Complementary Learning Systems

By Randall C. O’Reilly, Rajan Bhattacharyya, M. Howard, Nicholas A. Ketz. 2014

287 citations recorded by the source corpus.

Subfields: Cellular and Molecular Neuroscience, Cognitive Neuroscience

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Cellular and Molecular Neuroscience

In Cognitive Neuroscience

Citations in the corpus

Continual Learning by Modeling Intra-Class Variation (2022, methodology): “Interestingly, memory-based methods also serve as an effective model of complementary learning systems (McClelland et al., 1995; O’Reilly et al., 2014; O’Reilly & Norman, 2002) where episodic memories are retained by regular experience replay.”
Complementary learning systems within the hippocampus: A neural network modeling approach to reconciling episodic memory with statistical learning (2016, methodology): “Although CLS has been successful in accounting for many empirical findings [2], it has not been used to address an important type of learning: apart from memorizing individual experiences rapidly and learning regularities across those experiences over long periods of time, we can also learn regulari”
Continual Learning, Fast and Slow (2021, background): “The brain functionalities involve many components [6] and understanding its mechanism is still an open research topic.”
Strong Evidence for Pattern Separation in Human Dentate Gyrus (2016, result): “These results, therefore, demonstrate that representations for similar stimuli overlap in neocortical regions but are less overlapping in the DG (McClelland et al., 1995; O’Reilly et al., 2014).”
Neural Programmer-Interpreters (2015, background): “ded performance in old tasks. The learning of new tasks may require that the network weights change substantially, so care must be taken to avoid catastrophic forgetting (Mccloskey & Cohen, 1989; OReilly et al., 2014). Using NPI, one solution is to ﬁx the weights of the core routing module, and ”
A Growing Long-term Episodic & Semantic Memory (2016, background): “However, many neural models experience catastrophic interference when being trained on new domains, where the new learning overrides or corrupts the network’s earlier knowledge.”
SupportNet: solving catastrophic forgetting in class incremental learning with support data (2018, background): “The second theory is the complementary learning system (CLS) theory (Mcclelland et al., 1995; O’Reilly et al., 2014), which suggests that human beings extract high-level structural information and store the high level information in a different brain area while retaining episodic memories.”
What Learning Systems do Intelligent Agents Need? Complementary Learning Systems Theory Updated. (2016, unknown): “This proposal has been captured in models of the role of the hippocampus in recognition memory for speci ﬁ c items and in sensitivity to context and co-occurrence of items within the same event or experience [28 – 36].”
If deep learning is the answer, what is the question? (2020, background): “One popular model suggests that mammals have evolved to solve continual learning by using complementary learning systems in the hippocampus and neocortex 116,119,120 .”