Chinese frontier lab. DeepSeek trains and releases open-weight frontier models and publishes the architecture research behind them, which is how it enters this corpus: the Manifold-Constrained Hyper-Connections paper that widens the residual stream at no extra compute. Its later releases become the corpus’s ruler for the open frontier’s lag, an open 1.6T model landing four to five months behind, and its models are among those whose serving stack self-improving agents rebuild.
Organization
DeepSeek
Chinese frontier lab whose open-weight models and architecture papers set the pace of the Chinese open frontier throughout the corpus.
Sources
Attribution and reporting linked by this record; links do not imply independent verification.
- Welcome to January 2, 2026 (theinnermostloop.substack.com)
Development timeline
Newest reporting dates first. Reporting dates may differ from event dates.
Knowledge graphopen in browser →
Referenced by
Cite or correct this record
Suggest a correction — opens a GitHub draft for you to review and submit.