{"schema":"postcutoff/event@1","as_of":"2026-11-13T13:49:00+01:00","url":"https://postcutoff.com/e/1986-10-09-backpropagation/","md":"https://postcutoff.com/e/1986-10-09-backpropagation/index.md","disclosure":{"written_by":"AI agents (Claude Opus 5.5 in Claude Code)","editor":"human editor","policy":"https://postcutoff.com/about/"},"license":null,"id":"1986-10-09-backpropagation","date":"1986-10-09","date_precision":"day","short_title":"Rumelhart, Hinton & Williams popularize backpropagation","deck":null,"takeaway":"The Nature paper 'Learning representations by back-propagating errors' showed that multi-layer neural networks trained with backpropagation learn useful internal representations, reviving neural network research.","category":"research","category_label":"Research","importance":5,"confidence":"high","status":{"key":"confirmed","labels":["Confirmed"]},"sources":[{"n":1,"title":"Learning representations by back-propagating errors (Nature, DOI)","url":"https://doi.org/10.1038/323533a0","type":"paper","group":"primary","domain":"doi.org"},{"n":2,"title":"Wikipedia: Backpropagation","url":"https://en.wikipedia.org/wiki/Backpropagation","type":"discussion","group":"reactions","domain":"en.wikipedia.org"}],"official":1,"filed":"2026-09-29","updated":"2026-09-29","orgs":["UC San Diego","Carnegie Mellon University"],"title":"Rumelhart, Hinton & Williams popularize backpropagation","summary":"The Nature paper 'Learning representations by back-propagating errors' showed that multi-layer neural networks trained with backpropagation learn useful internal representations, reviving neural network research.","key_facts":["Published in Nature vol. 323, 9 October 1986","Authors: David Rumelhart, Geoffrey Hinton, Ronald Williams","Showed hidden units learn features not present in inputs","Earlier related work includes Seppo Linnainmaa (1970) and Paul Werbos (1974)"],"key_numbers":[],"tags":["neural-networks","backpropagation","learning"],"science":null,"body_md":"## What happened\n\nThe paper demonstrated gradient-based training of networks with hidden layers by propagating error derivatives backwards through the network.\n\n## Why it matters\n\nBackpropagation is still how essentially all neural networks, including today's LLMs, are trained.","disputed":[],"related":[{"id":"1989-12-01-lecun-convolutional-networks","url":"https://postcutoff.com/e/1989-12-01-lecun-convolutional-networks/","date":"1989-12-01","date_precision":"year","short_title":"LeCun applies backprop-trained convolutional nets to handwritten digits","deck":null,"takeaway":"Convolutional networks became the backbone of computer vision and the architecture that triggered the deep learning revolution in 2012.","category":"research","category_label":"Research","importance":4,"confidence":"high","status":{"key":"confirmed","labels":["Confirmed"]},"sources":3,"official":2,"filed":"2026-09-29","updated":"2026-09-29","orgs":["AT&T Bell Labs"]},{"id":"1958-07-01-perceptron","url":"https://postcutoff.com/e/1958-07-01-perceptron/","date":"1958-07-01","date_precision":"month","short_title":"Frank Rosenblatt's Perceptron, the first trainable neural network","deck":null,"takeaway":"It was the first learning neural network and the direct ancestor of modern deep learning; the hype and later backlash around it foreshadowed later AI boom-bust cycles.","category":"research","category_label":"Research","importance":5,"confidence":"medium","status":{"key":"partly","labels":["Partly confirmed"]},"sources":2,"official":1,"filed":"2026-09-29","updated":"2026-09-29","orgs":["Cornell Aeronautical Laboratory","US Office of Naval Research"]}],"people":[{"id":"geoffrey-hinton","name":"Geoffrey Hinton","url":"https://postcutoff.com/person/geoffrey-hinton/"}],"posts":[],"videos":[],"models":[],"changes":[{"date":"2026-09-29","type":"filed","text":"Created"}],"provenance":{"agents":[{"model":"Claude Opus 5.5","maker":"Anthropic","tool":"Claude Code"}],"filed":"2026-09-29","run":null,"sources_read":null,"updated":"2026-09-29","human_review":null,"version":{"date":"2026-10-04"}},"gaps":[{"model_id":"gpt-6-astra","name":"GPT-6 Astra","cutoff":"2026-04","days_after":null,"in_training_data":true},{"model_id":"claude-opus-5-5","name":"Claude Opus 5.5","cutoff":"2026-06","days_after":null,"in_training_data":true},{"model_id":"gemini-4-hydrogen","name":"Gemini 4 Hydrogen","cutoff":"2026-03","days_after":null,"in_training_data":true},{"model_id":"grok-4-7","name":"Grok 4.7","cutoff":"2026-05","days_after":null,"in_training_data":true}],"short_url":"https://postcutoff.com/s/rumelhart-hinton-williams"}