framework
active
framework:multiple-cortical-inputs-to-hippocampus-extensionMultiple Cortical Inputs to Hippocampus Extension
Extension of TEM-t to handle conjunctions of more than two brain regions with linear (not exponential) scaling in hippocampal neuron count.
Neighborhood — ranked by edge-count
Papers (1)
paper
Concepts (1)
concept
- Experimental evidence that hippocampal neurons respond to more than two task variables; motivates multi-input extension of TEM-t.
Frameworks (1)
framework
- TEM-Transformer (TEM-t)extendsThe transformer version directly analogous to TEM, introduced in this paper, offering dramatic performance improvements.
Related by similarity (8)
cosine ≥ 0.65 · no typed edgeEntities in the same semantic neighborhood but without a typed relation to this one — candidates for new edges or unrecognized duplicates.
- Fundamental theoretical claim motivating DAS, attributed to Smolensky/Rumelhart/McClelland.
- Theory that hippocampus provides an index binding together cortical patterns across different brain regions; TEM-t is shown to instantiate this.
- Graph representation model of hippocampus; related work category 2.
- Brain region corresponding to the memory/conjunction component in TEM; memory neurons resemble place cells.
- Cited as empirical evidence that confabulation is universal in biological cognition, not AI-specific
- Original discovery of place cells; the neural phenomenon TEM-t memory neurons resemble.
- Original paper on hippocampal indexing theory; TEM-t is shown to instantiate this theory.
- Speculative hypothesis about how cortical transformer instantiation avoids requiring hippocampus.