{"rewrite":{"id":"r_503725c970f02623e733e174","clusterId":"c_1bc81360a7855898a0decf98","slug":"anthropic-finds-ai-j-space-that-mirrors-human-conscious-processing","model":"deepseek-v4-flash","headline":"Anthropic Finds AI 'J-Space' That Mirrors Human Conscious Processing","summary":"Researchers at Anthropic have identified a distinct neural workspace in their Claude AI model, dubbed 'J-space,' that activates when the model performs complex reasoning tasks. This workspace emerged naturally during training, not from explicit design. Removing J-space severely degraded Claude's ability to perform multi-step reasoning, summarization, and poetry, while simple tasks remained unaffected. The discovery provides a new tool for studying AI cognition and its parallels to human conscious processing.","whyItMatters":"The discovery of J-space offers a concrete, testable model of 'access consciousness' in large language models, potentially bridging AI safety research and neuroscience.","webCardHtml":"\u003cp\u003eJ-space is not a programmed feature but a spontaneously formed structure within Claude\u0026#39;s neural network. It holds information beyond the immediate text, such as identifying errors in code or detecting prompt injection attacks. In experiments, manipulating the contents of J-space directly changed Claude\u0026#39;s reported thoughts. Anthropic notes that J-space is limited in capacity, holding only a few dozen concepts at a time, and accounts for less than a tenth of the model\u0026#39;s total activity. The company emphasizes that this does not mean Claude experiences consciousness like a human, but it may provide insights into access consciousness.\u003c/p\u003e","blueskyPost":"Anthropic researchers found a natural 'J-space' in Claude that handles complex reasoning. It's not designed but emerged. Removing it tanks advanced tasks. This could help study AI consciousness.","twitterPost":"Anthropic's Claude has a 'J-space' - a natural workspace for complex reasoning. It forms spontaneously, not by design. Removing it kills multi-step reasoning. A new angle on AI consciousness and safety.","threadsPost":null,"newsletterBlurb":"Anthropic researchers have identified a neural workspace in Claude, called J-space, that emerged naturally during training. This region handles complex reasoning and can be manipulated to change the model's outputs. The finding offers a new framework for exploring AI cognition and its parallels to human consciousness.","attributionJson":"[{\"source\":\"GIGAZINE\",\"url\":\"https://gigazine.net/news/20260707-ai-j-apace/\",\"title\":\"AIが指示を処理するとき「思考」に特化した「J空間」が活性化していることが判明\"}]","lintFlagsJson":null,"lintHits":0,"costUsd":0,"inputTokens":5467,"outputTokens":2278,"status":"published","repairAttempts":0,"nextRepairAt":null,"factsAttemptedAt":1783416032,"createdAt":"2026-07-07T09:10:40.000Z","publishedAt":"2026-07-07T09:13:32.000Z","updatedAt":"2026-07-07T09:10:40.000Z"},"cluster":{"id":"c_1bc81360a7855898a0decf98","canonicalTitle":"AIが指示を処理するとき「思考」に特化した「J空間」が活性化していることが判明","representativeArticleId":"a_88c50203ca0f0b3b981d914f","sourceCount":1,"writtenSourceCount":1,"writeAttempts":0,"isSolo":true,"entitiesJson":"{\"anime_titles\":[],\"manga_titles\":[],\"work_titles\":[],\"studios\":[],\"people\":[],\"type\":\"news\",\"domain\":\"other\",\"is_roundup\":false}","contentType":"news","status":"published","firstSeenAt":"2026-07-07T08:38:00.000Z","lastSeenAt":"2026-07-07T08:38:00.000Z","updatedAt":"2026-07-07T09:13:32.000Z"},"attribution":[{"source":"GIGAZINE","url":"https://gigazine.net/news/20260707-ai-j-apace/","title":"AIが指示を処理するとき「思考」に特化した「J空間」が活性化していることが判明"}],"entities":{"anime_titles":[],"manga_titles":[],"work_titles":[],"studios":[],"people":[],"type":"news","domain":"other","is_roundup":false},"keyFacts":null}