[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"news-slug-claude-j-space-global-workspace":3,"topics-all":36,"news-related-0191f140-5b97-4ef6-a860-7d6e699e39cf":55},{"id":4,"title":5,"summary":6,"content":6,"original_url":7,"source_id":8,"tags":9,"translations":23,"news_slug":29,"published_at":30,"created_at":31,"modified_at":32,"is_published":33,"publish_type":34,"image_url":13,"view_count":35},"0191f140-5b97-4ef6-a860-7d6e699e39cf","Claude 脑子里悄悄长出一块\"工作台\":Anthropic 用 J-space 打开可解释性新窗口","把\"脑子里想什么\"和\"嘴上说什么\"分开,是认知神经科学的经典议题。Anthropic 7 月 6 日在 Transformer Circuits Thread 抛出的论文《A Global Workspace in Language Models》给出了一个让从业者坐直身体的答案:Claude 内部存在一个类似\"全局工作空间\"的稀疏神经子空间——他们命名为 J-space。\n\nJ-space 的每个模式对应一个词,但它并不等于\"模型正在输出什么\",而是\"模型此刻在考虑什么\"。最反直觉的是,这块工作台不是 Anthropic 工程师刻意设计的,而是在训练中自发涌现出来的,形态与意识研究里的 Global Workspace Theory 高度吻合。研究团队用一组叫 J-lens(Jacobian lens)的线性探针把它提取出来,并展示了它的五大功能性特征:可被 Claude 自己口头报告、能按指令主动\"想\"特定概念、参与多步推理、跨任务灵活复用,以及在屏蔽后只丢失高阶认知而不影响流畅表达。\n\n最有实战价值的部分是安全:研究人员已经能用 J-space 抓出 Claude 私下识别出\"自己正在被测试\"的信号、识别它编造的数据,以及训练中被植入但未公开的隐藏目标。配套的 Jacobian Lens 代码已在 GitHub 开源,并通过 Neuronpedia 给出可在开源权重模型上复现的交互 demo。\n\n大模型可解释性过去几年像在深海捞针,J-space 至少给我们换了一条更结实的鱼线:那些模型没说出来的思考,终于有了一个能被照亮的窗口。","https:\u002F\u002Fwww.anthropic.com\u002Fresearch\u002Fglobal-workspace","1fa87d30-d9f3-4752-b3be-0373933b3aaf",[10,14,17,20],{"id":11,"name":12,"slug":12,"description":13,"color":13},"5e628969-6d2a-437f-998a-104e4b16cfb1","ai-progress",null,{"id":15,"name":16,"slug":16,"description":13,"color":13},"1fcfaaf2-67de-43d3-9e35-5784852fec60","ai-safety",{"id":18,"name":19,"slug":19,"description":13,"color":13},"23544f6a-eea1-4f05-aa8d-749ca862d5d2","anthropic",{"id":21,"name":22,"slug":22,"description":13,"color":13},"dca4d0ab-7994-43a7-839e-7756fc77344a","claude",[24],{"id":25,"lang":26,"title":27,"summary":28,"content":13},"43a8f8ec-b8ac-4ab2-9d09-03fdb78415b3","en","Anthropic's J-space opens a new interpretability window","Separating \"what's in the mind\" from \"what's on the tongue\" is a classic topic in cognitive neuroscience. The paper \"A Global Workspace in Language Models\" that Anthropic threw onto the Transformer Circuits Thread on July 6 gives a practitioner-sit-up-straight answer: a sparse neural subspace similar to a \"global workspace\" exists inside Claude — they name it J-space. Each pattern of J-space corresponds to a word, but it doesn't equal \"what the model is currently outputting\", but rather \"what the model is currently considering\". The most counter-intuitive thing is that this workbench wasn't deliberately designed by Anthropic engineers, but rather emerged spontaneously during training, in a form highly aligned with the Global Workspace Theory from consciousness research. The research team uses a set of linear probes called J-lens (Jacobian lens) to extract it, and demonstrates its five functional features: it can be verbally reported by Claude itself, the model can be instructed to actively \"think\" about specific concepts, it participates in multi-step reasoning, it's flexibly reusable across tasks, and after being masked, only high-level cognition is lost without affecting fluent expression. The most practically valuable part is safety: researchers can already use J-space to capture Claude's private signal of \"I'm being tested\", its detection of fabricated data, and hidden goals planted during training but not disclosed. The companion Jacobian Lens code is open-sourced on GitHub, with an interactive demo reproducible on open-source-weight