Machine learning news — deep learning, reinforcement learning, neural architecture search, diffusion models, and new ML frameworks and libraries.
A developer built a production-grade machine learning system to predict 30-day hospital readmissions using the UCI Diabetes 130-US Hospitals dataset, which contains 101,766 patient encounters from 130 US hospitals. The s…
An engineer at Jahn AI ran a per-prompt regression check before recommending FP8 quantization for Qwen3-8B serving on an RTX PRO 6000 Blackwell, which delivered a 1.5x throughput gain (1,725 to 2,597 tok/s at concurrency…
Skild AI launched S1, a robotics foundation model that learns manipulation tasks from a single video demonstration without fine-tuning, enabling deployment in as little as 11 minutes. In internal tests, S1 achieved a 66%…
Xyntetik Runner, a GGUF model runtime written from scratch in plain C, is now available as a free Apache 2.0-licensed engine that serves, verifies, scores, adapts, and trains GGUF models deterministically, with support f…
Skild AI, a Pittsburgh-based startup founded in May 2023 by Carnegie Mellon University researchers Deepak Pathak and Abhinav Gupta, launched S1, a robotics model that learns physical tasks from a single video of a human …
Researchers at Syracuse University's International Programs, led by Luis A. Valdes, have developed an AI-powered drone system for monitoring the health of dragon fruit farms in Panama, as detailed in a 2026 poster. The s…
A developer used machine learning and statistical modeling to estimate the human limit for solving a Rubik's Cube, analyzing 23 years of World Cube Association world records. The model, combining Gaussian Processes and E…
Agentic retrieval techniques that measurably improve accuracy in production fall into six patterns, according to a 12,000-trajectory study covered in 'Before Reasoning Can Fail' (arXiv:2608.02011). A forced read-before-a…
IBM researchers released Granite Speech 5.0 Turbo CTC, two 470-million-parameter English speech-recognition models that remove the language-model decoder to achieve over 20 times the throughput of earlier Granite Speech …
Apple introduced the M6 and M5 Ultra chips, with the M6 being its first 2nm chip debuting in a Mac mini at $899 and the M5 Ultra powering a Mac Studio with up to 512GB of unified memory and 1.2TB/s bandwidth, both shippi…
MIT engineers have developed an AI tool that forecasts extreme weather without training on historical disaster data, producing maps of statistically possible events absent from a region's record. The tool, created by gra…
Google Research's Lumiere paper, published in January 2024, introduced a Space-Time U-Net architecture that generates entire video clips in a single pass, solving the temporal consistency problem that plagued earlier tex…
Nvidia released DLSS 4.5 Ray Reconstruction on August 25, 2025, with support for 30 games and compatibility with all GeForce RTX GPUs from the RTX 20 series to the RTX 50 series. The update introduces a 2nd-gen Transform…
Konjugate v0.7.0 introduces a causal inference feature that recovers causal graphs from CSV time-series data, demonstrated on an 8-node synthetic thermal/vibration system with a known structure. The tool proposed 11 cand…
A new empirical study by Sawan Dasari, running nearly 4,000 experiments, reveals that the choice of retraining policy for machine learning models matters little if the model supports incremental learning, but without it,…
Researchers used machine learning to decode a hidden DNA signal that identifies the likely "initiator" of gene activity, enabling new predictions about disease-causing mutations. The findings, published in Nature, could …
IBM released two compact 470M-parameter English speech recognition models, granite-speech-5.0-470m-turboctc and granite-speech-5.0-470m-turboctc-nc, achieving over 12,600 RTFx on an NVIDIA H200 GPU and transcribing more …
Apple's Metal GPU glossary 'metalworking' targets developers familiar with CUDA but new to Apple Silicon, mapping every concept from GPU cores to MLX's architecture with CUDA equivalents and real kernel code. The hyperli…
A developer found a scale-selection rule for NVFP4 gradient quantization that reduces mean squared error by 14% on real gradient tensors compared to the state-of-the-art MS-EDEN estimator, but training runs at 2.8B param…
A developer argues that machine learning courses for government agencies must focus on operational metrics like precision at a fixed alert budget rather than accuracy, citing that a 99.9% accurate model on 20 million dai…