Leabra

Leabra is heavily influenced by and contributes to neural network designs and models, including emergent.

Version

v1 · 2026-09-28 · History

Domain-specific #

10341

Domain group

Interdisciplinary & Synthetic

Origin domain

Cognitive Science

Subdomains

Computational Cognitive Neuroscience, Neural Network Models → Cognitive Science

Core Idea

Leabra is treated here as the recurring computerscienceandinformation identity summarized by this source-grounded definition: Leabra is heavily influenced by and contributes to neural network designs and models, including emergent. Leabra stands for local, error-driven and associative, biologically realistic algorithm. It is a model of learning which is a balance between Hebbian and error-driven learning with other network-derived characteristics. This model is used to mathematically predict outcomes based on inputs and previous learning influences. Leabra is heavily influenced by and contributes to neural network designs and models, including emergent.

Scope of Application

• Background. It is the default algorithm in emergent (successor of PDP++) when making a new project, and is extensively used in various simulations.

• Background. The symmetric, midpoint version of GeneRec is used, which is equivalent to the contrastive Hebbian learning algorithm (CHL).

• Background. The activation function is a point-neuron approximation with both discrete spiking and continuous rate-code output.

• Background. Layer or unit-group level inhibition can be computed directly using a k-winners-take-all (KWTA) function, producing sparse distributed representations.

• Implementations. There is also an R version available, that can be easily installed via install.packages("leabRa") in R and has a short introduction to how the package is used.

Clarity

A clear use of Leabra names the carrier, the operative relation, and the conditions under which the source treats the identity as present. The minimal definition is Leabra is heavily influenced by and contributes to neural network designs and models, including emergent. The strongest recognition evidence in the frozen account is: Layer or unit-group level inhibition can be computed directly using a k-winners-take-all (KWTA) function, producing sparse distributed representations.

Manages Complexity

Leabra compresses multiple computerscienceandinformation details into a stable diagnostic relation. The source shows both the central mechanism—hebbian learning is performed using conditional principal components analysis (CPCA) algorithm with correction factor for sparse expected activity levels.—and the practical consequence—automatic scaling is performed to compensate for differences in expected activity level in the different projections. This compression makes cases comparable while leaving parameters, conventions, exceptions, and evidential quality explicit.

Abstract Reasoning

1. Type the carrier. Identify the computerscienceandinformation entities to which the claim applies.
2. State the relation. Use the source-grounded identity: Leabra is heavily influenced by and contributes to neural network designs and models, including emergent.
3. Check operation and conditions. Error-driven learning is performed using GeneRec, which is a generalization of the recirculation algorithm, and approximates Almeida–Pineda recurrent backpropagation.
4. Demand recognition evidence. Layer or unit-group level inhibition can be computed directly using a k-winners-take-all (KWTA) function, producing sparse distributed representations.
5. Test variation.

Knowledge Transfer

Within the home domain. Knowledge about Leabra transfers literally when a new case preserves the same carrier type, relation, and recognition test. It is the default algorithm in emergent (successor of PDP++) when making a new project, and is extensively used in various simulations. The symmetric, midpoint version of GeneRec is used, which is equivalent to the contrastive Hebbian learning algorithm (CHL). Beyond the home domain. No canonical parent is asserted for Leabra.

Neighborhood in Abstraction Space

Leabra sits in a sparse region of the domain-specific corpus (66^th percentile for distinctiveness): few abstractions share its structure, so a faithful description tends to retrieve it precisely.

Family — Unclustered & Miscellaneous (2551 abstractions)

Nearest neighbors

• BCM theory — 0.89
• Radial basis network — 0.87
• Entropy estimation — 0.85
• Downsampling (signal processing) — 0.83
• Residual neural network — 0.83

Computed from structural-signature embeddings · 2026-10-08