Исследователи компании Apple представили ReALM: искусственный интеллект, способный “видеть” и понимать контекст экрана.

 Apple Researchers Present ReALM: An AI that Can ‘See’ and Understand Screen Context

Natural Language Processing (NLP) and Reference Resolution in AI

Reference resolution is a big challenge in NLP. It’s about figuring out what a word or phrase refers to in a text. This is important for understanding different types of context, like in conversations or on-screen information.

Practical Solutions and Value

Researchers are improving large language models (LLMs) to better understand references, especially for non-conversational content. Models like MARRS and ReALM are tackling this by reconstructing the screen using parsed entities, tagging important parts, and fine-tuning the LLM to outperform existing models like GPT-3.5 and GPT-4. This provides a practical reference resolution system.

ReALM: An AI that Can ‘See’ and Understand Screen Context

Apple researchers have developed ReALM, an innovative approach that uses LLMs to understand references by encoding entity candidates as natural text. This model performs almost as well as the state-of-the-art LLM, GPT-4, despite having fewer parameters. It’s a great choice for practical reference resolution, even for on-screen references.

AI Solutions for Business Evolution

Use AI to transform your company by redefining work processes. Identify automation opportunities, set KPIs, choose AI solutions, and implement gradually for lasting impact on business outcomes.

AI Sales Bot from itinai.com/aisalesbot

Discover the AI Sales Bot designed to automate customer engagement 24/7 and manage interactions across all customer journey stages. It’s a practical AI solution for redefining sales processes and customer engagement.

List of Useful Links:

AI Lab in Telegram @aiscrumbot – free consultation

Apple Researchers Present ReALM: An AI that Can ‘See’ and Understand Screen Context


Twitter –  @itinaicom

Полезные ссылки: