Pith. sign in

REVIEW 4 cited by

VisualWebInstruct: Scaling up Multimodal Instruction Data through Web Search

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 2503.10582 v2 pith:JATOCAGP submitted 2025-03-13 cs.CV cs.AIcs.CL

classification cs.CVcs.AIcs.CL
keywords datadatasetmodelsmultimodalpairssearchtasksvisualwebinstruct
verification ladder T0 review T1 audit T2 compute T3 formal
0 comments
read the original abstract

Vision-Language Models have made significant progress on many perception-focused tasks. However, their progress on reasoning-focused tasks remains limited due to the lack of high-quality and diverse training data. In this work, we aim to address the scarcity of reasoning-focused multimodal datasets. We propose VisualWebInstruct, a novel approach that leverages search engines to create a diverse and high-quality dataset spanning multiple disciplines, including mathematics, physics, finance, and chemistry, etc. Starting with a meticulously selected set of 30,000 seed images, we employ Google Image Search to identify websites containing similar images. We collect and process HTML data from over 700K unique URLs. Through a pipeline of content extraction, filtering, and synthesis, we construct a dataset of approximately 900K question-answer (QA) pairs, with 40% consisting of visual QA pairs and the remaining comprising text-based QA pairs. Models fine-tuned on VisualWebInstruct demonstrate significant performance improvements: (1) fine-tuning on Llava-OV results in 10-20 absolute points improvement across benchmarks, and (2) fine-tuning from MAmmoTH-VL yields a 5 absolute points gain across benchmarks. Our best model, MAmmoTH-VL2, achieves state-of-the-art performance within the 10B parameter class on MMMU-Pro (40.7), MathVerse (42.6), and DynaMath (55.7). These results highlight the effectiveness of our dataset in enhancing the reasoning capabilities of vision-language models for complex multimodal tasks.

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 4 Pith papers

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

  1. IMPRINT: Image-Conditioned Query Enrichment for Long-Tail Object Goal Navigation

    cs.CV 2026-07 conditional novelty 6.0 of 10

    Web-image-enriched queries improve open-vocabulary map grounding and ObjectNav success, especially for fine-grained categories, with gains limited by downstream detectors.

  2. Sparse-LaViDa: Sparse Multimodal Discrete Diffusion Language Models

    cs.CV 2025-12 conditional novelty 6.0 of 10

    Sparse-LaViDa speeds up masked diffusion LLMs by truncating masked tokens, using register tokens and a step-causal attention mask, reaching ~2-2.8x speedup over LaViDa-O with comparable generation quality.

  3. Lavida-O: Elastic Large Masked Diffusion Models for Unified Multimodal Understanding and Generation

    cs.CV 2025-09 conditional novelty 6.0 of 10

    Lavida-O introduces an elastic mixture-of-transformers architecture that brings high-resolution text-to-image generation, object grounding, and image editing into a single masked diffusion model, using planning and se...

  4. Is Extending Modality The Right Path Towards Omni-Modality?

    cs.CL 2025-06 conditional novelty 6.0 of 10

    Fine-tuning LLMs on extra modalities improves some knowledge tasks but degrades reasoning and instruction-following; weighted model merging preserves language ability better than training one model on all modalities.

Pith tools