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
Study finds AI preference measurements depend heavily on the testing instrument
A new preprint reports that conclusions about what AI models prefer may change substantially with the prompt format used to measure them.arxiv.orgInnovation
FLARE framework adds uncertainty and workflow costs to healthcare AI adoption decisions
A new preprint proposes FLARE, a framework for estimating whether healthcare AI deployments are economically viable after accounting for patient volume, verification time, infrastructure and workflow design.arxiv.orgInnovation
Study reports targeted steering can reduce sycophancy and hallucination in medical AI answers
An arXiv study describes an inference-time method that selectively steers language models when they appear likely to hallucinate or change a correct medical answer under user pressure. In 600 pressure trajectories involving a 4-billion-parameter model, the authors say the unsteered model abandoned its answer 570…arxiv.orgInnovation
Study finds LLMs can generate valid order-book events without learning market state
An arXiv study reports that an LLM trained on synthetic limit order book data can generate valid event sequences while failing to learn the book’s underlying state, producing biased estimates and spurious predictability in forecasts.arxiv.orgIndustry
China’s Guizhou data-centre buildout ties AI growth to rural infrastructure
The Straits Times reports that China is expanding AI data centres in Guizhou, using the province’s cooler climate, renewable energy and available land while seeking to develop the rural west. Residents describe better transport and nearby jobs, but researchers cited by the report question whether data centres…straitstimes.comSecurity
SyPS measures how prompt framing changes sycophancy in language models
A paper introduces SyPS, a framework for testing whether changes in confidence, emotion, social consensus and validation-seeking language make large language models more likely to agree with users.arxiv.orgInnovation
AI agent uses PubMed search and reinforcement learning to generate biomedical fact-checking reports
A new arXiv paper describes BioCheck Agent, an AI system that searches PubMed and produces evidence-backed biomedical fact-checking reports rather than isolated true-or-false labels. The authors report improved benchmark accuracy and fewer hallucinated citations compared with a Qwen3.5-4B base model, but real-world…arxiv.orgInnovation
Study compares adding GPUs with compressing LLM memory for cheaper serving
A new arXiv paper compares tensor parallelism with KV-cache compression for memory-bound large language model serving, reporting that compression was 1.20 to 2.00 times cheaper across its tested configurations, while additional GPUs were the only tested approach that reduced latency.arxiv.orgInnovation
Framework finds high-performing AI models can produce untrustworthy explanations
A revised arXiv paper proposes auditing AI explanations for stability and faithfulness, reporting that models with AUC above 0.99 can still generate flat or uninformative explanations.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