feat: add audited evaluation scoring #70

Merged

TheComputerGuy merged 20 commits from feat/eval-regression-replay into master 2026-10-31 22:09:22 +00:00

Conversation 0 Commits 20 Files changed 15 +3136 -36

TheComputerGuy commented 2026-10-31 21:29:08 +00:00

Owner

Copy link

Closes LilithSemi/chock#44

Closes LilithSemi/chock#44

TheComputerGuy added 20 commits 2026-10-31 21:29:08 +00:00

feat: add audited evaluation scoring 0e5ac0d15e

feat: compare evaluation reports in CI 5bb00e25ae

feat: record and replay evaluation fixtures 6f1ec45e4e

feat: score exact eval state and subagents a6159d9c42

feat: emit composable evaluation rewards 1f19abe497

feat: export trainable RL trajectories 5d015ce07e

fix: score immutable evaluation results dfaa0cee16

feat: persist evaluation rollout artifacts 052971a491

feat: orchestrate grouped RL rollouts c1ff06fd8c

feat: pin evaluation workspace revisions ad3a31f3a7

feat: capture token-level RL traces fbe448c405

feat: record RL rollout provenance 146e89e808

feat: replay Chock-owned tool effects 93fcc1f934

feat: audit evaluation sandbox drivers c607cb8c9b

feat: fingerprint rollout environments 0801dc7716

test: prove replay effects are redacted cd2a49f6b1

feat: bind rollout artifact integrity d64b793e76

feat: bind live rollout provenance 9c741cec2a

fix: preserve live rollout training traces db6ee927a9

fix: make eval replay deterministic end to end 5652789c81

TheComputerGuy merged commit 5652789c81 into master 2026-10-31 22:09:22 +00:00

TheComputerGuy deleted branch feat/eval-regression-replay 2026-10-31 22:09:22 +00:00

Sign in to join this conversation.

Reviewers

No reviewers

Labels

Clear labels   

accessibility

Barrier affecting people with disabilities

  

bug

Something isn't working

  

documentation

Improvements or additions to documentation

  

duplicate

This issue or pull request already exists

  

enhancement

New feature or request

  

good first issue

Good for newcomers

  

help wanted

Extra attention is needed

  

invalid

This doesn't seem right

  

question

Further information is requested

  

wontfix

This will not be worked on

No labels

accessibility

bug

documentation

duplicate

enhancement

good first issue

help wanted

invalid

question

wontfix

Milestone

Clear milestone

No items

No milestone

Projects

Clear projects

No items

No project

Assignees

Clear assignees

No assignees

1 participant

Notifications

Subscribe

Due date

The due date is invalid or out of range. Please use the format "yyyy-mm-dd".

No due date set.

Dependencies

No dependencies set.

Reference

LilithSemi/chock!70

No description provided.