Action Selection Research Articles

Mixed representations of choice direction and outcome by GABA/glutamate cotransmitting neurons in the entopeduncular nucleus.

The basal ganglia (BG) are an evolutionarily conserved and phylogenetically old set of sub-cortical nuclei that guide action selection, evaluation, and reinforcement. The entopeduncular nucleus (EP) is a major BG output nucleus that contains a population of GABA/glutamate cotransmitting neurons (EP Sst+ ) that specifically target the lateral habenula (LHb) and whose function in behavior remains mysterious. Here we use a probabilistic switching task that requires an animal to maintain flexible relationships between action selection and evaluation to examine when and how GABA/glutamate cotransmitting neurons contribute to behavior. We find that EP Sst+ neurons are strongly engaged during this task and show bidirectional changes in activity during the choice and outcome periods of a trial. We then tested the effects of either permanently blocking cotransmission or modifying the GABA/glutamate ratio on behavior in well-trained animals. Neither manipulation produced detectable changes in behavior despite significant changes in synaptic transmission in the LHb, demonstrating that the outputs of these neurons are not required for on-going action-outcome updating in a probabilistic switching task.

An Exploratory Study of Contextual Control Modes in Teamwork.

This study investigated the relationship between human team member contextual control and team performance under time constraints. Contextual control modes, which describe different strategies for action selection in dynamic environments, characterize how humans maintain performance under variable demands. Control modes have not yet been studied in teamwork settings. Modeling of the cross-level interaction between team members' control modes and emerging team behaviors can improve understanding of effective teamwork in dynamic environments. A human-subjects study explored the relationship between individual contextual control and team performance. Questionnaires about contextual control were used to elicit individual control modes. Analysis compared team members' control modes and investigated how control modes changed under varying time pressures. Participant's control modes differed in their look ahead horizon, the extensiveness of prior action evaluation, and their prior experience. Many team members shifted control modes during trials, resulting in both convergence and divergence of paired control modes. No effects on communication rate were found due to changes in team members' control modes, but partially significant findings may suggest that the control mode divergence affects performance. Teams can operate in multiple control mode configurations that change dynamically according to context. Further research with an increased sample size is warranted to analyze how time constraints influence team members' control modes and overall teaming processes and whether divergence of team control mode is favorable under time pressures. Further study of contextual control in teams may help improve team design to better support teams in coping with time constraints in dynamic environments.

Computational joint action: Dynamical models to understand the development of joint coordination.

Coordinating with others is part of our everyday experience. Previous studies using sensorimotor coordination games suggest that human dyads develop coordination strategies that can be interpreted as Nash equilibria. However, if the players are uncertain about what their partner is doing, they develop coordination strategies which are robust to the actual partner's actions. This has suggested that humans select their actions based on an explicit prediction of what the partner will be doing-a partner model-which is probabilistic by nature. However, the mechanisms underlying the development of a joint coordination over repeated trials remain unknown. Very much like sensorimotor adaptation of individuals to external perturbations (eg force fields or visual rotations), dynamical models may help to understand how joint coordination develops over repeated trials. Here we present a general computational model-based on game theory and Bayesian estimation-designed to understand the mechanisms underlying the development of a joint coordination over repeated trials. Joint tasks are modeled as quadratic games, where each participant's task is expressed as a quadratic cost function. Each participant predicts their partner's next move (partner model) by optimally combining predictions and sensory observations, and selects their actions through a stochastic optimization of its expected cost, given the partner model. The model parameters include perceptual uncertainty (sensory noise), partner representation (retention rate and internale noise), uncertainty in action selection and its rate of decay (which can be interpreted as the action's learning rate). The model can be used in two ways: (i) to simulate interactive behaviors, thus helping to make specific predictions in the context of a given joint action scenario; and (ii) to analyze the action time series in actual experiments, thus providing quantitative metrics that describe individual behaviors during an actual joint action. We demonstrate the model in a variety of joint action scenarios. In a sensorimotor version of the Stag Hunt game, the model predicts that different representations of the partner lead to different Nash equilibria. In a joint two via-point (2-VP) reaching task, in which the actions consist of complex trajectories, the model captures well the observed temporal evolution of performance. For this task we also estimated the model parameters from experimental observations, which provided a comprehensive characterization of individual dyad participants. Computational models of joint action may help identifying the factors preventing or facilitating the development of coordination. They can be used in clinical settings, to interpret the observed behaviors in individuals with impaired interaction capabilities. They may also provide a theoretical basis to devise artificial agents that establish forms of coordination that facilitate neuromotor recovery.

