PyTorch patches CVE-2026-24747 (CVSS 8.8). The "weights_only" safety flag failed, allowing RCE via malicious models. Update to v2.10.0 now.
deep-learning
Google unveils Titans and MIRAS: AI architectures that “read while remembering” using a "surprise metric." They handle 2M tokens and outperform GPT-4 in memory tasks.
NVIDIA announces CUDA support for RISC-V processors, expanding its AI and GPU acceleration technology to a more diverse ecosystem and strengthening its position in emerging markets.
Google unveils an enhanced Gemini 2.5 Pro, now available for early trials. This powerful AI model offers superior performance and cost control for enterprises.
Microsoft's Aurora AI, an open-source model, achieves breakthrough accuracy in weather forecasting and cyclone prediction using deep learning.
CVE-2025-32434: Critical RCE vulnerability in PyTorch’s torch.load(weights_only=True) affects ≤2.5.1. Update to v2.6.0 to patch this major flaw