Oxford Centre for Computational Neuroscience

Professor Edmund T. Rolls

Computational neuroscience theories of brain function and behaviour


Brain Computations


Cerebral Cortex





The Noisy Brain



Emotion Explained



Memory, Attention, and Decision-Making



Neuroculture


Neural Networks and Brain Function

Overview: Rolls has developed with colleagues computational neuroscience theories that are closely linked to his discoveries made in neuronal recording, functional neuroimaging, and patient investigation studies. An aim of the theories has been to go beyond empirical discoveries about brain function to providing a  biologically based framework for understanding how the brain performs it computations.  This understanding in turn leads to better understanding of and potential treatments for many disorders of brain function. Rolls' computational neuroscience approach has included analyses of how information is represented in the brain (508), of stochastic dynamics in the brain (B9), and of how memory, visual face and object recognition, and emotion systems operate in our brains. Key summary descriptions are in B15 and B12.


Principles of Operation of the Cerebral Cortex (B12, 640, 639, 581).


Brain Computations: what computations are performed, and how they are performed, in different brain systems (B15). Rolls has made key contributions to understanding what computations are performed, and how they are performed, in primates including humans for many of the systems described in this book.


A theory of how neocortical pyramidal cells can implement the learning of new categories, attractor memory, and top-down recall and attention (640).

 

A theory of emotion, and pleasure, and reward; and the principles of their implementation in the brain (B5, B11, B13, 273, 520, 148, 364, 428, 509, 526, 533, 534, 552, 579).


A theory of motivation (557).


In this context, a key principle in primates including humans is that reward value and emotional valence are represented in the orbitofrontal cortex (and amygdala), whereas before that, the representations are about objects and stimuli independently of value, in the inferior temporal visual cortex, the primary taste cortex in the insula, and in the olfactory cortex (B11, B13, B14). This provides for the separation of emotion from perception.


In this framework, the dopamine neurons are seen as receiving their information from brain regions such as the orbitofrontal cortex, via the ventral striatum and habenula (572, B11, B13, B14). Further, orbitofrontal cortex neurons encode reward value and hence emotion, independently of goal-related actions. The orbitofrontal cortex provides reward-related information to the cingulate cortex for action-outcome learning, and to the basal ganglia for habit-based responses (B11, B13, B14, 579, 606).


The roles of the emotional and the reasoning systems in decision-making (497, 518, B5, B10, B11, B13) including in economic decision-making (B11, 600).

 

A biased activation theory of top-down attentional and cognitive control (339, 488, 520, 530, B11, B12, 603).

 

A theory and model of hippocampal operation and episodic memory (111, 125, 136, 163, 186, 200, 205, 258, 266, 268, 300, 306, 307, 309, 345, 370, 403, 411, 415, 433, 453, 479, 521, 527, 529, 539, 545, 550, 584, B12, B15). This remains the only quantitative theory of the storage and recall of memories in the hippocampo-cortical system.


A theory for how hippocampal spatial view cells are involved in memory and navigation (584, 594, 539, B12, 612, 633, B15).


A theory and model of the generation of time in the hippocampal memory system. Entorhinal cortex time ramping cells produce through a competitive network hippocampal time cells, providing neuronal mechanisms to encode the order of events (605). The theory shows how cells could be generated that show 'replay' and 'reverse replay' (605).


A theory and model of coordinate transforms in the dorsal visual system using a combination of gain modulation and slow or trace rule competitive learning. The theory starts with retinal position inputs gain modulated by eye position to produce a head centred representation, followed by gain modulation by head direction, followed by gain modulation by place, to produce an allocentric representation in spatial view coordinates useful for the idiothetic update of hippocampal spatial view cells (612). This is useful in navigation using spatial view cells when the view details are obscured (B15, 633).


A theory of how spatial view cells and hippocampal attractor networks are involved in the art of memory (using the method of loci) (571, 595).

 

A theory and model of invariant visual object recognition in the ventral visual system (162, 179, 192, 226, 245, 275, 277, 280, 283, 290, 304, 312, 396, 406, 414, 446, 455, 473, 485, 516, 535, 536, 554, 589639, B12, B15).

 

A theory of invariant global motion recognition in the dorsal visual system (413).

 

A theory of the utility of the stochastic dynamics of networks in the brain for decision-making and many other aspects of our behaviour, including creativity (404, 463, 477, 483, 491, 500, 502, 504, 507, 513, 518, 540, B9, B11, B12, B15). The stochastic dynamics arises from the almost random (Poisson) firing times of cortical neurons for a given mean firing rate, and gives rise to the concept of The Noisy Brain (B9).

 

Theories of how alterations in the stability of cortical attractor networks can account for the symptoms of schizophrenia (431, 436,  450, 490, 503, 629, 631), of obsessive compulsive disorder (449), of depression (559, 572, 626), of ADHD (629), of normal aging (540, 613), and for creativity (585). The theories (B12, B13, B14, B15) have implications for treatment, and are complemented by neuroimaging investigations (538, 541, 563, 564, 565, 583, 585, 587, 588, 590, 591, 592, 596, 602, 615, 626, 629, B14).


A theory of avalanches in the brain and the temporal variability of connectivity (630, 629).


A non-reward attractor theory of depression (559, 572, B13), supported by altered connectivity and activation of the orbitofrontal cortex in depression (564, 583, 588, 590, 591, 592, 596, 602, 615, 616, 623, 626, B13, B14, B15), and a model of non-reward computation in the orbitofrontal cortex (562).

 

The roles of cortical attractor networks in short-term memory and top-down attention (294, 295, B6, 47, 360, 372, 379, 391, 410, 520, B8, 523, 530, B12, B15, 640).

 

The design of neural networks in the cortex by genetic evolution (284, B12).

 

Separate limbic systems for emotion and memory, but no single limbic system (531). A conceptual framework for understanding the cingulate cortex, and how it forms a part of  different limbic systems (606).

The representation of information in the brain using a neuronal firing rate code (172, 196, 204, 225, 227, 321, 255, 419, 474, 508, 231, 265, 329, 348, 351, 369, 517, 193, 197, 257, 407, 553, 561, B12, B15).


The computational utility of diluted connectivity in attractor, pattern association, and competitive networks in the cerebral cortex (504, 515, 545, 550, B12, B15).


A possible implementation of syntax in the brain (537, B15).


A higher order syntactic thought (HOST) approach to consciousness (239, 341, 355, 398, 422, 432, 456, 493, 497, 525, 618, B11, B12, B13).

A theory of the relation between the mind and the brain utilizing levels of explanation in which causality operates within but not between levels (618, 632, 637).


Biological underpinnings of art and aesthetics (B10, 492, 509, 532, 556, 574).