Docs: disambiguate "overlap" in the dual-domain L3 examples

[ChaoWao]

Summary

dual_domain_overlap overlaps in domain membership — chip 1 belongs to
both domains — but its module docstring said only "with overlapping domains",
which reads equally as temporal overlap. Combined with the directory name,
that is a real trap: it led a reviewer to conclude the example demonstrates
concurrent/pipelined execution and to propose changes on that basis.

Two claims were also factually wrong:

• dual_domain_overlap's README credited the reduce→affine ordering to
  TensorMap's implicit same-buffer.addr edge. Those stages sit in separate
  synchronous worker.run calls, so the run boundary orders them and that
  edge carries no weight here. The per-domain isolation actually comes from
  reduce_out / affine_out being keyed by (domain, chip).
  ffn_tp_parallel is the example where the implicit edge is load-bearing,
  and is now linked as such.
• domain_rank_map's README pointed at dual_domain_overlap as "the case
  where both domains are live across the same DAG". It is not — it gives each
  domain its own run too. Both domains are live simultaneously only in
  domain_rank_map's own inspection pass, which runs no collective.

Changes:

• Rewrite the dual_domain_overlap module docstring: state that the overlap
  is structural rather than temporal, and name the actual sharp edge (chip 1
  writing a distinct result per domain).
• Add an Execution shape section to the README covering the three
  synchronous runs, the domain lifetime rule (release() on with exit is a
  non-blocking mark; the backend release lands after the run's completion
  wait), and why the allreduce must be submitted as a group.
• Correct the two claims above.

Doc-only; no behavior change.

Testing

• ast.parse on main.py; docstring content verified (it is user-visible
  via argparse(description=__doc__) with RawDescriptionHelpFormatter)
• README heading structure and the #execution-shape anchor checked
• Simulation tests pass — not run; no code changed
• Hardware tests pass — not applicable

Follow-up from the review of #1565.

[coderabbitai]

Important

Review skipped

Auto incremental reviews are disabled on this repository.

Please check the settings in the CodeRabbit UI or the .coderabbit.yaml file in this repository. To trigger a single review, invoke the @coderabbitai review command.

⚙️ Run configuration

Configuration used: Organization UI

Review profile: CHILL

Plan: Pro Plus

Run ID: 6ad6a631-28e9-4983-8a32-f46d2da833e0

You can disable this status message by setting the reviews.review_status to false in the CodeRabbit configuration file.

Use the checkbox below for a quick retry:

• 🔍 Trigger review

📝 Walkthrough

Walkthrough

Documentation updates clarify inspection passes, overlapping domain membership, synchronous execution, domain lifetimes, per-domain outputs, and collective peer grouping in the L3 examples.

Changes

L3 domain overlap documentation

Layer / File(s)	Summary
Domain overlap execution documentation examples/workers/l3/domain_rank_map/README.md, examples/workers/l3/dual_domain_overlap/README.md, examples/workers/l3/dual_domain_overlap/main.py	Documents non-collective inspection, overlapping membership with separate contexts and outputs, three synchronous runs, domain release timing, and allreduce and affine-stage grouping.

Estimated code review effort: 1 (Trivial) | ~5 minutes

Poem

A rabbit hops through domains two,
With notes on what the workers do.
No concurrent race, just steps in line,
Separate results neatly align.
Peer groups gather, then bunnies cheer!

🚥 Pre-merge checks | ✅ 5

✅ Passed checks (5 passed)

Check name	Status	Explanation
Title check	✅ Passed	The title clearly and concisely describes the documentation change clarifying that “overlap” refers to domain membership.
Description check	✅ Passed	The description directly explains the documentation clarifications, factual corrections, and testing performed for the changeset.
Docstring Coverage	✅ Passed	No functions found in the changed files to evaluate docstring coverage. Skipping docstring coverage check.
Linked Issues check	✅ Passed	Check skipped because no linked issues were found for this pull request.
Out of Scope Changes check	✅ Passed	Check skipped because no linked issues were found for this pull request.

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands.}

[coderabbitai]

Actionable comments posted: 1

🤖 Prompt for all review comments with AI agents

Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.

Inline comments:
In `@examples/workers/l3/dual_domain_overlap/main.py`:
- Around line 10-20: Update the module docstring to describe chip 1 as belonging
to both communication domains and receiving an independent ChipDomainContext in
each separate synchronous run. Remove wording such as “at once” and “holds two
independent” that implies simultaneous context liveness, while preserving the
membership and rank details.

🪄 Autofix (Beta)

Fix all unresolved CodeRabbit comments on this PR:

• Push a commit to this branch (recommended)
• Create a new PR with the fixes

---

ℹ️ Review info

⚙️ Run configuration

Configuration used: Organization UI

Review profile: CHILL

Plan: Pro Plus

Run ID: 2ff9c9e1-cc96-4936-abba-02ae02a2f66b

📥 Commits

Reviewing files that changed from the base of the PR and between 14f44a7 and 737c9f1.

📒 Files selected for processing (3)

• examples/workers/l3/domain_rank_map/README.md
• examples/workers/l3/dual_domain_overlap/README.md
• examples/workers/l3/dual_domain_overlap/main.py