Traffic signal phase control at urban isolated intersections: an adaptive strategy utilizing the improved D3QN algorithm

Abstract The intelligent control of traffic signals at urban single intersections has emerged as an effective approach to mitigating urban traffic congestion. However, the existing fixed phase control strategy of traffic signal lights lacks capability to dynamically adjust signal phase switching based on real-time traffic conditions leading to traffic congestion. In this paper, an adaptive real-time control method employed by the traffic signal phase at a single intersection is considered based on the improved Double Dueling Deep Q Network (D3QN) algorithm. Firstly, the traffic signal phase control problem is modeled as a Markov decision process, with its state, action, and reward defined. Subsequently, to enhance the convergence speed and learning performance of the D3QN algorithm, attenuation action selection strategy and priority experience playback technology based on tree summation structure are introduced. Then, traffic flow data from various traffic scenarios are utilized to train the traffic signal control model based on the improved D3QN to obtain the optimal signal phase switch strategy. Finally, the effectiveness and optimal performance of the improved D3QN-based traffic signal control strategy are validated across diverse traffic scenarios. The simulation results show that, compared with the control strategy based on AC, DQN, DDQN, D3QN, and C-D3QN algorithms, the cumulative reward of the proposed I-D3QN strategy is increased by at least 6.57%, and the average queue length and average waiting time are reduced by at least 9.64% and 7.61%, which can effectively reduce the congestion at isolated intersections and significantly improve traffic efficiency.

Cortical, striatal, and thalamic populations self-organize into a functionally connected circuit with long-term memory properties

The human brain is a complex organ with an intricate neuronal connectivity and diverse functional regions. Neurological disorders often disrupt the delicate balance among these anatomical compartments, resulting in severe impairments. The available therapeutic options constitute an incomplete solution as many patients respond partially, highlighting the need for continued research into causes and treatments. Bottom-up approaches, like in vitro models, offer insights into brain functions as they recreate the in vivo microenvironment that allows studying how specific features affect physiological and pathological conditions. In this work, we engineered the cortical-striatal-thalamic (CST) circuit, involved in many brain functions such as action initiation and selection, using a three-compartment polymeric device. We characterized the emerging spontaneous electrophysiological activity by using Micro-Electrode Arrays (MEAs). Cortical neurons exhibited complex bursting activity, which influenced the entire circuit. Striatal and thalamic neurons displayed predominantly tonic firing when isolated, while interconnections with the cortex synchronized and organized their neuronal activity, highlighting the cortical pivotal role in bursting activity and information processing. The CST circuit demonstrated self-organization abilities and displayed high entropy values, indicative of dynamic richness and information encoding potential. Furthermore, we proved the CST’s involvement in learning and memory. Our CST model provides a platform for further exploration into brain circuitry and potential therapeutic interventions, underscoring the necessity of realistic in vitro models to fully understand neurological diseases' pathophysiology.

Enhancing UAV Swarm Tactics with Edge AI: Adaptive Decision Making in Changing Environments

This paper presents a drone system that uses an improved network topology and MultiAgent Reinforcement Learning (MARL) to enhance mission performance in Unmanned Aerial Vehicle (UAV) swarms across various scenarios. We propose a UAV swarm system that allows drones to efficiently perform tasks with limited information sharing and optimal action selection through our Efficient Self UAV Swarm Network (ESUSN) and reinforcement learning (RL). The system reduces communication delay by 53% and energy consumption by 63% compared with traditional MESH networks with five drones and achieves a 64% shorter delay and 78% lower energy consumption with ten drones. Compared with nonreinforcement learning-based systems, mission performance and collision prevention improved significantly, with the proposed system achieving zero collisions in scenarios involving up to ten drones. These results demonstrate that training drone swarms through MARL and optimized information sharing significantly increases mission efficiency and reliability, allowing for the simultaneous operation of multiple drones.

Striatal lateral inhibition regulates action selection in a mouse model of levodopa-induced dyskinesia.

