Research map / Vision-language and multimodal / Multimodal LLMs
X-Prompt: Generalizable Auto-Regressive Visual Learning with In-Context Prompting
Zeyi Sun, Ziyang Chu, Pan Zhang, Tong Wu, Yuhang Zang, Xiaoyi Dong, Yuanjun Xiong, Dahua Lin, Jiaqi Wang
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 |
|---|---|---|---|
| add text prediction of dense image descriptions (ablation) | X-Prompt (Chameleon) · GenEval | overall: 0.49 -> 0.57 | Improved |
| add retrieval-augmented in-context example (RAIE) | X-Prompt · MagicBrush test | CLIPimg: 0.823 -> 0.862 | Improved |
| add retrieval-augmented in-context example (RAIE) | X-Prompt · MagicBrush test | CLIPdir: 0.067 -> 0.097 | Improved |
| add in-context example on unseen low-light task (ablation) | X-Prompt · LOL (novel task) | PSNR: 9.140 -> 17.22 | Improved |
| add in-context example on unseen derain task (ablation) | X-Prompt · Rain100H (novel task) | PSNR: 7.924 -> 18.91 | Improved |
| add in-context example on unseen depth palette (ablation) | X-Prompt · NYU-v2 novel color spectrum | RMSE: 0.745 -> 0.352 | Improved |
| add X-Prompt token compression (ablation) | X-Prompt · LOL (novel task) | PSNR: 17.00 -> 17.22 | Improved |
| add X-Prompt token compression (ablation) | X-Prompt · Rain100H (novel task) | PSNR: 18.10 -> 18.91 | Improved |
| add X-Prompt token compression (ablation) | X-Prompt · NYU-v2 novel color spectrum | RMSE: 0.390 -> 0.352 | Improved |
| add X-Prompt token compression (ablation) | X-Prompt · Image inpainting | LPIPS: 0.44 -> 0.34 | Improved |
| add X-Prompt token compression (ablation) | X-Prompt · Colorization | MSE: 0.58 -> 0.49 | Improved |
| use X-Prompt in-context vs OmniGen in-context | X-Prompt vs OmniGen · Rain100H (novel task) | PSNR: 13.14 -> 18.91 | Improved |
Closest papers by meaning
- IPO: Interpretable Prompt Optimization for Vision-Language Models (NeurIPS 2024 · similarity 0.91)
- LaViP: Language-Grounded Visual Prompting (AAAI 2024 · similarity 0.91)
- Concept-Guided Prompt Learning for Generalization in Vision-Language Models (AAAI 2024 · similarity 0.90)
- Learning to Prompt with Text Only Supervision for Vision-Language Models (AAAI 2025 · similarity 0.90)
- Embracing Collaboration Over Competition: Condensing Multiple Prompts for Visual In-Context Learning (CVPR 2025 · similarity 0.90)
- PromptKD: Unsupervised Prompt Distillation for Vision-Language Models (CVPR 2024 · 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.
