summaryrefslogtreecommitdiff
path: root/BASELINES.md
diff options
context:
space:
mode:
Diffstat (limited to 'BASELINES.md')
-rw-r--r--BASELINES.md5
1 files changed, 5 insertions, 0 deletions
diff --git a/BASELINES.md b/BASELINES.md
index 9c3a320..de7dfe8 100644
--- a/BASELINES.md
+++ b/BASELINES.md
@@ -15,6 +15,11 @@ weak in-house reimplementations from defining the state of the art.
The audited implementations and completed results are in `RESULTS.md`:
- BP, FA, and DFA share the SDIL forward architecture and data order;
+- convolutional hierarchical FA (`hfa`) mirrors the ResNet residual DAG with
+ independent random 3x3 feedback tensors, exact parameter-free shortcut
+ adjoints, and local ReLU/BatchNorm Jacobians; it never copies or reads a
+ downstream forward convolution and is the mandatory baseline for any learned
+ hierarchical extension;
- direct node perturbation uses a causal target on every hidden update and no
learned vectorizer;
- the no-traffic, `P=0` SDIL backbone is learned direct feedback by node