News
What happened
Results, releases, policy and industry moves worth recording. Every item links a primary source and says in one line why it matters. Dated by when the thing happened, not by when someone got round to filing it.
category
sort
-
August 20251
- N-0003 2025-08-02 EU AI Act obligations for general-purpose AI models take effect Regulation (EU) 2024/1689 — EUR-Lex The general-purpose AI provisions of Regulation (EU) 2024/1689 became applicable, imposing documentation, copyright-policy and training-data-summary duties on providers of GPAI models placed on the EU market, with additional obligations for models deemed to carry systemic risk. 2025-08-02 · N-0003 policy
-
January 20251
- N-0002 2025-01-22 DeepSeek-R1 shows reasoning emerging from reinforcement learning alone arXiv:2501.12948 Reports that long-form reasoning behaviour can be elicited by reinforcement learning on outcome rewards without a supervised fine-tuning stage, with weights released. The claim that matters is the negative one: the supervised reasoning traces widely assumed to be necessary appear not to be. 2025-01-22 · N-0002 research
-
July 20241
- N-0001 2024-07-31 Llama 3.1 405B released with open weights arXiv:2407.21783 — The Llama 3 Herd of Models Meta published the Llama 3 technical report alongside open weights for models up to 405B parameters. The significance for general-purpose work is distributional rather than architectural: capability near the contemporary frontier became something a research group could run and modify locally rather than only query through an API. 2024-07-31 · N-0001 release
no items match that filter