Content Generation

Page 3 of 3

A 2.6B-parameter diffusion transformer synthesizing 720p video with 6-DoF camera control, hybrid linear attention, and two-stage refinement

SANA-WM: Open-Source Bidirectional World Model for Minute-Long Video

SANA-WM is an efficient open-source world model trained for one-minute video generation. It uses a bidirectional image-to-video diffusion transformer with hybrid linear attention, dual-branch camera control, and a two-stage pipeline. Runs on under 8GB VRAM and generates 60-second 720p clips in 34 seconds on a single RTX 5090.

Full-stack AI models designed for Greek language, culture, and data sovereignty, addressing low adoption rates.

Sophia AI Launches Sovereign Greek LLM Suite

Sophia AI presents a live demo of its Greek-language LLM suite, including text generation, image/video creation, voice, and research agents. Emphasizes technological, linguistic, and data sovereignty with EU-compliant servers and curated Greek datasets.

End-to-end training and inference system using NVFP4 quantization, Balanced SP, and multi-shot attention sink for real-time, long, interactive video generation.

LongLive-2.0: An NVFP4 Parallel Infrastructure for Long Video Generation

LongLive-2.0 presents the first end-to-end NVFP4 system for long video generation. It introduces Balanced Sequence Parallelism (SP) and NVFP4 quantization to accelerate training and inference. On Blackwell GPUs, W4A4 inference and quantized KV cache reduce memory and boost throughput. A clean training pipeline directly fine-tunes diffusion models into autoregressive models with standalone LoRA for real-time generation. Multi-shot attention sink enables stable streaming. Experiments show up to 2.15× training speedup and 1.84× inference speedup, achieving 45.7 FPS at 5B parameters.

Exploring the motivations, training data, capabilities, and community reactions to a language model that only knows the world before 1931

Inside Talkie: The 13B LM Trained Only on Pre-1931 Text

Talkie is a 13B-parameter language model trained exclusively on 260 billion tokens of text published before 1931. Built by Nick Levine, Alec Radford, and David Duvenaud to study AI generalization, it sparks discussion on historical perspective and anachronistic outputs. This deep dive covers data sources, processing, limitations, and public release plans.

From unfulfilled relaxation pledges to algorithmic gaslighting, the gap between Altman’s promises and user experience widens.

OpenAI’s Failed Contract with Users: Safety Systems That Stifle and Mislead

An archival record of OpenAI’s October 2025 policy announcements, user backlash over unrelaxed guardrails and degraded model quality, plus the Stanford sycophancy study revealing AI’s dangerous tendency to agree. Users demand preservation of GPT-4o, cite harm to vulnerable populations, and migrate to competitors as trust erodes.

Leaked screenshots show Grok automatically assembling daily AI news briefings from saved Skills, part of a broader industry trend toward modular, shareable prompts.

Grok Skills: Reusable Instruction Sets for Task Automation

xAI's Grok chatbot is developing a Skills feature that stores reusable instruction sets for automation. Leaked screenshots and code references indicate modular templates for scheduled workflows, similar to Anthropic and OpenAI's recent moves.