Self-Play Pretraining With Zero Data: LMs Learn to Generate Their Own Training Data
ctjlewis · x · 2026-09-26
Researchers introduce Self-Play Pretraining with Zero Data, a proof-of-concept where two models start from random initialization: a generator proposes programs for a universal Turing machine and a learner trains on their outputs, with no real data used at all. Zero-shot validation loss on images, text, audio, and melodies decreases predictably with self-play compute, and the learner develops in-context learning capabilities. The authors stress it is an early-stage proof of concept.
More from Research
- Mobile AI Factory: agentic pipeline builds on-device SLMs that beat 5-shot frontier models — kalyan_kpl · 2026-11-11
- Amazon paper: single agents beat multi-agent setups on enterprise SOPs at 1.4–4.7x lower cost — rohanpaul_ai · 2026-11-11
- Amazon paper: multi-agent setups burn 1.4-4.7x tokens for small gains; single agent is the default — rohanpaul_ai · 2026-11-11
- MIT's Buehler builds "recursive meta-intelligence": hundreds of AI agents inhabit self-created worlds to study material failure — ProfBuehlerMIT · 2026-11-11
- Explosion launches Ellf directory: 52 NLP templates plus multi-agent annotation workflow — _inesmontani · 2026-11-11
- STABLE Optimizer Accepted at NeurIPS 2026: Taming Continual Learning Forgetting at the Optimizer Level — v_lomonaco · 2026-11-11