Anthropic, a Google-backed AI startup founded by former OpenAI engineers, has announced a major update to its Claude chatbot. The new Claude 2.1 can now process up to 200,000 tokens at a time, allowing it to analyze immense amounts of text, equivalent to over 500 pages of material.
This doubling of Claude’s previous token limit is an industry first. It allows Claude to drill down into entire code bases, long texts like the Iliad, or large corporate documents with ease.
A major improvement is the 50% reduction in hallucinations (invention information) from Claude 2.1. This should greatly improve reliability.
Our new Claude 2.1 model offers an industry-leading 200,000 token context window, a 2x decrease in hallucination rates, system prompts, tool usage, and updated pricing.
Claude 2.1 is available via API in our console and is powering our chat experience https://t.co/uLbS2JNczH. pic.twitter.com/T1XdQreluH
– Anthropic (@AnthropicAI) November 21, 2023
Users can now use tools to increase Claude’s capabilities. For example, ask Claude to search the web or perform a calculation. The chatbot automatically selects the best tools for the task based on the context. Additionally, you can also read about Claude Pro and ChatGPT Plus: Which AI Chatbot is Right for You? [Detail Guide]
Developers have more control with the updated console. The test window allows you to test indications to ensure performance. There’s also the ability to give Claude persistent, customizable instructions about personality, tone, etc.
The combination of vastly expanded context and reduced hallucinations cements Anthropic’s Claude as a leading contender in responsibly designed chatbots. With ethical AI coming under intense scrutiny today, they have clearly invested in security without sacrificing functionality.
Categories: Technology
Source: vtt.edu.vn