Loading
What changed between revision 1 and revision 2, compared over the article's markdown source with its frontmatter — so a change to status, evidence or reversibility shows up as the edit it is.
Older · revision 1
Import· initial corpus import, not an edit
Initial import of content/neural-decoding.md — the filesystem corpus, unchanged. Not an edit.
11.8 kB
Revision 1
Revision 2
lines 64–70 → 64–7063 unchanged lines not shown
Modern decoders are recurrent or transformer-based networks trained on many hours of data, oftenModern decoders are recurrent or transformer-based networks trained on many hours of data, oftenpooled across sessions. For [[speech-neuroprosthesis]] systems they emit phoneme sequences that apooled across sessions. For [[speech-neuroprosthesis]] systems they emit phoneme sequences that alanguage model then resolves into text. For surface recordings in [[ecog-interfaces]] the input[language model](https://aiwiki.ai/wiki/large_language_model) then resolves into text. For surface recordings in [[ecog-interfaces]] the inputfeatures differ but the architecture does not. At the opposite extreme, a sixteen-electrodefeatures differ but the architecture does not. At the opposite extreme, a sixteen-electrodeendovascular device such as the [[stentrode]] supports only a binary classifier, and the decodingendovascular device such as the [[stentrode]] supports only a binary classifier, and the decodingproblem shrinks accordingly — the design of a decoder is inseparable from the sensor feeding it.problem shrinks accordingly — the design of a decoder is inseparable from the sensor feeding it.removed, struck through added, underlinedLine numbers count the serialised markdown of each revision, frontmatter included.
191 lines are the same in both revisions, and 185 of them are not shown. Neither revision here is necessarily the current article — read Neural decoding for the text this wiki stands behind today.