dorsal anterior cingulate cortex (dACC) whose responses track these three variables. These neurons multiplexed prediction-related variables with a distinct and explicit representation of the prey's future position. Our results provide a clear demonstration that the brain can explicitly represent future predictions and highlight the critical role of anterior cingulate cortex for futureoriented cognition.

To make efficient foraging decisions, we must combine information about the values of available options with nonvalue information. Some accounts of ventromedial PFC (vmPFC) suggest that it has a narrow role limited to evaluating immediately available options. We examined responses of neurons in area 14 (a putative macaque homolog of human vmPFC) as 2 male macaques performed a novel foraging search task. Although many neurons encoded the values of immediately available offers, they also independently encoded several other variables that influence choice, but that are conceptually distinct from offer value. These variables include average reward rate, number of offers viewed per trial, previous offer values, previous outcome sizes, and the locations of the currently attended offer. We conclude that, rather than serving as specialized economic value center, vmPFC plays a broad role in integrating relevant environmental information to drive foraging decisions.

We have the capacity to follow arbitrary stimulus–response rules, meaning simple policies that guide our behavior. Rule identity is broadly encoded across decision-making circuits, but there are less data on how rules shape the computations that lead to choices. One idea is that rules could simplify these computations. When we follow a rule, there is no need to encode or compute information that is irrelevant to the current rule, which could reduce the metabolic or energetic demands of decision-making. However, it is not clear if the brain can actually take advantage of this computational simplicity. To test this idea, we recorded from neurons in 3 regions linked to decision-making, the orbitofrontal cortex (OFC), ventral striatum (VS), and dorsal striatum (DS), while macaques performed a rule-based decision-making task. Rule-based decisions were identified via modeling rules as the latent causes of decisions. This left us with a set of physically identical choices that maximized reward and information, but could not be explained by simple stimulus–response rules. Contrasting rule-based choices with these residual choices revealed that following rules (1) decreased the energetic cost of decision-making; and (2) expanded rule-relevant coding dimensions and compressed rule-irrelevant ones. Together, these results suggest that we use rules, in part, because they reduce the costs of decision-making through a distributed representational warping in decision-making circuits.

It remains unclear how and to what extent non-human animals make demanding on-the-fly predictions during pursuit. We studied this problem in a novel laboratory pursuit task that incentivizes prediction of future prey positions. We trained three macaques to perform joystick-controlled pursuit of prey that followed intelligent escape algorithms. Subjects reliably aimed towards the prey’s likely future positions, indicating that they generate internal predictions and use those predictions to guide behavior. We then developed a generative model that explains real-time pursuit trajectories and showed that our subjects use prey position, velocity, and acceleration to make predictions. We identified neurons in the dorsal anterior cingulate cortex (dACC) whose responses track these three variables. These neurons multiplexed prediction-related variables with a distinct and explicit representation of the prey’s future position. Our results provide a clear demonstration that the brain can explicitly represent future predictions and highlight the critical role of anterior cingulate cortex for future-oriented cognition.One-sentence summaryIn a dynamic pursuit environment, monkeys actively predict future prey positions and dACC neurons encode these future positions.

Neurofilament light chain, a putative measure of neuronal damage, is measurable in blood and cerebrospinal fluid and is predictive of cognitive function in individuals with Alzheimer Disease. There has been limited prior work linking neurofilament light and functional connectivity and no prior work has investigated neurofilament light associations with functional connectivity in autosomal dominant Alzheimer Disease. Here we assessed relationships between blood neurofilament light, cognition, and functional connectivity in a cross-sectional sample of 106 autosomal dominant Alzheimer Disease mutation carriers and 76 non-carriers. We employed an innovative network-level enrichment analysis approach in order to assess connectome-wide associations with neurofilament light. Neurofilament light was positively correlated with deterioration of functional connectivity within the default mode network and negatively correlated with connectivity between default mode network and executive control networks including the cingulo-opercular, salience, and dorsal attention networks. Further, reduced connectivity within the default mode network and between the default mode network and executive control networks was associated with reduced cognitive function. Hierarchical regression analysis revealed that neurofilament levels and functional connectivity within the default mode network and between the default mode network and the dorsal attention network explained significant variance in cognitive composite scores when controlling for age, sex, and education. A mediation analysis demonstrated that functional connectivity within the default mode network and between the default mode network and dorsal attention network partially mediated the relationship between blood neurofilament light levels and cognitive function. Our novel results indicate that blood estimates of neurofilament levels correspond to direct measurements of brain dysfunction, shedding new light on the underlying biological processes of Alzheimer Disease. Further, we demonstrate how variation within key brain systems can partially mediate the negative effects of heighted total serum neurofilament levels, suggesting potential regions for targeted interventions. Finally, our results lend further evidence that low-cost and minimally invasive blood measurements of neurofilament may be a useful marker of brain functional connectivity and cognitive decline in Alzheimer disease.

