Files
dela 3a322aa8dc feat: land superpaper v1 notes scaffold
Add the ledger schema, class router, lint codes, ingest/build
pipeline, and three work-tree examples: align derivation,
superfig delegation, and supertensor delegation.
2026-08-17 10:01:30 +08:00

9 lines
292 B
Markdown

# A One-Step Predictor (fixture)
We predict $\hat y = f_\theta(x)$ in one forward pass. The loss
$L(\hat y, y)$ is computed after the prediction; it is not a station
on the forward path.
The scale $1/\sqrt{d}$ is introduced so that the variance of the
inner product does not grow with $d$.