Researchers improve AI’s ability to learn new tasks without sacrificing performance
Peer-Reviewed Publication
Updates every hour. Last Updated: 13-Jun-2026 05:16 ET (13-Jun-2026 09:16 GMT/UTC)
A new framework allows AI models that have already been trained to learn new tasks without sacrificing performance when performing old tasks. The framework, called CHEEM, also improves an AI model’s operating efficiency by using fewer computational steps to perform simpler tasks.
A new study led by Dr Lin Su of Queen Mary University of London, published today in the Journal of the American Chemical Society, describes a new integrated solar reactor in which engineered Escherichia coli (E. coli) are grown directly inside the same liquid that converts CO₂ into a usable energy source using sunlight. In future, this technology may be used to make environmentally clean chemicals, plastics or even microbial protein.
A research paper by scientists from Nanjing Medical University developed a biomimetic nanoparticle (NP) system composed of macrophage-membrane-coated Prussian blue@KGN (KGN@PB@CM) loaded into a thermosensitive Pluronic@F127/hyaluronic acid (HA-F127) hydrogel for synergistic tendon healing.
The new research paper, published on Apr. 23 in the journal Cyborg and Bionic Systems, developed a biomimetic nanoparticle system that simultaneously tackles clinical challenges, accelerating tendon repair through a “dual‑modulation” approach.
In a new study published in Conservation Letters, scientists built the most comprehensive models ever of the ranges of 2,858 important fungal species and compared them with the world’s protected landscapes. The findings highlight a mismatch: Conservation has been built around the life we see, without regard to the hidden organisms below our feet that help aboveground life thrive. These range maps offer a roadmap to begin charting a new course.
A new study of the largest dam removal project in United States history on the Klamath River in Oregon and California is offering new insight into a long-running water conflict by finding that farmers and conservation groups share priorities that may help guide decision-making on future river restoration projects.
A research paper by scientists at Hangzhou Dianzi University presented a HER2+ breast cancer response prediction model based on hierarchical tissue specific modeling of pathological images.
The research paper, published on Apr 22, 2026 in the journal Cyborg and Bionic Systems.