feat(soxl): add isolated learning and completed-result consumption - #485
Merged
Merged
Conversation
Co-Authored-By: Codex <noreply@openai.com>
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
SOXL can run an explicitly requested, bounded learning experiment using its approved retained input package, and can consume an already completed aggregate result without touching the data, runner or model again. Source lineage stays separate from the pinned research dependencies. The adapter retains the baseline and risk limits and cannot produce promotion or live authority.
When parameter comparisons have identical aggregates at each fixed cost scenario, result consumption returns a terminal hold for the isolated learning profile. Different aggregates require review without proposing a parameter change; incomplete or altered results are rejected. The file reader binds the expected SHA to one bounded read and rejects symlinks.
Validation: 143 targeted tests passed, plus Ruff and diff checks. Coverage includes synthetic input and risk/event regressions, CLI/result validation, previously completed real three-parameter/nine-cost-scenario study (no improvement), and saved-cycle hold/reuse/defer integration. The real study was not rerun for this publication and is not an independent OOS or shadow result. No private inputs/results, production task identity, schedules, dependencies or trading configuration are added by this PR.