Amazon will invest an additional $4 billion in Anthropic, bringing its total commitment to $8 billion. The deal makes AWS Claude’s primary cloud partner and strengthens Amazon’s bet on its Trainium chips.
Anthropic updates Claude 3.5 Sonnet, introduces Claude 3.5 Haiku and makes available in beta a tool that lets the model use graphical interfaces with a cursor, clicks and a keyboard.
Anthropic is making prompt caching for Claude available in public beta. The feature reuses large parts of a request, cuts input costs by up to 90% and avoids processing the same context on every call.
Anthropic introduces Claude 3.5 Sonnet, a model that outperforms Claude 3 Opus on several tests despite costing far less. It also debuts Artifacts, a workspace for working with generated code and documents.
Anthropic used dictionary learning to identify 34 million internal features in Claude 3 Sonnet. Its Golden Gate Claude experiment shows that altering one of them visibly changes the model’s responses.
Anthropic unveils Claude 3, a family of three vision-capable models with a 200,000-token context window. Its Opus model leads GPT-4 on several benchmarks, although those tests do not fully measure its real-world usefulness.
Amazon announced today an investment of up to $4 billion in Anthropic, the lab behind Claude. AWS becomes its primary cloud provider, with Trainium and Inferentia chips underpinning the training and deployment of its models.
Anthropic opens Claude 2 to the public in the US and UK, with a 100,000-token context window and notable gains in coding, math and reasoning.
Anthropic has announced a $450 million funding round led by Spark Capital to develop Claude. The investment strengthens one of OpenAI’s leading rivals in language models, with a stated focus on safety.
Anthropic has unveiled Claude, a conversational assistant built on constitutional AI, a technique for training models to follow safety principles. Its arrival coincides with the launch of GPT-4 and intensifies competition among major language models.
This website uses cookies to improve the browsing experience. Cookie policy.
No results. Try different words.
