Skip to content

mind: retire jax-likelihood-smoke-pins-stale — fixed before it was filed - #335

Merged
Jammy2211 merged 1 commit into
mainfrom
claude/jax-likelihood-smoke-pins-stale-im98b0
Aug 26, 2026
Merged

mind: retire jax-likelihood-smoke-pins-stale — fixed before it was filed#335
Jammy2211 merged 1 commit into
mainfrom
claude/jax-likelihood-smoke-pins-stale-im98b0

Conversation

@Jammy2211

Copy link
Copy Markdown
Collaborator

Overview

/start_dev draft/bug/autolens/jax_likelihood_smoke_pins_stale.md found no reproducible bug left in the prompt. Retired unstarted: no GitHub issue, no active.md entry, no development.

Verified state

autolens_workspace_test main 8be3d59, Smoke Tests run 32922456742 (2026-08-26 02:24 UTC), green on Python 3.12 and 3.13:

Running 22 listed scripts
  scripts/imaging/jax_likelihood/lp.py ...              PASS (9.8s)
  scripts/interferometer/jax_likelihood/rectangular.py  PASS (21.1s)
  scripts/interferometer/jax_likelihood/mge.py ...      PASS (12.6s)
  ... 22/22 PASS

The prompt's headline — "every workspace_test smoke sweep reads 20/24 until this lands" — is now 22/22.

Disposition of each claim

Prompt claim Current state
imaging/jax_likelihood/lp.py pin -1.34797827e09 is double Repinned to -6.74165366e08 in 197ce6e (autolens_workspace_test#257), the same day the prompt was filed
fb1aefe0b sits unmerged in an orphan worktree Not orphaned — it is the head SHA of PyAutoLabs/PyAutoLens#700, merged 2026-08-17, two days before the prompt called it unmerged. 197ce6e is the paired re-pin the prompt asked for
interferometer/jax_likelihood/rectangular.py misses pin by 0.0124% Pin -3164.286252 unchanged and green (re-verified in 99d63b3 incl. the TransformerNUFFT cross-check). The 08-19 miss was control-run environment drift, not a stale pin
interferometer/jax_likelihood/mge.py stale pin Repinned by the same 197ce6e; a PositionsLH-doubling casualty, not independent drift
multi_dataset/jax_likelihood/mge.py stale pin Never a pin problem. Out of the smoke gate since #262 (2026-08-22) for the XLA compile stall tracked in autolens_workspace_test#245; its pin is untouched since the multimulti_dataset rename

Changes

  • complete/2026/08/jax-likelihood-smoke-pins-stale.md — the completion record, with the original prompt folded in
  • draft/bug/autolens/jax_likelihood_smoke_pins_stale.md — removed
  • complete/index.md — regenerated (1143 records)
  • dashboard.md / dashboard.html — regenerated; the prompt no longer renders as pickable backlog

lifecycle.py check and lifecycle.py index --check both report OK.

Findings kept in the record

  • Adjudicate an "orphan" worktree from the remote, not from the worktree. A local worktree over a merged-and-deleted branch is indistinguishable from one holding real unlanded work. Resolving the SHA against the remote's merged PRs settles it in one call.
  • Do not group symptoms into a "family" by folder. Three of the four scripts were one root cause (the 2x-last PositionsLH penalty). Acting on the "stale-pin family" framing would have widened tolerances on pins that were already correct.
  • A pass/total target is not a stable goal. Two entries left smoke_tests.txt for timing reasons, so the sweep reads 22/22 without any pin having been touched for that.

Follow-ups

  • Local debris only: ~/Code/PyAutoLabs-wt/positions-lh-penalty-accumulation is a stale worktree over a merged branch — a /repo_cleanup item, not code work.
  • Latent, owned elsewhere: multi_dataset/jax_likelihood/mge.py's pin has never been verified against a post-PyAutoLens#700 library. Re-derive it as part of autolens_workspace_test#245's re-enable. Deliberately not filed as its own prompt — it has no standalone trigger.

🤖 Generated with Claude Code

https://claude.ai/code/session_01Egf8PFc5UKJfVWY1jbBKGU


Generated by Claude Code

/start_dev research on draft/bug/autolens/jax_likelihood_smoke_pins_stale.md
found no reproducible bug left in it. Retired unstarted: no GitHub issue, no
active.md entry, no development.

Verified against autolens_workspace_test main 8be3d59, Smoke Tests run
32922456742 (2026-08-26 02:24 UTC), green on Python 3.12 and 3.13: 22/22 PASS,
including all three of the named scripts still in the gate. The prompt's
"20/24 until this lands" is now 22/22.

Per claim:

- imaging/jax_likelihood/lp.py     - repinned -1.34797827e09 -> -6.74165366e08
  in 197ce6e (autolens_workspace_test#257), the same day the prompt was filed.
- fb1aefe0b "orphan worktree"      - not orphaned. It is the head SHA of
  PyAutoLens#700, merged 2026-08-17, two days before the prompt called it
  unmerged; 197ce6e is the paired re-pin the prompt asked for.
- interferometer/.../rectangular.py - pin -3164.286252 unchanged and green
  (re-verified in 99d63b3 incl. the TransformerNUFFT cross-check). The 08-19
  miss was control-run environment drift, not a stale pin.
- interferometer/.../mge.py        - repinned by the same 197ce6e; a
  PositionsLH-doubling casualty, not independent drift.
- multi_dataset/.../mge.py         - never a pin problem. Out of the smoke
  gate since #262 (2026-08-22) for the XLA compile stall tracked in
  autolens_workspace_test#245; its pin is untouched since the multi ->
  multi_dataset rename.

The completion record keeps the reusable findings: adjudicate an "orphan"
worktree from the remote's merged PRs rather than from the worktree's own
state, and do not group symptoms into a "family" by folder — three of the four
scripts were one root cause.

Follow-ups are local debris (a stale worktree on the laptop, /repo_cleanup)
and one latent item owned by autolens_workspace_test#245's re-enable.

Dashboard regenerated in the same commit.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01Egf8PFc5UKJfVWY1jbBKGU
@Jammy2211
Jammy2211 merged commit 4965640 into main Aug 26, 2026
4 checks passed
@github-actions
github-actions Bot deleted the claude/jax-likelihood-smoke-pins-stale-im98b0 branch August 26, 2026 20:07
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants