REVIEW 2 cited by
Probabilistic Binary Neural Networks
Not yet reviewed by Pith; the record is open.
This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.
SPECIMEN: schema-true, not a live event
T0 review · schema-true
One-sentence machine reading of the paper's core claim.
pith:XXXXXXXX · record.json · timestamp
read the original abstract
Low bit-width weights and activations are an effective way of combating the increasing need for both memory and compute power of Deep Neural Networks. In this work, we present a probabilistic training method for Neural Network with both binary weights and activations, called BLRNet. By embracing stochasticity during training, we circumvent the need to approximate the gradient of non-differentiable functions such as sign(), while still obtaining a fully Binary Neural Network at test time. Moreover, it allows for anytime ensemble predictions for improved performance and uncertainty estimates by sampling from the weight distribution. Since all operations in a layer of the BLRNet operate on random variables, we introduce stochastic versions of Batch Normalization and max pooling, which transfer well to a deterministic network at test time. We evaluate the BLRNet on multiple standardized benchmarks.
Forward citations
Cited by 2 Pith papers
-
Energy-Efficient Supervised Learning with a Binary Stochastic Forward-Forward Algorithm
Binary stochastic forward-forward training reaches near-real-valued forward-forward accuracy on image benchmarks while estimating 10-100x energy savings in p-bit hardware.
-
A Principled Bayesian Framework for Training Binary and Spiking Neural Networks
A variational Bayesian framework with importance-weighted straight-through estimators trains binary and spiking networks without normalization layers, matching surrogate-gradient baselines on CIFAR-10, DVS Gesture, and SHD.
Discussion (0). Sign in to comment.