{"id":4584,"date":"2025-11-09T06:37:25","date_gmt":"2025-11-09T06:37:25","guid":{"rendered":"https:\/\/172-234-197-23.ip.linodeusercontent.com\/?page_id=4584"},"modified":"2025-11-09T06:37:25","modified_gmt":"2025-11-09T06:37:25","slug":"transformer-feature-fusion-for-iqfft","status":"publish","type":"page","link":"https:\/\/neurosphere-2.tail52f848.ts.net\/wordpress\/?page_id=4584","title":{"rendered":"Transformer Feature-Fusion for IQ+FFT"},"content":{"rendered":"\n<div data-wp-interactive=\"core\/file\" class=\"wp-block-file\"><object data-wp-bind--hidden=\"!state.hasPdfPreview\" hidden class=\"wp-block-file__embed\" data=\"https:\/\/neurosphere-2.tail52f848.ts.net\/wordpress\/wp-content\/uploads\/2025\/11\/Transformer-Feature-Fusion-for-IQFFT-bgilbert1984.pdf\" type=\"application\/pdf\" style=\"width:100%;height:600px\" aria-label=\"Embed of Transformer Feature-Fusion for IQ+FFT bgilbert1984.\"><\/object><a id=\"wp-block-file--media-6897ba2d-15da-41cd-a16e-0e33f2a9bb27\" href=\"https:\/\/neurosphere-2.tail52f848.ts.net\/wordpress\/wp-content\/uploads\/2025\/11\/Transformer-Feature-Fusion-for-IQFFT-bgilbert1984.pdf\">Transformer Feature-Fusion for IQ+FFT bgilbert1984<\/a><a href=\"https:\/\/neurosphere-2.tail52f848.ts.net\/wordpress\/wp-content\/uploads\/2025\/11\/Transformer-Feature-Fusion-for-IQFFT-bgilbert1984.pdf\" class=\"wp-block-file__button wp-element-button\" download aria-describedby=\"wp-block-file--media-6897ba2d-15da-41cd-a16e-0e33f2a9bb27\">Download<\/a><\/div>\n\n\n\n<p class=\"wp-block-paragraph\">We fuse per-timestep spectral context with temporal<br>I\/Q by repeating a pooled FFT magnitude vector across time and<br>concatenating it to the I\/Q channels. A small Transformer over<br>tokens (T= 128) learns cross-time interactions. We ablate the<br>fusion width W (spectral channels per timestep) and report p50<br>latency vs dmodel.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\"><\/p>\n","protected":false},"excerpt":{"rendered":"<p>We fuse per-timestep spectral context with temporalI\/Q by repeating a pooled FFT magnitude vector across time andconcatenating it to the I\/Q channels. A small Transformer overtokens (T= 128) learns cross-time interactions. We ablate thefusion width W (spectral channels per timestep) and report p50latency vs dmodel.<\/p>\n","protected":false},"author":2,"featured_media":4586,"parent":0,"menu_order":0,"comment_status":"closed","ping_status":"closed","template":"","meta":{"neve_meta_sidebar":"","neve_meta_container":"","neve_meta_enable_content_width":"","neve_meta_content_width":0,"neve_meta_title_alignment":"","neve_meta_author_avatar":"","neve_post_elements_order":"","neve_meta_disable_header":"","neve_meta_disable_footer":"","neve_meta_disable_title":"","footnotes":""},"class_list":["post-4584","page","type-page","status-publish","has-post-thumbnail","hentry"],"_links":{"self":[{"href":"https:\/\/neurosphere-2.tail52f848.ts.net\/wordpress\/index.php?rest_route=\/wp\/v2\/pages\/4584","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/neurosphere-2.tail52f848.ts.net\/wordpress\/index.php?rest_route=\/wp\/v2\/pages"}],"about":[{"href":"https:\/\/neurosphere-2.tail52f848.ts.net\/wordpress\/index.php?rest_route=\/wp\/v2\/types\/page"}],"author":[{"embeddable":true,"href":"https:\/\/neurosphere-2.tail52f848.ts.net\/wordpress\/index.php?rest_route=\/wp\/v2\/users\/2"}],"replies":[{"embeddable":true,"href":"https:\/\/neurosphere-2.tail52f848.ts.net\/wordpress\/index.php?rest_route=%2Fwp%2Fv2%2Fcomments&post=4584"}],"version-history":[{"count":0,"href":"https:\/\/neurosphere-2.tail52f848.ts.net\/wordpress\/index.php?rest_route=\/wp\/v2\/pages\/4584\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/neurosphere-2.tail52f848.ts.net\/wordpress\/index.php?rest_route=\/wp\/v2\/media\/4586"}],"wp:attachment":[{"href":"https:\/\/neurosphere-2.tail52f848.ts.net\/wordpress\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=4584"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}