Tag: AI Research
Sakana builds a reviewer that catches most claim errors
The Japanese lab's automated peer-review system flags 73 percent of faulty core claims in submitted papers.
Odyssey opens a world model you can walk through
The startup put Odyssey-3 into a free public preview that generates explorable environments in real time.
arXiv caps submissions as AI papers overwhelm its moderators
The preprint server will allow two submissions per person each month after AI-driven volumes doubled in two years.
Microsoft builds a biology world model to cut lab guesswork
Quine pairs a multimodal model of biological systems with a harness that runs experiments computationally first.
Google Research hands video generation a bandit-driven director
Four agentic frameworks target the drift and cascading failures that break long AI video, and one of them…
OpenAI spends most of its research on models it has not shipped
A senior researcher says 80 to 90 percent of the lab's effort targets GPT-7 and later because that…
Black Forest Labs releases an open robotics world model
FLUX 3 Action predicts the next move and the next state of the world from camera feeds alone.
Anthropic’s Claude agents flag a new enzyme family in viral DNA
Roughly 950 Claude agents searching bacterial virus sequences surfaced a previously uncharacterised enzyme system, the first result from…
OpenAI gives mathematicians a voice but hands them no brake
The new advisory group gives mathematicians influence over how OpenAI presents results, but no say over how fast…
SoL-Pi cuts coding agent token bills by rewriting the harness
The four mechanisms came out of automated research loops and trim token traffic by nearly half on a…
Anthropic confirms a Bay Area wet lab for model-run biology
The company says its models need real experiments, not only simulated ones, to matter in biology.
Anthropic puts a number on how much Claude runs its own research
A prototype index says Claude leads 26 percent of Anthropic's AI research and development work.