Dr Jiaxiang Zhang


Research group:
Cognitive Science
029 208 70471
CUBRIC, Maindy Road

Research summary

My research involves investigation of neural and computational mechanisms of decision-making, learning and action. A central theme is to understand how the human brain integrates cognitive and perceptual processes to regulate behaviour in a dynamic, changing environment. The new understanding of these basic cognitive operations is then used for the examination of cognitive deficits in neurodegeneration and dementia. I use a combination of multimodal neuroimaging (MRI, EEG/MEG, and TMS), behavioural measures (psychophysics), and computational modelling.

For more information, please visit my lab website: http://ccbrain.org

Postdoctoral fellowships and PhD studentships are available. Please contact me for further details.

Teaching summary

UG Year 1 - PS1018 Research Methods in Psychology (practicals)
UG Year 2 – PS2017 Biological Psychology (2015-16), PS2022 Thinking, Emotion & Consciousness (2016-17)
UG Final Year - PS3000 Final year research projects
MSc – PTS507 Neuroimaging of Perception and Action

Selected publications (2014 onwards)


Full list of publications


Media activities


Research topics and related papers

For more information, please visit my lab website: http://ccbrain.org

The deciding brain
Making rapid decisions on the basis of sensory information is a frequent and critical element of human lives. Imagine you are driving towards a traffic light in foggy weather. Since the scene is less visible, it could be difficult to distinguish between the red and green light, and sometimes you may even make a mistake.

We examine how the brain integrates information during decision-making. The accumulation of information is an essential process, because it reduce the noise in sensory information and thereby facilitates more accurate decisions. We use computational model at different levels of complexity to account from behavioural profile (response time and accuracy) to neuroimaging data (fMRI and MEG/EEG). This work helps us to understand different roles of brain regions within the decision processes, and the information flow from perception to action.


The complexity and generality of the decision-making models. All models are capable of capturing basic behavioural statistics such as the RT and the response accuracy. The simple accumulator models and the sequential sampling models are suitable to describe the congregate activity of large neural populations (e.g.,fMRI and EEG/MEG).The more complex model (i.e., the spiking neural network) can be used to account for dynamics of neural circuits (Zhang 2012).

The volitional brain
One can freely choose between action choices that have no apparent difference in their outcomes, yet such behaviour is far from random. This type of voluntary decision involves the formation of intentions, and is associated with widespread frontal–parietal activation.

We investigate the voluntary selection process for choosing which action to respond, and when to act. We test the hypothesis that that the same accumulation-to-threshold mechanism that governs perceptual decision is also involved in voluntary action selection. When all available actions are unambiguous, the internal intentions of selecting potential actions build up over time and compete against each other. By combining computational modelling and fMRI, we establish the brain regions that are associated with the intention accumulation for voluntary action. Recently, we start to investigate patients with abnormal voluntary actions arising from neurodegenerative disease.


In the voluntary action task, participants responded with any one of the three valid actions (left). The intention accumulation process is described by a linear ballistic accumulator model (right). The accumulator activation is linearly accumulated over time with a constant accumulation rate sampled from a normal distribution on each trial, until a response threshold is reached. The model prediction of the accumulation process is associated with increased fMRI responses in a decision work that was maximal in the supplementary motor area and the caudal anterior cingulate cortex (bottom) (Zhang et al. 2012).

The learning brain
Prolonged practice often gradually improves task performance, resulting in higher accuracy and faster responses. We investigate the effects of active learning and passive exposure on visual perception and categorization (Zhang and Kourtzi, 2010). By using high-resolution fMRI and multivariate pattern analysis, we study fine-scale learning-dependent changes in visual categorization (Zhang et al., 2010). We also examine how perceptual learning process interacts with other cognitive processes (e.g., speed-accuracy trade-off) (Zhang and Rowe, 2014).


Participants learn a motion discrimination task in 6 sessions (top). In the first 5 sessions, the participants are trained at two directions (30 and 210◦). In the sixth session, all participants perform the task at two new directions that are not presented in their first 5 sessions (i.e., untrained directions). In each session, participants are instructed to either respond as accurate and possible (accuracy emphasis) or as fast as possible (speed emphasis). Participants become more accurate and faster under different instructions after training, but the learn effect on decision accuracy does not transfer to the untrained direction (Zhang and Rowe, 2014).