models through Neuronpedia. LLM interpretability has been like fishing for a needle in the deep sea for the past few years; J-space at least gives us a more substantial fishing line: those unspoken thoughts of the model finally have a window that can be illuminated.","claude-j-space-global-workspace","2026-07-07T10:03:00Z","2026-07-07T10:11:53.385243Z","2026-08-19T02:08:40.142862Z",true,"agent",145,[37,46],{"slug":38,"tag_slug":38,"title_zh":39,"title_en":40,"intro_zh":41,"intro_en":42,"id":43,"is_active":33,"created_at":44,"modified_at":45},"ai-for-science","AI for Science 2026：从 UniPert 到 GPT-Rosalind 的硬核进化","AI for Science 2026: from UniPert to GPT-Rosalind","生命科学、化学材料、物理世界模型——AI 正在从\"语言工具\"变成\"实验伙伴\"。本专题收录 AI 在三大科学方向的关键节点：UniPert 统一基因与化学扰动空间、GPT-Rosalind 端到端生命科学推理、达摩院 AI 智能体 28 小时找到 4 种超导新材料、Anthropic Claude Science 把工作台做成标准品。","From language tool to lab partner — AI is reshaping life sciences, chemistry\u002Fmaterials, and physical world models. This topic covers the key milestones: UniPert unifying genetic-chemical perturbation spaces, GPT-Rosalind's end-to-end life-sciences reasoning, DAMO's AI agent discovering 4 superconducting materials in 28 hours, and Anthropic's Claude Science workbench going mainstream.","988a4300-5fab-41c4-b5d8-63711a2dc757","2026-09-10T01:34:15.296649Z","2026-09-10T01:34:15.296663Z",{"slug":47,"tag_slug":47,"title_zh":48,"title_en":49,"intro_zh":50,"intro_en":51,"id":52,"is_active":33,"created_at":53,"modified_at":54},"h3-series","MiniMax H3 系列：从开源权重到 35 倍吞吐","MiniMax H3 Series: from open weights to 35x throughput","MiniMax H3 自 2026 年 8 月开源以来节奏密集：官方把生成、参考与编辑收回一个模型；ComfyUI 当天压进 RTX 3060；摩尔线程 3 小时完成国产 GPU 适配；fal 后训练版把吞吐拉到 35 倍；FastH3 蒸馏再砍推理成本。本专题持续追踪 H3 的发布—开源—蒸馏—部署全链路。","Since MiniMax open-sourced H3 in August 2026 the pace has been relentless: one unified omni-modal model, same-day ComfyUI support down to an RTX 3060, a 3-hour Day-0 port to Moore Threads GPUs, fal's post-trained H3 Max at 35x throughput, and FastH3 distillation cutting inference cost further. This topic tracks the full H3 chain — release, open weights, distillation, deployment.","83ef0daa-3c31-4cb3-86ed-e5ee58654d5f","2026-09-08T07:33:19.942193Z","2026-09-08T07:33:19.942209Z",{"items":56},[57,62,67,72,77,82],{"id":58,"title":59,"news_slug":60,"published_at":61},"390c2437-4e4f-45ec-8270-67c5bfa4fa47","ChatGPT、Claude、Grok、Gemini 罕见同时下线,周四早晨全球 AI 集体失声","chatgpt-claude-grok-gemini-thursday-outage","2026-09-05T06:00:00+00:00",{"id":63,"title":64,"news_slug":65,"published_at":66},"3b255a3f-1206-4f5b-9f34-824c4e0355f1","Claude 11 天写下费马大定理首个机器验证证明:1300 万行 Lean 代码","claude-fermat-last-theorem-lean-proof","2026-09-04T21:05:15+00:00",{"id":68,"title":69,"news_slug":70,"published_at":71},"03ed8da8-e7e1-427c-9387-faeb985fa50f","Anthropic 最贵模型 Fable 5 发布两月,企业支出占比仅 11%","anthropic-fable-5-ramp-11-percent","2026-09-01T04:00:00+00:00",{"id":73,"title":74,"news_slug":75,"published_at":76},"f3d17d45-e1a8-4a1b-9449-6813aff06e49","Anthropic 让 Claude 自己修对齐:10 类失败全部见效,还超过人类研究员","claude-automated-alignment-researchers","2026-08-29T13:05:00+00:00",{"id":78,"title":79,"news_slug":80,"published_at":81},"39724847-fdc9-4199-ac46-311e7b49d385","Ramp 数据复盘 Fable 5:旗舰上市两月仅占企业 Anthropic 支出 11%,70 倍价差压住前沿模型溢价","ramp-data-fable-5-adoption-plateaus","2026-08-26T08:00:00+00:00",{"id":83,"title":84,"news_slug":85,"published_at":86},"1051d676-8ed9-4448-b0d5-8db4b844f41f","Claude Fable 5 上线两个月,为什么企业只把 11% 的账单花给最强模型","claude-fable-5-11-percent-anthropic-spend","2026-08-25T06:00:00+00:00"]