{"version":1,"type":"story","url":"https://digestai.news/story/anthropic-s-claude-begins-dreaming-to-reorganize-memory-in-tests","json":"https://digestai.news/story/anthropic-s-claude-begins-dreaming-to-reorganize-memory-in-tests.json","markdown":"https://digestai.news/story/anthropic-s-claude-begins-dreaming-to-reorganize-memory-in-tests.md","slug":"anthropic-s-claude-begins-dreaming-to-reorganize-memory-in-tests","headline":"Anthropic's Claude begins 'dreaming' to reorganize memory in tests","summary":"Anthropic announced that its Claude model can enter a post‑dialogue \"dream\" phase, using idle compute to replay the day’s interaction logs. In these dreams the model corrects basic errors, extracts strategy tips and builds a pre‑loaded Wiki that can be consulted in future sessions. The blog notes that the idea is separate from updating model weights and that any link to the upcoming \"Model 4\" or recursive self‑improvement is speculative, not an official claim.\n\nThe company demonstrated the approach with two experiments. In a Pokémon Red playthrough, the memory‑lean Fable 5 wrote only nine files (plus a global note) and earned four badges, while the older Sonnet 3.7 logged 413 files across 225 folders and secured just one badge. A separate corporate test fed Claude 350,000 internal documents (about 420 million tokens) from the software firm Stainless. Raising the number of dream sessions from two to 128 lifted the Wiki‑only scoring‑criteria satisfaction rate from 13% to 48%, and to 61% when combined with search tools. The work draws on the \"sleep‑time compute\" concept proposed by researchers at UC Berkeley.","keyPoints":["Claude uses \"dream\" sessions to review logs, correct errors and build an external Wiki","Increasing dream sessions from 2 to 128 on 350,000 files raised Wiki‑only scoring from 13% to 48% (61% with search)","In a Pokémon Red test, memory‑efficient Fable 5 wrote 9 files and earned 4 badges, versus Sonnet 3.7’s 413 files and 1 badge"],"whyItMatters":"Shows a practical method to extend AI memory without larger context windows, improving consistency on corporate documents and agent tasks, which could shape future AI agent designs.","category":{"slug":"research","name":"Research","url":"https://digestai.news/category/research"},"entities":{"companies":["Anthropic","UC Berkeley","Stainless"],"models":["Claude","Opus 5","Sonnet 3.7","Fable 5"],"people":["Xin Zhi Yuan","Aeneas","Letta"]},"firstPublishedAt":"2026-11-06T22:10:00Z","updatedAt":"2026-11-06T22:10:00Z","sourceCount":1,"hasPrimarySource":false,"sources":[{"outlet":"note.com","title":"Claude has started to 'dream'! How does 'Dreaming' change the memory of a forgetful AI?","url":"https://note.com/trans_n_ai/n/na9810ff65a7b?hl=en","publishedAt":"2026-11-06T22:10:00Z","type":"press","primary":false,"lead":true}],"sourceNotes":null,"discussions":[],"thread":null,"cite":{"text":"Digest AI, \"Anthropic's Claude begins 'dreaming' to reorganize memory in tests\", 6 November 2026, https://digestai.news/story/anthropic-s-claude-begins-dreaming-to-reorganize-memory-in-tests","publisher":"Digest AI","title":"Anthropic's Claude begins 'dreaming' to reorganize memory in tests","datePublished":"2026-11-06T22:10:00Z","url":"https://digestai.news/story/anthropic-s-claude-begins-dreaming-to-reorganize-memory-in-tests"},"generatedBy":"Written by Digest AI's editorial model from the linked sources; the sources are the record.","license":"Headlines, digests and key points are written by Digest AI and may be quoted with a link to the story page. Linked articles belong to their publishers. Terms: https://digestai.news/terms#reuse"}