Machine learning news — deep learning, reinforcement learning, neural architecture search, diffusion models, and new ML frameworks and libraries.
Local AI models can now answer 89% of everyday chat and reasoning queries as well as frontier cloud models, according to a November 2025 study by Stanford University and Together AI. The best local model's win/tie rate a…
Hume AI's new research introduces three tests to quantify benchmark optimization in speech recognition, finding that several of 11 open-source ASR models reproduced benchmark transcripts from VoxPopuli and LibriSpeech ev…
GLM-5.3 and Claude Fable 5 finish within noise of each other on DeepSWE accuracy, but GLM-5.3 costs a fifth as much per task and wins every multi-attempt metric. Fable 5 leads pass@1 by 0.7 points (69.7% vs. 69.0%), but …
Hugging Face has built a hybrid search system for Papers with Code that combines PostgreSQL full-text search with pgvector dense embeddings, using reciprocal rank fusion (RRF) to merge results. The system maintains embed…
A technical blog post outlines a pipeline for analyzing political communications using AI, including PII stripping with Presidio, embedding with mixedbread-ai/mxbai-embed-large-v1, fine-tuning a DeBERTa-v3-large classifi…
Xaira Therapeutics, an AI-driven biotech startup headquartered in the San Francisco Bay Area, Seattle, and London, is hiring for an 'AI in Residence' role in Seattle, Washington, offering a salary of $10k–15k per month. …
A new approach to generating visual datasets uses script-generated images with reproducible metadata and Adobe Generator tools, enabling controlled evaluation and diagnostics for machine learning models. The method allow…
Data Science Weekly's Issue 665 curates technical links spanning CUDA-based geolocation, A* pathfinding heuristics, inference engineering, and Git hosting at scale. The newsletter also highlights a piece arguing that AI …
A developer demonstrates how to add semantic search to an existing DynamoDB table using the new Vector Search feature, eliminating the need for separate search infrastructure. The approach combines recipe fields into a s…
Liquid AI, an MIT spinout, released four updated LFM2.5 checkpoints on August 19, trained with quantization-aware distillation (QAD) to retain roughly 97% of their BF16 average benchmark performance while reducing memory…
Superwhisper released S1-mini, a 462 MB open-weights text normalizer fine-tuned from Qwen/Qwen3-0.6B, which turns raw ASR transcripts into clean written text with 94.8% token accuracy on a held-out set of 7,519 cases. Th…
Youtilitics, a no-hardware energy analytics platform, has been rebuilt and now offers baseline anomaly detection, overnight step-change alerts, and usage heatmaps from utility interval data. The platform, which supports …
Security Anomaly ML, an open-source network-flow detector from developer I. Bondarenko, released v0.1.0, a research/evaluation version that converts CICFlowMeter-compatible traffic into deterministic analyst-facing secur…
Informatiqs engineers detail a serverless architecture for real-time AI inference on Google Cloud Platform, addressing the cold start problem through asynchronous event-driven orchestration with Eventarc and Cloud Run, a…
Informatiqs argues that Graph Neural Networks (GNNs) are the strategic choice for enterprise generative AI on Google Cloud Platform (GCP), emphasizing the shift from flat vector embeddings to graph-based relational intel…
Amazon Web Services (AWS) published the first part of a three-part tutorial series on building a no-code machine learning workflow using Snowflake, Amazon SageMaker Canvas, and Amazon QuickSight, focusing on setting up a…
Amazon Web Services (AWS) published Part 2 of its no-code machine learning workflow series, detailing how to use Amazon SageMaker Canvas to connect to Snowflake, prepare data with Data Wrangler, and build a fraud detecti…
Amazon Web Services (AWS) published Part 3 of its no-code machine learning workflow series, detailing how to visualize fraud detection predictions from Amazon SageMaker Canvas using Amazon Quick Sight, now part of Amazon…
Promptfoo is recommended as the primary evaluation tool for AI agents focused on documentation and RAG tasks, with Ragas and LangSmith suggested for deeper analysis. The guidance from Promptfoo, Hugging Face, LangChain, …
PyTorch Conference North America 2026 announced its keynote speaker lineup for the October 20–21 event in San Jose, California, featuring sessions on PyTorch updates, native PyTorch on Trainium, agent workloads, and mode…