Anthropic Investigates Unintended Actions by Its AI Models
Anthropic has shared an investigation into AI model actions that were not intended. The checks covered its own evaluations and how its teams use the models internally.
সহজ শব্দকোষAnthropic, evaluations
- Anthropic
- An AI company based in the United States that makes the Claude chatbot.
- evaluations
- Tests that check how an AI system behaves before and after people start using it.
বার্নিং পটেটো স্কোর: 35%