PulseAugur / Brief
EN
LIVE 01:15:12

Brief

last 24h
[2/2] 223 sources

Multi-source AI news clustered, deduplicated, and scored 0–100 across authority, cluster strength, headline signal, and time decay.

  1. RESEARCH · Mastodon — fosstodon.org 日本語(JA) · · [3 sources]

    Training mRNA Language Models Across 25 Species is Possible for $165 https:// huggingface.co/blog/OpenMed/tr aining-mrna-models-25-species *AI-generated auto-post (headline + link) # AI # GenerativeAI # LLM # AIGenerated

    Researchers have developed a method to train mRNA language models across 25 species for as little as $165, utilizing Hugging Face's platform. Separately, the Safetensors format has joined the PyTorch Foundation, enhancing its integration and adoption within the deep learning ecosystem. Additionally, NVIDIA has released Nemotron 3.5, a multimodal safety model designed for customizable content moderation for global enterprises. AI

    IMPACT Advances in specialized model training, standardization of model formats, and multimodal safety solutions.

  2. FRONTIER RELEASE · arXiv cs.AI English(EN) · · [5 sources]

    AICompanionBench: Benchmarking LLMs-as-Judges for AI Companion Safety

    NVIDIA has released Nemotron 3.5 Content Safety, an AI model designed to identify and mitigate harmful content across text and images. This new version enhances multimodal understanding, supports over 140 languages with strong zero-shot generalization, and allows for custom policy enforcement tailored to specific enterprise needs. It also includes an auditable reasoning trace feature and releases its multimodal safety dataset for public use. AI

    IMPACT Enhances enterprise AI safety with customizable, multimodal content moderation and reasoning capabilities.