Striatal medium spiny neurons (MSNs) integrate multiple external inputs to shape motor output. In addition, MSNs form local inhibitory synaptic connections with one another. The function of striatal lateral inhibition is unknown, but one possibility is in selecting an intended action while suppressing alternatives. Action selection is disrupted in several movement disorders, including levodopa-induced dyskinesia (LID), a complication of Parkinson's disease (PD) therapy characterized by involuntary movements. Here, we identify chronic changes in the strength of striatal lateral inhibitory synapses in a mouse model of PD/LID. These synapses are also modulated by acute dopamine signaling. Chemogenetic suppression of lateral inhibition originating from dopamine D2 receptor-expressing MSNs lowers the threshold to develop involuntary movements in vivo , supporting a role in motor control. By examining the role of lateral inhibition in basal ganglia function and dysfunction, we expand the framework surrounding the role of striatal microcircuitry in action selection.

Deconstructing Deep Active Inference: A Contrarian Information Gatherer.

Active inference is a theory of perception, learning, and decision making that can be applied to neuroscience, robotics, psychology, and machine learning. Recently, intensive research has been taking place to scale up this framework using Monte Carlo tree search and deep learning. The goal of this activity is to solve more complicated tasks using deep active inference. First, we review the existing literature and then progressively build a deep active inference agent as follows: we (1) implement a variational autoencoder (VAE), (2) implement a deep hidden Markov model (HMM), and (3) implement a deep critical hidden Markov model (CHMM). For the CHMM, we implemented two versions, one minimizing expected free energy, CHMM[EFE] and one maximizing rewards, CHMM[reward]. Then we experimented with three different action selection strategies: the ε-greedy algorithm as well as softmax and best action selection. According to our experiments, the models able to solve the dSprites environment are the ones that maximize rewards. On further inspection, we found that the CHMM minimizing expected free energy almost always picks the same action, which makes it unable to solve the dSprites environment. In contrast, the CHMM maximizing reward keeps on selecting all the actions, enabling it to successfully solve the task. The only difference between those two CHMMs is the epistemic value, which aims to make the outputs of the transition and encoder networks as close as possible. Thus, the CHMM minimizing expected free energy repeatedly picks a single action and becomes an expert at predicting the future when selecting this action. This effectively makes the KL divergence between the output of the transition and encoder networks small. Additionally, when selecting the action down the average reward is zero, while for all the other actions, the expected reward will be negative. Therefore, if the CHMM has to stick to a single action to keep the KL divergence small, then the action down is the most rewarding. We also show in simulation that the epistemic value used in deep active inference can behave degenerately and in certain circumstances effectively lose, rather than gain, information. As the agent minimizing EFE is not able to explore its environment, the appropriate formulation of the epistemic value in deep active inference remains an open question.

Online Unmanned Ground Vehicle Path Planning Based on Multi-Attribute Intelligent Reinforcement Learning for Mine Search and Rescue

Aiming to improve the efficiency of the online process in path planning, a novel searching method is proposed based on environmental information analysis. Firstly, a search and rescue (SAR) environmental model and an unmanned ground vehicle (UGV) motion model are established according to the characteristics of a mining environment. Secondly, an online search area path-planning method is proposed based on the gray system theory and the reinforcement learning theory to handle multiple constraints. By adopting the multi-attribute intelligent (MAI) gray decision process, the action selection decision can be dynamically adjusted based on the current environment, ensuring the stable convergence of the model. Finally, experimental verification is conducted in different small-scale mine SAR simulation scenarios. The experimental results show that the proposed search planning method can capture the target in the search area with a smoother convergence effect and a shorter path length than other path-planning algorithms.

Spike-based computational primitives to reproduce neural dynamics in the parietal cortex during motor preparation

Abstract Biologically plausible spiking neural network models of sensory cortices can be instrumental in understanding and validating their principles of computation.
 Models based on Cortical Computational Primitives (CCPs), such as Hebbian plasticity and Winner-Take-All (WTA) networks, have already been successful in this approach.
 However, the specific nature and roles of CCPs in sensorimotor cortices during cognitive tasks are yet to be fully deciphered.
 The evolution of motor intention in the Posterior Parietal Cortex (PPC) before arm-reaching movements is a well-suited cognitive process to assess the effectiveness of different CCPs.
 To this end, we propose a biologically plausible model composed of heterogeneous spiking neurons which implements and combines multiple CCPs, such as multi-timescale learning and soft WTA modules.
 By training the model to replicate the dynamics of in-vivo recordings from non-human primates, we show how it is effective in generating meaningful representations from unbalanced input data, and in faithfully reproducing the transition from motor planning to action selection.
 Our findings elucidate the importance of distributing spike-based plasticity across multi-timescales, and provide an explanation for the role of different CCPs in models of frontoparietal cortical networks for performing multisensory integration to efficiently inform action execution.

