Research
Softmax linear units \ Anthropic
Replacing standard MLP activations with SoLU substantially increases the fraction of interpretable neurons with no performance loss.
ResearchLLM-discovered 0 days \ Anthropicanthropic.com
ResearchMoral self-correction in large language models \ Anthropicanthropic.com
ResearchIntroducing the Anthropic Economic Index \ Anthropicanthropic.com
ResearchA small number of samples can poison LLMs \ Anthropicanthropic.com
ResearchScaling laws of learning from repeated data \ Anthropicanthropic.com
ResearchFinding bugs with Claude and property-based testing \ Anthropicanthropic.com
LearnThe Simplest Possible Neural Network, Explained - Deepgram Blog ⚡️Deepgram
BlogsNeo Integrations: MCP Servers and Cloud CLIsPulumi
ResourcesWhat is a Load/Store Unit?Modal
BlogsSome Things You Oughta Know When Working with Viewport UnitsCss Tricks
NewsHeterogeneous Processing Requires Data Parallelization: SYCL and DPC++ are a Good StartThenewstack
Products & ServicesNew this month: GitBook in Linear, Slack & GitHubGitbook
BlogsCSS fix for 100vh in mobile WebKitCss Tricks
Products & ServicesUsing GitBook Agent in Slack, Linear, and GitHubGitbook
Products & ServicesHow Linear approaches knowledge sharingGitbook
BlogsLinear Scaling Made Possible with Weight Streaming - CerebrasCerebras
JobsSenior / Staff Product Engineer - Linear CareersLinear
ResourcesDelete and archive issuesLinear
ResourcesThird-Party App ApprovalsLinear
ResourcesCycle graph – Linear DocsLinear
