The field of neuroscience is experiencing a rapid development of new neurotechnologies, which enable large-scale recordings of up to thousands of neurons simultaneously, possibly from multiple brain regions, with cell type specificity, during complex behavior, and the ability to selectively manipulate neurons in the population. While this has had significant benefits to the field, resulting in a wealth of new data and the ability to perform targeted manipulations of neural systems, it has revealed that many of the field’s former theories are unable to account for the new data or inform the reliable use of those manipulations. These former theories were, for the most part, developed to account for the activity of a small number of neurons, with selective attention to interpretable neurons, and population-level theories which involved idealized collections of these interpretable units, or highly abstract models/descriptions, have been unable to account for the complexity observed with the new methods. As a result, the field has seen the development of a number of candidate theories: Deep learning framework: Explanations of the neural computations underlying cognition should focus on objective functions, learning rules, and architectures (Richards et al., 2019) Hopfieldian view: Cognition can be well explained by transformations between or movement within representational spaces that are implemented by neural populations. To be contrasted with a “Sherringtonian view,” that cognition can be well explained by point-to-point communication between neurons organized into circuits (Barack & Krakauer, 2021) Neuroconnectionism: Artificial neural networks (ANNs) are a highly suitable computational language to model the brain computations underlying cognition: sufficiently abstract to be computationally tractable and reproduce cognitive functions, while still being close enough to biology to relate to, implement, and test neuroscientific hypotheses (Doerig et al., 2023) Phylogenetic refinement: Biologically plausible theories of behavior (including cognition) can be constructed by following a method of “phylogenetic refinement,” whereby they are progressively elaborated from simple to complex according to phylogenetic data on the sequence of changes that occurred over the course of evolution (Cisek, 2019) Inside-out view: Rather than appealing to cognitive terms, theories of neural phenomena should be framed in terms of intrinsic patterns which are selected and grounded by action and prediction (Buzsáki, 2019) A problem-centric view would suggest considering the coverage of each theory—what problems does each cover, and where is the overlap? It is interesting to note that these are all very general theories (“views” or conceptual “frameworks”), which specify how explanations should be framed but have little to say about specific phenomena (Levenstein et al., 2023). Thus, they have the potential for large and overlapping problem coverage (all refer to neural phenomena of “cognition”), with a high degree of explanatory depth but a low degree of empirical accuracy. In order to solve problems requiring a notable degree of accuracy, these theories would need to be combined with other, more specific, theories about the phenomena of relevance. For example, in addition to the standard systems consolidation theory (SCT, Box 1), there have been multiple developments and alternative theories that all pertain to phenomena of memory, sleep, and hippocampal-cortical interactions: Multiple trace transformation theory: Memory traces are formed in the hippocampus and neocortex at the time of encoding. Hippocampally dependent, episodic or context-specific memories transform into semantic or gist-like versions that are represented in extra-hippocampal structures. To the extent that episodic memories are retained, they will continue to require the hippocampus, but the hippocampus is not needed for the retrieval of semantic memories (Winocur et al., 2010) Memory indexing theory: The role of the hippocampus is to form and retain an index of neocortical areas activated by experiential events. The reactivation of the stored hippocampal memory index (e.g., during sleep) will reactivate the associated array of neocortical areas, resulting in a memorial experience and, through cortico-cortical plasticity, establishing a cortically based memory trace (Teyler & DiScenna, 1986) Synaptic homeostasis hypothesis: During sleep, spontaneous activity renormalizes synaptic strength across neuronal populations. Memories are consolidated through comprehensive sampling of statistical regularities incorporated in neuronal circuits over a lifetime (Tononi & Cirelli, 2014) This illustrates multiple kinds of pluralism relevant in neuroscience—the field maintains theories with different degrees of explanatory depth, in which broader theories have a wider domain but require more domain-specific theories, as well as theories with a similar degree of coverage of problems in highly overlapping domains.