Skip to main content
. Author manuscript; available in PMC: 2020 Jun 1.
Published in final edited form as: Trends Cogn Sci. 2019 Apr 25;23(6):510–524. doi: 10.1016/j.tics.2019.03.006

Figure 4 (Key Figure): Summary of functional cell types and their putative interactions during recognition memory encoding and retrieval.

Figure 4 (Key Figure):

Summarized are four cell types (filled colored circles): visually selective (VS) cells, memory selective (MS) cells, novelty-sensitive dopamine neurons, and choice neurons. There are two types of MS cells: Novelty and familiarity selective (NS and FS). Arrows indicate direction of information flow, but do not indicate monosynaptic connections. Anatomical areas are indicated in dashed boxes. The theoretical concepts/models (gray boxes) discussed are: i) the Hippocampus-VTA/SN loop model, which proposes that hippocampal novelty signals excite dopamine neurons in the VTA/SN, which in turn release dopamine in the hippocampus, which leads to long-lasting plasticity. ii) the sparse coding memory model, which suggests that distinct neurons encode semantic and episodic aspects of declarative memories in a sparse but distributed manner. iii) the temporal context model, which suggests that the neural state present at encoding is re-instated at retrieval. iv) mnemonic accumulation, which suggests that neurons exist that integrate memory signals to make choices. v) the balance-of-evidence model, which describes how the difference between two mnemonic integrators can be used to make metacognitive confidence judgments about memories. Jointly, the body of single-neuron experiments discussed provides evidence for key aspects predicted by these models, including novelty cells that exhibit rapid single-trial learning, representations of memory strength that predict subjective confidence judgments, sparse coding of semantic memories, reinstatement of neural state by VS cells during retrieval, and representations of memory-based choices that are putative mnemonic accumulators.