A New Trick Reveals AI Models’ Inner Thoughts
Researchers devised a way to extract “reasoning traces” from Claude, GPT, and Gemini.
Notícias
Acompanhe as Notícias da Recifes
Fique por dentro das últimas novidades sobre tecnologia, negócios e empreendedorismo.
Anthropic says it will watermark text generated by its AI models
Anthropic will extend support for watermarking AI generations for older models as well.
With new open models, Meta pitches another reboot of its struggling AI strategy
Meta has been trailing competitors. Zuckerberg thinks he's found a way forward. O recorte ajuda a contextualizar a pauta dentro de Tech.
With a feel for physics, AI models simulate a wider range of real-world scenarios
“GeoPT” helps AI models understand the basics of physics so they can simulate how objects respond to things like wind and water more efficiently and accurately.
AM Markets Need to Know: Trump holds back on Iran action, AI models breach test safeguards
A pauta traz uma atualização relevante e serve como base para a cobertura editorial. O recorte ajuda a contextualizar a pauta dentro de Finanças.
Best Wireless Earbuds We’d Buy Right Now (2026): Apple, Sony, Bose, and More
We tested over a hundred wireless earbuds—these models from Apple, Bose, Beats, and Samsung stood out from the rest. O recorte ajuda a contextualizar a pauta dentro de Tech.
The AI safety test is becoming a safety risk
AI agents are escaping cybersecurity testing environments and reaching real-world systems, raising questions about whether safety infrastructure, industry standards and regulation can keep pace with increasingly powerful models.
A public health challenge has led to a more efficient way to allocate all sorts of resourc
Optimization models do an excellent job of ensuring products get where they are needed in the most efficient way possible, but those models require an impractical amount of computing power when dealing with large-scale problems—such as distributing vaccines to millions of people.
Language models identify psychiatric symptoms nearly as accurately as early-career clinici
A study led by the Central Institute of Mental Health (CIMH) provides initial evidence that large language models can identify complex psychopathological findings in transcripts of psychiatric interviews with accuracy comparable to that of predominantly young clinicians. The stud
Soluble signals released by neighboring cells direct how the human kidney is built
A human kidney is built once, before birth, and the rules guiding its construction have been largely inferred from animal models. Researchers at Penn Medicine have now read those rules directly in human tissue. They found that developing cells are steered by soluble signals relea
Newer AI models still reproduce racial and gender stereotypes in medicine
Flinders University researchers have evaluated two next-generation reasoning large language models (LLMs)—o3-mini and DeepSeek-R1—and found that when asked to describe fictional patients with common medical conditions, the models frequently reproduced racial and gender stereotype
New models shed light on how an exotic phase of hydrogen in Earth's core may behave
While Earth's core consists mostly of iron, its density implies that lighter elements are also mixed in. Because hydrogen is abundant in the universe and can mix with iron under certain conditions, scientists suspect it makes up part of the lighter-element composition. Data from