ArticlePDF Available
... Seminal work by Rabbitt and colleagues in the mid-1960s [5][6][7][8] first raised the importance of a system that detected errors to adjust performance. Evidence for neural signals that processed errors first appeared in the early 1990s with the observation of negative electrical potential that occurred in the medial frontal region of the brain, about 50 to 100 ms after making an error, termed error-related negativity 9,10 . Combining error-related negativity measurement and functional magnetic resonance imaging (fMRI), a prominent error-detection network has been described in humans 11,12 . ...
... Areas including anterior cingulate cortex (ACC), anterior insular (operculum), ventral lateral prefrontal cortex, dorsal lateral prefrontal cortex and parietal lobe have been reported to contain signals sensitive to errors, with the ACC most consistently found to be involved in error detection. Using behavioral neurophysiology, various groups have also identified prominent error-detection signals in the ACC of non-human primates 10,11,[13][14][15][16][17][18][19][20] , consistent with human literature. ...
Article
Full-text available
Computational models proposed that the medial temporal lobe (MTL) contributes importantly to error-driven learning, though little direct in-vivo evidence for this hypothesis exists. To test this, we recorded in the entorhinal cortex (EC) and hippocampus (HPC) as macaques performed an associative learning task using an error-driven learning strategy, defined as better performance after error relative to correct trials. Error-detection signals were more prominent in the EC relative to HPC. Early in learning hippocampal but not EC neurons signaled error-driven learning by increasing their population stimulus-selectivity following error trials. This same pattern was not seen in another task where error-driven learning was not used. After learning, different populations of cells in both the EC and HPC signaled long-term memory of newly learned associations with enhanced stimulus-selective responses. These results suggest prominent but differential contributions of EC and HPC to learning from errors and a particularly important role of the EC in error-detection. Ku et al. recorded in the entorhinal cortex (EC) and hippocampus (HPC) of macaques during associative learning tasks in order to test the computational model prediction that they contribute to error-driven learning. They demonstrate that the EC and HPC have prominent but differential contributions to learning from errors, with the EC having a particularly prominent role in error-detection.
... Seminal work by Rabbitt and colleagues in the mid-1960s [5][6][7][8] first raised the importance of a system that detected errors to adjust performance. Evidence for neural signals that processed errors first appeared in the early 1990s with the observation of negative electrical potential that occurred in the medial frontal region of the brain, about 50 to 100 ms after making an error, termed error-related negativity 9,10 . Combining error-related negativity measurement and functional magnetic resonance imaging (fMRI), a prominent error-detection network has been described in humans 11,12 . ...
... Areas including anterior cingulate cortex (ACC), anterior insular (operculum), ventral lateral prefrontal cortex, dorsal lateral prefrontal cortex and parietal lobe have been reported to contain signals sensitive to errors, with the ACC most consistently found to be involved in error detection. Using behavioral neurophysiology, various groups have also identified prominent error-detection signals in the ACC of non-human primates 10,11,[13][14][15][16][17][18][19][20] , consistent with human literature. ...
Preprint
Full-text available
Computational models proposed that the medial temporal lobe (MTL) contributes importantly to error-driven learning, though little direct in-vivo evidence for this hypothesis exists. To test this, we recorded in the entorhinal cortex (EC) and hippocampus (HPC) as monkeys performed a task using an error-driven learning strategy, defined as better performance after error relative to correct trials. Error-detection signals were more prominent in the EC relative to the HPC. Early in learning hippocampal but not EC neurons signaled error-driven learning by increasing their population stimulus-selectivity following error relative to correct trials. This same pattern was not seen in another learning task where error-driven learning was not used. After learning, different populations of cells in both the EC and HPC signaled long-term memory with enhanced stimulus-selective responses. These results suggest prominent but differential contributions of EC and HPC to learning from errors and a particularly important role of the EC in error-detection.
... Indeed, when the monitoring system detects emerging conflicts in information processing that may lead to errors occurrence, the ACC orchestrates the need to allocate cognitive resources to task-relevant features and optimize flexible adjustments (Luu et al., 2004;Cohen and Donner, 2013). Neuro-electrical signatures of this process are typically recorded over the medial frontal sites of the scalp as Event-Related Potentials (ERPs) functionally related to conflict (i.e., the N200; Folstein and van Petten, 2008) and to error monitoring (i.e., the error-related negativity or ERN; the Positivity error or Pe; Falkenstein et al., 1991;Gehring et al., 1995;Falkenstein et al., 2000;Pezzetta et al., 2018Pezzetta et al., , 2021. Importantly, these frontocentral potentials reflect common spectral features in the theta rhythm (4-8 Hz), a further endogenous oscillatory biomarker generated by the ACC (Ishii et al., 1999;Luu and Tucker, 2001) that can be recorded over the midline of the frontal cortex during performance monitoring (Cavanagh and Frank, 2014). ...
