Anthropic Economic Index report: Cadences
…We refer to the primary output Claude produces in a conversation—a document, an explanation, a piece of code, an academic paper, and so on, whether presented in a chat window or…
Search
Measuring LLMs’ ability to develop exploits
…Over the last month, however, we have seen the development of two new, more challenging academic benchmarks: ExploitBench and ExploitGym . We collaborated with the researchers who produced these benchmarks to measure Mythos…
Project Fetch: Can Claude train a robot dog?
…We randomly divided eight Anthropic researchers (none of whom were robotics experts) into two teams—one with Claude access, one without—and asked them to program quadruped robots to fetch beach balls…
More details on Fable 5’s cyber safeguards and our jailbreak framework
…Our hope is to spark a helpful discussion across academia, industry, civil society, and government about how and where these lines should be drawn. We welcome feedback and critique on this framework…
Long-running Claude for scientific computing
…This post describes how to set up a similar pattern for scientific computing tasks using Claude Code, with a typical academic lab in mind. As a concrete example, I will walk through…
Project Glasswing: An initial update
…We’re supporting the development of other high-quality quantitative benchmarks through our External Researcher Access Program . Finally, Claude for Open Source supports maintainers and contributors, and we’re committing to scan…
Anthropic Economic Index report: Learning curves
…Some of the drop in coursework can be explained by academic calendars in countries where students were on winter break during our sample period. 4 At the same time, increasing signups beginning…
Vibe physics: The AI grad student
…But for me the real phase transition was running Claude Code with access to files, terminal commands, agents, skills, memory etc. It makes a big difference. Conclusions This paper started out as…
To show you the most relevant results, we’ve omitted some entries very similar to those already shown. Repeat the search with the omitted results included.