Framework for analyzing learning dynamics in low-rank RNNs via overlap space decomposition. Distinguishes loss-visible overlaps (determine activity/output/loss) from loss-invisible overlaps (encode training history). Enables understanding of why functionally equivalent networks learn differently. Activation: low-rank RNN learning, RNN overlap space, loss-visible invisible, RNN gradient descent dyn
no risky patterns detected
Type: Agent skill (SKILL.md) · Category: Science, Health & Education · Source: https://github.com/hiyenwong/ai_collection/tree/HEAD/collection/skills/ai-ml/low
| Capabilities (fewer powerful permissions = higher) | 50/50 |
| Maintenance | 21/35 |
| Documentation | 15/15 |
Always review files before installing. How scores work. Request a correction