Updated daily708 verified stories
Research News tracker.
Material AI papers, studies, datasets, evaluations, and scientific advances with direct links to the underlying research.
- 30stories this week
- 708sources cited
- 708verified timeline entries
Verified sourcing
Every story links to the strongest available evidence: original sources when available, otherwise clearly attributed reporting.
Plain English
What happened, why it matters, and what to watch — without the jargon.
No filler
When the signal is thin, we publish nothing rather than padding the feed.
More stories
9 storiesInnovation
Review finds foundation models have not generally replaced specialized machine-learning architectures
A review of 159 papers argues that language-based foundation models can be highly competitive in selected tasks, but finds no evidence of broad architectural replacement when researchers directly test whether models preserve and compute the underlying structure of data.arxiv.orgInnovation
Constrained LLM system reports safer kitchen-robot manipulation in small physical tests
A preprint describes an LLM-controlled robotic-arm system that rejects malformed commands and verifies planned motions against collision and kinematic constraints before acting.arxiv.orgInnovation
Study finds a single mean shift dominates LLM hallucination detection
A paper accepted to EMNLP 2026 reports that a simple linear probe detected hallucinations more reliably than twelve tested architectural alternatives in a controlled evaluation.arxiv.orgInnovation
Training-free method speeds up vision-language-action models on edge hardware, paper reports
AdaVLA is an online method that the authors say accelerates flow-matching vision-language-action models without retraining or access to training data. On a Jetson AGX Orin, they report speedups of 1.87× for π0.5 and 2.24× for X-VLA with negligible success-rate degradation.arxiv.orgInnovation
RoboPhys-3D benchmark tests whether embodied world models preserve 3D scenes and actions
A new preprint proposes evaluating video world models with 3D reconstruction and task-oriented metrics, reporting that visual judgments can miss failures tied to state understanding and executable actions.arxiv.orgInnovation
Paper introduces public AI benchmark for sorting shredded black industrial plastics
Researchers describe a machine-learning pipeline and public hyperspectral-imaging dataset designed to identify four polymers in shredded black plastic from end-of-life vehicles.arxiv.orgInnovation
Study finds apparent stroke-treatment gains from offline reinforcement learning are confounded
A 129,033-patient stroke registry study found that offline reinforcement-learning policies appeared to outperform physician decisions, but the estimated improvement weakened substantially after researchers removed baseline-severity information embedded in the reward.arxiv.orgInnovation
Study finds anytime-valid AI monitors can repeatedly trigger on real forecast streams
A new preprint reports that a statistical monitor behaved as expected on exchangeable synthetic data but fired in every tested clean run on five real forecasting streams, warning that deployment guarantees depend on assumptions that may fail in adaptive AI systems.arxiv.orgInnovation
Paper proposes a two-step test for choosing LLMs under uncertain evaluations
A new preprint argues that companies can sometimes certify the best assignment of large language models to recurring workloads even when model-quality estimates remain uncertain. It proposes a two-solve test and an evidence-gathering method called CASE.arxiv.org
One useful briefing each week
Keep up with AI without living in the feed.
Get the week’s verified AI news, original data, useful tools, learning picks, and fresh AI jobs.
Reach people who are learning AI
Hiring an AI professional or launching a useful AI product? Put it in front of people who came here to learn and act.
Post an AI jobSubmit an AI tool