Search

Google DeepMind Unveils AlphaEvolve, a Coding Agent Designed to Reduce AI Hallucinations

Google said AlphaEvolve is an evolutionary coding agent for general-purpose algorithm discovery and optimisation.

Advertisement
Highlights
  • The system adds automated evaluators that verify answers by AI models
  • Google said AlphaEvolve has enhanced the efficiency of its data centres
  • AlphaEvolve uses Gemini Flash and Gemini Pro in tandem
Google DeepMind Unveils AlphaEvolve, a Coding Agent Designed to Reduce AI Hallucinations

Google is working on a user interface for AlphaEvolve

Photo Credit: Google

Google DeepMind announced a new artificial intelligence (AI) coding agent on Wednesday that can enhance the capabilities of AI models. Dubbed AlphaEvolve, it is designed to discover and optimise algorithms across complex computing and mathematical tasks. The powerful AI system is built on the Mountain View-based tech giant's Gemini models, and it combines outputs generated by large language models with automated evaluators to ground the responses in reality and reduce the risk of hallucinations. Beyond this, the system is also said to have shown potential in solving and optimising mathematical problems.

Google DeepMind Introduces AlphaEvolve Coding Agent

In a blog post, DeepMind detailed the new technology it has been working on. AlphaEvolve is not an AI model, instead, it is a complex AI system with agentic capabilities. One of the primary functions the system performs is algorithm discovery and optimisation.

AI models, at a fundamental level, are a series of code. These code bases process and compile information, break it down, and use probabilistic algorithms to generate an output. However, since AI systems are highly complex, their code bases are massive. This large size often causes optimisation and efficiency-based issues. AlphaEvolve can help with that, the company said.

alphaevolve working Google AlphaEvolve

AlphaEvolve structure
Photo Credit: Google

 

AlphaEvolve uses automated evaluation metrics, and using these parameters, it verifies, runs, and scores responses generated by AI models. Google said this method allows the system to quantifiably assess responses from multiple AI models and reduce the risk of hallucinations. Additionally, the system can also fix and improve code that allows such hallucinations.

The tech giant said that AlphaEvolve has improved the efficiency of Google's data centres, chip design, and AI training processes. Interestingly, it was also able to improve the training of its own base LLM. In one case, it discovered a new scheduling method that recovers around 0.7 percent of Google's global compute resources — a massive gain when applied across the company's massive infrastructure.

Since AlphaEvolve works with code bases and algorithms, it is also said to have high potential in different areas of mathematical problem solving, the company said. It is said to have discovered a faster method to multiply 4x4 complex matrices, beating a solution that had stood for more than 50 years. In tests across 50 open mathematical problems, AlphaEvolve matched the current best solutions in most cases, and even improved on them in about 20 percent of problems, the post added.

For the latest tech news and reviews, follow Gadgets 360 on X, Facebook, WhatsApp, Threads and Google News. For the latest videos on gadgets and tech, subscribe to our YouTube channel. If you want to know everything about top influencers, follow our in-house Who'sThat360 on Instagram and YouTube.

 
Show Full Article
Please wait...
Advertisement

Related Stories

Popular Mobile Brands
  1. Vivo X200 FE Compact Smartphone Launched With 6,500mAh Battery
  2. Nothing Phone 3a Pro 5G Long Term Review: A Blend of Style, Speed, and Power
  3. Kubera OTT Release Reportedly Revealed: Where to Watch Dhanush Starrer Movie Online?
  4. AI+ Pulse, AI+ Nova 5G India Launch Timeline, Design and Colours Revealed
  5. Nothing Phone 3 Full Specifications Surface Ahead of Its July 1 Debut
  6. Samsung Galaxy Book 5 Pro Review: Your Work Companion
  7. Asus V470VA All-in-One PC Review: A Stylish AiO PC For Everyday Work
  8. Death Stranding 2: On the Beach Review - Ode to a Dying Planet
  9. Amazon Diagnostics Service Launched in These 6 Cities in India
  10. Tecno Spark Go 2 India Launch Date, Key Features Announced
  1. Samsung Exynos 2500 SoC With Up to 15 Percent Improved CPU Performance, Xclipse 950 GPU Launched
  2. Vivo X200 FE With 6,500mAh Battery, MediaTek Dimensity 9300+ SoC Launched: Specifications
  3. Tecno Spark Go 2 India Launch Date Set for June 24; Key Features, Colourways Teased
  4. Amazon Diagnostics Launched in India, Offers Doorstep Sample Collection in Six Cities
  5. Honor Magic V5 Teased to Measure 8.8mm Thick; Colours, RAM and Storage Details Revealed
  6. Apple Reportedly in Talks to Acquire Perplexity AI to Strengthen Its Artificial Intelligence Efforts
  7. AI+ Pulse, AI+ Nova 5G Confirmed to Launch in India in July; Design, Colour Options Revealed
  8. iPhone, iPad Pages Updated With Energy Labels to Comply With New EU Regulations
  9. China’s Restrictions on Rare Earth Mineral Exports to Reportedly Adversely Impact Supply Chain Jobs in India
  10. Boat Airdopes Prime 701 ANC With Up to 50 Hours Total Playback Time Launched in India: Price, Features
Gadgets 360 is available in
Download Our Apps
App Store App Store
Available in Hindi
App Store
© Copyright Red Pixels Ventures Limited 2025. All rights reserved.
Trending Products »
Latest Tech News »