Article
Full-text available
The performance monitoring system is fundamentally important for adapting one’s own behavior in conflicting and error-prone, highly demanding circumstances. Flexible behavior requires that neuronal populations optimize information processing through efficient multi-scale communication. Non-invasive brain stimulation (NIBS) studies using transcranial magnetic stimulation (TMS) and transcranial electrical stimulation (tES) fields to alter the cortical activity promise to illuminate the neurophysiological mechanisms that underpin neuro-cognitive and behavioral processing and their causal relationship. Here, we focus on the transcranial direct current stimulation (tDCS) and transcranial alternating current stimulation (tACS) that have been increasingly used in cognitive neuroscience for modulating superficial neural networks in a polarity (tDCS) and frequency/phase (tACS) fashion. Specifically, we discuss recent evidence showing how tDCS and tACS modulate the performance monitoring network in neurotypical samples. Emphasis is given to studies using behavioral tasks tapping conflict and error processing such as the Stroop, the Flanker, and the Simon tasks. The crucial role of mid-frontal brain regions (such as the medial frontal cortex, MFC; and the dorsal anterior cingulate cortex, dACC) and of theta synchronization in monitoring conflict and error is highlighted. We also discuss current technological limitations (e.g., spatial resolution) and the specific methodological strategies needed to properly modulate the cortical and subcortical regions.
... Various experimental manipulations have been found to alter the amplitude of the ERN. Notably, amplitude tends to increase when emphasis is placed on accuracy (Gehring et al., 1995) or when the error made is costlier (Gehring et al., 1993). The Pe is another error-related component, thought to reflect conscious awareness of the commission of an error. ...
Article
Full-text available
Exercise may influence components of executive functioning, specifically cognitive control and action monitoring. We aimed to determine whether high level exercise improves the efficacy of cognitive control in response to differing levels of conflict. Fitter individuals were expected to demonstrate enhanced action monitoring and optimal levels of cognitive control in response to changing task demands. Participants were divided into the highly active (HA) or low-active group based on self-reported activity using the International Physical Activity Questionnaire. A modified flanker task was then performed, in which the level of conflict was modulated by distance of distractors from the target (close, far) and congruency of arrows (incongruent, congruent). Electroencephalography (EEG) was collected during 800 trials; trials were 80% congruent, 20% incongruent, 50% close, and 50% far. The error-related negativity (ERN) and error positivity (Pe) were extracted from the difference wave of correct and incorrect response locked epochs, the N2 from the difference wave of congruent and incongruent stimulus locked epochs and the P3 from stimulus locked epochs. The HA group showed a larger Pe amplitude compared to the low-active group. Close trials elicited a larger N2 amplitude than far trials in the HA group, but not the low-active group, the HA group also made fewer errors on far trials than on close trials. Finally, the P3 was smaller in the lowest conflict condition in the HA, but not the low-active group. These findings suggest that habitual, high levels of exercise may influence the endogenous processing involved in pre-response conflict detection and the post-error response.
... One such ERP is the error-related negativity (ERN). This is a negative deflection in neural activity peaking approximately 50-100 ms post-error commission that is reflective of immediate error processing (Gehring et al., 1993(Gehring et al., , 1995(Gehring et al., , 2012Simons, 2010). The ERN may also be considered an index of error salience, such that larger (i.e., more negative) ERNs reflect a higher level of significance of or individual reactivity to that error (Hajcak et al., 2005). ...
Article
Romantic relationships involve a range of positive and negative experiences, from supportive and security-enhancing behaviors to unsupportive interactions involving criticism and dismissiveness. The present study aimed to examine the functional impact of these experiences on reactivity to mistakes, as error salience has key implications for adaptive functioning in areas such as goal-striving and appropriate risk-taking. To this end, a study was conducted in which participants completed the Eriksen Flanker Task (EFT) alone and under romantic partner observation while electrophysiological brain activity related to error salience (the error-related negativity (ERN)) was recorded. Findings indicated that unsupportive, but not supportive, partner behaviors were associated with changes in error salience, furthering the notion that negative relationship experiences have a stronger effect on functioning than do positive ones and highlighting the impact of relationship context on reactivity to mistakes.
... One particular ERP component of interest is the error-related negativity (ERN). The ERN appears as a negative deflection in the response-locked ERP that most commonly occurs when people make errors in reaction time tasks (Falkenstein, Hohnsbein, Hoorman, & Blanke, 1990, 1991Gehring, Coles, Meyer, & Donchin, 1995;Gehring, Goss, Coles, Meyer, & Donchin, 1993). Figure 1 shows the original ERN waveforms reported in Falkenstein et. ...
