Repository navigation
Publish skill prerequisites: access, reach and recovery - #38
Merged
Merged
Conversation
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Publish two native MuJoCo PiPER environments for skill prerequisites: guarded hinged-box access and a mounted rake that brings an out-of-workspace shoe closer. The reader and recorded 3D replay preserve all 16 frozen attempts, complete 2 ms contact traces, authored policies, source pin and artifact hashes.
The result is deliberately limited: the lid guard avoids two baseline contacts while completion remains 2/4; the rake reaches 3/4 intermediate retrieval checkpoints, but no held-out attempt retains the final lift. The paper diagnoses the grip failures, reports native CPU timing separately from export, and distinguishes privileged simulation state, pre-mounted tool capability and startup hazard detection from learning or physical qualification.
The new archive is additive. All 21 prior paper bodies and historical artifacts remain unchanged. Scoped CORS on immutable binary assets supports the Hugging Face replay; the browser replays recorded states, while Python runs fresh physics. The project now pins its deployment CLI.
Validation: 57 research tests passed; deterministic build and catalog/link validation passed; Wrangler deployment dry-run passed. Complete frozen lab: 74 files / 25,726,353 bytes, unchanged from source commit
756306a2beda60ac54567971626080a56227e781. Training source and benchmark are reviewed in Dvidia-Inference/dvidia-training#18.