According to Anthropic (@AnthropicAI), the company has announced the release of open-source interpretability tools, specifically designed to work with open-weights AI models. As detailed in their official communication, these tools enable developers and enterprises to better understand, visualize, and debug large language models, supporting transparency and compliance initiatives in AI deployment. The tools, accessible via their GitHub repository, offer practical resources for model inspection, feature attribution, and decision tracing, which can accelerate AI safety research and facilitate responsible AI integration in business operations (source: Anthropic on Twitter, May 29, 2025).