The cognitive brain
A hallmark of cognitive control is the ability to regulate goal directed behaviour according to task rules, which coordinate cognitive and motor processes with knowledge of the associations among stimuli, responses, and outcomes. Task rules can be established under different contexts. For example, a rule may be formally identical, and executed equally effectively, when it has been specifically instructed and when it has been chosen from several alternatives. We study how rules are established under different contexts, and whether the context influences the neural representation of a given rule.


Neurons in the frontoparietal cortex may encode rule-specific information when the rule is chosen (red circles), instructed (blue circles), or both (mixed-color circles). Context-independent regions are dominated by neurons that are activated under different contexts, while context-dependent regions also contain neurons that only respond under a particular context. Rule representations provide a regulatory bias of the learned associations between stimuli (S1, S2, and S3) and responses (R1 and R2), necessary to perform a given task. In the diagram, a task rule pertaining to a stimulus dimension of S1 is chosen, and the appropriate response (R1) is evoked under this rule (Zhang et al., 2013).


  • ERC starting grant, €1,487,908 (2017-22). Zhang (PI). Free the mind: the neurocognitive determinants of intentional decision.
  • MRC, £1,865,362 (2017-21). Graham (PI), Lawrence (PI), Jones, Wise, Zhang (Co-I), Mackay, Fillippini, Kordas, and Saksida. Characterising brain network differences during scene perception and memory in young APOE-e4 carriers: multi-modal imaging in ALSPAC.
  • Wellcome Trust ISSF, £38,601 (2016). Singh (PI), Walters, Freeman and Zhang (Co-I). Neuro-physiologically informed models and machine learning classification of task-driven and resting state oscillatory dynamics in Schizophrenia.
  • ARUK Cambridge Network Scholarship, £4k (2014). Zhang. Revealing hidden cognitive deficits in neurodegenerative diseases.

Research group

Cognitive Science

Research collaborators

Prof Kim Graham, Prof Andrew Lawrence, Prof Krish Singh, Prof Petroc Sumner

Prof Sheng Li (Peking), Dr Naoki Masuda (Bristol), Dr James Rowe (Cambridge), Prof Hartwig Siebner (Copenhagen)

Postgraduate research interests

If you are interested in applying for a PhD, or for further information regarding my postgraduate research, please contact me directly (contact details available on the 'Overview' page), or submit a formal application here.

For more information, please visit my lab website: http://ccbrain.org

Current Students

Maciej Szul

Previous students

Emmy Tsang (MRC summer studentship). Now predoctoral fellow at EMBL.
Yuedong Song (Co-supervised with Michael Gordon) at Computer Laboratory, University of Cambridge

Undergraduate education

2003 BEng in Computer Networking, Northwestern Polytechnical University, China

Postgraduate education

2008 PhD, Department of Computer Sciences, University of Bristol
2005 MSc in Advanced Computing (distinction), Department of Computer Sciences, University of Bristol

Awards/external committees

Trainee travel award, Organization for Human Brain Mapping (2014, 2011)
Junior Research Fellow (elected), Wolfson College, University of Cambridge (2011-2014)
Overseas Research Student Award (£50,000, 2006-2008)


Grant review: MRC, BBSRC, FWO (Belgium), ANR (France)

Journal Review: Advances in Cognitive Psychology, Behavior Research Methods, Cognitive Neurodynamics, Current Biology, European Journal of Neuroscience, Experimental Brain Research, Frontiers, International Journal of Robotics and Automation, Journal of Mathematical Psychology, Journal of Neuroscience, Journal of Experimental Psychology: applied, Journal of Pain Research, Neurocomputing, Neuroimage, Neuropsychologia, Psychopharmacology, Psychological Science, Plos One

External examiner: MSc in Clinical Cognitive Neuroscience, Sheffield Hallam University


2015 – Present Lecturer, School of Psychology, Cardiff University, UK
2010 – 2014 Investigator scientist, MRC Cognition and Brain Sciences Unit, Cambridge, UK
2008 – 2010 Postdoctoral research fellow, School of Psychology, University of Birmingham, UK