✨ Evaluate stochastic RL pass candidates

[flowerthrower]

🤖 AI text below 🤖

Description

Evaluates repeated candidates for Qiskit RL actions marked as stochastic and retains the candidate with the best configured figure of merit. Candidates are translated to the target basis before scoring; supported scoring failures fall back to minimizing inserted SWAPs.

PredictorEnv.stochastic_action_trials configures the number of attempts and defaults to 20. QiskitSabreMapping and the SabreSwap action introduced by #785 use this path. A failed candidate does not abort the episode: the remaining attempts continue, and an all-failed run leaves the circuit unchanged. Pass and scoring timeouts are re-raised so the timeout layer can truncate immediately rather than continuing unbounded trials. Set the trial count to 1 to retain single-attempt selection.

This is position 9 of the stack. It depends on the timeout PR #789 and is followed by the GNN PR #788.

Validated with repository lint, focused RL checks, and direct pass- and scoring-timeout smokes.

Fixes #667

Checklist

• The pull request only contains commits that are focused and relevant to this change.
• I have updated the documentation to reflect these changes.
• I have added entries to the changelog for noteworthy behavior changes.
• I have added migration instructions to the upgrade guide.
• The changes follow the project's style guidelines and introduce no new warnings.
• I have reviewed my own code changes.

If PR contains AI-assisted content:

• Any agent that created, edited, or submitted GitHub content was explicitly authorized for that scope, as required by our AI Usage Guidelines.
• Every agent-authored or agent-edited public text body begins with the visible disclosure 🤖 *AI text below* 🤖 (titles are exempt).
• I have disclosed AI assistance in the PR description.
• I confirm that I have personally reviewed and understood all AI-generated content, and accept full responsibility for it.

[codecov]

Codecov Report

❌ Patch coverage is 59.57447% with 19 lines in your changes missing coverage. Please review.

Files with missing lines	Patch %	Lines
src/mqt/predictor/rl/actions/qiskit_actions.py	64.2%	10 Missing ⚠️
src/mqt/predictor/rl/predictorenv.py	50.0%	9 Missing ⚠️

📢 Thoughts on this report? Let us know!

[flowerthrower]

🤖 AI text below 🤖

Superseded by #791 after reordering the stack; the implementation remains on the same head branch.