Gemma Scope: helping the safety community shed light on the inner workings of language models
Announcing a comprehensive, open suite of sparse autoencoders for language model interpretability.
WeiterlesenMonth: July 2024
Making Machines Move
We’re Fly.io, a global public cloud with simple, developer-friendly ergonomics. If you’ve got a working Docker image, we’ll transmogrify it into a Fly Machine: a VM...
WeiterlesenWhat we got wrong about HTTP imports
Designing a module system around HTTP imports was ambitious. Here are some issues we encountered and how we solved for them.
WeiterlesenBenchmarking AWS Lambda Cold Starts Across JavaScript Runtimes
When running production JavaScript in the cloud, performance is a critical consideration. Here’s how Deno’s cold start times compare against other JavaScript runtimes on AWS Lambda.
WeiterlesenAI achieves silver-medal standard solving International Mathematical Olympiad problems
Breakthrough models AlphaProof and AlphaGeometry 2 solve advanced reasoning problems in mathematics
Weiterlesen
Are We Ready for Multi-Image Reasoning? Launching VHs: The Visual Haystacks Benchmark!
Humans excel at processing vast arrays of visual information, a skill that is crucial for achieving artificial general intelligence (AGI). Over the decades, AI researchers have...
Weiterlesen
Google DeepMind at ICML 2024
Exploring AGI, the challenges of scaling and the future of multimodal generative AI
WeiterlesenAnnouncing Supabase on JSR
Supabase’s isomorphic JavaScript client library is now available on JSR.
WeiterlesenDeno 1.45: Workspace and Monorepo Support
Deno 1.45 introduces workspaces and monorepo support, improved Node.js compatibility, updates to `deno install`, the new `deno init –lib` command, deprecation of `deno vendor`, Standard Library...
Weiterlesen
The AI summer
My old boss Marc Andreessen liked to say that every failed idea from the Dotcom bubble would work now. It just took time – it took...
WeiterlesenIntroducing KV Backup for Deno Subhosting
This new feature allows Subhosting users to configure their KV databases to back up data to their own S3-compatible object storage via APIs.
Weiterlesen
The VR winter continues
It can feel a little odd to write about anything other than generative AI these days, but I sometimes remind people that all the things that...
WeiterlesenEntering text in the terminal is complicated
The other day I asked what folks on Mastodon find confusing about working in the terminal, and one thing that stood out to me was “editing...
WeiterlesenExtrinsic Hallucinations in LLMs
Hallucination in large language models usually refers to the model generating unfaithful, fabricated, inconsistent, or nonsensical content. As a term, hallucination has been somewhat generalized to...
Weiterlesen