Spinning Up Deep Rl
When to use
Use this skill for spinning up deep rl when you need to study a reinforcement learning concept through implementation. The expected result is an RL learning plan with experiments.
Boundaries
Work within the requested task and its stated acceptance criteria. Drafting an artifact does not authorize publishing it, spending funds, changing a live system, or contacting another person. Identify any such action separately before taking it.
Inputs
Inspect the target algorithm, environment, baseline, and evaluation metric. Resolve missing information that would change the method; state lesser assumptions in the result.
Method
- Diagnose. Identify the runtime boundary, data shape, dependencies, deployment path, and the constraint that dominates the design.
- Decide. Connect the mathematical objective to the code and test one assumption at a time.
- Produce. Build an RL learning plan with experiments from the inspected material; keep assumptions distinguishable from observed facts.
Decision rules
- Compare implementation choices on correctness, compatibility, performance, operability, and reversibility. Choose against the actual workload.
- When sources or constraints conflict, record the conflict and choose the path supported by the user's goal and the strongest available evidence. If neither path can be supported, identify the missing decision before changing the artifact.
Domain rules
- Check compatibility, performance, security, and operational cost for the chosen implementation.
- Include migration and rollback steps when contracts or persisted data change.
Verification
Compare training behavior with a reproducible baseline. Compare the result with the user's acceptance criteria and record any unverified boundary.
Supply the artifact, migration notes when needed, measured or tested behavior, and a rollback condition for risky changes.
Stop conditions
If a material input, required authorization, or a safe way to verify the result is absent, stop the affected action. Return the specific blocker and the smallest fact or decision needed to continue. Do not report an unrun check as passed.
Output
Provide an RL learning plan with experiments. Include the decisive evidence and actual verification result. Name any artifact location and unresolved issue that affects its use.
