Sakana AI's self-improving model reaches 70% win rate but struggles with shared blind spots
Original titleSakana AI's self-improving model gets better at solving and judging, but can't catch its own blind spots
- Outlet
- The Decoder
- Published
- Importance
- 4/5
Summary
Sakana AI's MASS method enables one language model to improve itself by solving tasks, judging results, and refining workflows, reaching nearly 70 percent win rate against the base model after two rounds. However, the model can reinforce errors it fails to detect across all three roles.
Also reported by 2 outlets
Original excerpt
Sakana AI's MASS method lets one language model improve itself by solving tasks, judging results, and refining workflows, reaching a win rate of nearly 70 percent against the base model after two rounds. But when the model misses the same errors across all three roles, it can mistake them for progress and reinforce them during training. The article Sakana AI's self-improving model gets better at solving and judging, but can't catch its own blind spots appeared first on The Decoder.