3 ms·
Outrageously Small Neural Networks: Emergent Basic Reasoning at 6,616 tok/SEC [pdf]
- gdiamos 1mo agoThe loss does not saturate. Across a 4.91B-token run, smoothed training loss falls monotonically within each curriculum phase and is still descending at the end
- gdiamos 1mo agoblog: https://gregdiamos.com/2026/09/07/outrageously-small-neural-networks.html https://gregdiamos.com/2026/09/07/outrageously-small-neural-... X discussion: https://x.com/GregoryDiamos/status/2096873745420075020?s=20 https://x.com/GregoryDiamos/status/2096873745420075020?s=20 I added some of the main points to the thread so they are easier to read.
- gdiamos 1mo ago[flagged]
- gdiamos 1mo agoData is doing more of the work than it used to. Every source in our mixture is a curated artifact built with large models Training a model this small on them is distillation When models of this size were last studied seriously such corpora did not exist
- deadcatfound 1mo ago[dead]