Two-Stage Optimization Model Based on Neo4j-Dueling Deep Q Network

To alleviate the power flow congestion in active distribution networks (ADNs), this paper proposes a two-stage load transfer optimization model based on Neo4j-Dueling DQN. First, the Neo4j graph model was established as the training environment for Dueling DQN. Meanwhile, the power supply paths from the congestion point to the power source point were obtained using the Cypher language built into Neo4j, forming a load transfer space that served as the action space. Secondly, based on various constraints in the load transfer process, a reward and penalty function was formulated to establish the Dueling DQN training model. Finally, according to the ε−greedy action selection strategy, actions were selected from the action space and interacted with the Neo4j environment, resulting in the optimal load transfer operation sequence. In this paper, Python was used as the programming language, TensorFlow open-source software library was used to form a deep reinforcement network, and Py2neo toolkit was used to complete the linkage between the python platform and Neo4j. We conducted experiments on a real 79-node system, using three power flow congestion scenarios for validation. Under the three power flow congestion scenarios, the time required to obtain the results was 2.87 s, 4.37 s and 3.45 s, respectively. For scenario 1 before and after load transfer, the line loss, voltage deviation and line load rate were reduced by about 56.0%, 76.0% and 55.7%, respectively. For scenario 2 before and after load transfer, the line loss, voltage deviation and line load rate were reduced by 41.7%, 72.9% and 56.7%, respectively. For scenario 3 before and after load transfer, the line loss, voltage deviation and line load rate were reduced by 13.6%, 47.1% and 37.7%, respectively. The experimental results show that the trained model can quickly and accurately derive the optimal load transfer operation sequence under different power flow congestion conditions, thereby validating the effectiveness of the proposed model.

Functional role of area 8A in the posterior dorsolateral frontal region of the human brain: A single case study.

Area 8A has traditionally been considered to be the frontal eye field (FEF), i.e. the area for the motor production of eye movements. However, recent research has shown that the FEF lies posteriorly in premotor area 6. Research in macaque monkeys has demonstrated that, in contrast to premotor area 6, which is involved in the production of motor actions, area 8A is implicated in the cognitive allocation of attention to stimuli based on instruction cues. The aim of the present study was to elucidate the specific cognitive role of area 8A by examining a unique patient with a lesion restricted to area 8A, i.e. sparing the premotor cortex. This right-handed male patient underwent neuropsychological assessment and testing on two conditional associative-learning tasks: the motor hand conditional associative-learning task (MCALT) assessing the selection of motor actions based on instruction cues and the visual conditional associative-learning task (VCALT) assessing the selection of visual stimuli based on instruction cues. The patient's performance on the VCALT was significantly impaired compared to control subjects, but performance on the MCALT and on an Eye Movement Control Task was preserved. The present study provides evidence that area 8A is critical for the cognitive process regulating the allocation of attention to different stimuli in the environment, but not in the production of eye movements. These findings enhance understanding of the functional organization of the posterior dorsolateral frontal region in the human brain and suggest a specialized role of area 8A in high-level cognitive processes.

Resource Allocation Approach of Avionics System in SPO Mode Based on Proximal Policy Optimization

Single-Pilot Operations (SPO) mode is set to reshape the decision-making process between human-machine and air-ground operations. However, the limited on-board computing resources impose greater demands on the organization of performance parameters and the optimization of process efficiency in SPO mode. To address this challenge, this paper first investigates the flexible requirements of avionics systems arising from changes in SPO operational scenarios, then analyzes the architecture of Reconfigurable Integrated Modular Avionics (RIMA) and its resource allocation framework in the context of scarcity and configurability. A “mission-function-resource” mapping relationship is established between the reconfiguration service elements of SPO mode and avionics resources. Subsequently, the Proximal Policy Optimization (PPO) algorithm is introduced to simulate the resource allocation process of IMA reconfiguration in SPO mode. The objective optimization process is transformed into a sequential decision-making problem by considering constraints and optimization criteria such as load, latency, and power consumption within the feasible domain of avionics system resources. Finally, the resource allocation scheme for avionics system reconfiguration is determined by controlling the probability of action selection during the interaction between the agent and the environment. The experimental results show that the resource allocation scheme based on the PPO algorithm can effectively reduce power consumption and latency, and the DRL model has strong anti-interference and generalization. This enables avionics resources to respond dynamically to the capabilities required in SPO mode and enhances their ability to support the aircraft mission at all stages.

