{"id":137,"date":"2026-05-29T01:28:17","date_gmt":"2026-05-29T05:28:17","guid":{"rendered":"https:\/\/millionminds.one\/notes\/?p=137"},"modified":"2026-05-29T01:31:27","modified_gmt":"2026-05-29T05:31:27","slug":"may-29-2026","status":"publish","type":"post","link":"https:\/\/millionminds.one\/notes\/may-29-2026\/","title":{"rendered":"May 29, 2026"},"content":{"rendered":"\n<p>I&#8217;ve been making progress..<\/p>\n\n\n\n<p>TextSensor is built and acts like a simple eye over text. It has multiple patches, moves with a saccade\/fixation rhythm, and emits sparse feature SDRs from local character windows.<\/p>\n\n\n\n<p>Layer-1 columns now receive those SDRs, maintain monotonic pose, learn temporal transitions with a graph-based memory, and settle into private object SDRs.<\/p>\n\n\n\n<p>The big recent shift was making layer 1 preserve retinotopic structure instead of collapsing local observations into a bag of features. That let columns form stable visual token objects like words, punctuation-attached words, repeated-letter forms, and digit\/symbol tokens.<\/p>\n\n\n\n<p>Also added stress tests to check what the representations actually encode. Exact token forms are now stable and distinct. The remaining abstraction gaps are also informative: for example, word, and word. are different layer-1 visual objects, while a higher layer may later learn their relationship.<\/p>\n\n\n\n<p>Most recently, I started wiring hierarchy. Layer-1 object SDRs now project upward through a sparse InterLayerRelay component into layer-2 columns. The wiring is not all-to-all and not one-to-one: it is sparse, overlapping, deterministic-random, and designed to become plastic later.<\/p>\n\n\n\n<p>It is still early, but the architecture is starting to take shape:<\/p>\n\n\n\n<ul class=\"wp-block-list\">\n<li>sensor-like input<\/li>\n\n\n\n<li>sparse distributed representations<\/li>\n\n\n\n<li>independent columns with private codes<\/li>\n\n\n\n<li>movement and pose<\/li>\n\n\n\n<li>prediction and surprise<\/li>\n\n\n\n<li>stable object formation<\/li>\n\n\n\n<li>lateral context<\/li>\n\n\n\n<li>sparse bottom-up hierarchy<\/li>\n<\/ul>\n\n\n\n<p>The next step is to characterize what layer 2 learns from the layer-1 object surface, without hardcoding concepts like \u201cphrase,\u201d \u201csentence,\u201d or \u201ctopic.\u201d Those names should be diagnostics, not mechanisms.<\/p>\n","protected":false},"excerpt":{"rendered":"<p>I&#8217;ve been making progress.. TextSensor is built and acts like a simple eye over text. It has multiple patches, moves with a saccade\/fixation rhythm, and emits sparse feature SDRs from local character windows. Layer-1 columns now receive those SDRs, maintain monotonic pose, learn temporal transitions with a graph-based memory, and settle into private object SDRs. [&hellip;]<\/p>\n","protected":false},"author":1,"featured_media":0,"comment_status":"closed","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[10],"tags":[],"class_list":["post-137","post","type-post","status-publish","format-standard","hentry","category-update"],"_links":{"self":[{"href":"https:\/\/millionminds.one\/notes\/wp-json\/wp\/v2\/posts\/137","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/millionminds.one\/notes\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/millionminds.one\/notes\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/millionminds.one\/notes\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/millionminds.one\/notes\/wp-json\/wp\/v2\/comments?post=137"}],"version-history":[{"count":2,"href":"https:\/\/millionminds.one\/notes\/wp-json\/wp\/v2\/posts\/137\/revisions"}],"predecessor-version":[{"id":141,"href":"https:\/\/millionminds.one\/notes\/wp-json\/wp\/v2\/posts\/137\/revisions\/141"}],"wp:attachment":[{"href":"https:\/\/millionminds.one\/notes\/wp-json\/wp\/v2\/media?parent=137"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/millionminds.one\/notes\/wp-json\/wp\/v2\/categories?post=137"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/millionminds.one\/notes\/wp-json\/wp\/v2\/tags?post=137"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}