#software engineering

New research shows that powerful long-context language models can reach 50.8% accuracy on the SWE-Bench software engineering benchmark without relying on complex tool scaffolding, simplifying LM agent design.

Windsurf launches SWE-1, a family of AI models engineered to support complete software development workflows with advanced context understanding and cost-effective performance.

Salesforce AI introduces SWERank, a novel retrieve-and-rerank framework that delivers precise and scalable software issue localization with significantly reduced costs compared to existing agent-based methods.