Thesis
http://deepblue.lib.umich.edu/bitstream/2027.42/63923/1/schwikert_shane_2009.pdf
... The error-related negativity (ERN), an eventrelated potential (ERP) component involved in error detection, measures one aspect of performance monitoring. The ERN is a fronto-centrally maximal negative deflection in the ERP waveform that differentiates erroneous from correct responses within 100 ms of response onset (Falkenstein et al., 1991;Gehring et al., 1995). It functions as an early alarm signal in an action monitoring network that indicates the need to adjust behaviour and increase executive control to remediate mistakes (Botvinick et al., 2001;Gehring et al., 1993;Holroyd & Coles, 2002). ...
Article
Life stress increases risk for multiple forms of psychopathology, in part by altering neural processes involved in performance monitoring. However, the ways in which these stress-cognition effects are influenced by the specific timing and types of life stressors experienced remains poorly understood. To address this gap, we examined how different social-psychological characteristics and developmental timing of stressors are related to the error-related negativity (ERN), a negative-going deflection in the event-related potential (ERP) waveform that is observed from 0 to 100 ms following error commission. A sample of 203 emerging adults performed an ERN-eliciting arrow flanker task and completed an interview-based measure of lifetime stress exposure. Adjusting for stress severity during other developmental periods, there was a small-to-medium effect of stress on performance monitoring, such that more severe total stress exposure, as well as more severe social-evaluative stress in particular, experienced during early adolescence significantly predicted an enhanced ERN. These results suggest that early adolescence may be a sensitive developmental period during which stress exposure may result in lasting adaptations to neural networks implicated in performance monitoring.
... In particular, it would be useful to monitor θ-tACS-induced changes in ERPs during performance tasks focused on cognitive functions. In the case of action monitoring, the suitable ERPs to study would be error-related negativity/error negativity (ERN; Gehring et al., 1995;Luu et al., 2000), or cognitive inhibition (N2; Yeung et al., 2004;Huster et al., 2013). Both, ERN and N2 are associated with the theta band (Luu et al., 2004;Cavanagh and Frank, 2014). ...
Article
Full-text available
Increased frontal midline theta activity generated by the anterior cingulate cortex (ACC) is induced by conflict processing in the medial frontal cortex (MFC). There is evidence that theta band transcranial alternating current stimulation (θ-tACS) modulates ACC function and alters inhibitory control performance during neuromodulation. Multi-electric (256 electrodes) high definition θ-tACS (HD θ-tACS) using computational modeling based on individual MRI allows precise neuromodulation targeting of the ACC via the medial prefrontal cortex (mPFC), and optimizes the required current density with a minimum impact on the rest of the brain. We therefore tested whether the individualized electrode montage of HD θ-tACS with the current flow targeted to the mPFC-ACC compared with a fixed montage (non-individualized) induces a higher post-modulatory effect on inhibitory control. Twenty healthy subjects were randomly assigned to a sequence of three HD θ-tACS conditions (individualized mPFC-ACC targeting; non-individualized MFC targeting; and a sham) in a double-blind cross-over study. Changes in the Visual Simon Task, Stop Signal Task, CPT III, and Stroop test were assessed before and after each session. Compared with non-individualized θ-tACS, the individualized HD θ-tACS significantly increased the number of interference words and the interference score in the Stroop test. The changes in the non-verbal cognitive tests did not induce a parallel effect. This is the first study to examine the influence of individualized HD θ-tACS targeted to the ACC on inhibitory control performance. The proposed algorithm represents a well-tolerated method that helps to improve the specificity of neuromodulation targeting of the ACC.
Article
Providing Reinforcement Learning (RL) agents with human feedback can dramatically improve various aspects of learning. However, previous methods require human observer to give inputs explicitly (e.g., press buttons, voice interface), burdening the human in the loop of RL agent’s learning process. Further, providing explicit human advise (feedback) continuously is not always possible or too restrictive, e.g., autonomous driving, disabled rehabilitation, etc. In this work, we investigate capturing human’s intrinsic reactions as implicit (and natural) feedback through EEG in the form of error-related potentials (ErrP), providing a natural and direct way for humans to improve the RL agent learning. As such, the human intelligence can be integrated via implicit feedback with RL algorithms to accelerate the learning of RL agent. We develop three reasonably complex 2D discrete navigational games to experimentally evaluate the overall performance of the proposed work. And the motivation of using ErrPs as feedbacks is also verified by subjective experiments. Major contributions of our work are as follows, (i) we propose and experimentally validate the zero-shot learning of ErrPs, where the ErrPs can be learned for one game, and transferred to other unseen games, (ii) we propose a novel RL framework for integrating implicit human feedbacks via ErrPs with RL agent, improving the label efficiency and robustness to human mistakes, and (iii) compared to prior works, we scale the application of ErrPs to reasonably complex environments, and demonstrate the significance of our approach for accelerated learning through real user experiments.
ResearchGate has not been able to resolve any references for this publication.