Announcement
Hydrogen-1.0 unveiled as an open-weight model for scientific research
Hydrogen-1.0's creators say they post-trained GLM-5.3 on internal scientific data, lab records and discovery trajectories.
TLDR
The creators report that Hydrogen-1.0 scored 38.3% on FrontierScience-Research, compared with 33.3% for GPT-6-Astra, and 93.0% on GPQA-Diamonds. They argue that open weights let researchers inspect and adapt models for scientific discovery.
Combined views
1.5K
9 Sources, first seen ago
