feat(intervene): read the steering direction from Results (composable…#112
Merged
Conversation
alirezabayatmk
force-pushed
the
feat/results-driven-intervene
branch
from
2057b18 to
6c8a081
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Summary 📝
Makes activation steering and ablation composable:
Intervenecan read its direction fromResultsat run time, so deriving a direction and applying it is one pipeline instead of two with a manual hand-off.Details
Intervenegetsdirection_key/mode/alpha: whendirection_keyis set it reads theSteeringResultan upstream step wrote and builds a steer or ablate intervention at run time. Thefn=API is unchanged (backward compatible).direction_keyis added to the step'sreads, so a pipeline that forgets to produce the direction fails pre-flight with a clearKeyErrorinstead of silently generating without the intervention.Checks