fix(eval): load engram students with the memory attached - #222
Merged
Merged
Conversation
The vanilla loader drops _engram.* (the PKM lesson: injected parameters need the injected loader) — the first eval of run-013 was scoring a backbone missing a component it trained with. That number stands as the no-memory CONTROL; the gate needs the table in the forward pass. load_student_with_engram attaches first, loads the full state strictly except fresh-zero engram params (loud on anything else); evaluate_der uses it for engram specs.
ronaldtse
force-pushed
the
fix/engram-eval-load
branch
from
September 29, 2026 08:31
848e029 to
d2740c1
Compare
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
The trained run-013 checkpoint's first eval dropped the memory module (vanilla loader: _engram.* UNEXPECTED — the PKM lesson applies). load_student_with_engram attaches before loading; the eval path uses it for engram specs. The dropped-table number is retained as the no-memory control; the gate needs the table live in the forward pass.