Repository navigation
Conversation
This was referenced Oct 26, 2026
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
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Experiment methods need a maintained home and reproducible inputs
Study methods, trial settings, generated artifacts, and analysis code need a clear
ownership boundary. Repeating an analysis should recover its actual configuration
and input identities without editing a maintained preset or reconstructing an old
working directory.
Add experiment conventions and a study-specific analyzer
Document
experiments/as the home for a question, methods, commands, andstudy-specific analysis. Keep reusable implementation in the shared package and
generated artifacts outside the checkout.
Add a command that builds a frozen-defender response report from retained
measurements using shared NumPy and Matplotlib dependencies. It checks retained producer and
artifact identities, validation selection, and paired held-out results before
writing tables and figures. Outputs retain resolved settings, source evidence,
input references, and hashes.
Research inputs are required for a historical comparison
The study README specifies the input files and reference values expected by the
analyzer. Research data is not included in this change. General reporting,
training launch, and shared training/evaluation capture remain separate work.
Related: #29.