Table 1.
Parameter values and definitions of parameters and abbreviations.
| Brief description | Value | |
|---|---|---|
| Basic BACON parameters | ||
| Nctx | Number of representable attributes (kinds of cortical attribute cells) | 1 K |
| Natr | Number of attributes per context | 100 |
| F | Number of attribute cells innervating each DG cell | 60 |
| K | Number of DG & CA3 winners during representation creation and CA3 winners during recall | 60 |
| Ndg | Number of DG cells (and of their dedicated CA3 followers) | 100 K |
| Bnew | BRep below which a new representation gets made | -5 |
| Badd | BRep at which newly observed attributes get associated with a representation | 15 |
| Bcnd | BRep at which conditioning becomes possible (i.e., conditionability is >0) | 2 |
| BmxCnd | BRep at which conditionability becomes maximal | 12 |
| BmxF | BRep at which fear expression is maximal | 4 |
| α | Increment in synaptic weight on amygdala cell due to US if conditionability is maximal | 1/60 |
| Parameters especially relevant to systems consolidation | ||
| Nptc | Number of attributes of a context that are particulars | 80 |
| Ncat | Number of attributes of a context that are categoricals | 20 |
| Natr = (Nptc + Ncat) | Total number of attributes per context | 100 |
| πptc = (Nptc/Natr) | Proportion of attributes of a context that are particulars | 0.8 |
| Bcat | BRep at which known categorical attributes become associated with a contextual representation | 5 |
| OcatHet | Proportion of cat attribute overlap of unrelated contexts in different categories | 0.2 |
| OptcHet | Proportion of ptc attribute overlap of unrelated contexts in different categories | 0.2 |
| OptcHom | Proportion of ptc attribute overlap in unrelated contexts in the same category | 0.2 |
| PoCat | Average proportion of attributes known at the time that an established context’s representation became cortical. | 0.85 |
| ZoRec | Average number of attributes of established reps known prior to systems consolidation | 85 |
| κ | Ceiling on proportion of the Nptc Unc’s (“unclassified” attributes—i.e., not known to be categoricals) that can get associated with a cortical representation | 0.3 |
| κo | Ceiling on proportion of Uncs that can get associated with a cortical representation if the context’s category cannot be determined. | 0.8 |
| ϵ | parameter controlling increase of kappa when a great deal is known about a context | 20 |
| γ | Factor determining weight of particulars in determining context similarity | 0.6 |
| δ | Amount by which BRep of winner must exceed that of runner-up if new rep is made when winner BRep falls below Bnew | 5 |
| Additional abbreviations (alphabetical) | ||
| BRep | Bayesian weight of evidence for a representation | |
| Cnd | Conditionability of a context (which is a function of BRep) | |
| Cur | A Current attribute | |
| Fef | Fear expression factor | |
| OptcAB | Proportion of overlap of particular attributes of contexts A and B or any other two contexts | |
| OcatAB | Proportion of overlap of categorical attributes of contexts A and B or any other two contexts | |
| OatrAB | Proportion of overlap of all attributes of contexts A and B or any other two contexts | |
| Rec | A Recalled attribute | |
| Rep A, B, etc. | Representation A, B, etc. | |
| Unc | An Uncategorized context–one whose category is not known | |
| Zcom | Number of attributes in common between a set of Current and Recalled attributes | |
| Zcur | Number of (Current) attributes so far observed in a session | |
| Zrec | Number of attributes of a context that are recalled. | |
| Zo | Number of current attributes observed at the time that a representation is created. |