News from the AI & ML world
DeeperML
Amber Hoak,@Microsoft Research
//
Microsoft Research is pioneering a new data science approach called Semantic Telemetry to gain a deeper understanding of how users interact with AI systems. This innovative project aims to classify human-AI interactions and understand user behavior, which is crucial for developing and supporting increasingly high-value use cases for generative AI. Semantic Telemetry is a new way to measure how people interact with AI systems.
Semantic Telemetry employs a unique methodology using large language models (LLMs) to generate meaningful categorical labels from chat log data. This process involves iteratively prompting an LLM to generate summaries of conversations and refine classification labels. The resulting classifications are then used to label new, unstructured chat log data, enabling a comprehensive analysis of user interactions and the value derived from AI tools.
ImgSrc: www.microsoft.c
References :
Semantic Telemetry employs a unique methodology using large language models (LLMs) to generate meaningful categorical labels from chat log data. This process involves iteratively prompting an LLM to generate summaries of conversations and refine classification labels. The resulting classifications are then used to label new, unstructured chat log data, enabling a comprehensive analysis of user interactions and the value derived from AI tools.
ImgSrc: www.microsoft.c
Share:
References :
- Source Asia: How real-world businesses are transforming with AI — with more than 140 new stories
- Microsoft Research: Semantic Telemetry: Understanding how users interact with AI systems
Classification:
- HashTags: #SemanticAI #ResponsibleAI #HumanAI
- Company: Microsoft
- Target: AI users, Businesses
- Product: AI Systems
- Feature: Semantic Telemetry
- Type: AI
- Severity: Informative