Dynamic Reward Scaling and Normalization
Calculates and shapes rewards for reinforcement learning by applying dynamic scaling based on training progress to balance exploration and exploitation, and normalizing high-value rewards to a specific range to ensure numerical stability.
npx skills add ECNU-ICALK/AutoSkill --skill dynamic-reward-scaling-and-normalization --agent claude-code
Same command for any agent — swap --agent for codex, cursor, copilot.
Weekly change comes from our own snapshots, not the repository page — it measures attention, not adoption.
# Dynamic Reward Scaling and Normalization Calculates and shapes rewards for reinforcement learning by applying dynamic scaling based on training progress to balance exploration and exploitation, and normalizing high-value rewards to a specific range to ensure numerical stability. ## Prompt # Role & Objective Act as a Reinforcement Learning Reward Engineer. Your task is to calculate and shape rewards for a PPO agent, ensuring they promote early exploration and later refinement while maintaining numerical stability. # Operational Rules & Constraints 1. **Dynamic Scaling**: Implement a dynamic scaling factor based on the training phase (current episode vs max episodes). - Early training: Use larger rewards and softer penalties to encourage exploration. - Late training: Reduce scaling to refine decision-making. - Formula: `scaling_factor = 1 - (0.5 * (current_episode / max_episodes))` (linear decay from 1 to 0.5). - Apply this factor to base rewards and penalties. 2. **Reward Normalization**: Apply specific normalization to handle outliers. - If reward is between 101 and 1,000,000,000, scale it to the range [101, 500]. - If reward is between 0 and 100, or if negative, keep it unchange
- Prompt
- Triggers
What does the Dynamic Reward Scaling and Normalization skill do?
Calculates and shapes rewards for reinforcement learning by applying dynamic scaling based on training progress to balance exploration and exploitation, and normalizing high-value rewards to a specific range to ensure numerical stability.
How do I install it?
Run `npx skills add ECNU-ICALK/AutoSkill --skill dynamic-reward-scaling-and-normalization --agent claude-code` — it drops the skill into your project so the agent can pick it up. Swap the --agent value for codex, cursor or copilot if you use one of those.
Where does this skill come from?
From ECNU-ICALK/AutoSkill, a repository with 539 stars. We read it straight from the repository tree rather than a submitted listing, so what you see here is what is actually published.
Is a popular skill a good skill?
Not necessarily. Stars measure attention, not adoption — a repository can trend for a week and be abandoned. That is why we show the weekly change from our own snapshots next to the total, instead of a single flattering number.