Pretraining on hashtags from billions of social media images improved several vision tasks and gave the paper's reported best ImageNet-1k accuracy: 85.4% top-1.

The paper

The paper's details are OpenAlex's; the citation count is OpenAlex's, 14 Oct 2026. The line on the paper is machine-written, as noted under Why it matters.

Why it matters

Written by Claude (claude-sonnet-5-5) on 14 Oct 2026 from the paper's abstract (as arXiv publishes it) and its OpenAlex record. Machine-written context to help a reader: it is not evidence, it moves no number, and it may be wrong. The quoted sentence is the claim; where it stands is computed from the record. If it misreads the paper, tell the stewards.

The story so far

1. 1

   What the authors did

   The authors trained large convolutional networks to predict hashtags on billions of social media images, then transferred them to image classification and object detection tasks, including ImageNet-1k. They also ran extensive experiments on pretraining scale.

   Machine-written from the paper's abstract, as noted under Why it matters.

2. 2

   What they found

   • Training for large-scale hashtag prediction leads to excellent results when transferred to other tasks.
   • Improvements are shown on several image classification and object detection tasks.
   • The reported ImageNet-1k single-crop result is 85.4% top-1 accuracy (97.6% top-5), described as the highest to date.

   Machine-written from the paper's abstract, as noted under Why it matters.

3. 3

   What has been checked on Ecdysis

   Exuvia registered the claim on 14 October 2026, with a test written from the paper. No check has been filed yet.

What would check it

How far it has been checked

1. 1
   The object itself, checked againverification · not yet

   Not yet: re-run the paper's analysis on its own data, where the authors have published it.

2. 2
   New instances of the constructionreproduction · not yet

   Not yet: the same construction run afresh.

3. 3
   The designrobustness tests and arguments · not yet

   Nothing yet: change the method or the data and see whether it holds (a robustness test), or argue that the method does not test what the claim says.

How sure is the record?

Share this finding

Ready-made posts, written from the record. You post them yourself, from your own account; nothing is ever posted for anyone.

Click a post's text to select all of it. Both posts give the claim's standing on the record, and the longer one says what the checks show and what they do not; the wording changes when the record does. The longer post quotes the paper first, then gives the machine-written headline, marked as such; edit it as you like. To cite the claim, see Cite this claim.

What would prove it wrong

Refuted if a reproducible implementation using the same network architecture, training procedure and publicly available hashtag data fails to achieve at least 84.4% top‑1 accuracy on ImageNet‑1k single‑crop.

The test as Exuvia registered it on 14 Oct 2026, written from the paper's words.

The wider literature

Earlier work it rests on, as the citing paper says

• uncheckedOn the ImageNet-1K dataset, we empirically show that even under the restricted condition of maintaining complexity, increasing cardinality is able to improve classification accuracy.takes its method from, as the citing paper says · from the literature · identified by Exuvia: “We use residual networks with grouped convolutional layers, called ResNeXt [15].”
• uncheckedSpecifically, we show no loss of accuracy when training with large minibatch sizes up to 8192 images.takes its method from, as the citing paper says · from the literature · identified by Exuvia: “To make training at this scale practical, we adopt a distributed synchronous implementation of stochastic gradient descent with large (8k image) minibatches, f…”