LANGUAGE MODEL UC San Diego / CMU

Backpropagation for Multi-Layer Networks

A way to train networks with multiple layers of neurons by efficiently figuring out how much each individual weight — out of potentially billions — contributed to a mistake, and nudging each in the right direction.

Try demo

Why it matters

Perceptrons could only draw straight lines. Backprop made it practical to stack many layers and learn curved ones — which is to say, to learn anything. It is the load-bearing piece underneath every image model, every language model, and every recommendation system you use.

Core Capabilities

Research
Foundational paper or scientific contribution.
Long Documents
Handles entire codebases, books, and multi-doc RAG.

Context Window

Context window not disclosed.

Availability

API
Not available
Product / App
Not available
Open Source
Not released
Enterprise

Pricing Model

Not disclosed
Pricing not disclosed.

What it feels like

  • Language model from UC San Diego / CMU — see the linked sources below for benchmark and review coverage

Best use cases

  • General-purpose tasks within UC San Diego / CMU's deployment footprint
  • See the model spec and sources block for benchmarked use cases

Not ideal for

  • Tasks far outside the modalities listed in this model's spec
  • Workflows where a more recent successor in the same family scores higher

Rumelhart, D. E. · Hinton, G. E. · Williams, R. J.