Positional Information

RASL Pull · Sprint S-05 — Parallel Learning

You're reading this because RASL is blocked on: why do token positions matter — how does the model know step 3 comes after step 2?

Positional encoding is why transformers can distinguish order. In RASL's state tracker, action step numbers serve the same role — step 4 failing after step 3 succeeded means something different than step 1 failing. Return to mini-llm-lab/tiny_llm/transformer.py and continue building. Read only what unblocks you.

🔨 Lab: positional encoding in transformer.py 🔁 Positional order → RASL state_tracker step ordering 📋 View full parallel plan →