PulseAugur
EN
LIVE 14:56:04

Feedback Alignment training method improved with new dimensionality techniques

Researchers have identified a key limitation in Feedback Alignment (FA), a method for training neural networks that bypasses the biological implausibility of backpropagation. They found that FA's error signals have a lower rank than those used in backpropagation, restricting the exploration of the parameter space and hindering its scalability to deeper architectures. To address this, the study proposes two mechanisms: an optimizer called Muon that orthogonalizes weight updates and hidden activity normalization, which promotes activation orthogonality. These methods significantly improve FA's performance on benchmarks like CIFAR100, suggesting that increasing the dimensionality of update geometry is crucial for scaling FA as an alternative to backpropagation. AI

IMPACT Introduces techniques to improve training efficiency and scalability for neural networks, potentially enabling more complex models.

RANK_REASON Academic paper detailing a novel method for improving neural network training.

Read on arXiv cs.LG →

AI-generated summary · Google Gemini · from 2 sources. How we write summaries →

COVERAGE [2]

  1. arXiv cs.LG TIER_1 English(EN) · Gauthier Boeshertz, Razvan Pascanu, Claudia Clopath ·

    Overcoming Rank Collapse in Feedback Alignment

    arXiv:2606.11123v1 Announce Type: new Abstract: Backpropagation (BP) is widely viewed as biologically implausible, in part because it requires feedback weights to be the transpose of forward weights for error propagation. Interestingly, when training a network with fixed random f…

  2. arXiv cs.LG TIER_1 English(EN) · Claudia Clopath ·

    Overcoming Rank Collapse in Feedback Alignment

    Backpropagation (BP) is widely viewed as biologically implausible, in part because it requires feedback weights to be the transpose of forward weights for error propagation. Interestingly, when training a network with fixed random feedback weights to circumvent this issue, learni…