Therapeutic ultrasound: an innovative approach for targeting neurological disorders affecting the basal ganglia.

The basal ganglia are involved in motor control and action selection, and their impairment manifests in movement disorders such as Parkinson's disease (PD) and dystonia, among others. The complex neuronal circuitry of the basal ganglia is located deep inside the brain and presents significant treatment challenges. Conventional treatment strategies, such as invasive surgeries and medications, may have limited effectiveness and may result in considerable side effects. Non-invasive ultrasound (US) treatment approaches are becoming increasingly recognized for their therapeutic potential for reversibly permeabilizing the blood-brain barrier (BBB), targeting therapeutic delivery deep into the brain, and neuromodulation. Studies conducted on animals and early clinical trials using ultrasound as a therapeutic modality have demonstrated promising outcomes for controlling symptom severity while preserving neural tissue. These results could improve the quality of life for patients living with basal ganglia impairments. This review article explores the therapeutic frontiers of ultrasound technology, describing the brain mechanisms that are triggered and engaged by ultrasound. We demonstrate that this cutting-edge method could transform the way neurological disorders associated with the basal ganglia are managed, opening the door to less invasive and more effective treatments.

A transient high-dimensional geometry affords stable conjunctive subspaces for efficient action selection

Flexible action selection requires cognitive control mechanisms capable of mapping the same inputs to different output actions depending on the context. From a neural state-space perspective, this requires a control representation that separates similar input neural states by context. Additionally, for action selection to be robust and time-invariant, information must be stable in time, enabling efficient readout. Here, using EEG decoding methods, we investigate how the geometry and dynamics of control representations constrain flexible action selection in the human brain. Participants performed a context-dependent action selection task. A forced response procedure probed action selection different states in neural trajectories. The result shows that before successful responses, there is a transient expansion of representational dimensionality that separated conjunctive subspaces. Further, the dynamics stabilizes in the same time window, with entry into this stable, high-dimensional state predictive of individual trial performance. These results establish the neural geometry and dynamics the human brain needs for flexible control over behavior.

A two-stage stochastic programming model for comprehensive risk response action selection: A case study in Industry 4.0

Effective project risk management is critical in environments where both micro-level and macro-level risks are present. Traditional models often focus on micro-level risks, neglecting broader macroeconomic uncertainties such as geopolitical instability and supply chain disruptions. This research introduces a two-stage stochastic programming model designed to optimize the selection of Risk Response Actions (RRAs) under uncertainty while addressing both types of risk. The model incorporates “here-and-now” decisions at the planning stage and “wait-and-see” decisions as uncertainties unfold, enabling adaptive risk management throughout the project lifecycle. To solve the model efficiently, we employ an evolutionary algorithm combined with Sample Average Approximation (SAA) to handle the computational complexity of multiple scenarios. The model is applied to a real-world case study involving the integration of IoT and ERP systems in a smart factory in Iran, a project characterized by significant macroeconomic and geopolitical risks. Our key contribution lies in providing a comprehensive risk response strategy selection model that simultaneously addresses micro- and macro-level risks while incorporating strategic flexibility through outsourcing decisions. The results demonstrate that our model outperforms traditional deterministic models, offering enhanced resilience against macro-level risks and improved project performance under uncertainty. These findings provide valuable insights for project managers aiming to increase resilience and adaptability in volatile environments. By integrating both internal and external risk factors, our model offers a robust tool for managing complex projects, enhancing decision-making and project outcomes in uncertain conditions

Difficulties in perceptual–motor coordination of reaching behavior in children with autism spectrum disorder

Increased risk of injury from collisions with objects is an important issue in children with autism spectrum disorder (ASD). The purpose of the present study was to examine whether impaired perceptual–motor coordination may underlie the high frequency of collisions. Specifically, we hypothesized that collisions are likely to occur frequently in children with ASD due to their difficulty with body-related spatial perception and their inadequate motor planning, due to their detail-focused processing style, i.e., a tendency to focus attention on specific details rather than on the overall context. To evaluate impairment of perceptual–motor coordination in children with ASD, an original action selection task was developed to assess (a) body-related spatial perception and (b) predictive attentional properties for planning the subsequent movements based on the measures of hand movement and gaze behavior. Twenty-six children (13 diagnosed as ASD and 13 with typical development [TD]) completed the task, and their parents completed a questionnaire regarding their child's susceptibility to injury. Results showed that children with ASD displayed inaccurate body-related spatial perception and inadequate motor planning under conditions that induced their detail-focused processing style. In addition, analyses of gaze behavior reflected the characteristics of a detail-focused processing in children with ASD. This was correlated with the severity of ASD and the measure of motor planning. The present findings suggest that difficulty with perceptual–motor coordination, resulting in part from the detail-focused processing style, might contribute to susceptibility to injury in children with ASD. We consider that our evaluation of the difficulty with perceptual–motor of individuals with ASD provides a framework for understanding their high rate of collision-related injuries and could inform strategies for preventing these injuries.

