Installs into .claude/skills of the current project.
Are you the author of Diayn Discriminator Intrinsic Reward?
Add the live security badge to your README. It updates with every re-scan.
[](https://www.skillsdirectory.com/skills/vectorspacelab-diayn-discriminator-intrinsic-reward)
---
name: diayn_discriminator_intrinsic_reward
description: Compute DIAYN discriminator log-probabilities, cross-entropy diagnostics, and intrinsic rewards from skill-conditioned state predictions.
---
# DIAYN Discriminator Intrinsic Reward
Use this skill when implementing or validating the DIAYN reward term `log q_phi(z|s) - log p(z)`. It is appropriate for tiny deterministic fixtures, reduced recovery experiments, and full implementations that need stable reward diagnostics. Do not use it to add downstream task rewards.
## Inputs
- A batch of discriminator logits or scores shaped as `batch x num_skills`.
- True sampled skill ids for each row.
- A fixed `log_prior`, usually `-log(num_skills)` for a uniform prior.
## Outputs
- Stable log-softmax values.
- Selected `log q_phi(z|s)` values.
- Intrinsic reward values.
- Mean reward, cross-entropy loss, and accuracy.
## Workflow
1. Validate batch and label dimensions.
2. Convert logits to log probabilities with a max-subtracted log-softmax.
3. Select log probability for the true skill on each transition.
4. Compute pseudo-reward by subtracting `log_prior`.
5. Report loss and accuracy for discriminator training diagnostics.
## Validation
Run `python scripts/diayn_reward.py --demo`. Tests verify that confident correct logits produce higher rewards and valid discriminator metrics.
## Limitations
This skill computes the reward and discriminator diagnostics only. It does not update neural network weights; pair it with a policy or discriminator optimizer in a larger recovery harness.
