Local AI - Page 10
Stories from Local AI come first, newest first, then stories from connected topics (llama.cpp, Ollama, NVIDIA RTX Spark, on-device AI, AI Models, Hugging Face), each labelled with where it comes from. This view is not indexed by search engines.
Latest
Hundreds of Scientists, Including Geoffrey Hinton and Yoshua Bengio, Call for an International Body on AI Containment and Incident Reporting
Hundreds of scientists, including Geoffrey Hinton and Yoshua Bengio, call for an international body to set AI containment standards and log incidents.
Noiz AI and HKUST Release Helix World, a Real-Time World Model Whose Sound Turns With the Camera
Noiz AI and HKUST release Helix World, a real-time world model that makes video and stereo sound together. All benchmarks are the authors' own.
Sperid Labs Releases Iris-3B, an Open Image Model With No VAE; Its Own Paper Finds No Clear Gain for Depth or Restoration
Sperid Labs released Iris-3B, a 3B image model with no VAE. Its paper finds no clear gain for depth or restoration. Code and weights are open.
LAION Releases Humaneness Voice Converter, an Open Voice-Conversion Model Built on Chatterbox; the Scores Are LAION’s Own
LAION released an open voice-conversion model built on Chatterbox. The scores are LAION's own, and match to the target voice fell from 0.821 to 0.773.
Google Open-Sources ML Drift, the GPU Engine Inside LiteRT, Under Apache 2.0
Google has open-sourced ML Drift, the GPU engine inside LiteRT, under Apache 2.0. The 40%, 30% and 12% figures are Google's own claims; the old TFLite GPU delegate gets no new features.
Underdog’s Saluki 27B Shrinks Qwen3.8-27B to 7.89 GB: Tool Calling Up on Its Own Test, Competition Maths Down
Underdog's Saluki 27B is a 7.89 GB, 2-bit version of Qwen3.8-27B. Underdog says it passes more tool-calling tasks than the full model on its own 120-task test but scores 15 to 18% lower on competition maths. What quantisation means, and what the numbers do not prove.
Harness Buys Augment Code Assets, Plus TRL 1.15 and Genkit Dart 1.0
Four AI developer-tool updates from 8–9 October 2026: Harness buys selected Augment Code assets (not the whole company, price undisclosed), Hugging Face's TRL v1.15.0 claims up to 6.9× longer training sequences, Google's Genkit Dart reaches 1.0, and NVIDIA shows its Boro kernel-review tool. What each says, and what it doesn't.
Two Open Models From 8 October: a Fast Coding Agent and a Gene Finder
JetBrains has released Mellum2.1, a small open model for coding agents, and Hugging Face has released Carbon-A, a 1.2B-parameter model that predicts protein-coding regions in DNA, plus a database of 566 million predicted loci. What each does, and which numbers are the developers' own.
Same Model, Different Score: Training AI Inside the Coding Tools People Actually Use
Hugging Face's FineEnvs team has released an open framework that turns unmodified coding-agent harnesses such as Claude Code, Codex and OpenCode into reinforcement-learning environments. In its own tests, one small model scored 62% in one harness and 33% in another. What a harness is, how the training works, and what the results do and do not show.
Microsoft Execution Containers and AWS Strands Box Fence In AI Agents
Microsoft made its Execution Containers generally available on Windows 11 as it rebuilds Windows for local and cloud "hybrid" AI, and opened Surface Laptop Ultra pre-orders from $2,599. AWS open-sourced Strands Box, a sandbox that enforces rules such as "no git push" whatever the agent decides. What an enforced sandbox does, in plain words.
Latest on TSN
Explore the TSN Atlas: Every topic we cover, mapped. Click a bubble to dive in.