Storylines / Organisation
Stanford University
6 developments involving Stanford University in 3 countries and bodies. The most recent is DeepSeek memory-saving technology may reduce model accuracy, researchers find, 12 Oct 2026.
KeyStoryAlso on another Storyline
- DeepSeek memory-saving technology may reduce model accuracy, researchers find
The paper is a rare challenge to the memory-saving technology that has helped DeepSeek close the gap with OpenAI and Anthropic at low cost, though the study has not been peer-reviewed.
- Stanford researchers introduce Next-Concept Modeling, generating text coarse to fine in parallel
It is a faster alternative to diffusion language models for parallel text generation, tested so far on research benchmarks at matched scale, not against frontier chat models.
- Google DeepMind publishes SynthID Bio watermarking for AI-designed proteins in Nature
DNA-synthesis screeners can use the open code to check whether an unfamiliar sequence came from a vetted model, and watermarked designs showed similar binding in lab tests.
- AI-designed bacteriophages kill E. coli strains resistant to a natural phage
AI-designed viruses that kill resistant bacteria point to new treatments, and sharpen the biosecurity question of what else such models can design.
- Google unveils Gemini-based ‘AI co-scientist’ with lab-validated hypotheses
Researchers in leukaemia, liver fibrosis and antimicrobial resistance got AI-generated hypotheses that lab work supported in leukaemia and liver fibrosis, and in antimicrobial resistance the system reproduced an unpublished, already-validated finding.
- Trial finds GPT-4 access did not improve doctors’ diagnoses, but GPT-4 alone scored higher
Hospitals giving clinicians chatbots had trial evidence that access alone did not raise diagnostic scores, while the model working alone scored higher than unaided doctors.