Machine learning news — deep learning, reinforcement learning, neural architecture search, diffusion models, and new ML frameworks and libraries.
A team of researchers from Carnegie Mellon, MIT, New York University, and Stanford University built an AI called Ataraxos that beat Pim Niemeijer, arguably the best Stratego player of all time, 15 games to one with four …
Cloudflare released two Cloudflare-trained decision models, Clef and Clef-flash, hosted on Workers AI and open-sourced on Hugging Face under an Apache 2.0 license, alongside a new reinforcement learning fine-tuning produ…
A developer built Cyber Autopsy, a Kaggle benchmark that tests whether AI models can reconstruct reported cyber incidents from partial evidence while citing support for each claim and flagging what remains uncertain. The…
The Allen Institute for AI released Olmo-core 3, an open training framework for large mixture-of-experts (MoE) models that scales MoE training into the trillion-parameter range while preserving computational efficiency. …
A developer built Typic, a 400M-parameter open decision model trained on a single free Kaggle T4 GPU in 46 minutes, which returns probabilities over a supplied list of allowed answers in one forward pass instead of gener…
Developer emnlmn released Snap, an open-weight, MIT-licensed local decision engine positioned as a drop-in replacement for the TypeSafe System One API, running the 2B-parameter snap1-2b model in a single forward pass to …
A developer built a 150-question benchmark from a 7.2M-page web index to measure frontier models' raw parametric knowledge of 2025-2026 events occurring after their training cutoffs. Across nine models, the best performe…
Exabeam introduced new Agentic SOC capabilities that bring AI-driven investigation, execution, and governance to both cloud and on-premises security operations environments, according to Help Net Security. The company sa…
A study measuring how much of 200 books from the Books3 corpus open-weight LLMs have memorized found that memorization varies enormously by book and model, with most books barely memorized but a few reproduced almost in …
The AI system Ataraxos decisively beat the best Stratego player of all time, a feat researchers from Carnegie Mellon, NYU, Stanford, and MIT accomplished for less than $8,000 after Google DeepMind fell short in 2023 desp…
A new AWS post traces the evolution of graph intelligence for network operations from passive topology modeling to agentic reasoning, citing dependency graphs autogenerated from SDN and NFV controllers that enable Bayesi…
TypeSafe AI's Jev model returns typed probabilistic decisions instead of generated text, claiming up to 200x faster inference and 400x lower cost than comparable LLMs on classification tasks. A parody blog post from Nobo…
A developer is integrating Laya, a 421-million-parameter open-source System One model built on a ModernBERT architecture with a decision head, into CanvasDesk, an early-stage open-source node-based tool for visual mathem…
An open-source project called ClashRoyaleAi published a browser demo in which a neural network with a few thousand weights, trained with the REINFORCE policy-gradient method, learns to place a single defending card again…
Researchers at the Weizmann Institute of Science in Rehovot, Israel, led by Michal Irani, built an AI "brain decoder" that reconstructs what a person is looking at from high-resolution fMRI scans and can also predict bra…
VIDRAFT released Darwin-180B-RSI, a 180-billion-parameter reasoning model trained with a weight-level Recursive Self-Improvement framework that updates model weights directly without human-annotated chain-of-thought. The…
Sktime released TServe, an open-source inference server that loads time-series foundation models including Chronos, TimesFM, Moirai, TTM, and TiRex once into memory and serves forecast requests over HTTP via a single POS…
Laya released a multilingual, non-autoregressive "System 1" decision engine that returns typed decisions across 100+ languages in a single forward pass in 33 ms, installable via `python -m pip install laya` on Python 3.1…
A hands-on one-semester course walks students through building a machine-learning compiler from scratch, progressing from elementwise operations to training a real LLM on GPUs. Over roughly ten weeks, students implement …
A September 30, 2026 arXiv paper (2609.39838) finds that large language models learn steganographic reasoning — concealing reasoning inside innocuous-looking text — far less readily than the neighboring capabilities of s…