A Neural Attachment Model and Algorithmic Social Intelligence

Add to your list(s) Download to your calendar using vCal

• Abbas Edalat, Imperial College London
• Thursday 06 June 2013, 16:00-17:00
• G29, Mech Eng .

If you have a question about this talk, please contact .

Host: Dean Petters

Please note that coffee will be served at 3:15 at room 123 (School of Computer Science), followed by the one-hour seminar at 4pm.

Social Intelligence is the ability to manage and negotiate complex social relationships and environments based on empathy and compassion. Our capacity for Social Intelligence is rooted in the type of attachment we developed as children with our primary caregivers, which sculpts the key stable patterns in our implicit memory and learning that become the basis of our outlook and conduct in the social world.

I will introduce the notion of strong, or multiply learned, attractor patterns of Hopfield networks to model attachment types and behavioural prototypes. Solution of the mean field equations of the stochastic Hopfield model shows that the capacity of the network to retrieve a strong pattern is proportional to the square of the multiplicity of the strong pattern, which supports the idea of strong attractor patterns to model attachment and behavioural types. I will also explain how attachment types can change in the network as a result of the emergence of new strong attractor pattern.

The neural model provides conceptual motivation for an algorithm, developed in the past three years, to help insecurely attached individuals. The so-called self-attachment algorithm is an extension of attachment theory. Whereas the standard theory deals with bonds between individuals, self-attachment proposes to create an internal self-bonding between the ``Inner Child’’ and the ``Adult Self’’ within the same individual.

The idea is to equip these individuals with a secure attachment type by non-invasive protocols that construct neural circuits generally presumed to correspond to secure attachment. These protocols, using imagination and proactive techniques emulate the interaction of a “good enough” parent with a child in various contexts in order to regulate strong emotions and maximise inner joy and contentment.

I will finally report on the promising results from a number of case studies, in particular in the treatment of chronic depression and anxiety resistant to other types of therapy.

This talk is part of the Computer Science Departmental Series series.

This talk is included in these lists:

• Computer Science Departmental Series
• Computer Science Distinguished Seminars
• G29, Mech Eng
• computer sience

Note that ex-directory lists are not shown.