8000
2 Factor Learning Rule doesn't handle graded spikes in the post-trace #909
Labels
Comments
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
{{ message }}
Describe the bug
When using a 2F learning rule with a neuron that produces graded spikes, the signal passed back is only ever the impulse value of the neuron and not the graded spike value. In lava/src/lava/magma/core/model/py/connection.py in the class LearningModelConnectionFloat, at approximately line 1364 in the function _process_post_spikes, the magnitude of the output value is ignored and only the impulse is added to the post trace.
This is in contrast to the function _process_pre_spikes in the same file (about line 1306), which permits adding graded spikes to the pre-trace.
To reproduce current behavior
Use the 2Factor learning rule with a neuron that produces a graded spike. Observing the eligibility trace shows that it outputs a constant value.
Expected behavior
The eligibility trace, y1, should take on the graded spike value.
Environment (please complete the following information):
Additional context
The text was updated successfully, but these errors were encountered: