Repository navigation
Conversation
measure.py builds both levels with Pi0Config(pi05=True); the labels pi0-tiny and pi0-base name openpi's model class, not the model. The 100 ms row is a full-size pi0.5 forward, as E43, E46 and experiments/README.md already say. Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
Member
Author
|
Superseded by the PR for docs/e10-pi05-amendment (same change, without attribution lines). |
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
E10's
measure.pybuilds both of its policy levels withPi0Config(pi05=True, action_horizon=10), but its findings, results and first amendment label thempi0-tinyandpi0-baseand never say thatpi05=Trueis set. That label invites reading the 100 ms full-size row as pi0.This adds correction 3 to
experiments/e10-vla-forward-cost/AMENDMENT.md, dated 2026-10-06: both rows are pi0.5 forwards, the full-size one with random weights. E43, E46 andexperiments/README.mdalready call it pi0.5. No number, prediction or verdict changes.PROTOCOL.mdandFINDINGS.mdare left as written, as the amendment rule requires.🤖 Generated with Claude Code