Machine learning news — deep learning, reinforcement learning, neural architecture search, diffusion models, and new ML frameworks and libraries.
Red Alice AI released the first official benchmark of its Version 2 architecture, reporting a 200x performance gain in the RedTensor engine. The upgrade introduces a PyTorch-backed TorchTensor backend with unified flat i…
A developer building a passive AI meeting assistant discovered that their speech recognition model was hallucinating the phrase 'Thank you' during applause, laughter, and silence. The model, trained on real-world audio w…
Poindexter shipped several improvements on June 24, 2026, including wrapping DeployCheckoutSync in a VBS helper to suppress flashing terminal windows on Windows. The team folded embedding hygiene jobs into a declarative …
Immunai announced a discovery collaboration with Boehringer Ingelheim valued at up to $15 million through 2027. The partnership will apply Immunai's single-cell AI platform to identify T-cell targets for immuno-oncology …
UN Women warned that AI systems reproduce historical gender stereotypes, citing a study of 133 AI systems where 44 percent exhibited gender bias. The warning, issued ahead of the UN Global Dialogue on AI Governance, high…
Amazon SageMaker AI now supports NVIDIA Blackwell GPUs with P6-B200 instances, enabling optimized training for large AI models through expanded memory and new precision formats. The integration allows larger batch sizes,…
ByteDance's open-source SeedVR2 video restoration model is now deployable on Amazon SageMaker AI, enabling scalable super resolution for low-resolution video libraries. The solution uses a three-tier AWS architecture wit…
A new open-source tool called mlx-chronos provides a standardized benchmark suite and community leaderboard for comparing local LLM inference engines on Apple Silicon Macs. The tool runs reproducible benchmarks, generate…
Prefill/decode disaggregation separates the two phases of LLM inference—prefill (compute-bound) and decode (memory-bound)—onto different GPUs to avoid the performance compromise of running both on the same hardware. The …
A developer built an Arxiv scraper on Apify that automatically monitors new machine learning research papers by keyword. The tool supports scheduled runs, structured output, and easy integration with platforms like Slack…
Researchers at the Allen Institute for AI compared their 7B transformer model Olmo 3 with the hybrid model Olmo Hybrid to determine which tokens each predicts better. The hybrid model excels on meaningful tokens like nou…
VibeDrift, a new open-source tool, detects and prevents AI coding drift by analyzing codebase consistency across sessions, offering local scans and deep analysis to catch pattern violations and semantic conflicts before …
Metaflow, an open-source ML/AI orchestration framework, ranks first in every category of the Cloud Native Computing Foundation's latest Technology Radar report, with 51% of surveyed users highly likely to recommend it. T…
Outerbounds engineers achieved a nine-fold speedup in OpenAI Whisper transcription on Kubernetes by identifying bottlenecks and making a two-line code change, demonstrating how default Kubernetes configurations can be su…
Avon and Somerset Police and Bristol City Council built the Think Family Database, scoring 170,000 people on exploitation risk using school, housing, and police data without consent. An audit found the models were the we…
Metaflow, an open-source ML platform, can be used to train large language models (LLMs) like Databricks' Dolly, enabling integration into production projects. Fine-tuning a foundation model such as Dolly costs around $50…
ARENA (Alignment Research Engineer Accelerator) announced its ninth iteration, a 4-5 week ML bootcamp focused on AI safety, running in-person at LISA in London from October 5 to November 6, 2026. Applications are due by …
A company reduced its AI costs by 80% through optimization techniques, enabling more efficient deployment of machine learning models.
Navatala Systems released Navatala GPU, an open-source cross-platform GPU compute runtime and kernel corpus supporting CUDA, HIP, Vulkan, OpenCL, and Metal backends. The alpha release includes Python bindings on PyPI and…
Loop engineering is a new paradigm for building self-prompting autonomous AI systems that iterate through cycles of task generation, execution, observation, and refinement until goals are achieved. The approach replaces …