HomeTagsJev

Jev - Page 11

Stories from Jev come first, newest first, then stories from connected topics (LLM, AI Models, AI, AI), each labelled with where it comes from. This view is not indexed by search engines.

Latest

Nesformer Compiles an NES Processor Into a Qwen-Style AI Model

Nesformer compiles the NES's 6502 processor into the weights of a Qwen3.5-style model with no training. What that means, and the caveats.

AI Reasoning Models Diagnose Better but Add 34% More False Positives

AI reasoning models made more accurate diagnoses but over 34% more false positives in a retrospective study; doctors checking them cut those by 44%.

SenseTime Open-Sources SenseNova Loop, a 9B Model That Reuses Layers

SenseNova Loop, SenseTime's new open 9B model, runs a 24-layer block twice with shared weights. It is Apache 2.0, with no technical report yet.

Microsoft Open-Sources Mage-VL-Cowork, a 4B Agent That Clicks or Codes

Microsoft's open Mage-VL-Cowork is a 4B computer-use agent that switches between clicking and code. Microsoft reports 66% on OSWorld-Verified.

Pika Opens H3 Flex Video Weights, but Licence Excludes UK and US

Pika's H3 Flex video model, built on MiniMax H3, has downloadable weights, but its licence excludes the US, UK, EU and South Korea.

OpenAI Rebuilds Its Model Spec Around Five Core Values for GPT

OpenAI has reorganised its Model Spec around five values, adding rules on human oversight and context and dropping most of its style guidance.

Meta Says LLM-Written Tests Helped Catch 74 Serious Privacy Bugs

Meta says its LLM-generated JiTTests, built to fail on risky code changes, helped catch 74 serious privacy bugs before they landed in H1 2026.

Mistral Opens Large 4 Weights: a 1-Trillion-Parameter Model

Mistral has released the weights of Large 4, a 1-trillion-parameter multimodal model, and warns its cyber skills carry a potential for real attack uplift.

Kakao Open-Sources Two Kanana-2 Embedding Models for Korean Search

Kakao has released 3bn and 1.3bn-parameter Kanana-2 embedding models on Hugging Face, topping its Korean retrieval tests while trailing Qwen3 in English.

Particle-1 Coding Model Uses About 45% Fewer Tokens, Particle Says

Suhail Doshi's Particle has launched Particle-1, a DeepSeek V4.1 Flash coding model it says uses about 45% fewer tokens with similar accuracy.

Latest on TSN

Explore the TSN Atlas: Every topic we cover, mapped. Click a bubble to dive in.