Successful pursuit and evasion require rapid and precise coordination of navigation with adaptive motor control. We hypothesize that the dorsal anterior cingulate cortex (dACC), which communicates bidirectionally with both the hippocampal complex and premotor/motor areas, would serve a mapping role in this process. We recorded responses of dACC ensembles in two macaques performing a joystick-controlled continuous pursuit/evasion task. We find that dACC carries two sets of signals, (1) world-centric variables that together form a representation of the position and velocity of all relevant agents (self, prey, and predator) in the virtual world, and (2) avatar-centric variables, i.e. self-prey distance and angle. Both sets of variables are multiplexed within an overlapping set of neurons. Our results suggest that dACC may contribute to pursuit and evasion by computing and continuously updating a multicentric representation of the unfolding task state, and support the hypothesis that it plays a high-level abstract role in the control of behavior.

24 25 SIGNIFICANCE STATEMENT 26 27One important part of our ability to adapt flexibly to the world around us is our ability to 28 implement arbitrary stimulus-response mappings, known as "rules". Many studies have shown 29 that when we follow a rule, its identity is encoded in neuronal firing rates. However, it remains 30 unclear how rules regulate behavior. Here, we report that rules warp the way that sensorimotor 31 information is represented in decision-making circuits: enhancing information that is relevant to 32 the current rule at the expense of information that is irrelevant. These results imply that rules are 33implemented as a kind of attentional gate on what information is available for decision-making. 34 35 36ABSTRACT 37 38We have the capacity to follow arbitrary stimulus-response rules, meaning policies that 39 determine how we will behave across circumstances. Yet, it is not clear how rules guide 40 sensorimotor decision-making in the brain. Here, we recorded from neurons in three regions 41 linked to decision-making, the orbitofrontal cortex, ventral striatum, and dorsal striatum, while 42 macaques performed a rule-based decision-making task. We found that different rules warped 43 the neural representations of chosen options by expanding rule-relevant coding dimensions 44relative to rule-irrelevant ones. Some cognitive theories suggest that warping could increase 45 processing efficiency by facilitating rule-relevant computations at the expense of irrelevant ones. 46To test this idea, we modeled rules as the latent causes of decisions and identified a set of "rule-47 free" choices that could not be explained by simple rules. Contrasting these with rule-based 48 choices revealed that following rules decreased the energetic cost of decision-making while 49warping the representational geometry of choice.

Keywords 19 generalized linear model, mixed selectivity, prey, predator, foraging theory 20 21Acknowledgements 22 We thank Alex Thomé for his critical role in designing the task, for devising the 23 training protocols, and for developing our joysticks. We thank Marc Mancarella for 24 his critical help with joystick training. We are grateful for helpful discussions from 25Habiba Azab, Steve Piantadosi, Marc Schieber, and Adam Rouse. 26 27 Author Contributions 28 SBMY and BYH conceptualized and designed the experiment. SBMY collected the 29 data. SBMY, JCT, and BYH developed the model and analyzed the data. SBMY 30 and BYH wrote the manuscript. 31 32Recent studies have begun to identify the computational processes underlying pursuit and 51 evasion behavior (hereafter shortened to pursuit) in several species [7,8]; nonetheless, the 52 neural bases of these behaviors remain almost wholly unexplored (but see 9). Despite this 53 relative paucity of scholarly interest, pursuit is an important problem in neuroscience 54 because it is common in mobile animals, and because it is highly determinative of 55 reproductive success and thus a likely driver of evolution. Moreover, it represents a 56 mathematically tractable form of continuous decision-making, which decision 57 neuroscience often ignores in favor of discrete decisions [10,11]. 58When foragers move in their environments, neurons in the hippocampus and 59 adjacent structures track the forager's own positions of using a firing rate code [12][13][14]. 60 They do so by means of an explicit cognitive map [15,16]. Specifically, each neuron in the 61 hippocampus or medial entorhinal cortex exhibits one or more preferred firing fields [13]. 62That is to say, entry by an animal into a specific location results in robust spiking activity. 63The hippocampal spatial map is allocentric, meaning that it is organized relative to external 64 space [12,17]. To employ this information to guide actions, however, foragers must use a 65 complementary egocentric coding system, that is, one that is relative to the self [17][18][19]. 66Egocentric spatial representations are related to action planning and are often associated 67 with the premotor cortex/primary motor cortex [20,21] and sometimes with the parietal and 68 4 posteromedial cortex [21,22]. Even when navigating virtual or abstract environments, 69 foragers can benefit from multiple reference frames. That is, they can make an abstract, or 70 world-centric representation, but when the time comes to perform an action, or to navigate 71 the virtual space, they may need to use a coordinate system that is aligned to the 72 framework of their response modality. 73In addition to monitoring one's place in space, pursuit requires the careful 74 coordination of two distinct processes: (1) the computation and dynamic updating of a 75 representation of the pursuit environment, including the kinematics of the prey and 76 predator; (2) the ability to select and quickly adjust behavior in response to changing 77 demands. In other words, purs...