Research map / Vision-language and multimodal / Multimodal LLMs
ProLaViT: Learning Progressive Latent Visual Thoughts in Structured Latent Space
Peiming Li, Yifan Wang, Xiaotian Zhang, Zhiyuan Hu, Shiyu Li, Zheng Wei, Yang Tang
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 |
|---|---|---|---|
| replace one-step latent prediction with progressive multi-step latent chain (ablation) | ProLaViT · avg of 4 benchmarks | accuracy (%): 69.28 -> 76.73 (+7.45) | Improved |
| progressive chain instead of one-step latent prediction (ablation) | ProLaViT · ChartQA | accuracy (%): 63.02 -> 78.99 (+15.97) | Improved |
| progressive chain instead of one-step latent prediction (ablation) | ProLaViT · BLINK-Jigsaw | accuracy (%): 66.00 -> 76.00 (+10.00) | Improved |
| add distance-weighted diversity loss (ablation) | ProLaViT · avg of 4 benchmarks | accuracy (%): 75.14 -> 77.34 | Improved |
| add distance-weighted diversity loss (ablation) | ProLaViT · BLINK-Jigsaw | accuracy (%): 71.33 -> 76.00 (+4.67) | Improved |
| add distance-weighted diversity loss (ablation) | ProLaViT · ChartQA | accuracy (%): 76.18 -> 78.99 (+2.81) | Improved |
| add margin-based diversity loss without distance weights (ablation) | ProLaViT · BLINK-Jigsaw | accuracy (%): 71.33 -> 67.33 (-4.00) | Hurt |
| add margin-based diversity loss without distance weights (ablation) | ProLaViT · avg of 4 benchmarks | accuracy (%): 75.14 -> 74.32 | Hurt |
| use native Qwen2.5-VL ViT teacher instead of DINOv2 (ablation) | ProLaViT · avg of 4 benchmarks | accuracy (%): 75.57 -> 77.34 | Improved |
| use native ViT teacher instead of SDXL-VAE (ablation) | ProLaViT · VisPuzzle | accuracy (%): 45.50 -> 74.25 | Improved |
| use canonical coarse-to-fine step order instead of random order (ablation) | ProLaViT · avg of 5 benchmarks | accuracy (%): 71.16 -> 77.40 (-6.24 for random) | Improved |
| use canonical order instead of random on CV-Bench (ablation) | ProLaViT · CV-Bench | accuracy (%): 69.14 -> 77.66 (-8.52 for random) | Improved |
Closest papers by meaning
- Latent Visual Reasoning (ICLR 2026 · similarity 0.92)
- Monet: Reasoning in Latent Visual Space Beyond Image and Language (CVPR 2026 · similarity 0.92)
- SLVR: Structured Latent Visual Reasoning via Human-like Reasoning Flows (NeurIPS 2026 · similarity 0.91)
- UniVLR: Unifying Text and Vision in Visual Latent Reasoning for Multimodal LLMs (NeurIPS 2026 · similarity 0.91)
- Fill the GAP: A Granular Alignment Paradigm for Visual Reasoning in Multimodal Large Language Models (NeurIPS 2026 · similarity 0.91)
- Vision-aligned Latent Reasoning for Multi-modal Large Language Model (ICML 2026 · similarity 0.90)
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.
