Back to timeline

Research · 1970

Reverse-mode automatic differentiation

Seppo Linnainmaa described how to compute the derivatives of a composite function in one backward pass over the computation graph.

Why it matters

The algorithm that makes training deep networks possible appeared in a problem about accumulated rounding error.

Linnainmaa wanted to estimate how the rounding errors of individual operations affect the final result. The solution is to expand the accumulated error as a Taylor series and traverse the graph in reverse. The cost of that pass is proportional to the forward computation regardless of the number of parameters, and that property is what makes training large models feasible.

Event record

Event date
1970
Timeline date
Event date
Verification
Sources gathered automatically · September 17, 2026
Lines
ID
evt-0117

The master's thesis was defended in 1970 at the University of Helsinki; it is in Finnish and is not available online. The published account of the method is the 1976 BIT paper.

Sources

Related events