Higher education
fromThe Verge
3 days agoAI-generated research papers are overwhelming peer review
A paper was cited excessively because many new studies used the same dataset and produced low-quality, AI-assisted predictions at scale.
Half of laboratory mice are not what scientists think they are, a genetic analysis of hundreds of strains that are distributed globally for animal research has found. The study, published today in Science, uncovered widespread inconsistencies between the reported names of mouse strains and their actual genetic makeup. The mismatches have the potential to compromise the reproducibility of mouse studies and undermine research conclusions, scientists say.
Calling nanoscientists: your field needs you to try to replicate a landmark finding that quantum dots can act as biosensors inside living cells. As part of the first large-scale effort in the physical sciences to tackle the reproducibility crisis, researchers in France and the Netherlands are offering funds and resources in exchange for a few months of work. "We are trying to use
A research team from Stanford University has released Paper2Agent, a framework that automatically converts scientific papers into interactive AI agents. The system, introduced in a recent paper, aims to make research methods more accessible by transforming traditional publications into dynamic entities that can execute analyses, reproduce results, and respond to new scientific queries through natural language interaction. Paper2Agent builds on the Model Context Protocol (MCP), a standard that allows large language models to connect with external tools and datasets.
Marimo notebooks provide an innovative approach to interactivity in coding, featuring sandboxing for environment management and enabling reproducibility in sharing notebooks.