LINKED LIST [txt mode] ▸ Augmented Lagrangian Predictive Codin...
home explore | log in

Augmented Lagrangian Predictive Coding: training 1000-layer networks without backpropagation

pub.sakana.ai · first added by @kmurph · 2026-10-01 · 1 upvotes

log in to save, upvote or flag this.


─── In 0 lists ─────────────────────────────────────────

(not in any lists yet)

* Dust: Pretraining Transformers Without Backpropagation (from the discussion)

─── Discussions ────────────────────────────────────────

* Backprop Alternative: Augmented Lagrangian Predictive Coding
128 pts · 45 comments · node

─── From the discussion ────────────────────────────────

* https://direct.mit.edu/neco/article/34/6/1329/110646/Predictive-Coding-Approximates-Backprop-Along
direct.mit.edu · node
* Training a Predictive Coding Network on ImageNet using Equilibrium Propagation
arxiv.org · node
* Augmented Lagrangian Predictive Coding
arxiv.org · node