docs: document reward logging contract

[TATP-233]

Summary

• Document the Manager-Based reward logging contract in the English and Chinese logging guides.
• Clarify that reward/<term> is the UniLab per-term diagnostic, Train/mean_reward is the native RSL-RL aggregate return metric, and Episode_Termination/<term> is a distinct termination diagnostic.
• Record that Episode_Reward/<term> is intentionally removed and may still appear in historical TensorBoard event files.
• Make no package version or runtime behavior changes.

Linked Work

• Issue: follow-up to Remove duplicate Episode_Reward per-term TensorBoard metrics #1569
• Merged implementation: fix: remove duplicate reward episode metrics #1570
• Parent roadmap (when applicable): not applicable
• Roadmap declared base (when applicable): not applicable
• Milestone: none
• Base branch: main

Validation

• make test-all passed on the final local head before this PR was created or updated
• Additional task-specific validation listed below

make test-all was not run because this is a docs-only change; the repository's docs-specific gate from docs/sphinx/AGENTS.md was used instead.

Commands actually run:

uv run pytest tests/scripts/test_check_docs.py -q
cd docs/sphinx
UNILAB_DOCS_SKIP_AUTODOC=1 uv run --no-project --with-requirements requirements.txt sphinx-build -b html -n source build/html
git diff --check

Results:

• 19 passed in 0.36s
• Sphinx HTML build succeeded
• No whitespace errors

Remote CI route:

• Base main: current-head CI is the remote gate for this PR.

Impact

• Backend impact: none
• Platform impact: none
• Training effect expected: no — documentation only

Artifacts

• W&B: none
• benchmark result: none
• video / screenshot: none
• ONNX / checkpoint: none

Checklist

• Added or updated tests where needed
• Updated docs if behavior or workflow changed
• Linked the driving issue
• Noted any follow-up work explicitly