Synthetic Persona Pretraining: Alignment from Token Zero
- Published
- Source
- arXiv
- Paper number
- 898
- Field
- Machine Learning
- arXiv ID
- 2608.13482
Key points
- It proposes persona binding, which attaches value reflections to pretraining documents to install a persona early and then binds it to the assistant identity during post-training.
- Injecting from the first token improves constitutional compliance and leads to more aligned choices even on moral dilemmas not covered during training.
- If the same reflection is injected only at the end of pretraining, the effect is weak and value priorities do not change.
- The effect grows as the pretraining budget increases, from 1.7B/100B to 3B/500B.
- Late injection alone was enough for jailbreak defense, but deep value alignment required early injection.
Paper links
External research summaries. These are not HDATF publications or measured product results.