Research map / Vision-language and multimodal / Human and embodied modeling
OBELICS: An Open Web-Scale Filtered Dataset of Interleaved Image-Text Documents
Hugo Laurençon, Lucile Saulnier, Leo Tronchon, Stas Bekman, Amanpreet Singh, Anton Lozhkov, Thomas Wang, Siddharth Karamcheti, Alexander M Rush, Douwe Kiela, Matthieu Cord, Victor Sanh
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| train on interleaved web documents (OBELICS) instead of image-text pairs (LAION) (ablation) | 9B Flamingo-style model · aggregated 4-shot over 8 benchmarks | average performance | Improved |
| train on interleaved documents instead of image-text pairs (ablation) | 9B model · captioning, classification, OCR | task performance | Hurt |
| mix OBELICS with image-text pairs (ablation) | 9B model · aggregated 4-shot | average performance | Improved |
| train on OBELICS instead of mmc4 (IDEFICS-9B vs OpenFlamingo-9B) | 9B model · aggregated 8 benchmarks | aggregate score: 55.8 -> 56.5 | Improved |
| IDEFICS-80B vs Flamingo-80B at 32 shots | IDEFICS · 8 multimodal benchmarks | task scores: COCO 113.8 -> 116.6; HatefulMemes 70.0 -> 52.5 | Mixed |
| stricter filtering thresholds | OBELICS filtering · dataset quality | document quality | No clear effect |
Closest papers by meaning
- OmniCorpus: A Unified Multimodal Corpus of 10 Billion-Level Images Interleaved with Text (ICLR 2025 · similarity 0.91)
- CoMM: A Coherent Interleaved Image-Text Dataset for Multimodal Understanding and Generation (CVPR 2025 · similarity 0.86)
- A Picture is Worth More Than 77 Text Tokens: Evaluating CLIP-Style Models on Dense Captions (CVPR 2024 · similarity 0.86)
- MINT-1T: Scaling Open-Source Multimodal Data by 10x: A Multimodal Dataset with One Trillion Tokens (NeurIPS 2024 · similarity 0.86)
- T-MARS: Improving Visual Representations by Circumventing Text Feature Learning (ICLR 2024 · similarity 0.86)
- Towards Natural Language-Based Document Image Retrieval: New Dataset and Benchmark (CVPR 2025 · similarity 0.86)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
