Customers can fine-tune Claude 3 Haiku, Amazon's fastest and most cost-effective model, to customize its knowledge and capabilities for their business. Fine-tuning involves preparing high-quality prompt-completion pairs and using the fine-tuning API, available in preview, to create a custom model. The model can be tested and refined using the Amazon Bedrock console or API.
AI | Rating: 58 | 2024-07-10 04:56:16 PM |
Anthropic has streamlined the process of crafting high-quality prompts for AI-powered applications, allowing users to generate, test, and evaluate prompts in the Anthropic Console. The Console features a built-in prompt generator powered by Claude 3.5 Sonne, enabling users to describe a task to Claude and generate prompts. The new features include automatic test cases and output comparison, allowing users to leverage Claude to generate the best responses for their needs.
AIAnthropic | Rating: 89 | 2024-07-09 06:47:23 PM |
A new initiative is introduced to fund evaluations developed by third-party organizations that can measure advanced capabilities in AI models. The initiative aims to elevate the field of AI safety by providing valuable tools for the ecosystem. The evaluations will focus on measuring advanced model capabilities, with a specific emphasis on safety-relevant evaluations.
AI | Rating: 56 | 2024-07-01 09:03:49 PM |
Anthropic's AI system, Claude, is now available in the AWS Marketplace for the US Intelligence Community and AWS GovCloud, offering a range of applications for government agencies. Claude can be used to improve citizen services, streamline document review and preparation, and enhance policymaking with data-driven insights.
AI | Rating: 90 | 2024-06-26 05:36:28 PM |
Claude.ai Pro and Team users can now organize their chats into Projects, bringing together curated sets of knowledge and chat activity in one place. This new functionality enables idea generation, strategic decision-making, and exceptional results. Projects are available on Claude.ai for all Pro and Team customers, powered by Claude 3.5 Sonnet, which outperforms its peers on various benchmarks.
AI | Rating: 81 | 2024-06-25 02:56:15 PM |
Claude.ai has launched Claude 3.5 Sonnet, a new model that outperforms competitors and Claude 3 Opus on various evaluations. It is available for free on Claude.ai and the Claude iOS app, with higher rate limits for subscribers. The model costs $3 per million input tokens and $15 per million output tokens, with a 200K token context window.
AI | Duplicated with: 1 | Rating: 89 | 2024-06-22 03:38:36 AM |
Researchers and AI developers employ a range of red teaming techniques to test their AI systems, identifying potential vulnerabilities and improving safety and security. This practice involves adversarially testing technological systems to identify weaknesses.
AI | Rating: 80 | 2024-06-12 03:35:31 PM |
Companies developing AI models aim to train them to avoid saying harmful things and to avoid assisting with harmful tasks, but also strive to create models that exhibit admirable traits such as curiosity, truthfulness, and patience.
AI | Rating: 81 | 2024-06-08 08:35:37 PM |
Since last summer, we've been testing and mitigating elections-related risks by developing a flexible process using in-depth expert testing and large-scale automated evaluations to identify potential risks and guide our responses. We've applied this process to various elections-related topics in different regions across the globe.
AI | Rating: 81 | 2024-06-06 01:15:44 PM |
Claude is now available in Canada, accessible via Claude.ai, the Claude iOS app, the Claude API, and the Claude Team plan. The Claude Pro subscription offers increased usage and access to all models for CA$28 + tax per month.
AI | Rating: 90 | 2024-06-05 03:06:09 PM |