Urgent.News

What's breaking now, across thousands of outlets.

AI

D2K-Bench: Can LLM Agents Turn Expert Designs into Efficient GPU Kernels?

GPU kernels generated by large language model (LLM) agents can remain less efficient than expert implementations, but runtime alone does not reveal how the gap relates to design discovery and implementation. We introduce D2K-Bench, a diagnostic benchmark of 26 tasks and 85 workloads that measures how effectively agents translate expert design guidance into efficient GPU kernels. The guidance…

We haven't written up this one. arXiv cs.AI has the full story — the link below goes straight to it.

Read the original at arxiv.org →

More in AI

OpenScout: An Autonomous Open-Weight AI Event Scout

This is a submission for the Hacktoberfest Weekend Challenge: Build for a Friend What I Built I built OpenScout – Personal Open Source Event Notifier , an autonomous AI-powered scout that monitors…

  • OpenScout is autonomous AI-powered event notifier for developers
  • Utilizes open-weight AI models like Google's Gemma (gemma2:2b / gemma3:4b) for local inference
  • Aggregates open-source opportunities, ranks matches using AI, and alerts users about relevant events

From Blender to AI Video: Building a Browser-Based Import Bridge

A rough 3D scene can express things that take a surprisingly long prompt to describe: where the camera sits, how objects overlap, how much space a subject occupies, and how a shot moves.

  • Blender add-on enables import of snapshots or animation previews into Klox canvas
  • Media transferred while preserving editable .blend scene on user's machine
  • Klox supports Snapshot, Animation, and Local file modes with 1080 pixel max resolution

More from Friday 2 October →