Pith. sign in

REVIEW 7 cited by

Big Transfer (BiT): General Visual Representation Learning

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

arxiv 1912.11370 v3 pith:UDS5LKDS submitted 2019-12-24 cs.CV cs.LG

Big Transfer (BiT): General Visual Representation Learning

classification cs.CV cs.LG
keywords transferclassdatasetsexamplestaskcifar-10componentsilsvrc-2012
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved
0 comments
Share X Bluesky LinkedIn Reddit HN
read the original abstract

Transfer of pre-trained representations improves sample efficiency and simplifies hyperparameter tuning when training deep neural networks for vision. We revisit the paradigm of pre-training on large supervised datasets and fine-tuning the model on a target task. We scale up pre-training, and propose a simple recipe that we call Big Transfer (BiT). By combining a few carefully selected components, and transferring using a simple heuristic, we achieve strong performance on over 20 datasets. BiT performs well across a surprisingly wide range of data regimes -- from 1 example per class to 1M total examples. BiT achieves 87.5% top-1 accuracy on ILSVRC-2012, 99.4% on CIFAR-10, and 76.3% on the 19 task Visual Task Adaptation Benchmark (VTAB). On small datasets, BiT attains 76.8% on ILSVRC-2012 with 10 examples per class, and 97.0% on CIFAR-10 with 10 examples per class. We conduct detailed analysis of the main components that lead to high transfer performance.

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Forward citations

Cited by 7 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score.

  1. Silhouette Loss: Differentiable Global Structure Learning for Deep Representations

    cs.LG 2026-03 unverdicted novelty 8.0

    Soft Silhouette Loss offers a batch-global differentiable metric to promote intra-class compactness and inter-class separation in learned representations, boosting performance when hybridized with cross-entropy and co...

  2. Swin Transformer: Hierarchical Vision Transformer using Shifted Windows

    cs.CV 2021-03 accept novelty 8.0

    Swin Transformer reaches 87.3% ImageNet accuracy and sets new records on COCO detection and ADE20K segmentation by replacing global self-attention with shifted-window local attention inside a hierarchical pyramid.

  3. Improving Neural Network Training by Decoupling the Magnitude and Direction of Weight Vectors

    cs.LG 2026-06 unverdicted novelty 6.0

    MD Decoupling factorizes weights into fixed-norm directions and learnable per-row/column magnitudes updated at independent rates, improving Adam and Muon training stability and scale transfer without weight decay or warmup.

  4. DexHoldem: Playing Texas Hold'em with Dexterous Embodied System

    cs.RO 2026-05 unverdicted novelty 6.0

    DexHoldem is a new benchmark providing 1,470 teleoperated demonstrations across 14 manipulation primitives, plus standardized tests for dexterous policy execution and agentic perception in a physical Texas Hold'em setting.

  5. Florence: A New Foundation Model for Computer Vision

    cs.CV 2021-11 unverdicted novelty 6.0

    Florence is a new vision foundation model that learns universal visual-language representations from web-scale data and reports state-of-the-art results on 44 benchmarks including 83.74% zero-shot ImageNet top-1 accuracy.

  6. Improving Neural Network Training by Decoupling the Magnitude and Direction of Weight Vectors

    cs.LG 2026-06 conditional novelty 5.0

    Splitting weight matrices into a fixed-norm direction and learnable per-row/column magnitudes improves LLM training over AdamW/Muon, removes weight decay and warmup, and transfers the optimal LR across width.

  7. A Large-Scale Study on the Accuracy vs Cost Trade-offs of Training and Evaluation Settings in Fine-Grained Image Recognition

    cs.CV 2026-05 unverdicted novelty 5.0

    Large-scale experiments demonstrate that data-aware augmentations applied only during training allow fine-grained image models to reach high accuracy without using discriminative crops at inference, lowering costs.