Why There Are Complementary Learning Systems in the Hippocampus and Neocortex.¶
McClelland, J. L., McNaughton, B. L., & O'Reilly, R. C. (1995). Why There Are Complementary Learning Systems in the Hippocampus and Neocortex.: Insights from the Successes and Failures of Connectionist Models of Learning and Memory. Psychological Review, 102(3), 419-457.
Cited by¶
3 citations across 3 artifacts.
Each citation links to the sentence it supports in the citing article.
Primes¶
- Memory Consolidation
- The experience-replay buffer in deep reinforcement learning is a direct port of hippocampal replay, and the continual-learning literature explicitly invokes the neuroscience consolidation framework; in the other direction, complementary-learning-systems theory used connectionist modelling to refine the neuroscience hypothesis about why a two-store architecture exists at all, so the port ran both ways.
This sourceComplementary-learning-systems theory using connectionist modelling to explain the two-store (fast hippocampal / slow neocortical) consolidation architecture.
- The experience-replay buffer in deep reinforcement learning is a direct port of hippocampal replay, and the continual-learning literature explicitly invokes the neuroscience consolidation framework; in the other direction, complementary-learning-systems theory used connectionist modelling to refine the neuroscience hypothesis about why a two-store architecture exists at all, so the port ran both ways.
- Pattern Completion (Filling the Incomplete)
- Multiple different prompts may converge to similar completions (generalization); similar prompts may diverge into different completions if they tap different parts of the learned distribution (multi-modality). - The gestalt closure principle: The training process biases completions toward fluent, coherent, semantically sensible text rather than random or incoherent continuations — a direct analog of the perceptual bias toward complete, well-formed patterns.
This sourceFoundational theory of complementary learning systems explaining how hippocampal pattern completion (rapid, episodic) and cortical pattern generalization (slow, semantic) coexist without catastrophic interference.
- Multiple different prompts may converge to similar completions (generalization); similar prompts may diverge into different completions if they tap different parts of the learned distribution (multi-modality). - The gestalt closure principle: The training process biases completions toward fluent, coherent, semantically sensible text rather than random or incoherent continuations — a direct analog of the perceptual bias toward complete, well-formed patterns.
- Two-Store Architecture
- In neuroscience the hippocampus (fast, episodic, interference-prone) couples to the neocortex (slow, integrated, statistical) with consolidation during sleep, the canonical complementary-learning-systems account; hippocampal damage leaves consolidated long-term knowledge intact while eliminating new acquisition.
This sourceCanonical complementary-learning-systems account of fast hippocampal and slow neocortical stores with consolidation.
- In neuroscience the hippocampus (fast, episodic, interference-prone) couples to the neocortex (slow, integrated, statistical) with consolidation during sleep, the canonical complementary-learning-systems account; hippocampal damage leaves consolidated long-term knowledge intact while eliminating new acquisition.
Verification¶
This reference passed the adversarial substantiation pipeline: it was checked to exist and to support the claim it is attached to. See how references were verified.
Registry ID ref:d46ad9777ae5 · see in the full table