173 stories tagged Machine Learning
NVIDIA's Kumo Relational AI model analyzes relational data without any training, outperforming traditional supervised machine learning on benchmarks. This zero-shot approach could make advanced data analysis accessible to non-experts.
Google and Meta have released TorchTPU, an open-source hybrid eager/compiled architecture that enables native PyTorch execution on Google TPUs with industry-competitive training and inference performance.
Allen Institute for AI's HiRO-ACEv2 model simulates extreme weather at unprecedented 3km resolution. It's 100x faster and 10x cheaper than traditional methods, making high-quality weather forecasting more accessible.
NVIDIA has released a 100-billion-parameter AI model trained on 2.8 trillion tokens of Indic languages. The model and its open-source data generation pipeline are available on Hugging Face, making advanced AI more accessible for these languages.
Meta has released Muse Spark Open, a powerful open-weight multimodal AI model that can process text, images, and audio. The model is freely available for anyone to use, modify, and distribute.
Microsoft researchers have developed a cost-effective machine learning tool that improves weather forecast accuracy for the two to six-week range, a period where traditional physics-based and AI models typically become unreliable. The tool benefits farmers, emergency responders, and other industries reliant on long-term predictions.
Researchers introduced EliteTable, a self-improving program-search algorithm guided by large language models (LLMs). In evaluations across six Gymnasium control tasks, median candidate scores rose between the first and final generations in all six tasks, demonstrating consistent improvement.
Researchers developed a new system that analyzes table headers to assess data quality, helping create more reliable knowledge graphs. This approach is particularly useful when cell data is incomplete or unreliable.
Researchers introduced Route-Verify-Vote (RVV), a new AI framework that helps language models reason across unfamiliar domains without additional training. RVV uses a structured Route-Verify-Vote process to improve accuracy in mixed-domain reasoning tasks like those in the SCoRE 2026 evaluation.
Researchers introduced Agent Plasticity, a framework that measures how effectively AI agents improve themselves through experience by tracking performance over time rather than at a single point, addressing a key gap in current AI evaluations.
Researchers introduced Auto-Diagnosis and Skill Discovery (ADSD), a framework that helps AI numerical solvers identify and fix their own performance issues. This approach could make AI-generated scientific code more reliable and efficient by automating algorithm improvement.
Researchers introduced MintFlow, a training-free constrained sampling framework for flow matching models that enforces constraints like physical laws without displacing samples from the pretrained data distribution.
OpenAI has published a detailed guide to help developers and startups effectively use the GPT-6 family of models. The guide covers model selection, tuning reasoning effort, prompt improvement, tool coordination, and workflow optimization for production.
Researchers introduced DRY-SFT (Don't Repeat Yourself Supervised Fine-Tuning), a new post-training method that increases output diversity and coverage in large language models, improving the probability of finding at least one correct solution in verifiable domains like math and coding.
Researchers introduced Causal Latent Revision (CaLR), a new AI framework that improves reasoning by combining strengths of autoregressive and diffusion models. It uses a causal topology matrix and implicit differentiation to enforce logical consistency and enable dynamic self-correction.
AI engineer Rahul argues that fine-tuning pre-trained models is becoming a critical skill, not because everyone needs a custom model, but because understanding how models learn from data enables developers to build things others cannot.
NVIDIA released NeMo Data Designer (NDD), an open-source framework for multimodal synthetic data generation. It lets users define dataset columns for text, code, images, embeddings, and more via a declarative config format, with a plugin system for extensibility.
Researchers released ZGCM-1, a fully open 7-billion-parameter dense foundation model trained from scratch for math and agentic search. It uses deliberate internal thinking and active external tool use to overcome its compact size, achieving extreme data, system, and algorithmic efficiency.
Researchers fine-tuned the Nemotron 3 Ultra model to solve complex math problems at the Olympiad level. The study provides an open pipeline for natural-language proof generation without external tools, showing significant improvements in performance.
IBM Research has released Granite Time Series PatchTST-FM-r2, a new AI model for time series forecasting that achieves state-of-the-art accuracy and is available under a commercial-friendly license, making it accessible for business use.
MegaTrain is a new tool that makes it possible to train massive AI models on just one GPU. This could democratize AI development, making it accessible to individuals and small teams who couldn't afford expensive hardware before.
Researchers released EXAONE Forecast for Finance (EXAONE Finance), a financial time series foundation model (TSFM) tailored to financial forecasting. It handles partially observed inputs and uses an efficient computational backbone to reduce quadratic cost, achieving strong zero-shot performance on financial data.
OpenAI released GPT-6 Astra, a new AI model with improved capabilities in coding, cybersecurity, and science. It sets new benchmarks in intelligence and alignment, making it more reliable and versatile.
Researchers introduced Evidence Sufficiency Boundary Training, a method that teaches AI models to abstain from answering questions when supplied evidence is insufficient. This approach improves reliability in multi-hop QA systems by ensuring models only respond when they have enough information.