Sakana AI Unveils MASS: Recursive Self-Improvement via Multi-Agent Self-Supervision, Boosting Win Rates to 69.9%
When model capabilities exceed human expert evaluative bandwidth, achieving Recursive Self-Improvement (RSI) without external verifiers or compilers becomes the paramount frontier of AGI. While reinforcement learning with verifiable rewards (RLVR) thrives in formal code and math, open-ended research and complex reasoning tasks lack scalar ground truth, and solitary models evaluating their own outputs suffer from homogeneous confirmation bias. Sakana AI introduces MASS (arXiv:2610.12176, project: sakana.ai/mass), an RSI framework anchored in Multi-Agent Self-Supervision. Driven by a single base model (such as Qwen3.6-27B), MASS tasks the model with concurrently acting as task solver, workflow optimizer, and self-evaluator. The framework alternates between evolutionary multi-agent workflow discovery and supervised fine-tuning (SFT) over synthesized trajectories. Across two successive MASS self-evolution cycles, performance per output token increases by 1.2x to 1.6x across open-ended research suites (MLR-Bench, ScienceAgentBench, DSBench, AstaBench). Head-to-head win rates against the base model surge from 53.9% in round 1 to 69.9% in round 2, while a student model trained on multi-agent traces decisively beats a single-agent baseline trained on 1.4x more tokens.
Key Takeaways
- Sakana AI introduces MASS, using multi-agent self-supervision on a single base model to achieve recursive self-improvement
- Win rate against base model climbs from 53.9% to 69.9% over two iterations, boosting token reasoning efficiency by 1.2x-1.6x
- Distillation from multi-agent trajectories outperforms single-agent baselines trained on 1.4x more tokens
