{"slug": "dory1-0-4b-a-looped-normalized-gpt", "title": "Dory1.0-4B: A Looped Normalized-GPT", "summary": "A 4-billion-parameter looped (latent-recurrent) language model called Dory1.0-4B was introduced, built on a normalized Transformer (nGPT) backbone with interleaved sliding-window and full-attention layers. Dory1.0-4B offers elastic depth, meaning a single checkpoint serves different inference compute budgets.", "body_md": "We introduce Dory1.0-4B, a 4-billion-parameter looped (latent-recurrent) language model with elastic depth: a single checkpoint serves different inference compute budgets. Dory1.0-4B is built on a normalized Transformer (nGPT) backbone with interleaved sliding-window and full-attention layers, organ", "url": "https://wpnews.pro/news/dory1-0-4b-a-looped-normalized-gpt", "canonical_source": "https://aiflash.com/news/145469/", "published_at": "2026-11-04 14:42:49+00:00", "updated_at": "2026-11-04 14:49:48.532259+00:00", "lang": "en", "topics": ["large-language-models", "artificial-intelligence", "machine-learning", "ai-research"], "entities": ["Dory1.0-4B"], "also_reported_by": [], "alternates": {"html": "https://wpnews.pro/news/dory1-0-4b-a-looped-normalized-gpt", "markdown": "https://wpnews.pro/news/dory1-0-4b-a-looped-normalized-gpt.md", "text": "https://wpnews.pro/news/dory1-0-4b-a-looped-normalized-gpt.txt", "jsonld": "https://wpnews.pro/news/dory1-0-4b-a-looped-normalized-gpt.jsonld"}}