Conference proceeding
Analyzing Composability in a Sparse Encoding Model of Memorization and Association
2008 IEEE 7TH INTERNATIONAL CONFERENCE ON DEVELOPMENT AND LEARNING, pp.180-185
IEEE International Conference on Development and Learning
01/01/2008
DOI: 10.1109/DEVLRN.2008.4640826
Abstract
A key question in neuroscience is how memorization and association are supported by the mammalian cortex. One possible model, proposed by Valiant[10], uses sparse encodings in a sparse random graph, but the composability of operations in this model (e.g. an association triggering another association) has not previously been evaluated. We evaluate composability by measuring the size of "items" produced by memorization and the propagation of signals through the "circuits" created by memorization and association. While the association operation is sound, the memorization operation produces "items" with unstable size and produces circuits that are extremely sensitive to noise. We therefore amend the model, introducing an association stage into memorization. The amended model preserves and strengthens the sparse encoding hypothesis and invites further characterization of properties such as capacity and interference.
Details
- Title: Subtitle
- Analyzing Composability in a Sparse Encoding Model of Memorization and Association
- Creators
- Jacob Beal - MIT CSAIL, Cambridge, MA 02139 USAThomas F. Knight - MIT CSAIL, Cambridge, MA 02139 USA
- Resource Type
- Conference proceeding
- Publication Details
- 2008 IEEE 7TH INTERNATIONAL CONFERENCE ON DEVELOPMENT AND LEARNING, pp.180-185
- Publisher
- IEEE
- Series
- IEEE International Conference on Development and Learning
- DOI
- 10.1109/DEVLRN.2008.4640826
- eISSN
- 2161-9476
- Number of pages
- 6
- Language
- English
- Date published
- 01/01/2008
- Academic Unit
- Electrical and Computer Engineering
- Record Identifier
- 9984627219702771
Metrics
1 Record Views