Research
Specific versus general principles for Constitutional AI
Models trained with Constitutional AI can generalize from a single broad principle to produce harmless assistants.
ResearchConstitutional Classifiers: Defending against universal jailbreaksAnthropic
ResearchAuditing language models for hidden objectivesAnthropic
ResearchEnabling independent research on how people use ClaudeAnthropic
ResearchProject Swap: What happens when agents trade for us?Anthropic
ResearchForecasting rare language model behaviorsAnthropic
ResearchDisempowerment patterns in real-world AI usageAnthropic
BlogsVisualizing your Tailscale network traffic with TSFlowTailscale
BlogsSelf-hosting DuckDB: the road to productionMotherduck
BlogsNotion’s Annual Craft & Values WeekNotion So
BlogsThe Best Note-taking Apps for WindowsZapier
BlogsReintroducing Serve and Funnel: even simpler sharing with your tailnet (or the world!)Tailscale
BlogsDuckDB vs Pandas vs Polars for Python DevelopersMotherduck
BlogsContinually improving our agent harness · CursorCursor
Resourcesvercel alertsVercel
Products & ServicesComing soon: auto-translate your docs with computed content in GitBookGitbook
Products & ServicesComing soon: Give every user their own docs experience with adaptive contentGitbook
ResourcesProfile enumFastly
BlogsTrain past the frontier: Training API now generally availableFireworks
BlogsDocker Desktop 4.24: Compose Watch, Resource Saver, and Docker EngineDocker
ResourcesLimits and Pricing for Image OptimizationVercel
