Pith Number

pith:7VITVVLL

pith:2026:7VITVVLL3U272IR42MMYWQ4ZLK

not attested not anchored not stored refs pending

Image Generators are Generalist Vision Learners

Howard Zhou, Huizhong Chen, Jean-Baptiste Alayrac, Jonathan T. Barron, Kaiming He, Karen Truong, Kyle Genova, Mandy Guo, Nithish Kannen, Oliver Wang, Paul Voigtlaender, Radu Soricut, Saining Xie, Shangbang Long, Sherry Ben, Shuyang Sun, Songyou Peng, Suhas Yogin, Thomas Funkhouser, Valentin Gabeur, Wenlei Zhou, Yanan Bao, Yandong Li, Yiming Gu, Zhicheng Wang

Image generation pretraining builds general visual representations that reach SOTA on perception tasks when outputs are cast as RGB images.

arxiv:2604.20329 v3 · 2026-04-22 · cs.CV · cs.AI

Open paper page JSON Open Graph Bundle Merged state Verified badge What is a Pith Number?

Add to your LaTeX paper

\usepackage{pith}
\pithnumber{7VITVVLL3U272IR42MMYWQ4ZLK}

Prints a linked badge after your title and injects PDF metadata. Compiles on arXiv. Learn more · Embed verified badge

Record completeness

1 Bitcoin timestamp

2 Internet Archive

3 Author claim open · sign in to claim

4 Citations open

5 Replications open

✓ Portable graph bundle live · download bundle · merged state

The bundle contains the canonical record plus signed events. A mirror can host it anywhere and recompute the same current state with the deterministic merge algorithm.

Claims

C1strongest claim

image generation training serves a role similar to LLM pretraining, and lets models learn powerful and general visual representations that enable SOTA performance on various vision tasks.

C2weakest assumption

That the reported SOTA results stem primarily from the generative pretraining rather than from the specific data mixture, evaluation protocol, or implicit leakage in the instruction-tuning stage.

C3one line summary

Image generation pretraining builds generalist vision models that reach SOTA on 2D and 3D perception tasks by reframing them as RGB image outputs.

Formal links

2 machine-checked theorem links

Cited by

13 papers in Pith

From RGB Generation to Dense Field Readout: Pixel-Space Dense Prediction with Text-to-Image Models

Vision as Unified Multimodal Generation

Prompting Diffusion Models for Zero-Shot Instance Segmentation

ImageWAM: Do World Action Models Really Need Video Generation, or Just Image Editing?

Modality Forcing for Scalable Spatial Generation

Receipt and verification

First computed	2026-06-05T00:13:46.609853Z
Builder	pith-number-builder-2026-05-17-v1
Signature	Pith Ed25519 (`pith-v1-2026-05`) · public key
Schema	pith-number/v1.0

Canonical hash

fd513ad56bdd35fd223cd3198b43995a84f4667c52311226ad625c8e25832ead

Aliases

arxiv: 2604.20329 · arxiv_version: 2604.20329v3 · doi: 10.48550/arxiv.2604.20329 · pith_short_12: 7VITVVLL3U27 · pith_short_16: 7VITVVLL3U272IR4 · pith_short_8: 7VITVVLL

Agent API

Resolver JSON Graph JSON Events JSON Schema Signing key

Verify this Pith Number yourself

curl -sH 'Accept: application/ld+json' https://pith.science/pith/7VITVVLL3U272IR42MMYWQ4ZLK \
  | jq -c '.canonical_record' \
  | python3 -c "import sys,json,hashlib; b=json.dumps(json.loads(sys.stdin.read()), sort_keys=True, separators=(',',':'), ensure_ascii=False).encode(); print(hashlib.sha256(b).hexdigest())"
# expect: fd513ad56bdd35fd223cd3198b43995a84f4667c52311226ad625c8e25832ead

Canonical record JSON

{
  "metadata": {
    "abstract_canon_sha256": "7811065d4485055e432efffe682bf6adfc096c79f04da95cd16c4bf70f5273f2",
    "cross_cats_sorted": [
      "cs.AI"
    ],
    "license": "http://arxiv.org/licenses/nonexclusive-distrib/1.0/",
    "primary_cat": "cs.CV",
    "submitted_at": "2026-04-22T08:23:48Z",
    "title_canon_sha256": "6e0eda6351a0995d80d3309c161471b9ef58e5ca7e667ed24766f606b6b05b72"
  },
  "schema_version": "1.0",
  "source": {
    "id": "2604.20329",
    "kind": "arxiv",
    "version": 3
  }
}