AlphaEvolve
AlphaEvolve is an artificial intelligence model developed by Google DeepMind.[1][2] It aims to autonomously discovers and refines algorithms through a combination of large language models (LLMs) and evolutionary computation.
Built on top of DeepMind’s existing Gemini models, Google reports to have successfully generated and made breakthroughs in complex algorithmic and scientific problems, such as for matrix multiplication, that have been deployed across Google's computing ecosystem.[3] Google claims that across a selection of 50 open mathematical problems, the model was able to rediscover state-of-the-art solutions 75% of the time and discovered improved solutions 20% of the time, for example advancing the Kissing Number Problem.[1]
Unlike domain-specific predecessors like AlphaFold or AlphaTensor, AlphaEvolve is designed as a general-purpose system. It can operate across a wide array of scientific and engineering tasks by automatically modifying code and optimizing for multiple objectives. Its architecture allows it to evaluate code programmatically, reducing reliance on human input and mitigating risks such as hallucinations common in standard LLM outputs.[2]
See Also
[edit]References
[edit]- ^ a b "AlphaEvolve: A Gemini-powered coding agent for designing advanced algorithms". Google DeepMind. 2025-05-14. Retrieved 2025-05-14.
- ^ a b Whitwam, Ryan (2025-05-14). "Google DeepMind creates super-advanced AI that can invent new algorithms". Ars Technica. Retrieved 2025-05-17.
- ^ Nuñez, Michael (2025-05-14). "Meet AlphaEvolve, the Google AI that writes its own code—and just saved millions in computing costs". VentureBeat. Retrieved 2025-05-14.