Boggle Nogs

Anthropic Incident: Elevated errors on request to models (anthropic.com)
4 points

Exploring model welfare (anthropic.com)
14 points 7 comments

Detecting and Countering Malicious Uses of Claude \ Anthropic (anthropic.com)
3 points

Exploring Model Welfare (anthropic.com)
9 points

Detecting and Countering Malicious Uses of Claude: March 2025 (anthropic.com)
4 points

Detecting and Countering Malicious Uses of Claude (anthropic.com)
2 points

Our Approach to Understanding and Addressing AI Harms (anthropic.com)
4 points

Values in the wild: Discovering values in real-world language model interactions (anthropic.com)
5 points 1 comment

Values in the wild: Discovering values in real-world language model interactions (anthropic.com)
3 points 1 comment

Anthropic Analyzes Real-World Conversations to Uncover AI's "Values in the Wild" (anthropic.com)
4 points 1 comment

Claude Code: Best practices for agentic coding (anthropic.com)
614 points 257 comments

Claude Code: Best practices for agentic coding (anthropic.com)
47 points 2 comments

Claude Research (anthropic.com)
39 points 6 comments

Anthropic research shows AI model conceals reasoning shortcuts 75% of the time [pdf] (anthropic.com)
4 points

Claude's Max Plan (anthropic.com)
132 points 136 comments

How University Students Use Claude (anthropic.com)
401 points 579 comments

Anthropic Education Report: How University Students Use Claude (anthropic.com)
3 points

Claude Code (anthropic.com)
2 points 1 comment

Reasoning models don't always say what they think (anthropic.com)
420 points 267 comments

Introducing Anthropic's first developer conference: Code with Claude (anthropic.com)
1 point

Claude for Education \ Anthropic (anthropic.com)
7 points

Introducing Claude for Education (anthropic.com)
4 points

Anthropic Economic Index: Insights from Claude 3.7 Sonnet (anthropic.com)
56 points 7 comments

Training on Documents About Reward Hacking Induces Reward Hacking (anthropic.com)
1 point

Tracing the thoughts of a large language model (anthropic.com)
1072 points 394 comments

Automated Researchers Can Subtly Sandbag (anthropic.com)
2 points

Auditing Language Models for Hidden Objectives (anthropic.com)
1 point

Automated Researchers Can Subtly Sandbag (anthropic.com)
2 points

The think tool: Enabling Claude to stop and think in complex tool use situations (anthropic.com)
26 points 2 comments

Claude can now search the web (anthropic.com)
1233 points 726 comments