Perplexity AI Agent Features Self-Learning Memory Layer
Decrypt · Crypto
Perplexity's new 'Brain' feature is a self-improving memory layer. It tracks past agent actions, successes, and failures. This data is then used to optimize subsequent tasks, making them faster and more cost-effective.