MIT and Sakana AI framework cuts evaluation costs for self-improving coding agents
🤖 AI-generated content — The title and summary were produced automatically by artificial intelligence, without human editorial review.
MIT and Sakana AI’s SIFT framework uses a separate language model to rank candidate coding agents before costly benchmark tests. One run reached 35.1% accuracy on Polyglot in under five hours, using 42 CPU hours and about $150 in API credits.
Continue in the app — vote & join in ➔Source: VentureBeat · via ahirlevel.hu