Role of the Nucleus Accumbens in Signaled Avoidance Actions.

Animals, humans included, navigate their environments guided by sensory cues, responding adaptively to potential dangers and rewards. Avoidance behaviors serve as adaptive strategies in the face of signaled threats, but the neural mechanisms orchestrating these behaviors remain elusive. Current circuit models of avoidance behaviors indicate that the nucleus accumbens (NAc) in the ventral striatum plays a key role in signaled avoidance behaviors, but the nature of this engagement is unclear. Evolving perspectives propose the NAc as a pivotal hub for action selection, integrating cognitive and affective information to heighten the efficiency of both appetitive and aversive motivated behaviors. To unravel the engagement of the NAc during active and passive avoidance, we used calcium imaging fiber photometry to examine NAc GABAergic neuron activity in ad libitum moving mice performing avoidance behaviors. We then probed the functional significance of NAc neurons using optogenetics and genetically targeted or electrolytic lesions. We found that NAc neurons code contraversive orienting movements and avoidance actions. However, direct optogenetic inhibition or lesions of NAc neurons did not impair active or passive avoidance behaviors, challenging the notion of their purported pivotal role in adaptive avoidance. The findings emphasize that while the NAc encodes avoidance movements, it is not required for avoidance behaviors, highlighting the distinction between behavior encoding or representation and mediation or generation.

An Efficient Node Selection Policy for Monte Carlo Tree Search with Neural Networks

Monte Carlo tree search (MCTS) has been gaining increasing popularity, and the success of AlphaGo has prompted a new trend of incorporating a value network and a policy network constructed with neural networks into MCTS, namely, NN-MCTS. In this work, motivated by the shortcomings of the widely used upper confidence bounds applied to trees (UCT) policy, we formulate the node selection problem in NN-MCTS as a multistage ranking and selection (R&S) problem and propose a node selection policy that efficiently allocates a limited search budget to maximize the probability of correctly selecting the best action at the root state. The value and policy networks in NN-MCTS further improve the performance of the proposed node selection policy by providing prior knowledge and guiding the selection of the final action, respectively. Numerical experiments on two board games and an OpenAI task demonstrate that the proposed method outperforms the UCT policy used in AlphaGo Zero and MuZero, implying the potential of constructing node selection policies in NN-MCTS with R&S procedures. History: Accepted by Bruno Tuffin, Area Editor for Simulation. Funding: This work was supported by the National Natural Science Foundation of China [Grants 72325007, 72250065, and 72022001], and a PKU-Boya Postdoctoral Fellowship 2406396158. Supplemental Material: The software that supports the findings of this study is available within the paper and its Supplemental Information ( https://pubsonline.informs.org/doi/suppl/10.1287/ijoc.2023.0307 ) as well as from the IJOC GitHub software repository ( https://github.com/INFORMSJoC/2023.0307 ). The complete IJOC Software and Data Repository is available at https://informsjoc.github.io/ .

Attention as selection for action defended

AbstractAttention has become an important focal point of recent work in ethics and epistemology, yet philosophers continue to be noncommittal about what attention is. In this paper, I defend attention as selection for action in a weak form, namely that selection for action is sufficient for attention. I show that selection for action in this conception captures how we, the folk, experience it and how the cognitive scientist studies it. That is, selection for action pulls empirical and folk‐psychology together. Accordingly, philosophers who take seriously either source have reason to work with selection for action as their starting conception of attention. This conception provides a way to bridge empirical and philosophical concerns where attention is central. The theoretical advantages of selection for action have been obscured by the common opinion that it is easily refuted. I defend the position against many of the published objections and then deploy it to provide a foundation for the intuitive, but inchoate, idea of attention being gradable, something of which there can be more or less. An analysis of the gradability of attention is then applied to consider recent work on the harms due to a surplus of attention.