Search Anthropic

A model's ability to work through problems logically rather than just pattern-matching. How much AI systems truly reason versus simulate reasoning is an open question with real implications for safety.
Testing Claude on real-world electronics design
We asked Claude Fable 5.2 to design three electronic devices from short prompts: a speakerphone, a voice-assistant hub, and a handheld virtual pet. Diode Computers built working prototypes from its files.
Claude reveals ECDSA curve origins
The decades-old mystery of the origins of one of the most popular cryptographic standards has been explained.
Strengthening our defenses against distillation attacks
Illicit distillation allows model developers to copy frontier AI capabilities without creating proportionally robust safeguards. Here, we describe how we’ve reduced the rate of distillation by more than 90%.
Yes, Claude can do Nine Loops
Guest writer and physicist Matt von Hippel shares what happened when he issued a challenge to AI companies to solve a problem in his former subfield of theoretical physics.
Formalizing Fermat's Last Theorem
We are sharing the first complete computer-checked proof of Fermat’s Last Theorem. Claude worked largely autonomously over 11 days to write the proof in the Lean programming language.
Learning more about Claude's mathematical capabilities
An unreleased research version of Claude has made strides on a problem related to the Riemann hypothesis.
A global workspace in language models
New interpretability research reveals an emergent mental workspace in Claude that holds internal thoughts that don’t appear in the model’s output.
Making Claude a chemist
Anthropic is working with world-class synthetic, computational, and analytical chemists to make Claude better at chemistry. In this post, we share our first work as part of this effort.
An update on recent Claude Code quality reports
We traced recent reports of Claude Code quality issues to three separate changes. Here's what happened and what we're changing.
Reasoning models don't always say what they think
Research from Anthropic on the faithfulness of AI models' chain-of-thought reasoning.
Anthropic Economic Index: Insights from Claude 3.7 Sonnet
The second update from the Anthropic Economic Index on how people are using AI.
Tracing the thoughts of a large language model
Today, we're sharing two new papers that represent progress on the development of an “AI microscope,” and the application of it to see new “AI biology.”
The "think" tool: Enabling Claude to stop and think in complex tool use situations
A blog post for developers, describing a new method for complex tool-use situations
Claude 3.7 Sonnet and Claude Code
Today, we’re announcing Claude 3.7 Sonnet, our most intelligent model to date and the first hybrid reasoning model generally available on the market.
Claude’s extended thinking
Discussing Claude's new thought process
Alignment faking in large language models
A paper from Anthropic's Alignment Science team on Alignment Faking in AI large language models
A statistical approach to model evaluations
Recommendations for statistical rigor in AI evaluations, including standard errors and paired-difference analysis when comparing models.
Sleeper Agents: Training deceptive LLMs that persist through safety training
We trained LLMs with backdoors and found standard safety training failed to remove them, sometimes teaching models to hide them better.
Measuring faithfulness in Chain-of-Thought reasoning
Does chain-of-thought reasoning reflect a model's actual decision process? We find faithfulness varies by task and declines in larger models.
Question decomposition improves the faithfulness of model-generated reasoning